AN INTRODUCTION TO

EXCEL VBA PROGRAMMING

with Applications in Finance and Insurance
 AN INTRODUCTION TO
EXCEL VBA PROGRAMMING
with Applications in Finance and Insurance

GUOJUN GAN
University of Connecticut
Storrs
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2017 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

Version Date: 20170217

International Standard Book Number-13: 978-1-1381-9715-2 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access
www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc.
(CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization
that provides licenses and registration for a variety of users. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
To all my students.
Contents

List of Figures xi

List of Tables xiii

Preface xv

I VBA Preliminaries 1

1 Introduction to VBA 3
1.1 Getting Started . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 VBA Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 The Excel Macro Recorder . . . . . . . . . . . . . . . . . . . . 13
1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

2 Excel Objects 17
2.1 Excel Object Model . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Application Object . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3 Workbook Objects . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4 Worksheet Objects . . . . . . . . . . . . . . . . . . . . . . . . 28
2.5 Range Object . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.6 WorksheetFunction Object . . . . . . . . . . . . . . . . . . . 37
2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3 Variables, Data Types, and Scopes 43

3.1 Variable Declaration and Data Types . . . . . . . . . . . . . . 43
3.2 Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.3 Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.4 Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.5 Dates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.6 Scopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

vii
viii Contents

4 Operators and Control Structures 75

4.1 Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2 Flow Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.3 Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

5 Functions, Events, and File IO 95

5.1 User-Defined Functions . . . . . . . . . . . . . . . . . . . . . 95
5.2 Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.3 File IO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

6 Error Handling and Debugging 109

6.1 Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2 Debugging VBA Code . . . . . . . . . . . . . . . . . . . . . . . 115
6.3 Best Practices of VBA Coding . . . . . . . . . . . . . . . . . . 120
6.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

II Applications 127

7 Generating Payment Schedules 129

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
7.2 Public Holidays in the United States . . . . . . . . . . . . . . 130
7.3 Julian and Gregorian Calendars . . . . . . . . . . . . . . . . . 131
7.4 Day Count Conventions . . . . . . . . . . . . . . . . . . . . . 133
7.5 Business Day Conventions . . . . . . . . . . . . . . . . . . . . 134
7.6 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.6.1 MDate Module . . . . . . . . . . . . . . . . . . . . . . 135
7.6.2 MHoliday Module . . . . . . . . . . . . . . . . . . . . 147
7.6.3 MSchedule Module . . . . . . . . . . . . . . . . . . . 151
7.6.4 MInterface Module . . . . . . . . . . . . . . . . . . . . 156
7.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

8 Bootstrapping Yield Curves 161

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
8.2 Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
8.3 Bootstrapping Yield Curves . . . . . . . . . . . . . . . . . . . 164
8.4 Finding Roots of an Equation . . . . . . . . . . . . . . . . . . 167
8.5 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 168
8.5.1 MCurve Module . . . . . . . . . . . . . . . . . . . . . 168
8.5.2 MInterface Module . . . . . . . . . . . . . . . . . . . . 179
Contents ix

8.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

9 Generating Risk-Neutral Scenarios 183

9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.2 Black–Scholes Model . . . . . . . . . . . . . . . . . . . . . . . 184
9.3 Generating Random Normal Numbers . . . . . . . . . . . . 186
9.4 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 187
9.4.1 MGenerator Module . . . . . . . . . . . . . . . . . . . 187
9.4.2 MInterface Module . . . . . . . . . . . . . . . . . . . . 195
9.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

10 Valuing a GMDB 199

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
10.2 Life Table Construction . . . . . . . . . . . . . . . . . . . . . 201
10.3 GMDB Valuation . . . . . . . . . . . . . . . . . . . . . . . . . 202
10.4 Greek Calculation . . . . . . . . . . . . . . . . . . . . . . . . . 205
10.5 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 205
10.5.1 MLifeTable Module . . . . . . . . . . . . . . . . . . . 206
10.5.2 MGMDB Module . . . . . . . . . . . . . . . . . . . . . 209
10.5.3 MInterface Module . . . . . . . . . . . . . . . . . . . . 214
10.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

11 Connecting to Databases 217

11.1 ActiveX® Data Objects . . . . . . . . . . . . . . . . . . . . . . 218
11.2 SQL Queries . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
11.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.3.1 MDatabase Module . . . . . . . . . . . . . . . . . . . 221
11.3.2 MInterface Module . . . . . . . . . . . . . . . . . . . . 229
11.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

12 Object-Oriented Programming 233

12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
12.2 Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
12.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 234
12.3.1 CZeroCurve Module . . . . . . . . . . . . . . . . . . . 235
12.3.2 CSwap Module . . . . . . . . . . . . . . . . . . . . . . 238
12.3.3 CBootstrapper Module . . . . . . . . . . . . . . . . . 241
12.3.4 OO Bootstrapping Tool . . . . . . . . . . . . . . . . . 246
12.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
x Contents

A Solutions to Selected Exercises 251

A.1 Introduction to VBA . . . . . . . . . . . . . . . . . . . . . . . 251
A.2 Excel® Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
A.3 Variables, Data Types, and Scopes . . . . . . . . . . . . . . . 253
A.4 Operators and Control Structures . . . . . . . . . . . . . . . 258
A.5 Functions, Events, and File IO . . . . . . . . . . . . . . . . . . 260
A.6 Error Handling and Debugging . . . . . . . . . . . . . . . . . 262
A.7 Generating Payment Schedules . . . . . . . . . . . . . . . . . 262
A.8 Bootstrapping Yield Curves . . . . . . . . . . . . . . . . . . . 266
A.9 Generating Risk-Neutral Scenarios . . . . . . . . . . . . . . . 268
A.10 Valuing a GMDB . . . . . . . . . . . . . . . . . . . . . . . . . . 270
A.11 Connecting to Databases . . . . . . . . . . . . . . . . . . . . 273

References 275

Index 278

Index of VBA Keywords 281

List of Figures

1.1 Excel 2013 interface in a Windows computer . . . . . . . . . . 4

1.2 Visual Basic Editor in a Windows computer . . . . . . . . . . . 5
1.3 Excel 2010 interface on a Mac computer . . . . . . . . . . . . 6
1.4 Ribbon dialog in Excel 2010 on a Mac computer . . . . . . . . 7
1.5 Visual Basic Editor on a Mac computer . . . . . . . . . . . . . 8
1.6 Macro dialog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.7 A module displayed in the Visual Basic Editor . . . . . . . . . 9
1.8 Inserting a module in the Visual Basic Editor . . . . . . . . . . 10
1.9 The Record Macro dialog . . . . . . . . . . . . . . . . . . . . . 14

2.1 The Excel object hierarchy . . . . . . . . . . . . . . . . . . . . . 18

2.2 The Object Browser in the Visual Basic Editor . . . . . . . . . 18
2.3 A screenshot of the US Life Tables 1999–2001 — Males . . . . 33
2.4 Worksheet functions in the Object Browser . . . . . . . . . . . 41

3.1 An illustration of an explicit variable declaration . . . . . . . 48

3.2 VBA’s built-in color constants . . . . . . . . . . . . . . . . . . . 56
3.3 VBA’s built-in constants for variable types . . . . . . . . . . . . 57

4.1 Conversion functions in the Visual Basic Editor . . . . . . . . 77

5.1 Workbook events in the code window . . . . . . . . . . . . . . 102

6.1 An example illustrating a runtime error . . . . . . . . . . . . . 110

6.2 An example illustrating another runtime error . . . . . . . . . 111
6.3 Checking values of VBA statements in the Immediate
window in Break mode . . . . . . . . . . . . . . . . . . . . . . . 117
6.4 Setting a watch expression in the Watch window . . . . . . . 118
6.5 The Watches window and checking values of VBA
statements in the Immediate window in Break mode . . . . . 118
6.6 The Locals window in Break mode . . . . . . . . . . . . . . . . 119
6.7 Three tiers of an application and their relationships . . . . . 122

xi
xii List of Figures

7.1 Interface of the payment schedule generator . . . . . . . . . . 129

7.2 Interface of a holiday generator . . . . . . . . . . . . . . . . . . 159

8.1 Interface of the yield curve bootstrapper. . . . . . . . . . . . . 161

8.2 Yield curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

9.1 Interface of the risk-neutral scenario generator . . . . . . . . 183

9.2 First few rows and columns of the risk-neutral scenarios . . . 197

10.1 Interface of the GMDB valuation program . . . . . . . . . . . 199

11.1 Interface of the variable annuity inforce tool . . . . . . . . . . 217

11.2 Inforce10k table in an Access® database . . . . . . . . . . . . 218
11.3 Checking the reference for Microsoft ActiveX Data
Objects 2.8 Library . . . . . . . . . . . . . . . . . . . . . . . . . 219
11.4 Output in the worksheet when the first button is clicked . . . 231
11.5 New table in the database when the second button
is clicked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
List of Tables

2.1 Commonly Used Properties of the Application Object . . . 21

2.2 Commonly Used Methods of the Application Object . . . . 22
2.3 File Extensions of Excel Files . . . . . . . . . . . . . . . . . . . 25
2.4 Commonly Used Properties of a Workbook Object . . . . . . . 25
2.5 Commonly Used Methods of a Workbook Object . . . . . . . . 26
2.6 Two Properties of the Workbooks Object . . . . . . . . . . . . 27
2.7 Some Methods of the Workbooks Object . . . . . . . . . . . . 27
2.8 Commonly Used Properties of a Worksheet Object . . . . . . 28
2.9 Commonly Used Methods of a Worksheet Object . . . . . . . 30
2.10 Some Methods of the Worksheets Collection . . . . . . . . . 31
2.11 Various Ways to Refer to a Range Object . . . . . . . . . . . . . 32
2.12 Some Properties of a Range Object . . . . . . . . . . . . . . . . 35
2.13 Some Methods of a Range Object . . . . . . . . . . . . . . . . . 36
2.14 Some Worksheet Functions . . . . . . . . . . . . . . . . . . . . 38

3.1 List of Built-In Data Types in VBA . . . . . . . . . . . . . . . . 45

3.2 Some Commonly Used String Functions in VBA . . . . . . . . 59
3.3 VBA’s Built-in Date and Time Functions . . . . . . . . . . . . . 66
3.4 Scopes of a Variable . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.5 Scopes of a Constant . . . . . . . . . . . . . . . . . . . . . . . . 72
3.6 Scopes of a Procedure . . . . . . . . . . . . . . . . . . . . . . . 73

4.1 VBA’s arithmetic operators . . . . . . . . . . . . . . . . . . . . . 75

4.2 VBA’s String Operator . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3 VBA’s Comparison Operators . . . . . . . . . . . . . . . . . . . 79
4.4 VBA’s Logical Operators . . . . . . . . . . . . . . . . . . . . . . 80
4.5 VBA’s Programming Constructs for Flow Control . . . . . . . . 81
4.6 VBA’s Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

5.1 Workbook Events . . . . . . . . . . . . . . . . . . . . . . . . . . 102

5.2 Worksheet Events . . . . . . . . . . . . . . . . . . . . . . . . . . 103

6.1 Usage of the On Error Statement . . . . . . . . . . . . . . . . . 113

xiii
xiv List of Tables

6.2 Usage of the Resume Statement . . . . . . . . . . . . . . . . . 113

6.3 Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . 120
6.4 Scope Specifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.5 Array Specifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.6 Data Type Specifiers . . . . . . . . . . . . . . . . . . . . . . . . 121

7.1 Holidays and Observances in the United States . . . . . . . . 131

7.2 Parameters for Calendar Conversion Algorithms . . . . . . . . 132
7.3 Range Names Used by the Payment Schedule Generator . . . 156
7.4 Dates of Easter Sunday from 2010 to 2020 . . . . . . . . . . . . 160

8.1 Range Names Used by the Yield Curve Bootstrapper . . . . . 162

8.2 Interest Rates of Various Tenors . . . . . . . . . . . . . . . . . . 162
8.3 US Swap Rates for Various Maturities on February 8, 2016 . . 165

9.1 Range Names Defined in the Interface of the Risk-Neutral

Scenario Generator . . . . . . . . . . . . . . . . . . . . . . . . . 190

10.1 Range Names Used by the GMDB Valuation Program . . . . . 206

11.1 Range Names Used by the Variable Annuity Inforce Tool . . . 217
Preface

Visual Basic for Applications (VBA) is a programming language created by

Microsoft® that can be used to automate operations in Microsoft Excel,
which is perhaps the most frequently used computer software program for
manipulating data and building models in banks and insurance companies.
One advantage of VBA is that it enables you to do anything that you can do
manually in Excel and do many things that Excel does not allow you to do
manually. As a powerful tool, VBA has been used by actuaries and financial
analysts to build actuarial and financial models.
In the spring of 2016, I was assigned to teach the course “Math3550:
Programming for Actuaries,” which is taken primarily by junior and senior
undergraduate students majoring in actuarial science at the University
of Connecticut. This course explores how an actuary uses computers to
solve common actuarial problems and teaches students how to design,
develop, test, and implement programs using Microsoft Excel with VBA.
Since existing books on Excel VBA do not have exercises or applications
related to actuarial science, I started to write lecture notes for this course.
This textbook has grown out of those lecture notes.
This textbook has been written for undergraduate students majoring in
actuarial science who wish to learn the basic fundamentals and applications
of Excel VBA. In doing so, this book does not assume that readers have any
prior programming experience. This book will also be of use to actuaries
and financial analysts working in insurance companies and banks who wish
to learn Excel VBA.
This textbook is divided into two parts: preliminaries of Excel VBA pro-
gramming and some applications of VBA in finance and insurance. The pre-
liminaries covered in the first part include how to run VBA programs, mod-
ules, best practices of VBA coding, the Excel object model, variables, control
statements, functions, and error handling, among many other things. The
applications of VBA introduced in the second part include generating reg-
ular payment schedules, bootstrapping yield curves, creating risk-neutral
scenarios, pricing a guarantee embedded in a variable annuity contract,
how to connect to databases, and object-oriented programming in VBA.

xv
xvi Preface

The best way to learn programming is by doing. I encourage readers to

practice the VBA code presented in the book. The book also contains many
exercises. Sample solutions of some exercises are given in the appendix
of this book. Readers should explore the exercises before looking at the
solutions.
Finally, I take this opportunity to express my thanks to my students,
friends, and colleagues from the University of Connecticut, who have read
and provided valuable feedback on the draft of this book.

Guojun Gan
Storrs, Connecticut, USA
 Part I

VBA Preliminaries
1
Introduction to VBA

VBA refers to Visual Basic for Applications and is a programming language

created by Microsoft® to automate operations in Microsoft Office® appli-
cations such as Excel® , Access® , Word® , PowerPoint® , and Outlook® . VBA
is a powerful tool that enables you to do whatever you need to do in your
job. With VBA, you can do anything that you can do manually and do many
things that Excel® does not allow you to do manually. In particular, VBA
allows you to

• automate a recurring task.

• automate a repetitive task.
• run a macro automatically if an event occurs.
• build custom functions (i.e., user-defined functions).
• build a customized interface.
• manipulate files and folders.
• manipulate Microsoft Office® applications.
• work with Windows® by calling its Application Programming Interfaces
(APIs).
• work with other applications through Dynamic-Link Libraries (DLLs).

In this chapter, we introduce some basic concepts of VBA programming

in Excel® . After studying this chapter, readers will be able to

• run VBA macros or programs.

• understand regular and class modules.
• use the Excel® macro record to create VBA macros.

3
4 Introduction to VBA

1.1 Getting Started

VBA is a programming language built into Microsoft® applications. If the
computer you use has the Office® suite or an individual application, then
you can use VBA in that computer. For example, if the computer has Mi-
crosoft Excel® , then you can use VBA in the computer.
Since VBA is a programming language built into Microsoft applications,
you cannot use VBA to create standalone applications. VBA programs, also
referred to as macros, exist within a host application, such as Excel and
the workbook containing the VBA code. To run a VBA macro in an Excel
workbook, we must open the workbook.
To create a VBA macro in Excel, we first need to open the Visual Basic
Editor. In a Windows® computer, for example, we can open the Visual
Basic Editor by clicking the icon “Visual Basic” in the left side of the tab
“DEVELOPER.”1 Figure 1.1 shows a screen shot of the Excel 2013 interface in
a Windows® machine. The Visual Basic Editor in a Windows® machine is
shown in Figure 1.2.

Figure 1.1: Excel 2013 interface in a Windows computer

If you use a Mac® computer, you can open the Visual Basic Editor by
clicking the menu item “Tools/Macro/Visual Basic Editor” (see Figure 1.3).
You can also click the “Editor” icon in the “Developer” tab to open the Visual
Basic Editor. If you do not see the “Developer” tab in your version of Excel,
1 New Ribbon interface has been used since Microsoft Office version 2007. In old versions

of Excel, you can open the Visual Basic Editor from the traditional menus.
 Getting Started 5

Figure 1.2: Visual Basic Editor in a Windows computer

you can show the tab by clicking “Excel/Preferences.../Ribon” and selecting

“Developer” in the Ribon dialog, Mac® shown in Figure 1.4. The Visual Basic
Editor in a computer looks different than that in a Windows® machine.
Figure 1.5 shows the Visual Basic Editor in a Mac® computer.
There is a shortcut to open the Visual Basic Editor if you do not want to
click menu items to open it. In a Windows® machine, you can open Excel
and hit “Alt+F11” (i.e., hold the “Alt” key and press the “F11” key) to open
the Visual Basic Editor. On a Mac® computer, you can open Excel and hit
“fn+option+F11” (i.e., hold the “fn” and “option” keys and press the “F11”
key) to open the Visual Basic Editor.
To demonstrate how VBA works in Excel, we can type the following VBA
code in the Visual Basic Editor:
1 Sub HelloWorld ()
2 ThisWorkbook . Worksheets (1) . Cells (1 , 1) = " Hello
World ! "
3 End Sub

The code is contained in a sub procedure, which starts with the keyword
sub and ends with the keyword End Sub. The name of the sub procedure
is “HelloWorld.” Line 2 is the VBA command that writes “Hello World!” to
the first cell of the first worksheet of the current workbook. Note that VBA
6 Introduction to VBA

Figure 1.3: Excel 2010 interface on a Mac computer

code is case insensitive. The Visual Basic Editor changes the first letter of a
keyword to uppercase.
There are several ways to run the VBA macro. The first way to run the
VBA macro is to click the VBA code and press the key “F5.” After the code is
executed, the content of the first cell of the first worksheet will be changed
to “Hello World!"
A second way to run the VBA macro is to open the Macro dialog by
clicking “Macros” button in the left side of the “DEVELOPER” tab. On a
Mac computer, you can open the macro dialog by clicking “Tools/Macro/-
Macros.” The macro dialog is shown in Figure 1.6. You can select the macro
and click the “Run” button to run the macro. The Macro dialog also allows
you to set a shortcut key to your macro (See Exercise 1.1).

Exercise 1.1. Suppose that a workbook contains at least two worksheets.

(a) Write a VBA sub procedure named HelloWorld2 that puts “Hello” in
the first cell in the first row of the second worksheet and “World” in the
second cell in the first row of the second worksheet.
(b) Assign a shortcut key ("Ctrl+Shift+A" on a Windows machine) to the sub
procedure and hit the shortcut key.

h
Getting Started 7

Figure 1.4: Ribbon dialog in Excel 2010 on a Mac computer

8 Introduction to VBA

Figure 1.5: Visual Basic Editor on a Mac computer

Figure 1.6: Macro dialog

 Modules 9

1.2 Modules
A module is a container for VBA code. A module can contain declarations
and procedures. An Excel application can call VBA code contained in one
or more modules to perform specified tasks.

Figure 1.7: A module displayed in the Visual Basic Editor

Figure 1.7 shows a module displayed in the Visual Basic Editor. The
name of the module is MGetStarted. You can change the name of the
module in the Properties window, which is shown in the bottom left of the
figure.
As we can see from the figure, the module contains two declarations
and one procedure. The two declarations are:
1 Option Explicit
2 Dim nCount As Integer

The first statement contains two keywords: Option and Explicit. This
statement will force the explicit variable declaration, that is, you have to
declare a variable explicitly before using the variable. We will see some
examples of this statement in Section 3.1. The second statement declares
an integer variable named nCount.
A module can contain two types of procedures: sub procedures and
function procedures. A sub procedure contains VBA code to perform some
10 Introduction to VBA

operations and does not return a particular result. A function procedure

performs some operations and returns a particular result. We will introduce
function procedures in Section 5.1.
To create a module, we can right click the Modules folder in the Project
window and then click "Insert/Module" (See Figure 1.8). The Project window
is shown in the top left of Figure 1.8. On a Mac computer, you need to hold
the Control key and click the Modules folder.

Figure 1.8: Inserting a module in the Visual Basic Editor

There are two types of modules: standard modules and class modules.
If a module contains procedures that are not associated with any particular
object, then module is a standard module. If a module is associated with a
particular object, then it is a class module.
Class modules are related to object-oriented programming and are
different from standard modules in the way their data is stored. Only one
copy of a standard module’s data is stored. For example, if one part of
your program changes a public variable in a standard module and another
part of your program subsequently reads that variable, then it will get the
same value. Class module data is stored separately for each object. We will
introduce how to use class modules in Section 1.3.
Modules have the following advantages:

• Modules can be exported and imported into other workbooks.

• Modules can be organized in a logical way. For example, you can ded-
 VBA Classes 11

icate a separate module for data access, spreadsheet formatting, and

utilities.

Exercise 1.2. Create a standard module and change its name to MTest.

1.3 VBA Classes

VBA supports object-oriented programming. A class is a core concept of
object-oriented programming. In this section, we introduce how to define
and use classes in VBA.
In VBA, a class is defined in a class module. To show how to create a
class in VBA, we first insert a class module named CStudent and type the
following code in the class module:
1 Option Explicit
2
3 Dim sFirstName As String
4 Dim sLastName As String
5 Dim sGrade As String
6
7
8 Public Property Get FirstName () As String
9 FirstName = sFirstName
10 End Property
11
12 Public Property Let FirstName ( ByVal FirstName As
String )
13 sFirstName = FirstName
14 End Property
15
16 Public Property Get LastName () As String
17 LastName = sLastName
18 End Property
19
20 Public Property Let LastName ( ByVal LastName As String
)
 12 Introduction to VBA

21 sLastName = LastName
22 End Property
23
24 Public Property Get Grade () As String
25 Grade = sGrade
26 End Property
27
28 Public Property Let Grade ( ByVal Grade As String )
29 sGrade = Grade
30 End Property
31
32 Public Sub ShowInfo ()
33 MsgBox sFirstName & " " & sLastName & " 's grade
is " & sGrade
34 End Sub

The name of the class is the same as the name of the class module. From the
above code, we see that the class CStudent has three properties: FirstName,
LastName, and Grade. The class module also contains six property proce-
dures to set and get the values stored in the properties. The sub procedure
ShowInfo is used to display the property values in a message box.
To show how to use the class CStudent, we consider the following sub
procedure:
1 Sub ClassDemo1 ()
2 Dim cS1 As CStudent
3 Dim cS2 As CStudent
4
5 Set cS1 = New CStudent
6 cS1 . FirstName = " Graig "
7 cS1 . LastName = " Moyle "
8 cS1 . Grade = " A "
9
10 Set cS2 = New CStudent
11 cS2 . FirstName = " Violet "
12 cS2 . LastName = " Keen "
13 cS2 . Grade = " B "
14
15 cS1 . ShowInfo
16 cS2 . ShowInfo
17 End Sub

In the above code, we create two objects of the class CStudent by using
the keyword New. We assign the two objects to two variables by using the
 The Excel Macro Recorder 13

keyword Set. Then we call the class’s method ShowInfo to display the
property values of the two objects.

Exercise 1.3. Create a VBA class named CBasket with the following two
properties: iCount and dPrice. The first property stores how many items
in the basket. The second property stores the price of each item. The class
also has a method named CalculateValue to calculate the total value of
the basket. Then write a sub procedure to test the class.

1.4 The Excel Macro Recorder

In previous sections, we created Excel macros manually. Using the Excel
macro recorder is another way to create macros. In this section, we in-
troduce how to use the Excel macro recorder to create macros, and the
advantages and drawbacks of the Excel macro recorder.
To record a macro using the Excel macro recorder, we first click the
“Record Macro” button in the left side of the “Developer” tab (See Figure 1.1).
A dialog named “Record Macro” will pop up. Figure 1.9 shows the dialog on
a Windows machine. We can change the name, create a shortcut key, select
where to store the macro, and input the description in the dialog. In our
example, we use the default settings and just write some description in the
Description textbox. Then we click the “OK” button to start recording. We
select the cell “D3,” type “Hello,” change the cell borders to bold lines, and
select the cell “D5.” Then we click the stop button located in the bottom left
of the Excel window.
The recorded macro is saved to a new module of this workbook. In our
example, the recorded macro was saved to Module2. If we click Module2 in
the Visual Basic Editor, we see the following sub procedure:
1 Sub Macro1 ()
2 '
3 ' Macro1 Macro
4 ' Excel macro recorder demo
5 '
6
 14 Introduction to VBA

Figure 1.9: The Record Macro dialog

7 '
8 Range ( " D3 " ) . Select
9 ActiveCell . FormulaR1C1 = " Hello "
10 Range ( " D3 " ) . Select
11 Selection . Font . Bold = True
12 Selection . Borders ( xlDiagonalDown ) . LineStyle =
xlNone
13 Selection . Borders ( xlDiagonalUp ) . LineStyle =
xlNone
14 With Selection . Borders ( xlEdgeLeft )
15 . LineStyle = xlContinuous
16 . ColorIndex = 0
17 . TintAndShade = 0
18 . Weight = xlMedium
19 End With
20 With Selection . Borders ( xlEdgeTop )
21 . LineStyle = xlContinuous
22 . ColorIndex = 0
23 . TintAndShade = 0
24 . Weight = xlMedium
25 End With
 The Excel Macro Recorder 15

26 With Selection . Borders ( xlEdgeBottom )

27 . LineStyle = xlContinuous
28 . ColorIndex = 0
29 . TintAndShade = 0
30 . Weight = xlMedium
31 End With
32 With Selection . Borders ( xlEdgeRight )
33 . LineStyle = xlContinuous
34 . ColorIndex = 0
35 . TintAndShade = 0
36 . Weight = xlMedium
37 End With
38 Selection . Borders ( xlInsideVertical ) . LineStyle =
xlNone
39 Selection . Borders ( x lIn si de Hor iz on tal ) . LineStyle =
xlNone
40 Range ( " D5 " ) . Select
41 End Sub

In the above code, Lines 2–5 and 7 start with a single quote. These lines are
comments in VBA code.
The Excel macro recorder is useful in that it helps us figure out what
VBA commands we can use to do certain things, such as formatting a range.
However, the Excel macro recorder has some drawbacks. First, the Excel
macro recorder cannot create functions for us. It only produces sub proce-
dures. Second, the VBA code created by the recorder is not compact. It tends
to record more statements than we need. For example, we can write the
following sub procedure that does the same thing as the recorded macro:
1 Sub Macro1Manual ()
2 ' This macro does the same thing as Macro1
recorded by the Excel macro recorder
3 Range ( " D3 " ) . Select
4 ActiveCell . Value = " Hello "
5 ActiveCell . Font . Bold = True
6 Selection . Borders . LineStyle = xlContinous
7 Selection . Borders . Weight = xlMedium
8 End Sub

The macro written manually is much shorter than the recorded macro.

Exercise 1.4. Suppose that the range “A1:B2” contains four numbers. Use
16 Introduction to VBA

the Excel macro recorder to create a macro that does the following opera-
tions: select the range “A1:B2,” copy and paste the range to the range “F1:G2,”
and change the number format of the range “F1:G2” to a percentage with
two decimal places.

Exercise 1.5. Use the Excel macro recorder to record the following steps:
open the file “USLifeTable1999-2001Male.xls” by clicking “File/Open,”
select the range “A25:B134,” copy and paste the range to the range
“A1:B110” of a worksheet in your own workbook, and close the workbook
“USLifeTable1999-2001Male.xls.”

1.5 Summary
In this chapter, we introduced some basic concepts of VBA programming.
In particular, we introduced how to run VBA programs. We also introduced
VBA modules, including standard modules and class modules. Finally, we
introduced the Excel macro recorder, which is a useful tool to create VBA
commands.
There are quite a few books on VBA programming. For a brief history
of VBA, readers are referred to Urtis (2015). Readers who are beginners of
VBA programming are referred to Martin (2012), Walkenbach (2013b), and
Urtis (2015). Advanced readers are referred to Getz and Gilbert (2001) and
Bovey et al. (2009). For financial computing with Excel and VBA, readers are
referred to Lai et al. (2010).
2
Excel Objects

Programming in Visual Basic for Applications (VBA) is related to manipulat-

ing the objects that make up Excel® and other Office applications. In this
chapter, we introduce the Excel object model and some commonly used
Excel objects. After studying this chapter, readers will be able to

• understand the Excel object model.

• use the properties and methods of some common objects, such as
Application, Workbook, Worksheet, Range, and WorksheetFunction.

2.1 Excel Object Model

Object-oriented programming (OOP) is a style of programming based on
the use of objects and their interactions. An object is a data structure that
contains properties and methods. Properties of an object are used to store
data. Methods of an object are used to access and modify the data stored in
properties. An object can contain and interact with other objects.
Figure 2.1 shows a hierarchy of Excel objects. The Application ob-
ject (i.e., Excel itself) is at the top of the hierarchy and contains other ob-
jects. The Workbook object and the WorksheetFunction object are exam-
ples of objects contained in the Application object. The Workbook object
also contains other objects, such as Worksheet, Name, and VBProject. The
Worksheet object contains Name, Range, Cells, and other objects. The
Range object can also contain the Cells object.
A special Excel object is the collection object, which contains objects
of the same type. For example, the Workbooks object contains all currently
open Workbook objects. The Worksheets object contains all Worksheet
objects contained in a particular Workbook object.
Figure 2.1 shows only a subset of the Excel objects. To see the complete
list of Excel objects and their members, we can use the Object Browser

17
18 Excel Objects

Excel Application

Addin Window Workbook WorksheetFunction

Chart Name Worksheet VBProject

Name Range Cells

Cells

Figure 2.1: The Excel object hierarchy

Figure 2.2: The Object Browser in the Visual Basic Editor

 Excel Object Model 19

shown in Figure 2.2. We can open the Object Browser by clicking the menu
item “View/Object Browser.”
We use the dot operator to access an object in the object hierarchy. To
reference a single object from a collection, we use the object’s name or index
number. In the following sub procedure, for example, we first assign the
text “A String” to the second cell in the first row of the first sheet. Then we
use the VBA function Debug.Print to display the value of that cell and the
name of the current workbook in the Immediate window.
1 Sub AccessObject ()
2 Application . ActiveWorkbook . Sheets (1) . Cells (1 , 2) .
Value = " A String "
3 Debug . Print Application . ActiveWorkbook . Sheets ( "
Sheet1 " ) . Cells (1 , 2) . Value
4 Debug . Print Application . ActiveWorkbook . Name
5 End Sub

If the sub procedure is executed, we see the following output in the Imme-
diate window:
1 A String
2 Book1

If the Immediate window is not shown in the Visual Basic Editor, you can
open the Immediate window by clicking the menu item “View/Immediate
Window.”
We can simplify the reference to an object in the object hierarchy by
omitting references to the assumed objects. If an object is the currently
active object, we can omit the reference to the object. For example, we can
rewrite the sub procedure AccessObject as follows:
1 Sub AccessObject2 ()
2 Sheets (1) . Cells (1 , 2) . Value = " A String "
3 Debug . Print ActiveWorkbook . Sheets ( " Sheet1 " ) . Cells
(1 , 2) . Value
4 Debug . Print ActiveWorkbook . Name
5 End Sub

In this version, we omit the reference to the object Application.

Exercise 2.1. Write a sub procedure named DisplayCount to display the

 20 Excel Objects

number of open workbooks and the number of worksheets of the current

workbook in the Immediate window.

Exercise 2.2. Write a sub procedure named CountRowColumn to display

the number of rows and the number of columns of a worksheet in your
Excel application. You can use the objects Rows, Columns, and Count in
your code.

As we mentioned before, an object contains properties and methods. A

method can modify an object’s properties or make the object do something.
The following sub procedure shows how to use the VBA function Copy to
copy the content of a cell to another cell.
1 Sub CopyValue ()
2 ThisWorkbook . WorkSheets (1) . Activate
3 Cells (2 , 2) . ClearContents
4 Debug . Print Cells (1 , 1) . Value
5 Debug . Print Cells (2 , 2) . Value
6 Cells (1 , 1) . Copy Cells (2 , 2)
7 Debug . Print Cells (2 , 2) . Value
8 End Sub

In the above code, we first activate the first worksheet by calling the method
Activate. Then we call the method ClearContents to clear the content of
the cell at the second row and the second column. In Lines 4–5, we call the
VBA function Debug.Print to display the contents of the two cells in the
Immediate window. In Lines 6–7, we copy the content of the first cell to the
second cell and display the content of the second cell again. If we execute
the above code, we see the following output in the Immediate window:
1 Hello World !
2
3 Hello World !

Exercise 2.3. Suppose that a workbook contains at least two worksheets.

Write a sub procedure named CopyValue2 to copy the content of the cell
“A1” in the first worksheet to the cell “B2” in the second worksheet.

h
 Application Object 21

2.2 Application Object

As mentioned in the previous section, the Application object is the top-
most object in the Excel object model. The Application object contains
application-wide settings and options and top-level objects and methods.

Table 2.1: Commonly Used Properties of the Application Object

Property Description
ActiveCell Returns a Range object that represents
the active cell
ActiveSheet Returns an object that represents the active sheet
ActiveWorkbook Returns a Workbook object that represents the
workbook in the active window
ThisWorkbook Returns a Workbook object that represents the
workbook where the current macro code is running
DisplayAlerts True if Excel displays certain alerts and messages
while a macro is running. Read/write boolean
Read/write boolean
Columns Returns a Range object that represents all the
columns on the active worksheet
Rows Returns a Range object that represents all the rows
on the active worksheet
Name Returns a string that represents
the name of the object
Sheets Returns a Sheets collection that represents all
the sheets in the active workbook
Workbooks Returns a Workbooks collection that represents
all the open workbooks
FileDialog Returns a FileDialog object that represents
an instance of the file dialog

The Application object contains many properties and methods. A

list of commonly used properties is shown in Table 2.1. The following sub
procedure illustrates how we can refer to some of these properties:
1 Sub ApplicationDemo ()
2 Debug . Print Application . ActiveWorkbook . Name
3 Debug . Print Application . ActiveSheet . Name
4 Debug . Print Application . ActiveCell . Address
5 Debug . Print Application . Name
6 End Sub
 22 Excel Objects

Executing the above code produces the following output in the Immediate
window:
1 Chapter1 . xlsm
2 Sheet1
3 $D$1
4 Microsoft Excel

Table 2.2: Commonly Used Methods of the Application Object

Method Description
Evaluate Converts a Microsoft Excel name to
an object or a value
FindFile Displays the Open dialog box
GetOpenFilename Displays the standard Open dialog box and
gets a file name from the user
without actually opening any files
InputBox Displays a dialog box for user input and returns
the information entered in the dialog box
OnKey Runs a specified procedure when a particular key
or key combination is pressed
OnTime Schedules a procedure to be run at
a specified time in the future
Run Runs a macro or calls a function

The following sub procedure shows how to use the methods InputBox
and Evaluate:
1 Sub ApplicationDemo2 ()
2 Debug . Print Application . InputBox ( prompt := " Input a
cell address " , Type :=2)
3 Application . ActiveSheet . Range ( " B10 " ) . Value = 3.14
4 Debug . Print Application . Evaluate ( " B10 " ) . Value
5 End Sub

If we execute the above code, an input box will show up on the screen. If
we type “A1” (without the double quotes) in the input box, we will see the
following output in the Immediate window:
1 A1
2 3.14
 Workbook Objects 23

The function Evaluate returned a Range object from the reference “B10.”

g
Exercise 2.4. Read the Application.InputBox Method (Excel) at https://
msdn.microsoft.com/en-us/library/office/ff839468.aspx and write
a sub procedure named InputBoxDemo to do the following: use the method
InputBox to let users input a number and then display the sum of the input
and 5 in the Immediate window.
Exercise 2.5. Read the Application.OnKey Method (Excel) at https://
msdn.microsoft.com/en-us/library/office/ff197461.aspx and write
a sub procedure named OnKeyDemo that assigns the key combination “Shift
+ Ctrl + A” to the following sub procedure:
1 Sub HelloWorld ()
2 MsgBox " Hello World ! "
3 End Sub

For a complete list of members of the Application object and how

to use them, readers can read the MSDN document at https://msdn.
microsoft.com/en-us/library/office/ff194565.aspx. For more in-
formation about the Application object, readers are referred to the De-
veloper’s Guide to the Excel 2010 Application Object at https://msdn.
microsoft.com/en-us/library/office/gg192737%28v=office.14%29.
aspx.

2.3 Workbook Objects

A Workbook object represents a Microsoft Excel workbook. A Workbook
object is a member of the Workbooks collection, which contains all the
Workbook objects currently open in Microsoft Excel.
In addition to the Workbooks property, the Application object has
other properties related to workbooks: ThisWorkbook and ActiveWorkbook.
We can get a Workbook object from the Workbooks collection. We can
also get a Workbook object from the properties ThisWorkbook and
ActiveWorkbook. The following sub procedure shows how to refer to a
workbook using three approaches:
 24 Excel Objects

1 Sub WorkbookDemo1 ()
2 Debug . Print Application . ThisWorkbook . FullName
3 Debug . Print Application . ActiveWorkbook . FullName
4 Debug . Print Application . Workbooks (1) . FullName
5 End Sub

Executing the above code produces the following output in the Immediate
window:
1 Scratch : gan : ResearchU : trunk : book : vbapfi : code : Chapter2
. xlsm
2 Scratch : gan : ResearchU : trunk : book : vbapfi : code : Chapter2
. xlsm
3 Scratch : gan : ResearchU : trunk : book : vbapfi : code : Chapter2
. xlsm

The above output was produced by running the code on a Mac® computer.
That is why we see colons in the file paths. If we run the code on a Windows®
machine, we see the following output:
1 C :\ Users \ Guojun Gan \ Documents \ ResearchU \ trunk \ book \
vbapfi \ code \ Chapter2 . xlsm
2 C :\ Users \ Guojun Gan \ Documents \ ResearchU \ trunk \ book \
vbapfi \ code \ Chapter2 . xlsm
3 C :\ Users \ Guojun Gan \ Documents \ ResearchU \ trunk \ book \
vbapfi \ code \ Chapter2 . xlsm

From the output, we see that all three approaches refer to the same work-
book.
There are several types of Excel workbooks, each of which has a differ-
ent file extension. Table 2.3 presents a list of Excel file extensions. Except
for the extension .xll, all file extensions were introduced with Excel 2007.
Using these extensions, we can save a workbook without macros or with
macros. We can also save a workbook as a template without macros or with
macros. Add-in workbooks are workbooks that contain code. Using add-in
workbooks allows the user to separate code from the users’ data.

Exercise 2.6. Read the Application.ThisWorkbook Property (Excel) at

https://msdn.microsoft.com/en-us/library/office/ff193227.aspx
and examine whether the following statements are true or false:
Workbook Objects 25

Table 2.3: File Extensions of Excel Files

Extension Description
.xlsx Workbook
.xlsm Macro-enabled workbook; same as .xlsx but may contain
macros and scripts
.xltx Template
.xltm Macro-enabled template; same as .xltx but may contain
macros and scripts
.xlam Add-in workbook

(a) ThisWorkbook is the only way to refer to an add-in workbook from

inside the add-in itself.
(b) ThisWorkbook always returns the workbook in which the code is run-
ning.
(c) ThisWorkbook always is the same as ActiveWorkbook.

Table 2.4: Commonly Used Properties of a Workbook Object

Property Description
ActiveChart Returns a Chart object that represents the active chart
ActiveSheet Returns an object that represents the active sheet
Charts Returns a Sheets collection representing all
the chart sheets
FullName Returns the name of the object, including its path
IsAddin True if the workbook is running as an add-in
Name Returns the name of the object
Path Returns a string representing the complete path
to the workbook
Sheets Returns a Sheets collection that represents
all the sheets in the workbook
Worksheets Returns a Sheets collection that represents all the
worksheets in the specified workbook

A Workbook object has many properties. Table 2.4 shows a list of some
commonly used properties of a Workbook object. We already saw how to
 26 Excel Objects

refer to the FullName property of a Workbook object. The following sub

procedure shows how to refer to other properties:
1 Sub WorkbookDemo2 ()
2 Debug . Print ThisWorkbook . Charts . Count
3 Debug . Print ThisWorkbook . Sheets . Count
4 Debug . Print ThisWorkbook . Worksheets . Count
5 Debug . Print ThisWorkbook . IsAddin
6 End Sub

Executing the above code produces the following output:

1 0
2 1
3 1
4 False

The output shows that the current workbook contains one worksheet and
no charts. The workbook is not an add-in workbook.

Table 2.5: Commonly Used Methods of a Workbook Object

Method Description
Activate Activates the first window associated with the workbook
Close Closes the object
Save Saves changes to the specified workbook
SaveAs Saves changes to the workbook in a different file

Table 2.5 shows a list of some commonly used methods of a Workbook

object. Using these methods we can activate, close, and save a workbook.
For example, executing the following sub procedure will save changes and
then close the current workbook:
1 Sub WorkbookDemo3 ()
2 ThisWorkbook . Save
3 ThisWorkbook . Close
4 End Sub

Note that a Workbook object does not have the Open method.

Exercise 2.7. Write a sub procedure named WorkbookDemo4 to save the

 Workbook Objects 27

workbook “Chapter2.xlsm” to a new file called “Chapter2Copy.xlsm” in the

same folder.

Table 2.6: Two Properties of the Workbooks Object

Property Description
Count Returns the number of objects in the collection
Item Returns a single object from the collection

Table 2.7: Some Methods of the Workbooks Object

Method Description
Add Creates a new workbook, which becomes
the active workbook
Close Closes all workbooks
Open Opens a workbook

The Workbooks object does not have many properties or methods. Table
2.6 and Table 2.7 present some properties and methods of the Workbooks
object. From Table 2.7, we see that the Workbooks object provides methods
to create and open a workbook. For example, we can open the workbook
"Chapter2Copy.xlsm" (See Exercise 2.7) as follows:
1 Sub WorkbookDemo5 ()
2 Debug . Print ActiveWorkbook . Name
3 Workbooks . Open " Chapter2Copy . xlsm "
4 Debug . Print Workbooks . Count
5 Debug . Print ActiveWorkbook . Name
6 End Sub

Executing the above code produces the following output:

1 Chapter2 . xlsm
2 2
3 Chapter2Copy . xlsm

From the output we see that the active workbook was changed to the work-
book opened by the Open method.
 28 Excel Objects

2.4 Worksheet Objects

A Worksheet object represents a worksheet and is a member of the
Worksheets collection and the Sheets collection. The Worksheets collec-
tion contains all the Worksheet objects in a workbook. The Sheets collec-
tion contains all the sheets in a workbook, including charts and worksheets.
The Charts collection contains all the charts in a workbook.

Table 2.8: Commonly Used Properties of a Worksheet Object

Property Description
Cells Returns a Range object that represents all the cells on
the worksheet
Columns Returns a Range object that represents all the columns on
the active worksheet
Index Returns the index number of the object within
the collection of similar objects
Name Returns or sets a String value that represents
the object name
Next Returns a Worksheet object that represents the next sheet
Range Returns a Range object that represents a cell or
a Range of cells
Rows Returns a Range object that represents all the rows on
the specified worksheet
Type Returns an XlSheetType value that represents
the worksheet type
UsedRange Returns a Range object that represents the used range on
the specified worksheet

A Worksheet object has many properties and provides many methods.

Some commonly used properties and methods are described in Table 2.8
and Table 2.9, respectively. All of these properties are read-only except for
the property Name. The following sub procedure shows how to refer to a
Worksheet object’s properties:
1 Sub WorksheetDemo1 ()
2 Debug . Print ActiveSheet . Index
3 Debug . Print ActiveSheet . Name
4 Debug . Print ActiveSheet . UsedRange . Address
5 ActiveSheet . Name = " SheetA "
6 Debug . Print ActiveSheet . Name
 Worksheet Objects 29

7 End Sub

The above code displays the index, name, and used range of the active sheet.
Then the code changes the name of the active sheet and displays the name
again. Executing the code produces the following output:
1 1
2 SheetA
3 $A$1 : $G$110
4 SheetA

From the output, we see that the used range is “A1:G110.” This means that
all the cells outside of “A1:G110” are not used. It is possible that some cells
in the used range are not used.

Exercise 2.8. What is wrong with the following code?

1 Sub WorksheetDemo2 ()
2 Debug . Print ActiveSheet . Index
3 ActiveSheet . Index = 2
4 Debug . Print ActiveSheet . Index
5 End Sub

The following sub procedure shows how to copy and select a worksheet:
1 Sub WorksheetDemo3 ()
2 Debug . Print ActiveSheet . Name
3 ActiveSheet . Copy After := ActiveSheet
4 Debug . Print ActiveSheet . Name
5 ActiveSheet . Name = " Sheet2 "
6 Debug . Print ActiveSheet . Name
7 Sheets ( " SheetA " ) . Select
8 Debug . Print ActiveSheet . Name
9 End Sub

Executing the above code produces the following output in the Immediate
window:
1 SheetA
2 SheetA (2)
 30 Excel Objects

Table 2.9: Commonly Used Methods of a Worksheet Object

Method Description
Activate Makes the current sheet the active sheet
Calculate Calculates all open workbooks, a specific worksheet,
or a specified range on a worksheet
Copy Copies the sheet to another location in the workbook
Delete Deletes a worksheet
Move Moves the sheet to another location in the workbook
Paste Pastes the contents of the Clipboard onto the sheet
PasteSpecial Pastes the contents of the Clipboard onto the sheet,
using a specified format
SaveAs Saves changes to the chart or worksheet
in a different file
Select Selects the object

3 Sheet2
4 SheetA

From the output, we see that the new worksheet becomes the active sheet.
The following sub procedure shows how to delete a worksheet:
1 Sub WorksheetDemo4 ()
2 Debug . Print ActiveSheet . Name
3 ActiveSheet . Copy After := ActiveSheet
4 Debug . Print ActiveSheet . Name
5 ActiveSheet . Delete
6 Debug . Print ActiveSheet . Name
7 End Sub

If you run the above code, you will see a prompt dialog. If you click the
“Delete” button in the dialog, you will see the following output in the Imme-
diate window:
1 SheetA
2 SheetA (2)
3 SheetA

From the output, we see that the active sheet changed to the previous
worksheet after the active sheet was deleted. If you do not want to see
the prompt dialog when deleting a worksheet, you can set the application
 Worksheet Objects 31

property DisplayAlerts to be false. For example, we can get rid of the

prompt in the following code:
1 Sub WorksheetDemo5 ()
2 Application . DisplayAlerts = False
3 Debug . Print ActiveSheet . Name
4 ActiveSheet . Copy After := ActiveSheet
5 Debug . Print ActiveSheet . Name
6 ActiveSheet . Delete
7 Debug . Print ActiveSheet . Name
8 Application . DisplayAlerts = True
9 End Sub

Executing the above code will not show the prompt dialog.

Table 2.10: Some Methods of the Worksheets Collection

Method Description
Add Creates a new worksheet, chart, or macro sheet, which
becomes the active sheet
Copy Copies the sheet to another location in the workbook
Delete Deletes the object
Move Moves the sheet to another location in the workbook

The Worksheets collection and the Sheets collection have properties

and methods that are similar to those of the Workbooks collection. Those
properties and methods are shown in Table 2.6 and Table 2.7, respectively.
Table 2.10 shows some methods of a Worksheets collection.

Exercise 2.9. Write a sub procedure named WorksheetDemo6 to do the

following: create a new worksheet, rename the new worksheet to “Data,” and
write the index of the new worksheet to the cell “A1” of the new worksheet.

h
32 Excel Objects

2.5 Range Object

Excel is all about cells. A Range object is a collection of cells and represents
a range contained in a Worksheet object. Like other objects, a Range object
has properties and methods.
A Range object can contain a single cell or all cells in a worksheet. We
can find out the number of cells in a worksheet by multiplying the number
of rows and the number of columns (see Exercise 2.2). In Excel 2013, for
example, a worksheet contains

1048576 × 16384 = 17, 179, 869, 184

cells.
We can refer to a Range object in different ways. Table 2.11 gives a list of
expressions that can be used to refer to a Range object. In this table, we only
show the expression to refer to a Range object in the ActiveWorksheet
object. To refer to a Range object in a worksheet that is different from the
active worksheet, we can include the parent object in the reference.

Table 2.11: Various Ways to Refer to a Range Object

Expression Description
Range("A3:C4") Cells in Columns A to C and
Rows 3 to 4
Range("A1") The cell in Column A and Row 1
Range("RName") The range named "RName"
Range("1:1") Cells in Row 1
Range("A:A") Cells in Column A
Range("A1:C3,H5:K6") Cells in noncontiguous ranges
Cells(1,1) Equivalent to Range("A1")
Range(Cells(3,1), Cells(4,3)) Equivalent to Range("A3:C4")
Range("A1").Offset(2,3) The cell two rows below A1 and
three columns to the right of A1

For example, the active workbook is Chapter2.xlsm, and the workbook

USLifeTable1999-2001Male.xls is also open. The second workbook con-
tains the mortality table of the male population in the US (see Figure 2.3).
The following sub procedure illustrates how to refer to the range that con-
tains the mortality rates:
 Range Object 33

1 Sub MaleTable ()
2 Debug . Print Workbooks ( " USLifeTable1999 -2001 Male .
xls " ) . Worksheets ( " Sheet1 " ) . Range ( " A25 : B134 " ) .
Cells . Count
3 End Sub

In the above code, we refer to the workbook and the worksheet that contain
the Range object. Executing the sub procedure gives the number of cells
contained in the range.

Figure 2.3: A screenshot of the US Life Tables 1999–2001 — Males

From Table 2.11, we see that we can also refer to a Range object by Cells.
As pointed out by Walkenbach (2013b), Cells is a property not an object
or collection. VBA evaluates a Cells property and returns a Range object.
The Cells property takes two arguments. The first argument represents
the row index, and the second argument represents the column index. The
Cells property provides a useful way to refer to a range if we know the
coordinates of the range. Suppose that we want to find the address of the
range in the intersection of Rows 10–12 and Columns 20–22. Then we can
use the following sub procedure:
1 Sub FindAddress ()
2 Debug . Print Range ( Cells (10 , 20) , Cells (12 , 22) ) .
Address
 34 Excel Objects

3 End Sub

Executing the above code gives the following output:

1 $T$10 : $V$12

Exercise 2.10. Write a sub procedure named FindAddress2 to display the

address of the cell in the 20th row and 51st column.

Exercise 2.11. Write a sub procedure named ReferCell to display the

address of the cell “E10.” Use three different ways to refer to the cell “E10.”

The Offset provides another useful way to refer to a Range object rela-
tive to another Range object. Like the Cells object, the Offset object also
takes two arguments, which represent the number of rows and the number
of columns to offset, respectively. For example, we can refer to the cell “B2”
as follows:
1 Sub ReferCell2 ()
2 Debug . Print Range ( " A1 " ) . Offset (1 , 1) . Address
3 Debug . Print Range ( " E4 " ) . Offset ( -2 , -3) . Address
4 End Sub |

Executing the code produces the following output:

1 $B$2
2 $B$2

Exercise 2.12. Determine the output of the following VBA code:

1 Sub OffsetRange ()
2 Debug . Print Range ( " A2 : C3 " ) . Offset (2 , 3) . Address
3 Debug . Print Range ( " G6 : I7 " ) . Offset ( -2 , -3) . Address
4 End Sub

Calculate the output before running the code in the Visual Basic Editor.
 Range Object 35

Table 2.12: Some Properties of a Range Object

Property Read-Only Description

Value No The value contained in a cell
Text Yes The string displayed in a cell
Count Yes The number of cells in a range
Column Yes The column index of a cell
Row Yes The row index of a cell
Address Yes The address of a range in absolute
reference
Font No Returns a Font object
Interior No Returns an Interior object
Formula No The formula in a cell
NumberFormat No A string that represents
the number format

A Range object has many properties. Table 2.12 presents a list of some
commonly used properties of a Range object. Some of these properties are
read-only. For read-only properties, VBA code can look at their values but
cannot change them. Examples of read-only properties include Text and
Count. If a property is not read-only, then you can get and set the value of
the property.
The following example shows how to set the value of a cell, change the
number format, and display the string as shown in the worksheet:
1 Sub ShowValueText ()
2 Range ( " A1 " ) . Value = 198.35
3 Range ( " A1 " ) . NumberFormat = " $ # ,##0 _ ) ;( $ # ,##0) "
4 Debug . Print Range ( " A1 " ) . Value
5 Debug . Print Range ( " A1 " ) . Text
6 End Sub

Executing the above code gives the following output:

1 198.35
2 $198

g
 36 Excel Objects

Exercise 2.13. Write a sub procedure named ChangeFont to do the follow-

ing: Set the value of cell “B2” to the string “World,” and change the font to
be bold and the font size to be 16.

We can change the background color of a range by using the Interior

object. In the following example, we change the background color of cell
“C3” to blue and the font color to white:
1 Sub ChangeCellColor ()
2 Range ( " C3 " ) . Value = " White "
3 Range ( " C3 " ) . Interior . Color = vbBlue
4 Range ( " C3 " ) . Font . Color = vbWhite
5 End Sub

If we execute the above sub procedure, cell “C3” will show the word “White”
in white and the background color will be changed to blue. Note that vbBlue
and vbWhite are VBA constants used to represent colors.

Table 2.13: Some Methods of a Range Object

Method Description
Select Select a range of cells
Copy Copy a range
PasteSpecial Paste a range from the Clipboard into a specified
range
Clear Delete contents of a range
Delete Delete a range and fill it up by shifting remaining cells

A Range object also has many methods. Table 2.13 shows some com-
monly used methods of a Range object. The following piece of code shows
how to copy and paste a range:
1 Sub CopyRange ()
2 Range ( " A1 : C3 " ) . Copy
3 Range ( " D5 " ) . PasteSpecial xlPasteAll
4 Debug . Print Selection . Address
5 End Sub |

After executing the above code, the range “A1:C3” will be copied to the
range “D5:F7.” Since we specify xlPasteAll, the format of the range is also
copied.
 WorksheetFunction Object 37

Note that a Range object does not have a method called paste. In fact,
paste is a method of a Worksheet object. For example, we can copy a range
as follows:
1 Sub CopyRange2 ()
2 Range ( " A1 : C3 " ) . Copy
3 Range ( " D5 " ) . Select
4 ActiveSheet . Paste
5 Debug . Print Selection . Address
6 End Sub

The sub procedure CopyRange and the sub procedure CopyRange2 produce
the same result. If we just want to copy values of a range, then we need to use
the PasteSpecial method by specifying xlPasteValues in the argument.

Exercise 2.14. Suppose that the current active workbook is “Chapter2.xlsm”

and the workbook “USLifeTable1999-2001Male.xls” is open. Write a sub pro-
cedure named CopyTable to copy the mortality table (in range “A25:B134”)
in the second workbook to the first worksheet of the active workbook. The
start cell of the destination range should be “F1.”

For a complete list of properties and methods of a Range object,

readers are referred to the MSDN (Microsoft Developer Network) docu-
mentation at https://msdn.microsoft.com/en-us/library/office/
ff838238.aspx.

2.6 WorksheetFunction Object

The WorksheetFunction object is a container for Excel worksheet func-
tions that can be called from VBA code. Excel 2013 contains more than 400
worksheet functions. Table 2.14 shows some of these functions.
You are familiar with using the worksheet functions in a cell formula. In
this section, we introduce how to call the worksheet functions in VBA code.
Let us consider the following sub procedure:
1 Sub W or ksh ee tFu nc ti o n D e m o 1 ()
 38 Excel Objects

Table 2.14: Some Worksheet Functions

Function Description
Ceiling Returns number rounded up to the nearest multiple
of significance
Count Counts the number of cells that contain numbers and
counts numbers within the list of arguments
CountIf Counts the number of cells within a range that meet
the given criteria
Floor Rounds number down to the nearest multiple of
significance
Ln Returns the natural logarithm of a number
Lookup Returns a value either from a one-row or one-column
range or from an array
Max Returns the largest value in a set of values
Min Returns the smallest number in a set of values
NormDist Returns the normal distribution for the specified mean
and standard deviation
NormInv Returns the inverse of the normal cumulative
distribution for the specified mean and
standard deviation
Pi Returns the number 3.14159265358979
Power Returns the result of a number raised to a power
RandBetween Returns a random integer number between
two specified numbers

2 Range ( " A1 " ) . Value = Worksheet Function . RandBetween

(0 , 100) / 100
3 Range ( " A2 " ) . Value = Worksheet Function . RandBetween
(0 , 100) / 100
4 Range ( " A3 " ) . Value = Worksheet Function . RandBetween
(0 , 100) / 100
5 Range ( " A4 " ) . Value = Worksheet Function . RandBetween
(0 , 100) / 100
6 Range ( " A5 " ) . Value = Worksheet Function . RandBetween
(0 , 100) / 100
7 Debug . Print Works heetFunc tion . CountIf ( Range ( " A1 :
A5 " ) , " >0.5 " )
8 Debug . Print Works heetFunc tion . Max ( Range ( " A1 : A5 " ) )
9 Debug . Print Works heetFunc tion . Min ( Range ( " A1 : A5 " ) )
10 End Sub
 WorksheetFunction Object 39

The above sub procedure first generates five random numbers and puts
them into the first five cells of Column A. Then it displays the number
of values that are larger than 0.5, the maximum of the five values, and
the minimum of the five values. Executing the above code produces the
following output:
1 3
2 1
3 0

If you execute the above code, you may get different results, as the func-
tion RandBetween returns a random integer between two numbers you
specified.

Exercise 2.15. Write a sub procedure named RandBetweenDemo to generate

a random number between 0 and 1 that has four decimal places. Display
the random number in the Immediate window.

In the above example, we apply worksheet functions to values in a range.

We can also apply worksheet functions to numbers directly:
1 Sub W o r k s h e e t F u n c t i o n D e m o 2 ()
2 Debug . Print Works heetFunc tion . Power (2 , 10)
3 Debug . Print Works heetFunc tion . Pi
4 Debug . Print Works heetFunc tion . Ln (100)
5 Debug . Print Works heetFunc tion . Ceiling (0.51 , 1)
6 Debug . Print Works heetFunc tion . Floor (0.51 , 1)
7 Debug . Print Works heetFunc tion . NormInv (0.95 , 0 , 1)
8 Debug . Print Works heetFunc tion . NormDist (0 , 0 , 1 ,
True )
9 End Sub

Executing the above code produces the following output:

1 1024
2 3.14159265358979
3 4.60517018598809
4 1
5 0
6 1.64485362695147
7 0.5
40 Excel Objects

The function NormDist(x, µ, σ, p 4 ) calculates the density (if p 4 is false)

and probability (if p 4 is true) of a normal distribution with mean µ and
standard deviation σ at x. The normal density function is defined as

(x − µ)2
µ ¶
1
f (x, µ, σ) = p exp − . (2.1)
2πσ σ2

If p 4 is false, the function NormDist(x, µ, σ, p 4 ) returns f (x, µ, σ). If p 4 is

true, the function NormDist(x, µ, σ, p 4 ) returns the following integrated
value: Z x
f (s, µ, σ)ds.
−∞
The function NormIvs(x, µ, σ) returns the value of y, such that
Z y
x= f (s, µ, σ)ds.
−∞

Exercise 2.16. Write a sub procedure named WorksheetFunctionDemo3

to calculate the following quantities:

(a) The circumference of a circle that has a diameter of 2 meters. (The

formula is πd , where d is the diameter.)
(b) The probability of a standard normal variable less than 1.5.
(c) The value of y, such that the probability of a standard normal variable
less than y is 0.025.

Note that a standard normal variable has a mean of 0 and a standard devia-
tion of 1.

A complete list of worksheet functions contained in the WorksheetFunction

object can be found in the Object Browser of the Visual Basic Editor (see
Figure 2.4). The usage of worksheet functions can be found in the MSDN
document “WorksheetFunction Methods (Excel)” that is available at https:
//msdn.microsoft.com/en-us/library/office/dn301180.aspx.
Summary 41

Figure 2.4: Worksheet functions in the Object Browser

2.7 Summary
In this chapter, we introduced the Excel object model, which contains
hundreds of objects and thousands of properties and methods. We also in-
troduced some commonly encountered objects, such as the Application,
Range, and Cells objects. For an overall view of the Excel object model,
readers are referred to (Roman, 2002, Chapter 15). For more information
about object-oriented programming, readers are referred to (Urtis, 2015,
Lesson 5).
3
Variables, Data Types, and Scopes

A variable is a named storage location in a computer’s memory. Variables

allow us to store values and then come back to these values at a later time.
In this chapter, we introduce how to declare variables and some common
data types. After studying this chapter, readers will be able to

• know the common data types supported by Visual Basic for Applications
(VBA).
• know how to declare and use variables and constants.
• know how to declare and use arrays.
• know how to manipulate dates and strings.
• understand the concept of scopes.

3.1 Variable Declaration and Data Types

To declare a variable in VBA code, we first need to come up with a name for
the variable. The name of a variable should satisfy the following conditions:

• The first character must be a letter.

• The following symbols cannot be used in a name: space, period (.),
exclamation mark (!), and the characters @, &, $, #.
• A name cannot exceed 255 characters in length.
• A name cannot be a VBA keyword.

Exercise 3.1. Determine the legal variable names in the following list of
names:

ws1, 1ws, my.book, Sub, w b, Tmp, sub

43
 44 Variables, Data Types, and Scopes

The following sub procedure shows how to declare a variable in VBA

code:
1 Sub VariableDemo1 ()
2 Dim x As Double
3
4 Debug . Print x
5 x = 1
6 Debug . Print x
7 x = x + 10
8 Debug . Print x
9 End Sub

The syntax of declaring a variable in VBA is shown in Line 2 of the above

code. We use the VBA keyword Dim to declare variables. In this example,
we declare a double-type variable named x. Then we display the value of x
after some operations. Executing the above code gives the following output:
1 0
2 1
3 11

Table 3.1 shows a list of VBA’s built-in data types. Regarding how to
choose a data type for a variable, you should in general choose a data type
that can hold all the data you want to store and uses the smallest number of
bytes. Most data types shown in Table 3.1 are easy to understand except for
Object and Variant.
We declare an Object variable using the Dim statement as above. How-
ever, we need to use the Set statement to assign an object to an object
variable. The following sub procedure shows how to declare Object vari-
ables and assign objects to the variables:
1 Sub VariableDemo3 ()
2 Dim wb As Object
3 Dim ws As Object
4
5 Set wb = ThisWorkbook
6 Debug . Print wb . Name
7
8 Set ws = ThisWorkbook . ActiveSheet
9 Debug . Print ws . Name
10 End Sub
 Variable Declaration and Data Types 45

Table 3.1: List of Built-In Data Types in VBA

Data Type Bytes Used Range of Values

Boolean 2 True or False
Byte 1 0 to 28 − 1
Integer 2 −215 to 215 − 1
Long 4 −231 to 231 − 1
Single 4 −3.4E+38 to −1.4E-45 for negative values,
1.4E-45 to 3.4E+38 for positive value
Double 8 −1.79E+308 to −4.94E-324 for
negative values,
4.94E–324 to 1.79E+308 for
positive values
Currency 8 −922,337,203,685,477.5808 to
922,337,203,685,477.5807
Date 8 1/1/100 to 12/31/9999
String 1 per character Varies
Object 4 Any defined object
Variant Varies Varies

In the above code, we declare two Object variables and assign a Workbook
object and a Sheet object to the variables, respectively. Executing the above
code produces the following output:
1 Chapter1 . xlsm
2 Sheet1

Exercise 3.2. What is wrong with the following VBA code?

1 Sub VariableDemo4 ()
2 Dim wb As Workbook
3
4 wb = Workbooks (1)
5 Debug . Print wb . Name
6 End Sub

Exercise 3.3. What is wrong with following code?

1 Sub VariableDemo5 ()
 46 Variables, Data Types, and Scopes

2 Dim wb As Workbook
3
4 Set wb = ThisWorkbook . ActiveSheet
5 Debug . Print wb . Name
6 End Sub

Exercise 3.4. Does the following code contain any errors?

1 Sub VariableDemo6 ()
2 Dim myObj As Object
3
4 Set myObj = ThisWorkbook
5 Debug . Print myObj . Name
6
7 Set myObj = ThisWorkbook . ActiveSheet
8 Debug . Print myObj . Name
9 End Sub

Exercise 3.5. Write a sub procedure named WorkbookDemo6 to create a new

workbook, save the new workbook to a file named “MyWorkbook.xlsx,” and
then close the new workbook.

Exercise 3.6. Write a sub procedure named WorksheetDemo6 to create a

new worksheet and rename the new worksheet to "SheetB."

A Variant variable can contain any kind of data except fixed-length

String data. The following example illustrates how to use a Variant vari-
able:
1 Sub VariableDemo7 ()
2 Dim x As Variant
3
4 x = 1
5 Debug . Print x
6 x = 3.54
7 Debug . Print x
8 Set x = ThisWorkbook
9 Debug . Print x . Name
10 End Sub

Executing the above code produces the following output:

 Variable Declaration and Data Types 47

1 1
2 3.54
3 Chapter1 . xlsm

From the output, we see that a Variant variable can be used to store differ-
ent types of data without modifying the data.
VBA is not a strictly typed programming language, that is, we can use a
variable without explicitly declaring it. In the following example, we assign
values to two variables and use the function VarType to get the data type of
the variables.
1 Sub VariableDemo2 ()
2 x = 1
3 Debug . Print VarType ( x )
4 Debug . Print vbInteger
5
6 y = 3.14
7 Debug . Print VarType ( y )
8 Debug . Print vbDouble
9 End Sub

Executing the above code produces the following output:

1 2
2 2
3 5
4 5

From the output, we see that VBA automatically treated x as an Integer

variable and y as a Double variable. In the above code, vbInteger and
vbDouble are VBA constants that denote the data types.
It is not a best practice to use a variable without explicitly declaring
it. To force explicit variable declaration, we can put Option Explicit in
the declaration section of a module (see Section 1.2). For example, if we
put Option Explicit at the beginning of the module and run the sub
procedure VariableDemo2, we see the compile error shown in Figure 3.1.
 48 Variables, Data Types, and Scopes

Figure 3.1: An illustration of an explicit variable declaration

3.2 Arrays
In Section 3.1, we introduced how to declare variables, which contain a
single value. In this section, we introduce how to declare arrays, which can
contain multiple values. In particular, we will introduce one-dimensional,
multidimensional, and dynamic arrays.
An array is a group of variables that share a common name. We can
refer to a specific variable in an array by using the array name and an index
number in parentheses. The following example shows how to declare one-
dimensional arrays:
1 Sub ArrayDemo1 ()
2 Dim arA (0 To 9) As Integer
3 Dim arB (9) As Integer
4 Dim arC (1 To 10) As Integer
5
6 Debug . Print UBound ( arA )
7 Debug . Print LBound ( arA )
8 Debug . Print UBound ( arB )
9 Debug . Print LBound ( arB )
10 Debug . Print UBound ( arC )
 Arrays 49

11 Debug . Print LBound ( arC )

12 End Sub

In the above sub procedure, we declare three arrays, each of which has ten
elements. Inside the parentheses of arA, 0 means the lower index and 9
means the upper index of the array. The functions UBound and LBound are
used to get the upper and lower bounds of the array index. If we execute the
above code, we see the following output:
1 9
2 0
3 9
4 0
5 10
6 1

From the above output we see that arA and arB are identical in that the
bounds of the two arrays are the same. However, the bounds of arC are 1
and 10.

Exercise 3.7. Let arD be an array defined in the following code:

1 Sub ArrayDemo2 ()
2 Dim arD (5 To 100) As Integer
3 End Sub

What are the bounds of the array arD, and how many elements does the
array contain?

From the above example we see that if you omit the lower index in
an array declaration, VBA assumes that it is 0. However, you can change
the default lower index to 1 by putting Option Base 1 in the declaration
section of the module. Suppose that the underlying module is MArray, and
we have Option Base 1 in the declaration section of the module. Then
executing the following code
1 Sub ArrayDemo3 ()
2 Dim arE (5) As Integer
3
4 Debug . Print LBound ( arE )
 50 Variables, Data Types, and Scopes

5 Debug . Print UBound ( arE )

6 End Sub

will produce the following output:

1 1
2 5

Exercise 3.8. What is wrong with the following code?

1 Option Explicit
2 Option Base 10

Exercise 3.9. What is wrong with the following code?

1 Sub ArrayDemo4 ()
2 Dim arF (5 To 10) As Integer
3
4 Debug . Print LBound ( arF )
5 Debug . Print UBound ( arF )
6 End Sub

Once we declare an array, we can access its elements by using the array
name and an index number in parentheses. The following example shows
how to get and set values of an array:
1 Sub ArrayDemo5 ()
2 Dim arA (1 To 5) As Double
3 arA (1) = 1.1
4 arA (3) = arA (1) + 2
5 arA (5) = arA (3) + 2
6
7 Debug . Print arA (1) , arA (2) , arA (3) , arA (4) , arA
(5)
8 End Sub

Executing the above code gives the following output:

1 1.1 0 3.1 0 5.1
 Arrays 51

We can also declare a multidimensional array by specifying the bounds

of every dimension. In the following example, we declare a two-dimensional
array:
1 Sub ArrayDemo6 ()
2 Dim arM (1 To 3 , 1 To 3) As Double
3
4 arM (1 , 1) = 1
5 arM (1 , 2) = 3
6 arM (1 , 3) = 4
7 arM (2 , 1) = arM (1 , 2)
8 arM (2 , 2) = 4
9 arM (2 , 3) = 5
10 arM (3 , 1) = 4
11 arM (3 , 2) = arM (2 , 3)
12 arM (3 , 3) = 6
13
14 Debug . Print arM (1 , 1) , arM (1 , 2) , arM (1 , 3)
15 Debug . Print arM (2 , 1) , arM (2 , 2) , arM (2 , 3)
16 Debug . Print arM (3 , 1) , arM (3 , 2) , arM (3 , 3)
17 End Sub

The above example also shows how to access elements in a two-dimensional

array. Executing the above code produces the following output:
1 1 3 4
2 3 4 5
3 4 5 6

Exercise 3.10. Write a sub procedure named ArrayDemo7 that declares a

2 × 2 × 2 array, assigns a value (e.g., the sum of the indices) to each element,
and displays all the elements.

We can also use the functions LBound and UBound to get the bounds of
a multidimensional array. However, we need to specify which dimension in
the function calls. The following example shows how to get the bounds of a
multidimensional array:
1 Sub ArrayDemo8 ()
2 Dim arM (1 To 10 , 0 To 9 , 1 To 10) As Double
 52 Variables, Data Types, and Scopes

3
4 Debug . Print LBound ( arM , 1) , LBound ( arM , 2) ,
LBound ( arM , 3)
5 Debug . Print UBound ( arM , 1) , UBound ( arM , 2) ,
UBound ( arM , 3)
6 End Sub

Executing the above code produces the following output:

1 1 0 1
2 10 9 10

VBA allows us to create dynamic arrays, which do not contain a pre-

set number of elements. The following example shows how to declare a
dynamic array and change the size of the array:
1 Sub ArrayDemo9 ()
2 Dim arV () As Integer
3
4 ReDim arV (1 To 3)
5 arV (1) = 1
6 arV (2) = 2
7 arV (3) = 3
8 Debug . Print arV (1) , arV (2) , arV (3)
9
10 ReDim arV (1 To 5)
11 Debug . Print arV (1) , arV (2) , arV (3) , arV (4) , arV
(5)
12 End Sub

In the above code, we use the ReDim statement to change the size of the
array. We can use the ReDim statement as many times as we want. Executing
the above code gives the following output:
1 1 2 3
2 0 0 0 0 0

From the output, we see that the second ReDim statement destroyed all the
old values contained in the array arV.
To keep the old values while using the ReDim statement, we need to
use the Preserve statement after the ReDim statement, as shown in the
following example:
1 Sub ArrayDemo10 ()
2 Dim arV () As Integer
 Arrays 53

3
4 ReDim arV (1 To 3)
5 arV (1) = 1
6 arV (2) = 2
7 arV (3) = 3
8 Debug . Print arV (1) , arV (2) , arV (3)
9
10 ReDim Preserve arV (1 To 5)
11 Debug . Print arV (1) , arV (2) , arV (3) , arV (4) , arV
(5)
12 End Sub

Executing the above code produces the following output:

1 1 2 3
2 1 2 3 0 0

From the output, we see that the old values are preserved after changing
the size of the array.

Exercise 3.11. What is wrong with the following code?

1 Sub ArrayDemo11 ()
2 Dim arV () As Integer
3
4 ReDim arV (0 To 2)
5 Debug . Print arV (0) , arV (1) , arV (2)
6 arV (0) = 3
7 arV (1) = 1
8 arV (2) = 2
9
10 ReDim arV (1 To 3)
11 Debug . Print arV (1) , arV (2) , arV (3)
12
13 ReDim Preserve arV (0 To 3)
14 Debug . Print arV (0) , arV (1) , arV (2) , arV (3)
15 End Sub

h
 54 Variables, Data Types, and Scopes

3.3 Constants
A variable’s value often changes during execution of a program. If we want
to refer to a value or string that never changes during execution of the
program, we should use a constant. A constant is a meaningful name that
takes the place of a number or string that does not change during execution
of a program.
The following example illustrates how to define constants:
1 Sub ConstantDemo1 ()
2 Const conPi As Double = 3.1415
3 Const conDaysInWeek As Integer = 7
4 Const conVersion As String = " v1 .0 "
5
6 Debug . Print conPi
7 Debug . Print conDaysInWeek
8 Debug . Print conVersion
9 End Sub

In the above sub procedure, we define three constants using the Const
statement: a real constant, an integer constant, and a string constant. Ex-
ecuting the above code produces the following output in the Immediate
window:
1 3.1415
2 7
3 v1 .0

Exercise 3.12. What is wrong with the following code?

1 Sub ConstantDemo2 ()
2 Const conMonths As Single = 12
3 Const conQuarters As Byte = 4
4 Const conDebug As Boolean = False
5 Const conBook as Workbook = Thisworkbook
6 Const conCell as Range = Thiswookbook . cells (1 ,1)
7 Const conMode = " Release "
8 Const conCellsInRow As Integer = ThisWorkbook .
ActiveSheet . Rows (1) . Cells . Count
9 End Sub
 Constants 55

We can also define constants in terms of previously defined constants,

as shown in the following example:
1 Sub ConstantDemo3 ()
2 Const conPi As Double = 3.1415
3 Const conPi2 As Double = 2 * conPi
4
5 Debug . Print conPi2
6 End Sub

In the above example, we defined the constant conPi2 based on the con-
stant conPi.

Exercise 3.13. What is wrong with the following code?

1 Sub ConstantDemo4 ()
2 Const conE2 As Double = conE * 2
3 Const conE As Double = 2.718
4
5 Debug . Print conE2
6 End Sub

Exercise 3.14. What is wrong with the following code?

1 Sub ConstantDemo5 ()
2 Const conPi As Double = 3.14
3 conPi = conPi + 0.00159
4
5 Debug . Print conPi
6 End Sub

Since we define the constants in a procedure, these constants can only

be used in the procedure. If we want a constant to be available to many pro-
cedures, we can define the constant in the declaration section of a module
(see Section 1.2). For example, we can move the above three constants to
the declaration section of the module:
 56 Variables, Data Types, and Scopes

Figure 3.2: VBA’s built-in color constants

1 Option Explicit
2
3 Const conPi As Double = 3.1415
4 Const conDaysInWeek As Integer = 7
5 Const conVersion As String = " v1 .0 "

Then we can use these constants in all procedures in the module. In Section
3.6, we will discuss more about the scope of a constant.
We have discussed how to define our own constants. VBA also contains
many built-in constants. For example, Figure 3.2 shows a list of VBA’s built-
in color constants, and Figure 3.3 shows a list of VBA’s built-in constants for
variable types.

Exercise 3.15. Write a sub procedure named ConstantDemo6 to display the

values contained in the following built-in constants: vbGreen, vbYellow,
vbByte, and vbError.

h
 Strings 57

Figure 3.3: VBA’s built-in constants for variable types

3.4 Strings
In many situations, we need to handle string (or text) data. VBA provides
a useful set of functions for handling strings. In this section, we introduce
how to create strings and use existing functions to manipulate them.
There are two types of strings in VBA: fixed-length strings and dynamic
strings. The following example shows how to declare a fixed-length string:
1 Sub StringDemo1 ()
2 Dim strFixed As String * 15
3
4 Debug . Print Len ( strFixed )
5 Debug . Print strFixed
6 strFixed = " I am a fixed string "
7 Debug . Print Len ( strFixed )
8 Debug . Print strFixed
9 End Sub

In the above code, we declare a string variable strFixed that contains

exactly 15 characters. When VBA creates this variable, it fills the variable
 58 Variables, Data Types, and Scopes

with 15 spaces. The variable always contains 15 characters, no matter how

many characters you place into it. Executing the above code produces the
following output:
1 15
2
3 15
4 I am a fixed st

From the above output, we see that the function Len always returns the
fixed size of the string.
Unlike fixed-length strings, dynamic strings do not have a fixed size.
Creating a dynamic string is straightforward, as shown in the following
example:
1 Sub StringDemo2 ()
2 Dim strDynamic As String
3
4 Debug . Print Len ( strDynamic )
5 Debug . Print strDynamic
6 strDynamic = " I am a fixed string "
7 Debug . Print Len ( strDynamic )
8 Debug . Print strDynamic
9 End Sub

When we add or remove characters from a dynamic string, VBA will take
care of allocating or deallocating memory as necessary to store the text data.
If we execute the above code, we see the following output:
1 0
2
3 19
4 I am a fixed string

From the output, we see that the function Len returns the actual number of
characters contained in the string.
A fixed-length string cannot contain more than 65,526 characters. A
dynamic string can contain as many as two billion characters. Dynamic
strings are a bit slower to use because they require more processing effort
from VBA. However, we need to use trim functions to remove extra spaces
of a fixed-length string every time we use the string. As a result, we will use
fixed-length strings only when it is necessary.
Now let us introduce VBA’s built-in functions for manipulating strings.
Strings 59

Table 3.2: Some Commonly Used String Functions in VBA

Function Description
Filter Returns a zero-based array containing a subset of
a String array based on specified filter criteria
Format Returns a string formatted according to a specified format
InStr Returns the start position of the first occurrence of
one string within another
InStrRev Returns the position of the first occurrence of one string
within another, starting from the right side of the string
Join Returns a string created by joining substrings contained
in an array
LCase Returns a string or character converted to lowercase
Left Returns a string containing a specified number of
characters from the left side of a string
Len Returns the number of characters in a string
LSet Returns a left-aligned string containing
the specified string adjusted to the specified length
LTrim Returns a string containing a copy of a specified string
with no leading spaces
Mid Returns a string containing a specified number of
characters from a string
Replace Returns a string where a specified substring has been
replaced with another substring for a specified
number of times
Right Returns a string containing a specified number of
characters from the right side of a string
RSet Returns a right-aligned string containing
the specified string adjusted to the specified length
RTrim Returns a string containing a copy of a specified string
with no trailing spaces
Space Returns a string consisting of the specified number of
spaces
Split Returns a zero-based, one-dimensional array of substrings
StrComp Returns −1, 0, or 1 based on the result of
a string comparison
Trim Returns a string containing a copy of a specified string
with no leading or trailing spaces
UCase Returns a string or character containing the specified
string converted to uppercase
 60 Variables, Data Types, and Scopes

Table 3.2 gives a list of commonly used string functions provided by VBA.
We first introduce how to compare two strings.
In Excel VBA, we have two methods of comparing strings: the binary
mode and the text mode. In the binary mode, strings are compared based
on their binary representation. In the binary mode, “A” and “a” are different.
In the text mode, strings are treated case insensitively. In this mode, “A” is
the same as “a.” We can specify the method of comparison in a module
by using the Option Compare statement in the declaration section. For
example, if we want to compare strings in the text mode in a module, we
can put the following line
1 Option Compare Text

in the declaration section of the module. If we do not specify the Option

Compare Text statement, VBA uses Option Compare Binary.
The following example shows how to compare two strings:
1 Sub StringDemo3 ()
2 Dim strA As String
3 Dim strB As String
4
5 strA = " Hello "
6 strB = " hello "
7 Debug . Print srtA < strB
8 Debug . Print strA = strB
9 Debug . Print strA > strB
10 Debug . Print StrComp ( strA , strB , vbBinaryCompare )
11 Debug . Print StrComp ( strA , strB , vbTextCompare )
12 End Sub

Suppose that VBA uses the default comparison mode (i.e., the binary mode)
in the module where the above code is saved. Then, executing the above
code gives the following output:
1 True
2 False
3 False
4 -1
5 0

The above output shows that strA is less than strB in the binary mode.
However, the two strings are the same if compared in the text mode. In the
StrComp function, we can specify the mode of comparison.
 Strings 61

Leading or trailing spaces are included in string comparison, as shown

in the following example:
1 Sub StringDemo4 ()
2 Debug . Print StrComp ( " Hello " , " Hello " ,
vbTextCompare )
3 Debug . Print StrComp ( " Hello " , " Hello " ,
vbTextCompare )
4 End Sub

Executing the above code gives the following output:

1 -1
2 -1

The output shows that leading and trailing spaces in a string affect the result
of string comparison. We use the Trim function to remove the leading and
trailing spaces in the string as follows:
1 Sub StringDemo5 ()
2 Debug . Print StrComp ( " Hello " , Trim ( " Hello " ) ,
vbTextCompare )
3 Debug . Print StrComp ( Trim ( " Hello " ) , " Hello " ,
vbTextCompare )
4 End Sub

The output of the above code is

1 0
2 0

The output shows that the two strings are the same when the spaces are
removed.

Exercise 3.16. The following function returns an array of 1000 words gen-
erated randomly:
1 Function RandomWords ()
2 Dim arV (1 To 1000) As String
3 Dim i As Integer
4 Dim j As Integer
5 Dim k As Integer
6 Dim strW As String
7
 62 Variables, Data Types, and Scopes

8 Rnd ( -1)
9 Randomize (1)
10 For i = 1 To 1000
11 strW = " "
12 For j = 1 To 4
13 k = Math . Round ( Rnd * ( Asc ( " e " ) - Asc ( " a " )
+ 1) , 0) + Asc ( " a " )
14 strW = strW & Chr ( k )
15 Next j
16 arV ( i ) = strW
17 Next i
18
19 RandomWords = arV
20 End Function

Write a sub procedure named StringDemo6 to find out the frequency of

each word. Which words are the most frequent words in the array?

Now, we introduce how to search a string in another string. To do that,

we can use the InStr function, which returns the position of the first occur-
rence of a string in another. Suppose that we want to search the string “car”
or “Car” in the string “This car is beautiful.” Then, we can use this function
as follows:
1 Sub StringDemo7 ()
2 Dim strA As String
3 strA = " This car is beautiful "
4
5
6 Debug . Print InStr (1 , strA , " car " , vbTextCompare )
7 Debug . Print InStr (1 , strA , " Car " , vbTextCompare )
8 Debug . Print InStr (1 , strA , " Car " , vbBinaryCompare
)
9 Debug . Print InStr ( strA , " car " )
10 Debug . Print InStr ( strA , " Car " )
11 End Sub

The function InStr can take four arguments. The first and the last argu-
ments are optional. The default value of the first argument is 1, and the
default value of the last argument is vbBinaryCompare. Executing the above
code gives the following output:
 Strings 63

1 6
2 6
3 0
4 6
5 0

To replace a substring in a string with another string, we can use the

Replace function. The following example shows how to replace a substring
in a string:
1 Sub StringDemo8 ()
2 Dim strA As String
3 Dim strB As String
4 strA = " This car is beautiful "
5
6 strB = Replace ( strA , " car " , " house " )
7
8 Debug . Print strB
9 End Sub

In the above example, we use the function Replace to replace the substring
“car” with “house.” Executing the above code gives the following output:
1 This house is beautiful

Exercise 3.17. The function Replace has six arguments. Read the Microsoft
Developer Network (MSDN) document of this function and write a sub
procedure named StringDemo9 to replace only the first occurrence of “car”
in the following string with “house”:

This car is beautiful. That car is big.

To extract a substring from a string, we can use the Mid function. The
following example shows how to extract a substring from a string:
1 Sub StringDemo10 ()
2 Dim strA As String
3 Dim strB As String
4 Dim intA As Integer
5 strA = " This car is beautiful . That car is big . "
 64 Variables, Data Types, and Scopes

6
7 intA = InStr ( strA , " . " )
8 strB = Mid ( strA , 1 , intA )
9
10 Debug . Print strB
11 End Sub

In the above example, we find the first dot in the string and then use the
Mid function to extract the substring. The Mid function can take three ar-
guments: the first argument is a string, the second one is the start position,
and the third one is the length of the substring to be extracted. Executing
the above code produces the following output:
1 This car is beautiful .

Exercise 3.18. Write a sub procedure named StringDemo11 to replace only

the second occurrence of “car” in the following string with “house”:

This car is beautiful. That car is small. Your car is big.

Display the lengths of the original string and the resulting string to make
sure the difference of the lengths is 2. (The word “house” has two more
letters than “car.”)

Exercise 3.19. Write a sub procedure named StringDemo12 to extract

words enclosed by “td>” and “<” in the following string:

“<td>name</td><td>procedure</td><td>word</td>”

Display the extracted words in the Immediate window.

To split a string with a delimiter, we can use the Split function. The
following example shows how to split a string:
1 Sub StringDemo13 ()
2 Dim strA As String
3 Dim arB () As String
4
5 strA = " Apple and Banana "
6 arB = Split ( strA , " and " )
 Dates 65

7
8 Debug . Print LBound ( arB ) , UBound ( arB )
9 Debug . Print arB (0) & " -" & arB (1)
10 End Sub

In the above code, the Split function takes two arguments: the first ar-
gument is the string to be split, and the second argument is the delimiter.
Executing the above code produces the following output:
1 0 1
2 Apple - Banana

Exercise 3.20. Write a sub procedure named StringDemo14 to extract the

numbers in the following string into an array of double numbers

"1.1,0.2,3.14,4.2,0.5,4.8,1.3,6.2,1"
that consists of numbers separated by commas.

3.5 Dates
In many VBA programs, we need to deal with dates at one point or another.
VBA provides some built-in functions for creating and manipulating date
and time. Table 3.3 shows a list of VBA’s built-in functions for manipulating
date and time. In this section, we introduce how to use these function to
manipulate date and time.
The following example shows how to get the current date and time and
extract a portion from a time value:
1 Sub DateDemo1 ()
2 Debug . Print Now
3 Debug . Print Date
4 Debug . Print Year ( Now )
5 Debug . Print Month ( Now )
6 Debug . Print Day ( Now )
7 Debug . Print Hour ( Now )
 66 Variables, Data Types, and Scopes

Table 3.3: VBA’s Built-in Date and Time Functions

Function Description
Date Returns the current date
DateAdd Add and subtract dates
DateDiff Returns the difference in dates
DatePart Returns a portion of a date
DateSerial Returns a Date value representing a specified year,
month, and day, with the time set to midnight
DateValue Returns a Date value containing the date information
represented by a string, with the time set to midnight
Day Returns an Integer value from 1 through 31
representing the day of the month
Format Formats a date according to a specified format
Hour Returns the hour (0 to 23) of a time value
Minute Returns the minute (0 to 59) of a time value
Month Returns the month (1 to 12) of a date value
MonthName Returns a string representing the month given
a number from 1 to 12
Now Returns the current date and time
WeekDay Returns a number representing the day of the week
WeekDayName Returns a string representing the day of the week given
a number from 1 to 7
Year Returns a four-digit year (1900 to 9999) of a date value

8 Debug . Print Minute ( Now )

9 Debug . Print Year ( Date )
10 Debug . Print Month ( Date )
11 Debug . Print Day ( Date )
12 End Sub

Executing the above code produces the following output:

1 1/23/2016 2:28:58 PM
2 1/23/2016
3 2016
4 1
5 23
6 14
7 28
8 2016
9 1
 Dates 67

10 23

The above example shows how to get the current date. We can also
create a date variable and assign a value to it. The following sub procedure
shows how to create a date value:
1 Sub DateDemo2 ()
2 Dim datA As Date
3
4 datA = #12/1/2005#
5 Debug . Print datA
6 Debug . Print Format ( datA , " yyyy / mm / dd " )
7 Debug . Print Format ( datA , " yy / m / d " )
8 End Sub

In the above example, we assign the value “12/1/2005” to the variable datA.
We notice that the date value “12/1/2005” is surrounded by the # sign.
We also use the Format function to display the date in specified formats.
Executing the above code gives the following output:
1 12/1/2005
2 2005/12/01
3 05/12/1

Exercise 3.21. Write a sub procedure named DateDemo3 to get the year,
month, and day of the following date value:

Jan 30, 1990

To add an interval to a date value, we can use the DateAdd function, as

shown in the following example:
1 Sub DateDemo4 ()
2 Dim datA As Date
3
4 datA = #1/23/2009#
5 Debug . Print datA
6 Debug . Print DateAdd ( " d " , 2 , datA ) ' add 2 days
7 Debug . Print DateAdd ( " m " , 3 , datA ) ' add 3 months
8 Debug . Print DateAdd ( " ww " , 1 , datA ) ' add 1 week
 68 Variables, Data Types, and Scopes

9 Debug . Print DateAdd ( " q " , 1 , datA ) ' add 1 quarter

10 Debug . Print DateAdd ( " yyyy " , 1 , datA ) ' add 1 year
11 End Sub

Executing the above code produces the following output:

1 1/23/2009
2 1/25/2009
3 4/23/2009
4 1/30/2009
5 4/23/2009
6 1/23/2010

To find the number of time intervals between two Date values, we use
the DateDiff function. The following example shows how to use this func-
tion:
1 Sub DateDemo5 ()
2 Dim datA As Date
3 Dim datB As Date
4
5 datA = #3/20/2005#
6 datB = #1/22/2016#
7
8 Debug . Print DateDiff ( " d " , datA , datB ) ' days
between
9 Debug . Print DateDiff ( " ww " , datA , datB ) ' calendar
weeks between
10 Debug . Print DateDiff ( " w " , datA , datB ) ' weeks
between
11 Debug . Print DateDiff ( " yyyy " , datA , datB ) ' years
between
12 End Sub

Executing the above code gives the following output:

1 3960
2 565
3 565
4 11

Exercise 3.22. Write a sub procedure named DateDemo6 to create an array

 Scopes 69

of dates starting from January 1, 2016, to May 31, 2016, at an interval of one
week.

Exercise 3.23. Write a function procedure named MLK with one Integer-
type argument, such that MLK(y) returns a Date value representing the
Martin Luther King Day (the third Monday of January) of year y. What is the
output of the following code?
1 Sub DateDemo7 ()
2 Dim i As Integer
3
4 For i = 2010 To 2020
5 Debug . Print MLK ( i )
6 Next i
7 End Sub

Exercise 3.24. Write a function procedure named IsMLK with one Date-
type argument, such that IsMLK(dat) returns True if dat is a Martin Luther
King day, and False if dat is not a Martin Luther King day. What is the
output of the following code?
1 Sub DateDemo8 ()
2 Debug . Print IsMLK ( DateSerial (2016 , 1 , 18) )
3 Debug . Print IsMLK ( DateSerial (2015 , 1 , 19) )
4 Debug . Print IsMLK ( DateSerial (2010 , 1 , 20) )
5 End Sub

3.6 Scopes
In this section, we introduce the scopes of variables, constants, and proce-
dures.
A scope of a variable determines which modules and procedures can
use the variable. The scope of a variable is determined at the time when the
variable is declared. In VBA, there are three scopes available for variables:
procedure-only, module-only, and public. A procedure-only variable can
only be used in the procedure where the variable is declared. A module-only
 70 Variables, Data Types, and Scopes

variable is declared in the declaration section of a module and is available

to all procedures in the module. A public variable is also declared in the
declaration section of a module and can be used in all procedures in all
modules.

Table 3.4: Scopes of a Variable

Scope Declaration
Procedure-only Use a Dim or Static statement in the procedure
Module-only Use a Dim or Private statement before the
first procedure in the module
Public Use a Public statement before the first procedure
in a module

Table 3.4 lists approaches to declare a variable that has different scopes.
In the following sub procedure, we declare two procedure-only variables:
1 Sub ScopeDemo1 ()
2 Dim intA As Integer
3 Static intCount As Integer
4
5 intA = intA + 1
6 intCount = intCount + 1
7
8 Debug . Print intA
9 Debug . Print intCount
10 End Sub

The variables intA and intCount can only be used in the procedure
ScopeDemo1. The difference between the two variables is that one is de-
clared with the Dim statement we used before, and the other is declared
with the Static statement. The variable intA will be reset every time when
the procedure is called. However, the variable intCount remains in exis-
tence the entire time VBA is running. A static variable is reset when any of
the following occur:

• The macro generates certain run-time errors.

• VBA is halted.
• You quit Excel.
• You move the procedure to a different module.
• You reset the VBA program.
 Scopes 71

To see how to declare module-only variables, let us consider the follow-

ing module named MScope2:
1 Private dSum As Double
2 Dim dMean As Double
3 Dim intCount As Integer
4
5 Sub ScopeDemo2 ()
6 dSum = dSum + Work sheetFun ction . RandBetween (1 ,
100)
7 intCount = intCount + 1
8
9 dMean = dSum / intCount
10 End Sub
11
12 Sub ScopeDemo3 ()
13 Debug . Print dSum
14 Debug . Print dMean
15 Debug . Print intCount
16 End Sub

The module contains three module-only variables: dSum, dMean, and

intCount. Module-only variables can be declared with a Private state-
ment or a Dim statement. The three variables are available to all procedures
in this module. These variables are not available to procedures in other
modules.
The module also contains two sub procedures: ScopeDemo2 and
ScopeDemo3, respectively. If we run the procedure ScopeDemo2 three times
and run the procedure ScopeDemo3 once, we see the following output:
1 90
2 30
3 3

Now, let us introduce how to declare public variables that are available
to all procedures in all modules. Suppose that a module named MScope3
contains the following content:
1 ' Module MScope3
2
3 Public strMainFile As String
4
5 Sub ScopeDemo4 ()
6 Debug . Print strMainFile
7 End Sub
 72 Variables, Data Types, and Scopes

Then we can use the variable strMainFile in other modules. For example,
we can write a sub procedure in the module MScope1 as follows:
1 Sub scopedemo5 ()
2 ScopeDemo3 . strMainFile = " Chapter1 . xlsm "
3 End Sub

In the above code, we set a value to the public variable strMainFile by

referring to the variable with its module’s name.
Like variables, constants have scopes too. There are also three scopes
available to a constant: procedure-only, module-only, and public. Table 3.5
presents the three scopes of a constant and how to define them. We already
introduced how to define procedure-only constants in Section 3.3.

Table 3.5: Scopes of a Constant

Scope Declaration
Procedure-only Use a Const statement in the procedure
Module-only Use a Const or Private Const statement before the
first procedure in the module
Public Use a Public Const statement before the first
procedure in a module

Exercise 3.25. Let MS1 be the name of a module containing the following
content:
1 ' MS1
2 Public Const conPi As Doulbe = conPi2 / 2
3
4 Sub PrintPi ()
5 Debug . Print conPi
6 End Sub

Let MS2 be the name of a module containing the following content:

1 ' MS2
2 Public Const conPi2 As Doulbe = conPi * 2
3
 Summary 73

4 Sub PrintPi ()
5 Debug . Print conPi2
6 End Sub

What is wrong with the above code?

For procedures, there are two scopes available to them: module-only

and public. A module-only procedure can be called by other procedures in
the same module. A public procedure can be called by procedures in other
modules. Table 3.6 shows the scopes of a sub procedure and how to define
them. The scopes of a function procedure can be defined similarly.

Table 3.6: Scopes of a Procedure

Scope Declaration
Module-only Use a Private Sub to define the procedure
Public Use a Sub or Public Sub statement to define the
procedure

3.7 Summary
In this chapter, we introduced how to declare variables, arrays, and con-
stants in VBA. A variable is a named storage location in a computer’s mem-
ory. An array is a group of variables that share a common name. A constant is
a meaningful name that takes the place of a number or string that does not
change during execution of a program. We also introduced how to manipu-
late strings and dates. Finally, we gave a brief introduction to the concept
of scopes for variables, constants, and procedures. For more information
about manipulating strings and dates in VBA, readers are referred to Getz
and Gilbert (2001).
4
Operators and Control Structures

In this chapter, we introduce Visual Basic for Applications (VBA)’s built-in

operators that can be used in our Excel® VBA code. We also introduce some
control structures that can be used to direct the flow of a VBA program.
After studying this chapter, readers will be able to

• use VBA’s built-in operators.

• use the flow control statements, such as If-Then and Select-Case.
• use loops such as For-Next and Do-While.

4.1 Operators
There are four types of operators in VBA: arithmetic operators, string opera-
tors, comparison operators and logical operators.

Table 4.1: VBA’s Arithmetic Operators

Operator Description Precedence

ˆ Exponentiation 1
* Multiplication 2
/ Division 2
Mod Modulus operator 3
+ Addition 4
- Subtraction or negation 4

Precedence 1 has the highest precedence in the absence of brackets.

Table 4.1 gives a list of VBA’s arithmetic operators. The default prece-
dences of the operators are also listed and are applied in the absence of
brackets. All the operators are apparent to most people except the modulus
operator. The modulus operator finds the remainder after the division of

75
 76 Operators and Control Structures

one number by another. The following sub procedure shows how to use
these operators:
1 Sub OperatorDemo1 ()
2 Debug . Print 1 + 2 * 3
3 Debug . Print (1 + 2) * 3
4 Debug . Print 2 ^ 3 * 3
5 Debug . Print 10 Mod 3
6 Debug . Print 8 - 4 / 2
7 Debug . Print (8 - 4) / 2
8 End Sub

Executing the above code produces the following output:

1 7
2 9
3 24
4 1
5 6
6 2

Exercise 4.1. Write a sub procedure named OperatorDemo2 to calculate

the following expression:

(1 − 2)2 1
− − ln(2π) − ln 3.
2 × 32 2
Exercise 4.2. What is the output of the following code?
1 Sub OperatorDemo3 ()
2 Debug . Print 216 + 2 * 30 ^ 2
3 End Sub

If you multiply two large integers, VBA will raise an overflow error. For
example, if you execute the following code:
1 Sub OperatorDemo4 ()
2 Debug . Print 1000 * 1000
3 End Sub
 Operators 77

you will see a run-time error “Overflow.” The reason is that VBA treats the
two numbers as integers. Since the product of the two integers is outside the
range of an integer in VBA, VBA raised the run-time error. To fix this problem,
we can convert the integer to double before doing the multiplication:
1 Sub OperatorDemo5 ()
2 Debug . Print 1000# * 1000
3 End Sub

The symbol # after 1000 means that 1000 is a double, not an integer. We can
also use the function CDbl to convert a number to a double:
1 Sub OperatorDemo6 ()
2 Debug . Print CDbl (1000) * 1000
3 End Sub

For a complete list of conversion functions, readers are referred to the

Conversion module in the Visual Basic Editor (see Figure 4.1).

Figure 4.1: Conversion functions in the Visual Basic Editor

Exercise 4.3. Write a sub procedure named OperatorDemo7 to calculate

 78 Operators and Control Structures

the following quantity:

v v
u u s
u u r
u t q p
t
1 + 2015 1 + 2016 1 + 2017 1 + 2018 1 + 2019 1 + 2020(2015 + 7)

Table 4.2: VBA’s String Operator

Operator Description
& Concatenation operator
+ Concatenation operator

Table 4.2 shows two string operators. VBA’s main string operator is the
concatenation operator &, which is used to concatenate two strings. The
following example shows how to use the & operator:
1 Sub OperatorDemo8 ()
2 Debug . Print " Abc " & " 123 "
3 Debug . Print " Abc " & CStr (123)
4 Debug . Print ThisWorkbook . Name & " is " & " " & " a
workbook "
5 End Sub

In the above code, we use the function CStr to convert the number 123 to a
string. Executing the above code produces the following code:
1 Abc123
2 Abc123
3 Chapter4 . xlsm is a workbook

The plus sign + can also be used to concatenate two strings. If we apply
+ to two strings, we get a concatenated string. If we apply + to two numbers,
we get the sum of the numbers. If we apply + to a number and a string,
we may get an error. The following example shows the application of + in
various situations:
1 Sub OperatorDemo9 ()
2 Debug . Print " abc " + " 123 "
3 Debug . Print " 123 " + 456
4 Debug . Print " 123 " + " 456 "
 Operators 79

5 Debug . Print 123 + 456

6 End Sub

Executing the above code produces the following output:

1 abc123
2 579
3 123456
4 579

From the output, we see that the string "123" was converted to the number
123 automatically.

Exercise 4.4. What is wrong with the following code?

1 Sub OperatorDemo10 ()
2 Debug . Print " abc " + 123
3 End Sub

Exercise 4.5. What is wrong with the following code?

1 Sub OperatorDemo11 ()
2 Debug . Print " abc " & 123
3 End Sub

Table 4.3: VBA’s Comparison Operators

Operator Description
= Equality
<> Inequality
< Less than
> Greater than
<= Less than or equal to
>= Greater than or equal to
Is Compare two object reference variables
IsNot Compare two object reference variables
 80 Operators and Control Structures

Table 4.4: VBA’s Logical Operators

Operator Description
Not Logical opposite
And Logical conjunction
Or Logical disjunction or inclusion
Xor Logical exclusion

Table 4.3 and Table 4.4 give a list of comparison operators and a list of
logical operators, respectively. These operators are usually used together.
The following example shows how to use these operators:
1 Sub OperatorDemo12 ()
2 Debug . Print 1 = 2
3 Debug . Print Not 1 = 2
4 Debug . Print 1 <> 2
5 Debug . Print 1 < 2
6 Debug . Print 1 > 2
7 Debug . Print 1 <> 2 Or False
8 Debug . Print 1 <> 2 And False
9 Debug . Print True Xor False
10 Debug . Print True Xor True
11 Debug . Print False Xor False
12 End Sub

Executing the above code produces the following output:

1 False
2 True
3 True
4 True
5 False
6 True
7 False
8 True
9 False
10 False

Exercise 4.6. What is the output of the following code?

 Flow Control 81

1 Sub OperatorDemo13 ()
2 Debug . Print (1 > 2 Or 2 = 2) And ( False Xor True )
3 Debug . Print ( Not 1 <> 2) Xor (2 = 3 And 3 = 3)
4 End Sub

4.2 Flow Control

The procedures we created in previous sections start at the code’s begin-
ning and progress line by line to the end. All these procedures never divert
from such top-to-bottom flow. In this section, we introduce some meth-
ods to control the flow of VBA programs. In particular, we introduce some
programming constructs that allow us to skip over some statements.

Table 4.5: VBA’s Programming Constructs for Flow Control

Construct Description
GoTo statement Jump to a particular statement
If-Then statement Does something if a condition is true
Select Case statement Does one of several things according to
something’s value

Table 4.5 shows several programming constructs that we can use to

control the flow of VBA programs. A GoTo statement is a simple way to
change the flow of a program. The following example shows how a GoTo
statement works:
1 Sub GoToDemo1 ()
2 Debug . Print 1
3 Debug . Print 2
4 GoTo LabelA
5 Debug . Print 3
6
7 LabelA :
8 Debug . Print 4
9 End Sub
 82 Operators and Control Structures

In the above code, we defined a label named “LabelA” in Line 7. From the
code, we see that a label is just a string followed by a colon. Since we used a
GoTo statement in Line 4, the code in Line 5 is skipped. Executing the above
code produces the following output:
1 1
2 2
3 4

We should avoid using GoTo statements in VBA programs because using

GoTo statements can make a program too complicated to understand (see
Exercise 4.8). We can always use other constructs to replace GoTo state-
ments.
g

Exercise 4.7. What is the output of the following code?

1 Sub GoToDemo2 ()
2 Debug . Print 1
3 GoTo LabelA
4 Debug . Print 2
5 LabelB :
6 Debug . Print 3
7 GoTo LabelC
8 Debug . Print 4
9
10 LabelA :
11 Debug . Print 5
12 GoTo LabelB
13 Debug . Print 6
14
15 LabelC :
16 Debug . Print 7
17 End Sub

Exercise 4.8. What is wrong with following code?

1 Sub GoToDemo3 ()
2 Debug . Print 1
3 GoTo LabelA
4 Debug . Print 2
5 LabelB :
6 Debug . Print 3
7
8 LabelA :
 Flow Control 83

9 Debug . Print 4
10 GoTo LabelB
11 Debug . Print 5
12
13 End Sub

An If-Then statement can also be used to change the flow of a program.

The following example shows how to use an If-Then statement:
1 Sub IfThenDemo1 ()
2 If TypeOf ThisWorkbook Is Workbook Then
3 Debug . Print " This workbook is a workbook "
4 Else
5 Debug . Print " This workbook is not a workbook "
6 End If
7
8 If TypeOf ThisWorkbook Is Worksheet Then
9 Debug . Print " This workbook is a Worksheet "
10 Else
11 Debug . Print " This workbook is not a Worksheet
"
12 End If
13 End Sub

Exercise 4.9. The VBA function CDbl(Time) converts 24-hour time to a

real number between 0 and 1, with 0 corresponding to 12:00am and 0.5
corresponding to 12:00pm. Write a sub procedure named IfThenDemo2 to
display “Good Morning,” “Good Afternoon,” and “Good Evening” based on
the time.

The If-Then statements can be nested, that is, an If-Then statement

can be embedded into another one. The following example shows how to
use nested If-Then statements:
1 Sub IfThenDemo3 ()
2 Dim intDay As Integer
3 Dim intM As Integer
 84 Operators and Control Structures

4 Dim intWeekDay As Integer

5
6 intM = Month ( Date )
7 intDay = Day ( Date )
8 intWeekDay = Weekday ( Date )
9
10 ' In Canada , Family day is the third Monday in
February
11 If intM = 2 Then
12 If intDay > 14 And intDay < 22 Then
13 If intWeekDay = 2 Then
14 Debug . Print " Today is Family day "
15 Else
16 Debug . Print " Today is not Family day "
17 End If
18 Else
19 Debug . Print " Today is not Family day "
20 End If
21 Else
22 Debug . Print " Today is not Family day "
23 End If
24 End Sub

In the above code, we have three If-Then statements. The first If-Then
statement checks whether today is in February; the second statement
checks whether today is in the third week of the month; and the last state-
ment checks whether today is a Monday. The function Date returns today’s
date. We also use the VBA function Month, Day, and WeekDay to get the
month, day, and week day from today’s date, respectively.

Exercise 4.10. Labor Day is the first Monday in September. Write a sub
procedure named IfThenDemo4 to check whether today is Labor Day.

Exercise 4.11. What is the output of the following code?

1 Sub IfThenDemo5 ()
2 Dim intTmp As Integer
3 LabelA :
4 intTmp = intTmp + 2
5
6 If intTmp > 10 Then
7 GoTo LabelB
8 Else
 Flow Control 85

9 GoTo LabelA
10 End If
11 LabelB :
12 Debug . Print intTmp
13 End Sub

If we have several cases, then it is tedious to use the If-Then state-

ment. In such cases, we can use the Select-Case statement. The following
example shows how to use the Select-Case statement in VBA:
1 Sub SelectDemo1 ()
2 Dim intDay As Integer
3
4 intDay = Day ( Date )
5
6 Select Case intDay
7 Case 1 To 15
8 Debug . Print " Today is in the first half
of the month "
9 Case 16 To 31
10 Debug . Print " Today is in the second half
of the month "
11 End Select
12 End Sub

The above sub procedure displays “Today is in the first half of the month” if
the day is between 1 and 15 (inclusive), or displays “Today is in the second
half of the month” if the day is between 16 and 31 (inclusive). We can also
use a single number in a case, as shown in the following example:
1 Sub SelectDemo2 ()
2 Dim intWeekDay As Integer
3
4 intWeekDay = Weekday ( Date )
5
6 Select Case intWeekDay
7 Case 1
8 Debug . Print " Today is Sunday "
9 Case 2 To 6
10 Debug . Print " Today is a WeekDay "
11 Case 7
12 Debug . Print " Today is Saturday "
 86 Operators and Control Structures

13 End Select
14 End Sub

We can also use strings in cases:

1 Sub SelectDemo3 ()
2 Dim strM As String
3
4 strM = Format ( Date , " mmm " )
5 Select Case strM
6 Case " Jan "
7 Debug . Print " January "
8 Case " Feb "
9 Debug . Print " February "
10 Case Else
11 Debug . Print " Not January or February "
12 End Select
13 End Sub

In the above code, we use the function Format to convert a date into a
three-letter month symbol. Then we display the full name of the month.
Since there are twelve months, we did not write down all the cases. Instead,
we used the Case Else to include all other cases.

Exercise 4.12. The VBA function VarType returns the data type (e.g.,
vbDouble, vbCurrency, vbString, etc; see VarType in the Object Browser
for a complete list) of a variable. Write a sub procedure named SelectDemo4
to display the data type of the value of Cell “A1” in the active sheet.

4.3 Loops
In this section, we introduce VBA’s programming constructs that can be
used to execute a series of statements multiple times. These constructs are
referred to as loops.
Table 4.6 shows three types of loops supported by the VBA programming
language: the For-Next loop, the Do-While loop, and the Do-Until loop.
 Loops 87

Table 4.6: VBA’s Loops

Construct Description
For-Next loop Executes some statements for a specified
number of times
Do-While loop Executes some statements as long as something
remains true
Do-Until loop Executes some statements until something
becomes true

Let us first introduce the For-Next loop, which can be used to execute a
series of statements for a specified number of times.
The following example shows how to use the For-Next loop:
1 Sub ForNextDemo1 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 For n = 1 To 100
6 dSum = dSum + n
7 Next n
8
9 Debug . Print dSum
10 End Sub

In the above code, we use a For-Next loop to calculate the sum of integers
from 1 to 100. Executing the code gives the following output:
1 5050

Exercise 4.13. Write a sub procedure named ForNextDemo2 to calculate

the sum of integers from 1 to 10,000 and display the sum in the Immediate
window every 1000 steps.

Exercise 4.14. Write a sub procedure named ForNextDemo3 to calculate

the following sum:
1000
X¡ 2
i + 3i 3 .
¢
i =10
 88 Operators and Control Structures

In a For-Next loop, we can also specify the step size by adding a Step
statement after the For statement. The following example shows how to
specify a step size:
1 Sub ForNextDemo4 ()
2 Dim vA (1 To 100) As Double
3 Dim n As Integer
4
5 For n = LBound ( vA ) To UBound ( vA ) Step 2
6 vA ( n ) = 1
7 Next n
8
9 Debug . Print vA (1) ; vA (2) ; vA (3) ; vA (4) ; vA (5)
10 End Sub

In the above code, we declare an one-dimensional array with 100 elements

and use a For-Next loop to assign 1 to the elements with a step size of 2.
Then, we display the first five elements of the array. Executing the code
produces the following output:
1 1 0 1 0 1

Note that the semicolon “;” after the Print function means printing imme-
diately after one another.

Exercise 4.15. Write a sub procedure named ForNextDemo5 to create the

following one-dimensional array

(0, 1, 0, 0, 1, 0, 0, 1, 0, . . . , 0, 1, 0),

which contains 99 elements.

Exercise 4.16. Write a sub procedure named ForNextDemo6 to create the

following one-dimensional array,

(2, 1, 3, 2, 1, 3, 2, 1, 3, . . . , 2, 1, 3),

which contains 30 elements.

 Loops 89

Exercise 4.17. Write a sub procedure named ForNextDemo7 to create the

following 100 × 100 array,
 
0 1 0 ··· 0
1 0 1 · · · 0 
 
0 1 0 · · · 0  ,
 
 .. .. .. . . .
 
. . . . .. 
0 0 0 ··· 0
which has ones in its subdiagonal elements and zeros in other elements.
h

Like the If-Then statement, the For-Next statement can be nested.

Suppose that we want to calculate the following sum:
10 X
X 10
(i + j )( j + 1).
i =1 j =1

Then, we can use nested For-Next statements as follows:

1 Sub ForNextDemo8 ()
2 Dim dSum As Double
3 Dim i As Integer
4 Dim j As Integer
5
6 For i = 1 To 10
7 For j = 1 To 10
8 dSum = dSum + ( i + j ) * ( j + 1)
9 Next j
10 Next i
11
12 Debug . Print dSum
13 End Sub

g
Exercise 4.18. Write a sub procedure named ForNextDemo9 to create the
following 100 × 100 array:
 
0 1 2 3 · · · 99
1 2 3 4 · · · 100
 
2 3 4 5 · · · 101 .
 
 .. .. .. .. .. .. 
 
. . . . . . 
99 100 101 102 · · · 198
 90 Operators and Control Structures

Then put the array into the range “A1:CV100” of the active worksheet.

Exercise 4.19. Write a sub procedure named ForNextDemo10 to calculate

the following sum:
X20 X 10 i3
.
i =1 j =1 i j + 5

Exercise 4.20. What is wrong with the following code?

1 Sub ForNextDemo11 ()
2 Dim dSum As Double
3 Dim i As Integer
4 Dim j As Integer
5
6 For i = 1 To 20
7 For j = 1 To 10
8 dSum = dSum + i * j
9 Next i
10 Next j
11
12 Debug . Print dSum
13 End Sub

We can terminate a For-Next loop by using an Exit For statement

inside the loop. If we want to find the smallest integer N , such that

N
i 2 > 1000,
X
i =1

we can use the following sub procedure:

1 Sub ForNextDemo12 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 For n = 1 To 1000
6 dSum = dSum + n ^ 2
7 If dSum > 1000 Then
8 Exit For
9 End If
10 Next n
11
 Loops 91

12 Debug . Print n
13 Debug . Print dSum
14 End Sub

Executing the above sub procedure produces the following output:

1 14
2 1015

Exercise 4.21. The IsEmpty function can be used to check whether a cell is
empty or not. Write a sub procedure named ForNextDemo13 to display the
row index of the first empty cell in the first column of the active worksheet.

Now, let us introduce the Do-While loop, which relies on a condition

to terminate the loop. The Do-While loop is suitable for situations when
you do not know how many iterations you want to execute the code. For
example, we can use a Do-While loop to find the smallest integer N , such
that
N
i 2 > 1000.
X
i =1

We used a For-Next loop with an Exit statement to find such an integer.

However, we specified the maximum number of iterations in the For-Next
loop. If we use a Do-While loop, we do not need to specify the maximum
number of iterations. The following sub procedure shows how a Do-While
loop works:
1 Sub DoWhileDemo1 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 Do While dSum <= 1000
6 n = n + 1
7 dSum = dSum + n ^ 2
8 Loop
9
10 Debug . Print n
11 Debug . Print dSum
12 End Sub
 92 Operators and Control Structures

In the above code, we do not specify a maximum number of iterations.

However, we increase the value of n by 1 at each iteration.
In a Do-While loop, we need to make sure the condition will become
false after a fixed number of iterations. Otherwise, the loop becomes an
infinite loop and will not stop (see Exercise 4.22).

Exercise 4.22. What is wrong with the following code?

1 Sub DoWhileDemo2 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 Do While dSum <= 1000
6 dSum = dSum + n ^ 2
7 Loop
8
9 Debug . Print n
10 Debug . Print dSum
11 End Sub

Exercise 4.23. Write a sub procedure named DoWhileDemo3 that uses a

Do-While loop to calculate the following sum:
100
X
i.
i =1

The Do-Until statement is similar to the Do-While statement in terms

of syntax. However, the two loop statements are different in how they handle
the testing condition. A program continues to execute statements inside a
Do-While loop while the condition remains true. A program continues to
execute statements in a Do-Until loop until the condition is true.
Let us consider the example of finding the smallest integer N , such that

N
i 2 > 1000.
X
i =1

We can find the integer using a Do-Until loop as follows:

 Summary 93

1 Sub DoUntilDemo1 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 Do Until dSum > 1000
6 n = n + 1
7 dSum = dSum + n ^ 2
8 Loop
9
10 Debug . Print n
11 Debug . Print dSum
12 End Sub

Executing the above code produces the following output:

1 14
2 1015

From the above example, we see that the Do-Until statement and the
Do-While statement are exchangeable by just changing the condition state-
ment.

4.4 Summary
In this chapter, we introduced VBA’s built-in operators and some control
structures. VBA provides four types of operators: arithmetic operators, string
operators, comparison operators, and logical operators. The control struc-
tures in VBA include If-Then, Select-Case, and GoTo. VBA loops include
For-Next, Do-While, and Do-Until.
 5
Functions, Events, and File IO

As we mentioned before, a module contains two types of procedures: sub

procedures and function procedures. We already introduced sub proce-
dures. In this chapter, we introduce how to create function procedures. We
also introduce event handlers and methods for reading and writing files.
After studying this chapter, readers will be able to

• create functions.
• create event handlers.
• read and write text files.

5.1 User-Defined Functions

The difference between a sub procedure and a function procedure is that
the former does not return a result to the caller, but the later returns a result
to the caller. In the following example, we create a function named MySum
with two double-type arguments:
1 Function MySum ( x As Double , y As Double )
2 MySum = x + y
3 End Function

The function MySum can be called by other functions or procedures. For

example, we can use the above function as follows:
1 Sub FunctionDemo1 ()
2 Debug . Print MySum (1 , 2)
3 Debug . Print MySum ( MySum (2 , 3) , 4)
4 End Sub

Executing the above code produces the following output in the Immediate
window:

95
 96 Functions, Events, and File IO

1 3
2 9

The function can also be used in cell formulas. For example, we can use the
function in Cell “C1” by typing “=MySum(A1,B1)” in Cell “C1.” Then, the
value of Cell “C1” will be the sum of the two numbers in Cells “A1” and “B1.”

Exercise 5.1. Write a function procedure named MySum2 with one Long-
type argument, such that MySum2(n) returns the following value:
n
i 2.
X
i =1

What is the output of the following code?

1 Sub FunctionDemo2 ()
2 Debug . Print MySum2 (10)
3 Debug . Print MySum2 (100)
4 End Sub

Exercise 5.2. Write a function procedure named DNorm with three Double-
type arguments, such that DNorm(x ,µ,σ) calculates the following value

(x − µ)2
µ ¶
1
f (x, µ, σ) = p exp − ,
2πσ 2σ2

where exp(·) is the exponential function, i.e., exp(y) = e y . What is the output
of the following code?
1 Sub FunctionDemo3 ()
2 Debug . Print DNorm (0 , 0 , 1)
3 Debug . Print DNorm (0 , 2 , 3)
4 Debug . Print Works heetFunc tion . NormDist (0 , 0 , 1 ,
False )
5 Debug . Print Works heetFunc tion . NormDist (0 , 2 , 3 ,
False )
6 End Sub

You can use VBA’s built-in function EXP to calculate exp(·).

 User-Defined Functions 97

In the above examples, the arguments are single numbers. In fact, we

can pass arrays to a function procedure. The following function takes an
array-type argument:
1 Function MySum3 ( x () As Double )
2 Dim dSum As Double
3 Dim i As Integer
4
5 For i = LBound ( x ) To UBound ( x )
6 dSum = dSum + x ( i )
7 Next i
8
9 MySum3 = dSum
10 End Function

The above function calculates the sum of all elements of the input array. We
can test this function using the following sub procedure:
1 Sub FunctionDemo4 ()
2 Dim arA (1 To 100) As Double
3 Dim i As Integer
4
5 For i = 1 To 100
6 arA ( i ) = i
7 Next i
8
9 Debug . Print MySum3 ( arA )
10 End Sub

Executing the above code gives the following output:

1 5050

Exercise 5.3. The sample mean and the sample standard deviation of a vec-
tor of numbers can be calculated incrementally as follows. Let y 1 , y 2 , . . . , y n
be n numbers. Let {M 1,i : i = 1, 2, . . . , n} and {M 2,i : i = 1, 2, . . . , n} be two
sequences defined as follows:

y k − M 1,k−1
M 1,1 = y 1 , M 1,k = M 1,k−1 + , k = 2, 3, . . . , n,
k
 98 Functions, Events, and File IO

and
k −1
M 2,1 = 0, M 2,k = M 2,k−1 + (y k − M 1,k−1 )2 , k = 2, 3, . . . , n.
k
Then,
ȳ = M 1,n
and s
M 2,n
sy = , n > 1,
n −1
where ȳ and s y denote the sample mean and the sample standard deviation
of y 1 , y 2 , . . . , y n .
Write two function procedures named IncMean and IncStd with one
array-type argument, such that IncMean(y) and IncStd(y) calculate the
sample mean and the sample standard deviation of the array y incremen-
tally, respectively. What is the output of the following code?
1 Sub FunctionDemo5 ()
2 Dim vA (1 To 100) As Double
3 Dim i As Integer
4 For i = 1 To 100
5 vA ( i ) = Rnd () * 10000
6 Next i
7
8 Debug . Print IncMean ( vA )
9 Debug . Print IncStd ( vA )
10 Debug . Print Works heetFunc tion . Average ( vA )
11 Debug . Print Works heetFunc tion . StDev ( vA )
12 End Sub

Function procedures can be recursive, that is, a function procedure

can call itself. For example, we consider the Fibonacci numbers F 0 , F 1 , . . .
defined as follows:

F n = F n−1 + F n−2 , n ≥ 2, F 0 = 0, F 1 = 1. (5.1)

We can write a nested function procedure to calculate the Fibonacci number

at an arbitrary location n as follows:
 User-Defined Functions 99

1 Function Fibonacci ( n As Long )

2 If n = 0 Then
3 Fibonacci = 0
4 ElseIf n = 1 Then
5 Fibonacci = 1
6 Else
7 Fibonacci = Fibonacci ( n - 1) + Fibonacci ( n -
2)
8 End If
9 End Function

The above function Fibonacci(n) returns the Fibonacci number F n . To

test the above function, we run the following code:
1 Sub FunctionDemo6 ()
2 Debug . Print Fibonacci (10)
3 Debug . Print Fibonacci (20)
4 End Sub

which produces the following output:

1 55
2 6765

Exercise 5.4. Write a function procedure named Factorial with one Long-
type argument, such that Factorial(n) calculates the following value
n
Y
n! = 1 × 2 × · · · n = i.
i =1

We define 0! = 1. What is the output of following code?

1 Sub FunctionDemo7 ()
2 Debug . Print Factorial (0)
3 Debug . Print Factorial (10)
4 Debug . Print Factorial (20)
5 End Sub

Exercise 5.5. Recursive functions are usually not efficient. We can al-
most always substitute a loop for recursion. Write a nonrecursive function
procedure named Fibonacci2 with one Long-type argument, such that
 100 Functions, Events, and File IO

Fibonacci2(n) calculates the Fibonacci number Fn defined in Equation

(5.1).

The function we discussed so far returns a single value. In fact, a func-

tion can also return an array. The following code shows a function that
returns an array.
1 Function RandNumbers ( n As Integer )
2 Dim arA () As Double
3 Dim i As Integer
4
5 ReDim arA (1 To n )
6 Rnd ( -1)
7 Randomize (0)
8 For i = 1 To n
9 arA ( i ) = Rnd ()
10 Next i
11
12 RandNumbers = arA
13 End Function

In the above code, the function RandNumbers has one Integer-type argu-
ment and returns an array of random numbers. We use the VBA functions
Rnd and Randomize to generate random numbers. The function Randomize
is used to fix the seed of the random number generator so that we can repeat
the random numbers every time we call the function RandNumbers. To test
this function, we use the following sub procedure:
1 Sub FunctionDemo8 ()
2 Dim arA As Variant
3
4 arA = RandNumbers (10)
5 Debug . Print LBound ( arA ) , UBound ( arA )
6 Debug . Print arA (1) , arA (2) , arA (3)
7 End Sub

Executing the above sub procedure, we see the following output:

1 1 10
2 0.332 84288644 7906 0.3 20361137 390137
0.087 23753690 7196
 Events 101

Exercise 5.6. What is the difference of the following two functions?

1 Function RandNumbers ( n As Integer )
2 Dim arA () As Double
3 Dim i As Integer
4
5 ReDim arA (1 To n )
6 Rnd ( -1)
7 Randomize (0)
8 For i = 1 To n
9 arA ( i ) = Rnd ()
10 Next i
11
12 RandNumbers = arA
13 End Function
14
15 Function RandNumbers2 ( n As Integer )
16 Dim arA () As Double
17 Dim i As Integer
18
19 ReDim arA (1 To n )
20 Randomize (0)
21 For i = 1 To n
22 arA ( i ) = Rnd ()
23 Next i
24
25 RandNumbers2 = arA
26 End Function

5.2 Events
A Visual Basic for Applications (VBA) sub procedure can be executed au-
tomatically when a particular event occurs. In this section, we introduce
some Excel® event types and where to place event-handler VBA code.
Table 5.1 gives a list of workbook events. To add a workbook event-
handler, we need to double click ThisWorkbook in the VBAProject window
102 Functions, Events, and File IO

Table 5.1: Workbook Events

Event Description
Activate The workbook is activated
BeforeClose The workbook is closed
BeforeSave The workbook is saved
Deactivate The workbook is deactivated
NewSheet A new sheet is added to the workbook
Open The workbook is opened
SheetActivate A sheet in the workbook is activated
SheetChange A cell in the workbook is changed

Figure 5.1: Workbook events in the code window

 Events 103

of the Visual Basic Editor, select Workbook in the left drop-down list of the
code window, and then select the event type in the right drop-down list of
the code window (see Figure 5.1). To add an event-handler to the NewSheet
event, for example, we select the NewSheet event from the drop-down list.
The following code is an example of the event-handler for the NewSheet
event:
1 Private Sub Workbook _NewShee t ( ByVal Sh As Object )
2 MsgBox " A new sheet named " & " " & Sh . Name & "
is added "
3 End Sub

The event-handler is a sub procedure with one argument. The argument

is the new sheet added to the workbook. If we add a new sheet to the
workbook, we will see a message box that shows the sheet’s name.

Exercise 5.7. Write an event-handler for the NewSheet event, such that a
new sheet will be deleted without alert as soon as it is created.

Exercise 5.8. Add a piece of VBA code to a workbook, such that the total
number of sheets (including charts and worksheets) will not exceed 5.

Exercise 5.9. Write a piece of VBA code for your workbook, such that when-
ever a workbook is saved, a time stamp is written to the cell “D1” of the
sheet “Sheet1” of your workbook.

Table 5.2: Worksheet Events

Event Description
Activate The worksheet is activated
Change A cell in the worksheet is changed
Deactivate The worksheet is deactivated
SelectionChange The selection is changed

From the above example, we see that event-handlers for workbook

events have to sit in the workbook object. Similarly, event-handlers for
worksheet events have to sit in a worksheet object. Table 5.2 shows some
worksheet events.
 104 Functions, Events, and File IO

Suppose that the cell “C3” of the worksheet “Sheet1” is a positive num-
ber that is an input for some program. We need to make sure the number
in the cell is positive. We can add an event-handler for the Change event.
To do that, we double click the worksheet named “Sheet1” and select the
Change event from the right drop-down list in the code window. Then, we
add the following code:
1 Private Sub Worksheet_Change ( ByVal Target As Range )
2 If Target . Parent . Name = " Sheet1 " And Target .
Address = " $C$3 " Then
3 If Target . Value <= 0 Then
4 MsgBox " Invalid number "
5 End If
6 End If
7 End Sub

The above code is executed when the name of the sheet is “Sheet1” and the
address of the changed cell is $C$3. If we input −1 in the cell C3 of the sheet
“Sheet1,” we will see a message box.

Exercise 5.10. Add a piece of code to a worksheet, such that whenever

the worksheet is activated, a message box with the worksheet’s name is
displayed.

5.3 File IO
In this section, we introduce how to use VBA to read and write text files.
Let us introduce how to write text files using VBA code. Suppose that the
range “A1:B110” of the sheet “Sheet1” contains a mortality table. To write
the mortality table to a Comma Separated Values (CSV) file, we can use the
following VBA code:
1 Sub FileDemo1 ()
2 Dim strFileName As String
3 Dim ws As Worksheet
4 Dim i As Integer
 File IO 105

5
6 Set ws = ThisWorkbook . Sheets ( " Sheet1 " )
7 strFileName = " MaleTable . csv "
8 Open strFileName For Output As #1
9 For i = 1 To 110
10 Write #1 , ws . Cells (i , 1) , ws . Cells (i , 2)
11 Next i
12 Close #1
13
14 End Sub

In the above code, we open a file for output in Line 8. The file object is
denoted by #1, as it is the first file we opened. In Lines 9–11, we write 110
lines to a CSV file named “MaleTable.csv.” In Line 12, we close the CSV file.
Executing the above code will produce a CSV file containing the numbers
in the range “A1:B110.”
If we open the file “MaleTable.csv” created by the above VBA code, we
see that all numbers are surrounded by double quotes. To write a CSV file
without double quotes, we can use the following VBA code:
1 Sub FileDemo2 ()
2 Dim strFileName As String
3 Dim ws As Worksheet
4 Dim i As Integer
5
6 Set ws = ThisWorkbook . Sheets ( " Sheet1 " )
7 strFileName = " MaleTable . csv "
8 Open strFileName For Output As #1
9 For i = 1 To 110
10 Print #1 , ws . Cells (i , 1) & " ," & ws . Cells (i ,
2)
11 Next i
12 Close #1
13
14 End Sub

The difference between the sub procedures FileDemo2 and FileDemo1 is

that we use the Print function in the former. The Print function writes a
string to the file. If we write the CSV file using FileDemo2, we will not see
the double quotes around the numbers.
In the sub procedure FileDemo2, we concatenate the values using the &
operator and write the resulting string as a new line to the file. In fact, we
 106 Functions, Events, and File IO

do not have to concatenate the individual components before writing a line.

We can write each component of a line as follows:
1 Sub FileDemo3 ()
2 Dim strFileName As String
3 Dim ws As Worksheet
4 Dim i As Integer
5
6 Set ws = ThisWorkbook . Sheets ( " Sheet1 " )
7 strFileName = " MaleTable . csv "
8 Open strFileName For Output As #1
9 For i = 1 To 110
10 Print #1 , ws . Cells (i , 1) ;
11 Print #1 , " ," ;
12 Print #1 , ws . Cells (i , 2)
13 Next i
14 Close #1
15 End Sub

In the above code, the semicolon “;” after the Print command tells the
command not to write a new line break.

Exercise 5.11. The following VBA sub procedure

1 Sub GenerateMatrix ()
2 Dim i As Integer
3 Dim j As Integer
4
5 Rnd ( -1)
6 Randomize (0)
7 For i = 1 To 100
8 For j = 1 To 10
9 Cells (i , j ) . Value = Rnd ()
10 Next j
11 Next i
12 End Sub

writes a 100×10 matrix to the range “A1:J100.” Write a sub procedure named
FileDemo4 that writes the matrix to a CSV file named matrix.csv, such
that each line of the file contains a row of the matrix.

h
 File IO 107

Now, let us introduce how to read text files in VBA. The following exam-
ple shows how to read a CSV file:
1 Sub FileDemo5 ()
2 Dim strFile As String
3 Dim strV () As String
4 Dim j As Integer
5 Dim n As Integer
6 strFile = ThisWorkbook . Path & " : matrix . csv "
7 ' strFile is a full file name with path in a Mac
computer
8
9 n = 1
10 Open strFile For Input As #1
11 Do Until EOF (1)
12 Line Input #1 , Line
13 strV = Split ( Line , " ," )
14 For j = 0 To UBound ( strV )
15 Cells (n , j + 1) . Value = strV ( j )
16 Next j
17 n = n + 1
18 Loop
19 Close #1
20 End Sub

Let us examine the above code line by line. In Lines 2–5, we declare variables.
In Line 6, we create the full file name by combining the path of the workbook
and “matrix.csv,” which is the intended file name. By doing that, we save
the file “matrix.csv” into the same directory as this workbook. In Line 10
of the above code, we open the file for input. In Lines 11–18, we use a loop
to read the CSV file line by line. The function EOF returns a boolean value
indicating whether the end of the file is reached. Inside the loop, we split the
line by commas and write the individual component to cells of the active
worksheet.

Exercise 5.12. Suppose that the file words.txt contains the following two
lines:
1 Microsoft Excel is an extremely powerful tool
2 You can use to manipulate and present data

Write a sub procedure named FileDemo6 to do the following: read words

108 Functions, Events, and File IO

from the file words.txt, write the words in the first line to the first row of
your active worksheet, and write the words in the second line to the second
row of your active worksheet. Each cell contains one word.

Exercise 5.13. The file keywords.html is an HTML file that contains VBA
keywords, which are enclosed by the strings “(v=vs.90).aspx">” and “<.”

(a) Write a function procedure named ExtractKeywords with one String-

type argument, such that ExtractKeywords(line) returns a string that
contains keywords in line separated by spaces.
(b) Write a sub procedure named FileDemo7 to extract all keywords in the
file keywords.html and write each keyword to a cell in the first column
of your active worksheet.

5.4 Summary
In this chapter, we introduced how to create VBA functions, how to add a
handler to an event, and how to read and write files. A function allows us to
return some result to the caller of the function. An event handler enables
us to execute some VBA code automatically. The file Input-Output (IO)
functions allow us to import and export data from text files in a controlled
way. For more information about file IO, readers are referred to (Getz and
Gilbert, 2001, Chapter 12).
 6
Error Handling and Debugging

There are three categories of programming errors in general: compilation

errors, runtime errors, and logic errors. Compilation errors are errors that
prevent your program from running. When you run your Visual Basic for
Applications (VBA) code, the Visual Basic compiler compiles the VBA code
into binary language. If the compiler does not understand some of your VBA
code, it issues compilation errors. For example, misspelling a VBA keyword
is an example of compilation errors. Runtime errors are errors that occur
while your program is actually running. For example, division by zero and
taking square root of a negative number are runtime errors. Logic errors are
errors that prevent your VBA program from doing what you intended it to
do. For example, forgetting to terminate an infinite loop is a logic error.
Compilation errors are the easiest to handle, as they can be detected
by the Visual Basic Editor. Logic errors are the hardest to find and fix. In
this chapter, we introduce techniques for handling runtime errors and
debugging tools for finding logic errors. We will also introduce some best
practices of VBA coding that can help minimize the number of errors. After
studying this chapter, readers will be able to

• add error-handling code to VBA programs.

• use the debugging tools provided by VBA to debug VBA programs.
• know some best practices of VBA coding.

6.1 Error Handling

In this section, we introduce some techniques for handling runtime errors.
To illustrate runtime errors, we consider the following sub procedure:
1 Sub ErrorDemo1 ()
2 Dim dVal As Double
3

109
 110 Error Handling and Debugging

4 dVal = InputBox ( " Input a number " )

5 MsgBox " The square root of " & dVal & " is " & Sqr
( dVal )
6 End Sub

The above code asks the user for a number and then shows the square root
of the number in a message box. The above code is free of compilation
errors. However, it contains runtime errors because calculating the square
root of a negative number is illegal. For example, if we run the code and
input −1 in the input box, as shown in Figure 6.1(a), then we see the runtime
error shown in Figure 6.1(b).

(a)
(b)

Figure 6.1: An example illustrating a runtime error

The error message shown in Figure 6.1(b) is not helpful for many people.
A better way to handle the runtime error is to anticipate the error by adding
error-handling code. For example, we can add error-handling code to the
above sub procedure as follows:
1 Sub ErrorDemo2 ()
2 Dim dVal As Double
3
4 dVal = InputBox ( " Input a number " )
5 If dVal < 0 Then
6 MsgBox " You must input a non - negative number . "
7 Exit Sub
8 End If
9 MsgBox " The square root of " & dVal & " is " & Sqr
( dVal )
10 End Sub

In Lines 5–8, we use an If-Then statement to check whether the input

number is negative. If the number is negative, we display a message and
use the statement Exit Sub to terminate the sub procedure.
 Error Handling 111

The sub procedure ErrorDemo2 improves the sub procedure ErrorDemo1

by adding the error-handling code. However, it is still not perfect. For ex-
ample, if we input “a” in the input box (see Figure 6.2(a)), we will see the
runtime error shown in Figure 6.2(b).

(a)
(b)

Figure 6.2: An example illustrating another runtime error

To improve the code further, we add another piece of error-handling

code as follows:
1 Sub ErrorDemo3 ()
2 Dim dVal As Variant
3
4 dVal = InputBox ( " Input a number " )
5 If Not IsNumeric ( dVal ) Then
6 MsgBox " You must input a number "
7 Exit Sub
8 End If
9 If dVal < 0 Then
10 MsgBox " You must input a non - negative number . "
11 Exit Sub
12 End If
13 MsgBox " The square root of " & dVal & " is " & Sqr
( dVal )
14 End Sub

In the above code, we change the type of the variable dVal to Variant. Then
we use the function IsNumeric to check whether the input is a number. If
it is not a number, we display a message and exit the sub procedure.

Exercise 6.1. Consider the following function procedure:

1 Sub CalculateSpeed ( dMiles As Double , dTime As Double )
 112 Error Handling and Debugging

2 CalculateSpeed = dMiles / dTime

3 End Sub

Does the above function contain runtime errors? If so, improve the function
by adding error-handling code.

Exercise 6.2. Consider the following sub procedure:

1 Sub ErrorDemo4 ()
2 Dim dVal As Variant
3
4 dVal = InputBox ( " Input a number " )
5 If Not IsNumeric ( dVal ) Then
6 MsgBox " You must input a number "
7 Exit Sub
8 End If
9 If dVal < 0 Then
10 MsgBox " You must input a non - negative number . "
11 Exit Sub
12 End If
13 ActiveSheet . Cells (1 , 1) . Value = Sqr ( dVal )
14 End Sub

Does the above code contain runtime errors? If so, what are the possible
runtime errors?

Since it is hard for us to anticipate all possible runtime errors (see

Exercise 6.2), how can we handle every possible runtime error? Fortunately,
we can use the On Error statement to handle all errors. Table 6.1 gives
some On Error statements for error-handling.
For example, we can use the On Error statement as follows:
1 Sub ErrorDemo5 ()
2 Dim dVal As Double
3 On Error GoTo BadInput
4
5 dVal = InputBox ( " Input a number " )
6 MsgBox " The square root of " & dVal & " is " & Sqr
( dVal )
7
8 Exit Sub
9 BadInput :
10 MsgBox " Error number " & Err . Number & " occurred . "
 Error Handling 113

Table 6.1: Usage of the On Error Statement

Statement Description
On Error GoTo label After executing this statement, VBA
resumes execution at the line with the label
On Error GoTo 0 After executing this statement, VBA
resumes its normal error-checking behavior
On Error Resume Next After executing this statement, VBA ignores
all errors and resumes execution with
the next statement

11 End Sub

The above code is able to trap all errors. If an error occurs, the code will
display the error number in a message box. Here we use the Err.Number to
get the error number. Note that we use the Exit Sub statement right before
the BadInput label. This statement is necessary because we do not want to
execute the error-handling code if no errors occur.

Table 6.2: Usage of the Resume Statement

Statement Description
Resume After executing this statement, VBA resumes
with the statement that caused the error
Resume Next After executing this statement, VBA ignores all errors
and resumes execution with the next statement
Resume label After executing this statement, VBA resumes
execution at the line with the label

In the sub procedure ErrorDemo5, we add error-handling code to dis-

play an error message and exit the sub procedure. In some cases, we want
to recover from the error. In such cases, we use the Resume statement. Table
6.2 gives the usage of the Resume statement. The following example shows
how to use the Resume statement:
1 Sub ErrorDemo6 ()
2 Dim dVal As Double
3 Dim intAns As Integer
4 Dim strMsg As String
5
 114 Error Handling and Debugging

6 TryAgain :
7 On Error GoTo BadInput
8
9 dVal = InputBox ( " Input a number " )
10 MsgBox " The square root of " & dVal & " is " & Sqr
( dVal )
11
12 Exit Sub
13 BadInput :
14 strMsg = " Error number " & Err . Number & " occurred
."
15 strMsg = strMsg & vbNewLine & " Do you want to try
again ? "
16 intAns = MsgBox ( strMsg , vbYesNo + vbCritical )
17 If intAns = vbYes Then
18 Resume TryAgain
19 End If
20 End Sub

In Line 18 of the above code, we use the Resume TryAgain statement,

where TryAgain is a label defined in Line 6. The difference between Resume
and GoTo is that the Resume statement clears the error condition before
continuing, while the GoTo statement does not clear error conditions.

Exercise 6.3. Consider the following code:

1 Sub ErrorDemo7 ()
2 Dim i As Integer
3
4 For i = 1 To 100
5 Cells (i , 2) . Value = Sqr ( Cells (i , 1) . Value )
6 Next i
7 End Sub

The above sub procedure calculates the square root of numbers in the
first column of the active worksheet and places the results into the second
column. Improve the sub procedure by adding error-handling code so that
the code can run through without showing error dialogs.

h
Debugging VBA Code 115

6.2 Debugging VBA Code

In this section, we introduce some tools that can help us identify logic errors,
which are also referred to as bugs.
Common methods for debugging VBA code include examining the code,
inserting MsgBox statements, insert Debug.Print statements, and using
VBA’s built-in debugging tools. The first method requires knowledge and
experience. The second and third methods are similar. In previous sections,
we have used the Debug.Print to display information in the Immediate
window. In this section, we focus on the last method, which are much more
powerful than the first three methods. In particular, we introduce how to
set breakpoints in VBA code, use the Watch window, and use the Locals
window.
Let us first introduce how to set breakpoints in VBA code. Setting a
breakpoint allows us to halt a routine’s execution, take a look at the value
of any variables, and then continue execution. There several ways to set a
breakpoint:

• Move the cursor to the statement where we want execution to stop and
then press “F9.”
• Move the cursor to the statement where we want execution to stop and
then click the “Debug/Toggle Breakpoint.”
• Click the gray margin to the left of the statement where we want execu-
tion to stop.
• Right-click the statement where we want execution to stop and click
“Toggle/Breakpoint” in the pop-up menu.

When we execute a VBA procedure with a breakpoint, VBA goes into

Break mode before the line with the breakpoint is executed. In Break mode,
VBA stops running and the current statement is highlighted in bright yellow.
In Break mode, we can do several things:

• Check values of variables in the Immediate window.

• Step through the code one line at a time by pressing “F8.”
• See a variable’s value by moving the mouse pointer over the variable.
• Skip the next statement or even go back a couple of statements.
• Edit a statement and then continue.
 116 Error Handling and Debugging

To illustrate how to debug VBA code in Break mode, we consider the

following sub procedure:
1 Sub CopyTable ()
2 Dim strSourceFile As String
3 Dim wbSource As Workbook
4 Dim wsDest As Worksheet
5 Dim i As Integer
6
7 strSourceFile = Application . GetOpenFilename ()
8 If VarType ( strSourceFile ) <> vbString Then '
Cancel clicked
9 Exit Sub
10 End If
11
12 Set wbSource = Workbooks . Open ( strSourceFile )
13 wbSource . Sheets ( " Sheet1 " ) . Range ( " A25 : B134 " ) .
Select
14
15 Set wsDest = ThisWorkbook . Sheets ( " Sheet1 " )
16 For i = 1 To Selection . Rows . Count
17 wsDest . Cells (i , 1) . Value = Selection . Cells (i ,
1) . Value
18 wsDest . Cells (i , 2) . Value = Selection . Cells (i ,
2) . Value
19 Next i
20
21 wbSource . Close
22 End Sub

The above sub procedure first gets a file name of a workbook from a dialog
and then copies a range from the workbook to the active sheet of this
workbook.
Suppose that we set a breakpoint at Line 19 of the code. If we run
the code, VBA will enter Break model and the line with the breakpoint
is highlighted, as shown in Figure 6.3. In Break mode, we can type the
question mark "?" and a VBA statement to see the result. Figure 6.3 also
shows the results of four VBA statements. From the figure we see that the
value of the variable i is 1.
In some cases we want to see the values of variables near the end of a
loop. If the loop has 100 iterations and we want to see the values of variables
at the 90th iteration, then we need to respond to 90 prompts before the
 Debugging VBA Code 117

Figure 6.3: Checking values of VBA statements in the Immediate window in

Break mode

code finally gets to the 90th iteration. A more efficient method to do this is
to set a watch expression.
Suppose that we want to examine the code at the 20th iteration of the
Fot-Next loop. We can set a watch expression that puts the procedure into
Break mode. Figure 6.4 shows how to set such a watch expression. If we
execute the code, VBA will enter into Break mode when the value of the
variable i is equal to 20. In Break mode, we can check the values of variables
in the Immediate window, as shown in Figure 6.5.
Another useful debugging tool is the Locals window, which can be used
to look at the values of local variables in Break mode. To view the Locals
window, we click the “View/Locals window.” Figure 6.6 shows the Locals
window in Break mode, which was triggered by the watch expression i=20.
In the Locals window, we see the value and type of the variables.

Exercise 6.4. Consider the following sub procedure:

1 Sub GenerateRandNum ()
118 Error Handling and Debugging

Figure 6.4: Setting a watch expression in the Watch window

Figure 6.5: The Watches window and checking values of VBA statements in
the Immediate window in Break mode
 Debugging VBA Code 119

Figure 6.6: The Locals window in Break mode

2 Dim i As Integer
3 Dim dSum As Double
4 Dim dTemp As Double
5
6 Rnd ( -1)
7 Randomize (0)
8 For i = 1 To 1000
9 dTemp = Rnd ()
10 dSum = dSum + dTemp
11 Next i
12 End Sub

Do not modify the code. What is the value of the variable dTemp at the 900th
iteration?

Debugging is an integrated part of the software development. It is diffi-

cult, if not impossible, to completely eliminate bugs in your VBA programs.
Here are some tips that can help us keep bugs to a minimum (Walkenbach,
2013b):

• Use the Option Explicit statement at the beginning of modules.

• Format the VBA code with consistent indentation.
• Be careful with the On Error Resume Next statement.
• Use comments.
120 Error Handling and Debugging

• Write simple Sub and Function procedures.

• Use the macro recorder to help identify properties and methods.
• Understand VBA’s debugging tools.

6.3 Best Practices of VBA Coding

VBA best practices are useful for us to create high quality and reliable code.
In this section, we present some best practices for Excel® VBA development.
This section follows closely the third chapter of (Bovey et al., 2009) and
presents best practices in the following three broad categories: naming
convention, application structure, and application development.
Let us first present best practices for naming convention, which refers
to the system we use to name various parts of our application. Table 6.3
gives some naming conventions for common elements of VBA programs.

Table 6.3: Naming Conventions

Element Naming Convention

Variable <scope><array><data type>[DescriptiveName]
Constant <scope><data type>[DESCRIPTIVE_NAME]
User-defined type Type [DESCRIPTIVE_NAME]
<data type>[DescriptiveName]
End Type
Enumeration type Enum <project prefix>[GeneralDescr]
<project prefix>[GeneralDescrSpecificName1]
<project prefix>[GeneralDescrSpecificName2]
End Enum
Module M[GeneralDescr]
Class C[GeneralDescr]
UserForm F[GeneralDescr]

There are three scopes, and the scope specifiers are given in Table 6.4. If
a variable is an array, the array specifier <array> is a. Otherwise, the array
specifier is omitted (see Table 6.5). The specifiers of various data types are
given in Table 6.6.
The following gives some examples of names based on the naming
conventions:
Best Practices of VBA Coding 121

Table 6.4: Scope Specifiers

Scope Specifier
Public g
Module-level m
Procedure-level nothing

Table 6.5: Array Specifiers

Array Specifier
Array a
Not an array nothing

Table 6.6: Data Type Specifiers

Data Type Specifier Data Type Specifier

Boolean b Worksheet ws
Byte byt ComboBox cbo
Currency cur CheckBox chk
Date dte CommandButton cmd
Decimal dec Frame fra
Double d Label lbl
Integer i ListBox lst
Long l OptionButton opt
Object obj SpinButton spn
Single sng TextBox txt
String s Variant v
User-defined type u User-defined class variable c
Chart cht UserForm variable frm
Range rng VBA collection col
Workbook wb
122 Error Handling and Debugging

gsErrMsg A public String variable used to store an error message.

aiInd A procedure-level array of integers to hold an array of indices.

gdPI A public Double constant that holds the π value.

MDataAccess A module for data access.

We will see more examples in later chapters when we introduce case studies.

The Business
Logic Tier

The User The Data Access

Interface Tier Storage Tier

Figure 6.7: Three tiers of an application and their relationships

Next, let us introduce some best practices related to application struc-

ture and organization. Most Excel® applications have three logical tiers,
shown in Figure 6.7. The user interface tier consists of all the code and
visible elements your application uses to interact with users. The business
logic tier is completely code-based and performs the core operations of
your application. The data access and storage tier is responsible for storage
and retrieval of data required by the application.
As we can see from Figure 6.7, the user interface tier and the data access
and storage tier are not linked. The three tiers should by loosely coupled so
that a significant change in one tier does not lead to significant changes to
other two tiers.
VBA code should be organized into modules in a logical fashion. Pro-
cedures that perform similar functions should be grouped into the same
code module. Each procedure should be responsible for a well-defined task
that is easy to understand, validate, document, and maintain. Important
guidelines for creating procedures include:

Encapsulation A procedure should completely encapsulate the logical

operation it performs.
 Best Practices of VBA Coding 123

No duplicate code If a piece of code is used in multiple places to perform

the same operation, then you should extract the piece of code and place
it into a separate procedure.

Procedure size reduction If a procedure contains more than 100 lines of

code, it should be factored into multiple single-purpose procedures.

Limiting the number of arguments A procedure should accept no more

than five arguments.

Finally let us introduce some best practice for general application devel-
opment in the following areas: code commenting, code readability, use of
module directives, use of variables and constants, defensive coding, version
control, and change documentation.
Good code commenting is an important practice in programming. In
VBA programming, there are three categories of comments: module-level
comments, procedure-level comments, and internal comments. A good
module-level comment should be placed at the top of the module and look
something like the following example:
1 '
2 ' Description : A brief description of the purpose of
3 ' the code in this module .
4 '
5 ' Revision history :
6 ' Date Developer Action
7 '
------------------------------------------------------

8 ' 01/30/2016 Guojun Gan Created

9 '
10 Option Explicit

Procedure-level comments are usually the most detailed comments in your

application. The comment of a procedure is placed at the top of the pro-
cedure and contains the following content: the purpose of the procedure,
usage, a detailed list of arguments, a description of expected return values if
the procedure is a function procedure, and changes made to the procedure.
Internal comments are used within the body of the code itself to describe
the purpose of any code that is not self-evident. It is important to make sure
that comments are updated when the underlying code is changed.
Writing readable code is another important practice in programming.
Although code layout does not make any difference to computers, good
124 Error Handling and Debugging

visual layout of code allows us to infer lots of information about the logic
structure of the program. Related code should be grouped together, and
unrelated code should be separated by blank lines. Indentation should be
used to indicate a related block of code.
In addition to good code commenting and writing readable code, the
following is a list of best practices in other areas, such as use of directives
and use of variables and constants:

Option Explicit Alway use the Option Explicit in every module. This
statement forces you to explicitly declare all the variables you use.

Option Private Module The Option Private Module statement makes

all procedures within the module unavailable to other modules or the
Excel user interface. This statement should be used to hide procedures
that should not be called from outside of your application.

Option Base 1 The Option Base 1 statement causes the lower bound of
all array variables whose lower bound has not been specified to have a
lower bound of 1. This statement should not be used.

Option Compare Text The Option Compare Text statements causes all
string comparisons within the module to be text-based rather than
binary. This statement should be avoided.

Reuse variables Avoid reusing variables. Each variable should serve one
purpose only.

Variant data type Avoid using the Variant data type for the following rea-
sons: first, variants are not efficient; second, data stored in a variant can
behave unexpectedly.

Type conversion VBA automatically converts one data type to another data
type. Do not mix variables of different data types in your VBA expres-
sions without using the explicit casting functions (e.g., CStr and CDbl).

Array bounds Never hand-code the array bounds in a loop. Always use the
LBound function and the UBound function to get the bounds of an array.

ByRef and ByVal There are two conventions to declare arguments of a pro-
cedure: ByRef and ByVal. If an argument is declared with the ByRef
convention, then the memory address of the variable rather than the
value of the variable is passed. In such cases, modification of the vari-
able will be visible in the calling procedure. If an argument is declared
Summary 125

with the ByVal convention, the value of the variable is passed. Always
explicitly declare procedure arguments as ByRef or ByVal.

Argument validation Always validate arguments before using them in pro-

cedures so that erroneous input is caught as soon as possible.

Infinite loops Always add a guard counter to terminate an infinite loop

when the number of loops executed exceeds the highest number that
should ever occur in practice.

Use CodeNames to reference sheet objects Always reference worksheets

and chart sheets by their CodeNames. It is risky to identify sheets by their
tab names, because the tab names can be changed during execution of
the program.

Validate the data types of selection Always check the object type of the
selection using the TypeName function or the TypeOf-Is structure.

Saving versions Use a version control system that allows you to recover an
earlier version of your project if you have encountered problems with
your current version.

For detailed discussion of the above best practices, readers are referred to
(Bovey et al., 2009, Chapter 3).

6.4 Summary
In this chapter, we introduced how to handle runtime errors and how to
use VBA’s debugging tools to identify potential logic errors. We also intro-
duced some best practices of VBA coding. For more information about best
practices, readers are referred to (Bovey et al., 2009).
 Part II

Applications
7
Generating Payment Schedules

In this chapter, we implement a simple Visual Basic for Applications (VBA)

program that can generate regular payment dates of an interest rate swap
according to a business calendar. Figure 7.1 shows the interface of the
payment schedule generator.

Figure 7.1: Interface of the payment schedule generator

From 7.1, we see that the interface contains six input cells highlighted
in yellow and one button. The output is shown in the first two columns. The
first column shows the payment dates, and the second column shows the
year fractions between these dates. The goal of this chapter is to implement
the payment generator.

129
130 Generating Payment Schedules

7.1 Introduction
A derivative (Garrett, 2013) is a concept in finance and refers to a contract
that derives its value from the performance of an underlying entity, such
as an asset, index, or interest rate. Derivatives have been used for risk man-
agement purposes. For example, insurance companies have used interest
rate swaps to hedge the interest rate risks associated with variable annuity
contracts.
An interest rate swap (IRS) is a financial derivative in which two parties
agree to exchange interest rate cash flows, based on a specified notional
amount. Under an IRS, each counterpart agrees to pay either a fixed or
floating rate denominated in a particular currency to the other counterpart.
When pricing an IRS, one of the first steps is to generate a schedule of
dates, based on the holiday calendar and the date rolling rules specified
in the contract. The schedule of dates is used to project cash flows of the
IRS. Projecting cash flows is also dependent on the day count convention
specified in the swap contract.
In this chapter, we introduce how to generate a schedule of dates based
on the US holiday calendar. We will also introduce some day count conven-
tions and date rolling rules.

7.2 Public Holidays in the United States

There are ten public holidays and observances in the United States. The
New York Stock Exchange (NYSE) markets observe these holidays. Theses
holidays are listed in Table 7.1.
The holidays can be divided into two groups in terms of how they are
determined. The first group includes New Year’s Day, Independence Day,
Veterans Day, and Christmas Day. The holidays in the first group occur
on a specified date. If a specified date falls on a weekend, then either the
previous Friday or the next Monday is observed. The second group includes
the remaining holidays. The holidays in the second group are determined
based on a particular weekday.
Julian and Gregorian Calendars 131

Table 7.1: Holidays and Observances in the United States

Name Date
New Year’s Day January 1
(January 2 if January 1 is a Sunday,
December 31 if January 1 is a Saturday)
Martin Luther King The third Monday of January
Jr. Day
Washington’s Birthday The third Monday of February
Memorial Day The last Monday of May
Independence Day July 4
(July 5 if July 4 is a Sunday,
July 3 if July 4 is a Saturday )
Labor Day The first Monday of September
Columbus Day The second Monday of October
Veterans Day November 11
(November 12 if November 11 is a Sunday,
November 10 if November 11 is a Saturday)
Thanksgiving Day The fourth Thursday of November
Christmas Day December 25
(December 26 if December 25 is a Sunday,
December 24 if December 25 is a Saturday)

7.3 Julian and Gregorian Calendars

A calender is a system for assigning a unique date to every day so that the
order of the dates is obvious and the number of days between two dates
can be calculated easily (Richards, 2012). There are about forty calendars
in use in the world today, and most calendars are based on astronomical
cycles. In this section, we introduce the Julian and Gregorian calendars and
algorithms for converting a date between the two calendars.
The Julian calendar was introduced by Julius Caesar in 46 BC. The Julian
calendar was the predominant calendar in the Roman world and most of
Europe until it was replaced by the Gregorian calendar. The average length
of the year in the Julian calendar is 365.25 days.
The Gregorian calendar is a solar calendar, which maintains accurate
synchrony with the astronomical year but does not keep the months in syn-
chrony with the lunation (the cycle of the phases of the moon) of the moon.
The Gregorian calendar is the official calendar of the United Kingdom since
132 Generating Payment Schedules

1752 and is used throughout the world for secular purpose. The average
length of the year in the Gregorian calendar is 365.2425 days.
A date consists of three components: a year, a month, and a day of the
month. The three components correspond to the three principle astronomi-
cal cycles: the revolution of the Earth around the sun (a year), the revolution
of the moon around the Earth (a month), and the rotation of the Earth on its
axis (a day). The complexity of calendars comes from the fact that a month
and a year do not comprise an integral number of days, and they are neither
constant nor perfectly commensurable with each other.

Table 7.2: Parameters for Calendar Conversion Algorithms

Parameter Value Parameter Value Parameter Value

y 4716 p 1461 t 2
j 1401 q 0 w 2
m 2 v 3 A 184
n 12 u 5 B 274277
r 4 s 153 C −38

There are algorithms for interconverting dates in the Gregorian calen-

dar and the Julian day number (Richards, 2012). Let D/M/Y be a date in
the Gregorian calendar. Then, the corresponding Julian day number J is
calculated as follows:

(a) h = M − m
(b) g = Y + y − (n − h)\n
(c) f = mod (h − 1 + n, n)
(d) e = (pg + q)\r + D − 1 − j
(e) J = e + (s f + t )\u − (3 ∗ ((g + A)\100))\4 −C ,

where, mod (A, B ) represents the remainder when A is divided by B , and \

denotes integer division in which any remainder is ignored.
To convert a Julian day number J to a date D/M/Y in the Gregorian
calendar, we use the following algorithm:

(a) f = J + j + (((4J + B )\146097) × 3)\4 +C

(b) e = r f + v
(c) g = mod (e, p)\r
(d) h = ug + w
Day Count Conventions 133

(e) D = mod (h, s)\u + 1

(f) M = mod (h\s + m, n) + 1
(g) Y = e\p − y + (n + m − M)\n

The parameters used in the above two algorithms are given in Table 7.2.

7.4 Day Count Conventions

A day count convention determines how interest accrues over time for
interest-earning investments. A day count convention is also used to quan-
tify periods of time when discounting a cash flow to its present value.
In this section, we introduce the definition of the following day count
conventions: ACT/ACT, ACT/365, ACT/360, and 30/360. Since there is not
a unique definition of day count conventions, we follow the International
Swaps and Derivatives Association (ISDA), which has documented some
day count conventions (ISDA, 2006).
The ACT/ACT day count convention is defined as follows:
Days not in leap year Days in leap year
∆= + , (7.1)
365 366
where the days in the numerators are calculated on a Julian day difference
basis, and a leap year is a year that contains 366 days by extending February
to 29 days. In the Gregorian calendar, a year Y is a leap year if Y satisfied one
of the following two conditions:

(a) Y is divisible by 4 and is not divisible by 100

(b) Y is divisible by 400

Exercise 7.1. Which of the following years are leap years?

1900, 1996, 2000, 2016, 2100.

The ACT/365 day count convention is defined as follows:

Days in the calculation period
∆= , (7.2)
365
134 Generating Payment Schedules

where the days in the numerator are calculated on a Julian day difference
basis. Similarly, the ACT/360 day count convention is defined as follows:

Days in the calculation period

∆= , (7.3)
360
where the days in the numerator are calculated on a Julian day difference
basis.
The 30/360 day count convention is defined as follows:

360(Y2 − Y1 ) + 30(M 2 − M 1 ) + (D 2 − D 1 )
∆= , (7.4)
360
where

• Y1 is the year in which the first day of the calculation period falls;
• Y2 is the year in which the day immediately following the last day in-
cluded in the calculation period falls;
• M 1 is the month in which the first day of the calculation period falls;
• M 2 is the month in which the day immediately following the last day
included in the calculation period falls;
• D 1 is the first day of the calculation period. If the first day is 31, then
D 1 = 30;
• D 2 is the day immediately following the last day included in the calcula-
tion period. If D 1 = 30 and the day immediately following the last day is
31, then D 2 = 30.

7.5 Business Day Conventions

When a payment date falls on a holiday or weekend according to a specified
business calendar, we need to adjust the payment date forward or backward
in time, such that it falls on a business day in the same business calendar. A
business day convention refers to a convention for adjusting any payment
date that falls on a nonbusiness day (e.g., holidays or weekends). There are
three business day conventions (ISDA, 2006): following, modified following,
and preceding.
If “following” is specified, the payment date is adjusted to the next busi-
ness day. If “modified following” is specified, the payment date is adjusted
 Implementation 135

to the next business day unless next business day is in the next calendar
month, in which case the payment date is adjusted to the previous business
day. If “preceding” is specified, the payment date is adjusted to the previous
business day.

Exercise 7.2. The 2016 holidays observed by the NYSE markets are: January
1, January 18, February 15, March 25, May 30, July 4, September 5, November
24, and December 26. Adjust the following payment dates according to
various business day conventions:

Payment Date Following Modified Following Preceding

January 16, 2016
May 30, 2016
November 25, 2016

7.6 Implementation
In this section, we implement the payment schedule generator with the
interface shown in Figure 7.1. We would like to put the VBA code into three
modules: MInterface, MDate, and MHoliday. The module MInterface
contains VBA code for the button. The module MDate contains VBA code
for manipulating dates. The module MHoliday contains code for handling
business calendars and generating payment date schedules.

7.6.1 MDate Module

Let us first implement the module for manipulating dates. We name this
module as MDate. At the declarations section of this module, we define
some constants given in Table 7.2 as follows:
1 Option Explicit
2
3 Const cony As Long = 4716
4 Const conj As Long = 1401
5 Const conm As Long = 2
 136 Generating Payment Schedules

6 Const conn As Long = 12

7 Const conr As Long = 4
8 Const conp As Long = 1461
9 Const conq As Long = 0
10 Const conv As Long = 3
11 Const conu As Long = 5
12 Const cons As Long = 153
13 Const cont As Long = 2
14 Const conw As Long = 2
15 Const conA As Long = 184
16 Const conB As Long = 274277
17 Const conC As Long = -38

In the above code, we named the constants with the prefix “con” plus the
parameter names given in Table 7.2 so that we can easily connect the VBA
constants to the parameters. These constants are used to interconvert Gre-
gorian dates and Julian day numbers.
Once we have defined the constants in the module, we can implement
the function for converting a Gregorian date to a Julian day number. Fol-
lowing the algorithm given in Section 7.3, we implement the function as
follows:
1 Function ToJulian ( ByVal datIn As Date ) As Long
2 Dim lD As Long
3 Dim lM As Long
4 Dim lY As Long
5
6 Dim lh As Long
7 Dim lg As Long
8 Dim lf As Long
9 Dim le As Long
10
11 lD = Day ( datIn )
12 lM = Month ( datIn )
13 lY = Year ( datIn )
14
15 lh = lM - conm
16 lg = lY + cony - ( conn - lh ) \ conn
17 lf = ( lh - 1 + conn ) Mod conn
18 le = ( conp * lg + conq ) \ conr + lD - 1 - conj
19 ToJulian = le + ( cons * lf + cont ) \ conu - (3 *
(( lg + conA ) \ 100) ) \ 4 - conC
20 End Function
 Implementation 137

In the above code, we used the integer division operator \.

Similarly, we can follow the algorithm to implement the function for
converting a Julian day number to a Gregorian date as follows:
1 Function FromJulian ( ByVal lJ As Long ) As Date
2 Dim lD As Long
3 Dim lM As Long
4 Dim lY As Long
5
6 Dim lh As Long
7 Dim lg As Long
8 Dim lf As Long
9 Dim le As Long
10
11 lf = lJ + conj + (((4 * lJ + conB ) \ 146097) * 3)
\ 4 + conC
12 le = conr * lf + conv
13 lg = ( le Mod conp ) \ conr
14 lh = conu * lg + conw
15 lD = ( lh Mod cons ) \ conu + 1
16 lM = (( lh \ cons + conm ) Mod conn ) + 1
17 lY = le \ conp - cony + ( conn + conm - lM ) \ conn
18 FromJulian = DateSerial ( lY , lM , lD )
19 End Function

Before we continue to implement other functions, let us test the func-

tions ToJulian and FromJulian. We can write a sub procedure to test these
two functions as follows:
1 Sub Te s t Ju l ia n G re g or i an ()
2 Dim datA As Date
3 Dim lJ As Long
4
5 datA = #1/1/1998#
6 lJ = ToJulian ( datA )
7
8 Debug . Print lJ
9 Debug . Print FromJulian ( lJ )
10
11 datA = #5/1/2016#
12 lJ = ToJulian ( datA )
13
14 Debug . Print lJ
15 Debug . Print FromJulian ( lJ )
16 End Sub
 138 Generating Payment Schedules

In the above sub procedure, we first converted a date to a Julian day num-
ber and then converted the Julian day number back to a date. Executing
the above sub procedure produces the following output in the Immediate
window:
1 2450815
2 1/1/1998
3 2457510
4 5/1/2016

From the above output, we see that we can recover the dates exactly from
their Julian day numbers.
Now, let us implement functions for adding intervals (e.g., days, months,
years) to a date. Let us first implement the function AddDays, which returns
a date by adding a given number of days to an input date. We can implement
the function as follows:
1 Function AddDays ( ByVal datIn As Date , ByVal lCount As
Long ) As Date
2 Dim lJ As Long
3
4 lJ = ToJulian ( datIn ) + lCount
5 AddDays = FromJulian ( lJ )
6 End Function

To add days to a date, we first convert the date to a Julian day number and
add the given number of days to the Julian day number. Then we convert
the resulting Julian day number to a Gregorian date.
To test the function AddDays, we use the following sub procedure:
1 Sub TestAddDays ()
2 Debug . Print AddDays (#2/1/2016# , 28)
3 Debug . Print AddDays (#2/1/2016# , 365)
4 Debug . Print AddDays (#2/1/2016# , -365)
5 End Sub

Executing the above code gives the following output:

1 2/29/2016
2 1/31/2017
3 2/1/2015

The above output shows that the function AddDays works as expected.
 Implementation 139

Adding months or years to a date is more complicated than adding days,

in that the number of days in a month is not the same. For example, adding
12 months or 1 year to 2/29/2016 is not 2/29/2017 because the year 2017 is
not a leap year. We need a function to get the maximum number of days in
a month. We can implement the function as follows:
1 Function DaysInMonth ( ByVal datIn As Date ) As Integer
2 Dim lY As Long
3 Dim lM As Long
4
5 lY = Year ( datIn )
6 lM = Month ( datIn )
7 DaysInMonth = DateSerial ( lY , lM + 1 , 1) -
DateSerial ( lY , lM , 1)
8 End Function

The function DaysInMonth returns the number of days in the month of

the input date. The number of days is calculated by subtracting two dates
created by the function DateSerial, which is a VBA built-in function. We
can test the function as follows:
1 Sub TestDaysInMonth ()
2 Dim i As Integer
3 For i = 1 To 12
4 Debug . Print DaysInMonth ( DateSerial (2016 , i ,
1) ) ;
5 Next i
6 Debug . Print " "
7 For i = 1 To 12
8 Debug . Print DaysInMonth ( DateSerial (2017 , i ,
1) ) ;
9 Next i
10 End Sub

In the above test function, we used the Fot-Next loop to display the number
of days in each month of years 2016 and 2017. Executing the above code
gives:
1 31 29 31 30 31 30 31 31 30 31 30 31
2 31 28 31 30 31 30 31 31 30 31 30 31

From the output, we see that the function calculates the number of days
correctly.
Then we can implement the function for adding months as follows:
 140 Generating Payment Schedules

1 Function AddMonths ( ByVal datIn As Date , ByVal lCount

As Long ) As Date
2 Dim lD As Long
3 Dim lM As Long
4 Dim lY As Long
5 Dim lDMax As Integer
6 Dim datTmp As Date
7
8 lD = Day ( datIn )
9 lM = Month ( datIn )
10 lY = Year ( datIn )
11
12 lM = lM + lCount
13 Do While lM > 12 ' addition of months
14 lY = lY + 1
15 lM = lM - 12
16 Loop
17
18 Do While lM < 1 ' subtraction of months
19 lY = lY - 1
20 lM = lM + 12
21 Loop
22
23 ' need to truncate days in month
24 datTmp = DateSerial ( lY , lM , 1)
25 lDMax = DaysInMonth ( datTmp )
26 If lD > lDMax Then
27 lD = lDMax
28 End If
29 AddMonths = DateSerial ( lY , lM , lD )
30 End Function

The function AddMonths returns a date by adding a given number of months

to an input date. We truncate some days, if necessary. To test this function,
we use the following sub procedure:
1 Sub TestAddMonths ()
2 Debug . Print AddMonths ( DateSerial (2015 , 11 , 30) ,
3)
3 Debug . Print AddMonths ( DateSerial (2015 , 2 , 28) ,
12)
4 Debug . Print AddMonths ( DateSerial (2015 , 3 , 30) ,
-5)
5 End Sub
 Implementation 141

Executing the above code gives

1 2/29/2016
2 2/28/2016
3 10/30/2014

The above output shows that the function AddMonths works correctly.
Similarly, we can implement the function for adding years as follows:
1 Function AddYears ( ByVal datIn As Date , ByVal lCount
As Long ) As Date
2 Dim lD As Long
3 Dim lM As Long
4 Dim lY As Long
5 Dim lDMax As Integer
6 Dim datTmp As Date
7
8 lD = Day ( datIn )
9 lM = Month ( datIn )
10 lY = Year ( datIn ) + lCount
11
12 ' handle Feb 29
13 datTmp = DateSerial ( lY , lM , 1)
14 lDMax = DaysInMonth ( datTmp )
15 If lD > lDMax Then
16 lD = lDMax
17 End If
18 AddYears = DateSerial ( lY , lM , lD )
19 End Function

The function AddYears returns a date by adding a given number of years to

an input date. It is similar to the function AddMonths. But this function is
simpler than the function AddMonths, as we only need to take care of leap
years. We can test this function as follows:
1 Sub TestAddYears ()
2 Debug . Print AddYears ( DateSerial (1998 , 1 , 1) , 18)
3 Debug . Print AddYears ( DateSerial (2000 , 2 , 29) , 15)
4 Debug . Print AddYears ( DateSerial (2016 , 2 , 29) , 1)
5 End Sub

Executing the above sub procedure gives

1 1/1/2016
2 2/28/2015
3 2/28/2017
 142 Generating Payment Schedules

From the output, we see that the function AddYears works well.
We can combine the functions for adding intervals into a single function
as follows:
1 Function OffsetDate ( ByVal datIn As Date , ByVal lCount
As Long , ByVal sPeriod As String ) As Date
2 Select Case sPeriod
3 Case " M "
4 OffsetDate = AddMonths ( datIn , lCount )
5 Case " Y "
6 OffsetDate = AddYears ( datIn , lCount )
7 Case Else
8 Err . Raise Number :=1003 , Source := " MHoliday
. GenerateDate " , Description := " Invalid
period string "
9 End Select
10 End Function

The function OffsetDate returns a date by adding a given number of pe-

riods to an input date. In the current version, we can only add months or
years to a date. If the input period is not "M" or "Y," the function will raise
an error.

Exercise 7.3. The function OffsetDate can handle only two frequencies:
"M" and "Y." Modify the function OffsetDate so that it can handle the
following additional frequencies: "D," "W," and "Q," where D, W, and Q refer
to day, week, and quarter.

Now, let us implement functions for calculating the year fractions based
on the four day count conventions. We first implement a function that tells
whether a year is a leap year as follows:
1 Function IsLeapYear ( ByVal lY As Long ) As Boolean
2 Dim bRes As Boolean
3
4 bRes = False
5 If lY Mod 4 = 0 And lY Mod 100 <> 0 Then
6 bRes = True
 Implementation 143

7 End If
8 If lY Mod 400 = 0 Then
9 bRes = True
10 End If
11
12 IsLeapYear = bRes
13 End Function

The function IsLeapYear returns true if an input year is a leap year in

the Gregorian calendar. We can test this function using the following sub
procedure:
1 Sub TestIsLeapYear ()
2 Dim i As Integer
3
4 For i = 2000 To 2009
5 Debug . Print IsLeapYear ( i ) & " " ;
6 Next i
7 End Sub

Executing the above sub procedure gives

1 True False False False True False False False True
False

The output shows that the function IsLeapYear works as expected.

Then, we implement a function to calculate the number of days in leap
years between two given dates. We can implement this function as follows:
1 Function DaysInLeapYears ( ByVal datA As Date , ByVal
datB As Date ) As Long
2 ' If datA is before datB , we switch the order
3 If ToJulian ( datA ) > ToJulian ( datB ) Then
4 DaysInLeapYears = DaysInLeapYears ( datB , datA )
5 Exit Function
6 End If
7
8 Dim l As Long
9 Dim lYA As Long
10 Dim lYB As Long
11 Dim lDays As Long
12 ' Then we can assume datA is before datB
13 lYA = Year ( datA )
14 lYB = Year ( datB )
15
 144 Generating Payment Schedules

16 For l = lYA To lYB

17 If IsLeapYear ( l ) Then
18 lDays = lDays + _
19 ToJulian ( DateSerial ( l + 1 , 1 , 1) ) - _
20 ToJulian ( DateSerial (l , 1 , 1) )
21 End If
22 Next l
23 If IsLeapYear ( lYA ) Then
24 lDays = lDays + ToJulian ( DateSerial ( lYA - 1 ,
12 , 31) ) - ToJulian ( datA )
25 End If
26
27 If IsLeapYear ( lYB ) Then
28 lDays = lDays + ToJulian ( datB ) - ToJulian (
DateSerial ( lYB , 12 , 31) )
29 End If
30
31 DaysInLeapYears = lDays
32 End Function

The idea behind the above implementation is that we add all the days in
leap years between the years of the two given dates. Then, we subtract the
days outside the two given dates. To this this function, we use the following
sub procedure:
1 Sub Te s t Da y sI n L ea p Ye a rs ()
2 Debug . Print DaysInLeapYears (#1/1/2016# ,
#1/31/2016#)
3 Debug . Print DaysInLeapYears (#1/1/2016# ,
#1/31/2018#)
4 Debug . Print DaysInLeapYears (#1/1/2015# ,
#1/31/2018#)
5 Debug . Print DaysInLeapYears (#1/1/2015# ,
#12/31/2015#)
6 End Sub

Executing the above sub procedure gives

1 30 365 366 0

The output shows that the function calculates the number of days in leap
years correctly.
Using the function DaysInLeapYears, we can calculate the year frac-
tion based on the ACT/ACT convention as follows:
 Implementation 145

1 Function ActAct ( ByVal datA As Date , ByVal datB As

Date ) As Double
2 Dim dDaysNonLeap As Double
3 Dim dDaysLeap As Double
4 Dim lYA As Long
5 Dim lYB As Long
6 Dim l As Long
7
8 dDaysLeap = DaysInLeapYears ( datA , datB )
9 dDaysNonLeap = Abs ( ToJulian ( datA ) - ToJulian ( datB
) ) - dDaysLeap
10
11 ActAct = dDaysLeap / 366 + dDaysNonLeap / 365
12 End Function

In the above function, the number of days in nonleap years is calculated as

the total number of days minus the number of days in leap years.
To calculate the year fraction between two dates based on the 30/360
day count convention, we implement the following function:
1 Function Thirty360 ( ByVal datA As Date , ByVal datB As
Date ) As Double
2 Dim lY1 As Long
3 Dim lM1 As Long
4 Dim lD1 As Long
5 Dim lY2 As Long
6 Dim lM2 As Long
7 Dim lD2 As Long
8
9 If ToJulian ( datA ) > ToJulian ( datB ) Then
10 Thirty360 = Thirty360 ( datB , datA )
11 Exit Function
12 End If
13
14 lY1 = Year ( datA )
15 lM1 = Month ( datA )
16 lD1 = Day ( datA )
17 lY2 = Year ( datB )
18 lM2 = Month ( datB )
19 lD2 = Day ( datB )
20 If lD1 = 31 Then
21 lD1 = 30
22 End If
23 If lD1 = 30 And lD2 = 31 Then
 146 Generating Payment Schedules

24 lD2 = 30
25 End If
26 Thirty360 = lY2 - lY1 + ( lM2 - lM1 ) / 12# + ( lD2
- lD1 ) / 360#
27 End Function

The sign # after an integer tells VBA to treat the integer as a double.
We can combine the above two functions in the following function:
1 Function TFrac ( ByVal datA As Date , ByVal datB As Date
, ByVal sDCC As String ) As Double
2 Dim lJA As Long
3 Dim lJB As Long
4 lJA = ToJulian ( datA )
5 lJB = ToJulian ( datB )
6
7 If lJA > lJB Then
8 TFrac = TFrac ( datB , datA , sDCC )
9 Exit Function
10 End If
11
12 Select Case sDCC
13 Case " ACTACT "
14 TFrac = ActAct ( datA , datB )
15 Case " ACT365 "
16 TFrac = ( lJB - lJA ) / 365#
17 Case " ACT360 "
18 TFrac = ( lJB - lJA ) / 360#
19 Case " Thirty360 "
20 TFrac = Thirty360 ( datA , datB )
21 Case Else
22 Err . Raise Number :=1007 , Source := " MDate .
TFrac " , Description := " Unknown day
count convention . "
23 End Select
24 End Function

The function TFrac calculates the year fraction between two dates accord-
ing to a specified day count convention. The function will raise an error if an
unknown day count convention is encountered. We can test this function
as follows:
1 Sub TestTFrac ()
2 Dim datA As Date
3 Dim datB As Date
 Implementation 147

4
5 datA = #12/31/2015#
6 datB = #12/31/2016#
7
8 Debug . Print TFrac ( datA , datB , " ACTACT " )
9 Debug . Print TFrac ( datA , datB , " ACT365 " )
10 Debug . Print TFrac ( datA , datB , " ACT360 " )
11 Debug . Print TFrac ( datA , datB , " Thirty360 " )
12 End Sub

Executing the above test function gives the following output:

1 1
2 1.0027397260274
3 1.01666666666667
4 1

The output shows that the function works as expected.

7.6.2 MHoliday Module

In this subsection, we implement a module with functions for generat-
ing payment schedules based on the US business calendar. We name this
module MHoliday.
The first function we would like to implement is one that checks whether
a date is a weekend. We implement this function as follows:
1 Function IsWeekend ( ByVal datIn As Date ) As Boolean
2 Dim iWd As Integer
3 iWd = Weekday ( datIn )
4 If iWd = vbSaturday Or iWd = vbSunday Then
5 IsWeekend = True
6 Else
7 IsWeekend = False
8 End If
9 End Function

The function IsWeekend returns true if the input date is a weekend. We can
test this function as follows:
1 Sub TestIsWeekend ()
2 Dim i As Integer
3 For i = 1 To 7
4 Debug . Print IsWeekend ( DateSerial (2016 , 5 , i ) )
& " ";
 148 Generating Payment Schedules

5 Next i
6 End Sub

Executing the above sub procedure gives

1 True False False False False False True

The output shows that the function works as expected.

The second function checks whether a date is a US holiday. We imple-
ment this function as follows:
1 Function IsHolidayNY ( ByVal datIn As Date ) As Boolean
2 If IsWeekend ( datIn ) Then
3 IsHolidayNY = False
4 Exit Function
5 End If
6
7 Dim iWd As Integer
8 Dim iM As Integer
9 Dim iDoM As Integer
10 Dim bRet As Boolean
11
12 bRet = False
13 iWd = Weekday ( datIn )
14 iM = Month ( datIn )
15 iDoM = Day ( datIn )
16
17 ' January 1
18 If ( iM = 1 And iDoM = 1) Or _
19 ( iM = 12 And iDoM = 31 And iWd = vbFriday ) Or
_
20 ( iM = 1 And iDoM = 2 And iWd = vbMonday ) Then
21 bRet = True
22 End If
23 ' Martin Luther King , Third Monday of January
24 If iM = 1 And iWd = vbMonday And iDoM > 14 And
iDoM < 22 Then
25 bRet = True
26 End If
27 ' Washington ' s Birthday , The third Monday of
February
28 If iM = 2 And iWd = vbMonday And iDoM > 14 And
iDoM < 22 Then
29 bRet = True
30 End If
 Implementation 149

31 ' Memorial Day , The last Monday of May

32 If iM = 5 And iWd = vbMonday And iDoM > 24 Then
33 bRet = True
34 End If
35 ' Independence Day , July 4
36 If ( iM = 7 And iDoM = 4) Or _
37 ( iM = 7 And iDoM = 3 And iWd = vbFriday ) Or _
38 ( iM = 7 And iDoM = 5 And iWd = vbMonday ) Then
39 bRet = True
40 End If
41 ' Labor Day , The first Monday of September
42 If iM = 9 And iWd = vbMonday And iDoM < 8 Then
43 bRet = True
44 End If
45 ' Columbus Day , The second Monday of October
46 If iM = 10 And iWd = vbMonday And iDoM > 7 And
iDoM < 15 Then
47 bRet = True
48 End If
49 ' Veterans Day , November 11
50 If ( iM = 11 And iDoM = 11) Or _
51 ( iM = 11 And iDoM = 10 And iWd = vbFriday ) Or
_
52 ( iM = 11 And iDoM = 12 And iWd = vbMonday )
Then
53 bRet = True
54 End If
55 ' Thanksgiving Day , The fourth Thursday of
November
56 If iM = 11 And iWd = vbThursday And iDoM > 21 And
iDoM < 29 Then
57 bRet = True
58 End If
59 ' Christmas Day , December 25
60 If ( iM = 12 And iDoM = 25) Or _
61 ( iM = 12 And iDoM = 24 And iWd = vbFriday ) Or
_
62 ( iM = 12 And iDoM = 26 And iWd = vbMonday )
Then
63 bRet = True
64 End If
65
66 IsHolidayNY = bRet
67 End Function
 150 Generating Payment Schedules

The function IsHolidayNY returns true if the input date is a public holiday
in the United States. Since a weekend cannot be a holiday, the function
returns false if the input date falls on a weekend. This function is imple-
mented according to the public holidays given in Table 7.1. We can test this
function as follows:
1 Sub TestIsHolidayNY ()
2 Dim datA As Date
3 Dim i As Integer
4
5 For i = 0 To 365
6 datA = AddDays ( DateSerial (2016 , 1 , 1) , i )
7 If IsHolidayNY ( datA ) Then
8 Debug . Print datA
9 End If
10 Next i
11
12 End Sub

The above sub procedure displays all the holidays in 2016. Executing the
sub procedure gives the following output:
1 1/1/2016
2 1/18/2016
3 2/15/2016
4 5/30/2016
5 7/4/2016
6 9/5/2016
7 10/10/2016
8 11/11/2016
9 11/24/2016
10 12/26/2016

If we check the US holidays in the Internet, we see that the above list of
holidays is correct.
The third function in this module checks whether a date is a business
day based on a calendar. We implement this function as follows:
1 Function IsBusinessDay ( ByVal datIn As Date , ByVal
sCalendar As String ) As Boolean
2 Dim bRet As Boolean
3 bRet = False
4
5 Select Case sCalendar
6 Case " NY "
 Implementation 151

7 bRet = ( Not IsWeekend ( datIn ) ) And ( Not

IsHolidayNY ( datIn ) )
8 Case Else
9 Err . Raise Number :=1001 , Source := " MHolidy .
IsBusinessDay " , Description := " Unknown
calendar . "
10 End Select
11
12 IsBusinessDay = bRet
13 End Function

The function IsBusinessDay returns true if the input date is a business

day according to a given business calendar. In the current version, only one
business calendar (i.e., NY) is implemented. If the input business calendar
is not "NY," the procedure will raise an error by calling the VBA function
Err.Raise. To test this function, we use the following sub procedure:
1 Sub Tes tIsBusin essDay ()
2 Debug . Print IsBusinessDay (#2/15/2016# , " NY " )
3 Debug . Print IsBusinessDay (#12/26/2016# , " NY " )
4 Debug . Print IsBusinessDay (#12/27/2016# , " NY " )
5 End Sub

Executing the above test procedure gives

1 False
2 False
3 True

The output shows that the function works as expected.

7.6.3 MSchedule Module

The module MSchedule contains functions for generating payment sched-
ules according to a specified business calendar, a specified payment fre-
quency, and a specified payment horizon.
Since a payment date should be a business day, we need a function for
adjusting a payment date based on a given business day convention. We
can implement this function as follows:
1 Function RollDate ( ByVal datIn As Date , ByVal sBDC As
String , ByVal sCalendar As String ) As Date
2 Dim datTmp As Date
3
 152 Generating Payment Schedules

4 datTmp = datIn
5 Select Case sBDC
6 Case " Follow "
7 Do While Not IsBusinessDay ( datTmp ,
sCalendar )
8 datTmp = AddDays ( datTmp , 1)
9 Loop
10 Case " Modified "
11 Do While Not IsBusinessDay ( datTmp ,
sCalendar )
12 datTmp = AddDays ( datTmp , 1)
13 Loop
14 If Month ( datTmp ) <> Month ( datIn ) Then
15 datTmp = datIn
16 Do While Not IsBusinessDay ( datTmp ,
sCalendar )
17 datTmp = AddDays ( datTmp , -1)
18 Loop
19 End If
20 Case " Preceding "
21 Do While Not IsBusinessDay ( datTmp ,
sCalendar )
22 datTmp = AddDays ( datTmp , -1)
23 Loop
24 Case Else
25 Err . Raise Number :=1006 , Source := " MHoliday
. RollDate " , Description := " Unknown
business day convention . "
26 End Select
27
28 RollDate = datTmp
29 End Function

The function RollDate adjusts a payment date that falls on a nonbusiness

day (i.e., a weekend or holiday) according to a business calendar and a busi-
ness day convention. In the current version, three business day conventions
are implemented. The function will raise an error if an unknown business
day convention is encountered. To test this function, we use the following
sub procedure:
1 Sub TestRollDate ()
2 Debug . Print RollDate (#2/15/2016# , " Follow " , " NY " )
3 Debug . Print RollDate (#2/15/2016# , " Preceding " , "
NY " )
 Implementation 153

4 Debug . Print RollDate (#4/30/2016# , " Modified " , " NY

")
5 End Sub

Executing the above sub procedure gives the following output:

1 2/16/2016
2 2/12/2016
3 4/29/2016

From the above output, we see that the function produces correct results.
In our program, payment frequencies and horizons are represented by
strings that contain an integer representing a number of periods and a letter
representing a period. We need a function to extract the number of periods
and the period symbol from a single string, such as “1M” and “30Y.” We can
implement this function as follows:
1 Sub ExtractPeriod ( ByVal sIn As String , ByRef iNum As
Integer , ByRef sPeriod As String )
2 Dim iLen As Integer
3 Dim sNum As String
4 iLen = Len ( sIn )
5
6 If iLen < 2 Then
7 Err . Raise Number :=1003 , Source := " MHoliday .
ExtractPeriod " , Description := " Invalid
period string "
8 End If
9 sPeriod = Right ( sIn , 1)
10 sNum = Left ( sIn , iLen - 1)
11 If IsNumeric ( sNum ) Then
12 iNum = CInt ( sNum )
13 Else
14 Err . Raise Number :=1003 , Source := " MHoliday .
ExtractPeriod " , Description := " Invalid
period string "
15 End If
16 End Sub

The function ExtractPeriod has three arguments. The first argument is

a ByVal argument. The last two arguments are ByRef arguments used to
store the extracted number of periods and the extracted period symbol,
respectively. Since the last two arguments are ByRef arguments, the proce-
 154 Generating Payment Schedules

dure can change their values. To test this function, we use the following sub
procedure:
1 Sub Tes tExtract Period ()
2 Dim iNum As Integer
3 Dim sPeriod As String
4
5 ExtractPeriod " 2 M " , iNum , sPeriod
6 Debug . Print iNum & " --- " & sPeriod
7
8 ExtractPeriod " 30 Y " , iNum , sPeriod
9 Debug . Print iNum & " --- " & sPeriod
10 End Sub

Executing the above code gives

1 2 --- M
2 30 --- Y

The output shows that the function works correctly.

Now, we are ready to implement the main function for generating pay-
ment schedules. We implement this function as follows:
1 Function GenerateSchedule ( ByVal datSet As Date , ByVal
sFreq As String , ByVal sHorizon As String , ByVal
sCalendar As String , ByVal sBDC As String ) As
Variant
2 Dim iFreqNum As Integer
3 Dim sFreqPeriod As String
4 Dim iHorizonNum As Integer
5 Dim sHorizonPeriod As String
6 Dim lCount As Long
7 Dim sPeriod As String
8
9 Dim vDates () As Date
10 Dim lInd As Long
11
12 ' obtain and validate input
13 If Not IsBusinessDay ( datSet , sCalendar ) Then
14 Err . Raise Number :=1002 , Source := " MHolidy .
GenerateSchedule " , Description := " The
settlement date is not a business day . "
15 End If
16 ExtractPeriod sIn := sFreq , iNum := iFreqNum , sPeriod
:= sFreqPeriod
 Implementation 155

17 ExtractPeriod sIn := sHorizon , iNum := iHorizonNum ,

sPeriod := sHorizonPeriod
18
19 sPeriod = sFreqPeriod & sHorizonPeriod
20
21 If StrComp ( sPeriod , " MM " , vbTextCompare ) = 0 Or _
22 StrComp ( sPeriod , " YY " , vbTextCompare ) = 0 Then
23 lCount = iHorizonNum \ iFreqNum
24 ElseIf StrComp ( sPeriod , " MY " , vbTextCompare ) = 0
Then
25 lCount = iHorizonNum * 12 \ iFreqNum
26 ElseIf StrComp ( sPeriod , " YM " , vbTextCompare ) = 0
Then
27 lCount = iHorizonNum \ ( iFreqNum * 12)
28 Else
29 Err . Raise Number :=1004 , Source := " MHoliday .
GenerateSchedule " , Description := " Unknown
period combination . "
30 End If
31
32 If lCount <= 0 Then
33 Err . Raise Number :=1005 , Source := " MHoliday .
GenerateSchedule " , Description := " Invalid
number of periods . "
34 End If
35
36 ReDim vDates (0 To lCount )
37
38 For lInd = 0 To lCount
39 vDates ( lInd ) = RollDate ( OffsetDate ( datSet ,
lInd * iFreqNum , sFreqPeriod ) , sBDC ,
sCalendar )
40 Next lInd
41
42 GenerateSchedule = vDates
43 End Function

The function GenerateSchedule takes five arguments: the settlement date,

the payment frequency, the horizon of the last payment, the business cal-
endar, and the business day count convention. This function generates
an array of dates according to the given inputs and returns the array as
a variant. In this function, we only considered four combinations of the
payment frequency and the payment horizon. Since this is a main function,
we will use the interface to test the function.
 156 Generating Payment Schedules

7.6.4 MInterface Module

The VBA code contained in the module MInterface is responsible for pass-
ing the inputs to the main function and writing the outputs from the main
function to the worksheet.
Before we implement the macro for the button, let us define some
names for the inputs. By doing so, we do not hardcode the addresses of
input cells. Table 7.3 shows a list of range names we defined. On a Windows®
machine, we can click “Name Manager” in the “FORMULAS” tab to see the
names in the name manager window.

Table 7.3: Range Names Used by the Payment Schedule Generator

Name Refers To
BusinessDayConvention =Schedule!$E$6
Calendar =Schedule!$E$5
DayCountConvention =Schedule!$E$7
Frequency =Schedule!$E$3
Horizon =Schedule!$E$4
SetDate =Schedule!$E$2

Then, we implement macro for the button as follows:

1 Sub Button1_Click ()
2 Dim datSet As Date
3 Dim sFreq As String
4 Dim sHorizon As String
5 Dim sCalendar As String
6 Dim sBDC As String
7 Dim sDCC As String
8 Dim vSchedule As Variant
9
10 Dim lInd As Long
11 Dim lCount As Long
12
13 If IsDate ( Range ( " SetDate " ) . Value ) Then
14 datSet = Range ( " SetDate " ) . Value
15 Else
16 MsgBox " Please input a valid settlement date .
"
17 Exit Sub
18 End If
19
 Implementation 157

20 sFreq = Range ( " Frequency " ) . Value

21 sHorizon = Range ( " Horizon " ) . Value
22 sCalendar = Range ( " Calendar " ) . Value
23 sBDC = Range ( " B u s i n e s s D a y C o n v e n t i o n " ) . Value
24 sDCC = Range ( " Day Co un tCo nv en tio n " ) . Value
25
26 vSchedule = GenerateSchedule ( datSet := datSet ,
sFreq := sFreq , sHorizon := sHorizon , sCalendar :=
sCalendar , sBDC := sBDC )
27
28 Sheet1 . Range ( Cells (3 , 1) , Cells (3 , 1) . End ( xlDown )
) . ClearContents
29 For lInd = LBound ( vSchedule ) To UBound ( vSchedule )
30 Sheet1 . Cells ( lInd + 3 , 1) . Value = vSchedule (
lInd )
31 Next lInd
32
33 Sheet1 . Range ( Cells (4 , 2) , Cells (4 , 2) . End ( xlDown )
) . ClearContents
34 lCount = Sheet1 . Cells (3 , 1) . End ( xlDown ) . Row - 2
35 With Sheet1
36 For lInd = 1 To lCount - 1
37 . Cells ( lInd + 3 , 2) . Value = TFrac (. Cells (
lInd + 3 , 1) . Value , . Cells ( lInd + 2 ,
1) . Value , sDCC )
38 Next lInd
39 End With
40 End Sub

When the button is clicked, the sub procedure Button1_Click will run.
This sub procedure is an entry point of our program. Let us look at this
sub procedure. In Lines 2–9, we declare some variables. In Lines 13–18, we
make sure the value in the cell named “SetDate” is a date. In Lines 20–24,
we extract the inputs from relevant named cells. In Line 26, we call the
GenerateSchedule function, which is defined in the module MSchedule.
In Lines 28 and 33, we clear the content in the output area. In Lines 29–
31, we write the schedule of dates to the output area. In Lines 35–39, we
calculate the year fractions and write them to the output area.
To test the program, we use the inputs as shown in Figure 7.1. If we click
the button, we see that outputs as shown in Figure 7.1.

Exercise 7.4. Write a VBA function with two arguments that returns the
 158 Generating Payment Schedules

holiday date and the observed holiday date of a specified holiday in a

specified year. The VBA function should have the following specification:
1 Function GetHolidayNY ( ByVal lY As Long , ByVal
sHoliday As String ) As Variant

The argument sHoliday can take the following values:

Value Holiday
"NewYear" New Year’s Day
"MLK" Martin Luther King Jr. Day
"President" Present’s Day
"Memorial" Memorial Day
"Independence" Independence Day
"Labor" Labor Day
"Columbus" Columbus Day
"Veterans" Veterans Day
"Thanksgiving" Thanksgiving Day
"Christmas" Christmas Day

Exercise 7.5. Create the interface shown in Figure 7.2 and write a sub proce-
dure for the button so that clicking the button will generate the US holiday
dates and the observed holiday dates for any given year. When the year is
2017, the output should be identical to what shown in Figure 7.2.

Exercise 7.6. The date of the Easter Sunday in a year Y is a day in March of
the year Y . Suppose that March S, Y is the date of Easter Sunday in year Y .
Then the day S is determined as follows:

(a) N = 7 − mod (Y + Y \4 − Y \100 + Y \400 − 1, 7)

(b) G = 1 + mod (Y , 19)
(c) E = mod (11 ×G − 10, 30)
(d) H = Y \100
(e) SOL = H − H \4 − 12
(f) LU N = (H − 15 − (H − 17)\25)\3
(g) V = E \24 − E \25 + (G\12) × (E \25 − E \26)
(h) E = mod (11G − 10, 30) − mod (SOL − LU N , 30) + V
(i) R = 45 − E if E < 24; R = 75 − E if E ≥ 24
(j) C = 1 + mod (R + 2, 7)
 Summary 159

Figure 7.2: Interface of a holiday generator

(k) S = R + mod (7 + N −C , 7),

where \ denotes integer division. Write a VBA function to calculate the date
of Easter Sunday in a given year. The function should have the following
specification:
1 Function EasterSunday ( ByVal lY As Long ) As Date

Use your function to calculate Easter Sunday from 2010 to 2020. If your
function is implemented correctly, the dates should match the dates given
in Table 7.4.
h

7.7 Summary
Generating a schedule of payment dates is usually the first step of valuing an
interest rate derivative, such as swaps. In this chapter, we introduced how
160 Generating Payment Schedules

Table 7.4: Dates of Easter Sunday from 2010 to 2020

Year Easter Sunday

2010 4/4/2010
2011 4/24/2011
2012 4/8/2012
2013 3/31/2013
2014 4/20/2014
2015 4/5/2015
2016 3/27/2016
2017 4/16/2017
2018 4/1/2018
2019 4/21/2019
2020 4/12/2020

to generate a schedule of payment dates according to a business calendar

and a business day convention. We also introduced how to calculate the
year fractions between two dates according to a day count convention.
We only implemented the business calendar of the United States in this
chapter. However, readers can implement other business calendars in a
similar way. For information on calendars, readers are referred to Richards
(2012). The algorithms for interconverting Gregorian dates and Julian day
numbers, and the algorithm for calculating dates of Easter Sundays are
obtained from Richards (2012).
8
Bootstrapping Yield Curves

In this chapter, we implement a simple bootstrapper than can be used to

construct a yield curve from a series of swap rates. Figure 8.1 shows the
interface of the bootstrapper.

Figure 8.1: Interface of the yield curve bootstrapper.

The inputs of the bootstrapper include swap rates of various tenors

(i.e., maturities), specifications of the swaps, the yield curve day count
convention, and the output type. The output includes the dates, discount
factors, zero rates, forward rates, and day count factors. The yield curve
generator also uses two named ranges, which are given in Table 8.1.

161
162 Bootstrapping Yield Curves

Table 8.1: Range Names Used by the Yield Curve Bootstrapper

Name Refers To
Param =Curve!$E$3:$E$12
Rate =Curve!$B$3:$C$10

8.1 Introduction
The yield curve, also known as the term structure of interest rates, is the
curve that shows the relation between the interest rate and the time to
maturity. For example, Table 8.2 gives the interest rates on different times
to maturities (or tenors). If we plot the interest rates against the tenors, we
see the yield curve as shown in Figure 8.2.

Table 8.2: Interest Rates of Various Tenors

Tenor Rate
1 Year 0.69%
2 Year 0.77%
3 Year 0.88%
4 Year 1.01%
5 Year 1.14%
7 Year 1.39%
10 Year 1.68%
30 Year 2.21%

The interest rates shown in Table 8.2 and Figure 8.2 are continuously
compounded spot rates or zero-coupon rates. The continuously com-
pounded spot rate with maturity T is defined as

ln P (T )
R(T ) = − , (8.1)
τ(0, T )

where P (T ) is the price at time 0 of a zero-coupon bond, and maturity T

and τ(0, T ) is the year fraction between time 0 and time T according to
some day count convention (see Section 7.4). A zero-coupon bond with
maturity T is a contract that guarantees the holder a cash payment of one
unit at time T .
Introduction 163

2.0

l
1.5

l
1.0

0 5 10 15 20 25 30

Year

Figure 8.2: Yield curve

Let 0 ≤ T < S. Then, the forward rate for the period [T, S] is defined as

ln P (T ) − ln P (S)
R(T, S) = , (8.2)
τ(T, S)

where τ(T, S) is the year fraction between time T and time S according to
some day count convention. The simply compounded forward rate for the
period [T, S] is defined as

P (T )
µ ¶
1
F (T, S) = −1 . (8.3)
τ(T, S) P (S)

The discount factor is the factor that is used to multiply a future cash
flow in order to obtain the present value. The discount factor used to ob-
tained the present value of a cash flow at time T is defined as

DF (T ) = exp(−τ(0, T )R(T )) = P (T ). (8.4)

164 Bootstrapping Yield Curves

8.2 Interpolation
Interest rate products in the market have only a few different maturities. In
order to construct a yield curve, we need to know the interest rates at all
tenors within a time horizon (e.g., 30 years). Interpolation is used to estimate
the interest rates at other maturities. Interpolation is also connected to the
bootstrapping method described in Section 8.3.
There are several interpolation methods (Hagan and West, 2006, 2008):
linear on rates, linear on the log of rates, linear on discount factors, linear on
the log of discount factors, piecewise linear forward, and splines. The linear
on the log of discount factors method, also known as the raw interpolation
method, is stable and straightforward to implement. In this program, we
use the raw interpolation method.
Let Ti < Ti +1 be two maturities of which we know the discount factors.
Let T ∈ [Ti , Ti +1 ]. Then, in the raw interpolation method, the discount
factor DF (T ) at time T is interpolated from the discount factors DF (Ti )
and DF (Ti +1 ) as follows:

τ(T, Ti +1 ) ln DF (Ti ) + τ(Ti , T ) ln DF (Ti +1 )

ln DF (T ) = . (8.5)
τ(Ti , Ti +1 )

By Equation (8.4), the above equation is equivalent to

τ(T, Ti +1 )Ti R(Ti ) + τ(Ti , T )Ti +1 R(Ti +1 )

R(T ) = . (8.6)
τ(Ti , Ti +1 )T

For other interpolation methods, readers are referred to (Hagan and

West, 2006) and (Hagan and West, 2008).

8.3 Bootstrapping Yield Curves

In this section, we show how to construct a yield curve from a series of swap
rates using the bootstrapping method. One advantage of using swap rates
in the construction of yield curves is that we observe swap rates from a wide
number of maturities. For example, Table 8.3 gives the US swap rates for
various maturities on February 8, 2016.
To construct a yield curve from the swap rates for the maturities shown
in Table 8.3, we bootstrap the discount factors for the periods corresponding
Bootstrapping Yield Curves 165

Table 8.3: US Swap Rates for Various Maturities on February 8, 2016. These
rates were obtained from http://www.federalreserve.gov

Tenor Rate
1Y 0.69%
2Y 0.77%
3Y 0.88%
4Y 1.01%
5Y 1.14%
7Y 1.38%
10Y 1.66%
30Y 2.15%

to these maturities. Discount factors for other periods can be interpolated

from those discount factors.
To describe the bootstrapping method, let us introduce how interest
rate swaps are priced. As we mentioned in Chapter 7, an interest rate swap
(IRS) is a financial derivative in which two parties agree to exchange in-
terest rate cash flows, based on a specified notional amount. A typical IRS
consists of two legs of cash flows: one leg of incoming cash flows, and one
leg of outgoing cash flows. Usually, the cash flows of one leg are calculated
according to a fixed rate, and the cash flows of another leg are calculated
according to a floating rate. In such cases, we say that the swap has a fixed
leg and a floating leg. Some swaps consist of two floating legs.
The swap rates given in Table 8.3 are the rates of swaps that consist of a
fixed leg and a floating leg. In addition, the swap rates are determined in
such a way that the corresponding swaps have initial values of zero. In other
words, the swap rate is determined such that the present value of the fixed
leg cash flows is equal to that of the floating leg cash flows.
Let T0 be the settlement date, and T1 , T2 , . . . , Tn be the payment dates of
an IRS. In the US, the settlement date is two business days after the trade
date. An IRS usually has semiannual payments. If a swap matures in one
year, then it has two cash flows: one in 6 months and one in the maturity. If
a swap matures in 30 years, then it has 60 cash flows.
Suppose that the notional amount is one unit. For i = 1, 2, . . . , n, the
fixed cash flow at the date Ti is calculated as

τ(Ti −1 , Ti )r,
166 Bootstrapping Yield Curves

where τ(Ti −1 , Ti ) is the year fraction between dates Ti −1 and Ti , and r is the
swap rate. The floating cash flow at the date Ti is calculated as

τ(Ti −1 , Ti )F (Ti −1 , Ti ),

where F (Ti −1 , Ti ) is the simply compounded forward rate for the period
[Ti −1 , Ti ] (see Equation (8.3)). The swap rate is determined to satisfy the
following equation:
n n
τ(Ti −1 , Ti )r DF (Ti ) = τ(Ti −1 , Ti )F (Ti −1 , Ti )DF (Ti ),
X X
(8.7)
i =1 i =1

where DF (Ti ) = P (Ti ) is the discount factor for the period [0, Ti ]. Since
n n µ P (T ¶
i −1 )
τ(Ti −1 , Ti )F (Ti −1 , Ti )DF (Ti ) =
X X
− 1 P (Ti )
i =1 i =1 P Ti
X n
= [P (Ti −1 ) − P (Ti )]
i =1
= P (T0 ) − P (Tn ),

Equation (8.7) is reduced to

n
τ(Ti −1 , Ti )DF (Ti ) = DF (T0 ) − DF (Tn ).
X
r (8.8)
i =1

According to Equation (8.8), we can solve the swap rate r given a yield
curve, because we can obtain the discount factors DF (T0 ), DF (T1 ), . . . , DF (Tn )
from the yield curve. The bootstrapping procedure reverses this process,
that is, we solve the discount factors DF (T0 ), DF (T1 ), . . . , DF (Tn ) given the
swap r observed in the market. However, there are many unknown discount
factors and only one equation (i.e., Equation (8.8)). How can we solve the
discount factors based on one equation? To solve this problem, we just treat
DF (Tn ) as an unknown in the equation and interpolate other unknown dis-
count factors from DF (Tn ) and another known discount factor, which can
be 1 (i.e., the discount factor at time 0) or a discount factor bootstrapped
from a swap rate for a shorter maturity.
Let T0 be the settlement date. Let Ti denote the payment date after 6i
months, for i = 1, 2, . . . , 60. In particular, the bootstrapping procedure works
as follows:
(a) Bootstrap the discount factor DF (T2 ) using the 1Y swap rate. In this step,
we need to interpolate the discount factors at T0 (settlement date) and
T1 (6 months) from 1 (the discount factor at time 0) and DF (T2 ). We use
a method described in Section 8.4 to solve DF (T2 ).
Finding Roots of an Equation 167

(b) Bootstrap the discount factor DF (T4 ) using the 2Y swap rate. In this
step, we already know the discount factor DF (T0 ), DF (T1 ), and DF (T2 )
from the previous step. We only need to interpolate the discount factor
DF (T3 ) from DF (T2 ) and DF (T4 ). We can use the same method to solve
DF (T4 ).
(c) Repeat the above process to bootstrap discount factors DF (T6 ), DF (T8 ),
DF (T10 ), DF (T14 ), DF (T20 ), and DF (T60 ) from the swap rates for matu-
rities 3Y, 4Y, 5Y, 7Y, 10Y, and 30Y.

8.4 Finding Roots of an Equation

Bootstrapping discount factors from swap rates requires solving nonlinear
equations. For example, when we bootstrap the discount factor DF (T2 )
from the 1Y swap rate, we need to solve the following equation:

r τ(T0 , T1 )DF (T1 ) + r τ(T1 , T2 )DF (T2 ) = DF (T0 ) − DF (T2 ),

where DF (T0 ) and DF (T1 ) are interpolated from 1 and DF (T2 ). If we treat
DF (T2 ) as an unknown x in the equation, then the above equation can be
written as
r τ(T0 , T1 )g (T1 , x) + r τ(T1 , T2 )x = g (T0 , x) − x,
or
r τ(T0 , T1 )g (T1 , x) + r τ(T1 , T2 )x − g (T0 , x) + x = 0,
where g (·, ·) denotes the interpolation function. Solving the above equation
for x gives the discount factor DF (T2 ).
In this section, we introduce Newton’s method, also known as the
Newton–Raphson method, to find a root of the following equation:

f (x) = 0.

Newton’s method is an iterative method (Press et al., 2002, Chapter 9). Let
x 0 and x 1 be two initial guesses of the root of the above equation. Then,
Newton’s method proceeds as follows:
f (x i )
x i +1 = x i − (x i − x i −1 ) , i = 1, 2, . . . . (8.9)
f (x i ) − f (x i −1 )
We stop the iterative process until some condition is met. For example, we
can stop the iterative process when |x i +1 − x i | < 1e − 10.
 168 Bootstrapping Yield Curves

8.5 Implementation
As we can see from Figure 8.1, the interface of the bootstrapping tool con-
tains inputs, outputs, and one button. When we click the button, the yield
curve will be bootstrapped from the swap rates given in the input. The
Visual Basic for Applications (VBA) code behind the interface is organized
into five modules: MDate, MHoliday, MSchedule, MCurve, and MInterface.
The first three modules, MDate, MHoliday, and MSchedule, are imported
from the payment schedule generator implemented in Chapter 7. We will
implement the modules MCurve and MInterface in this section.

8.5.1 MCurve Module

The module MCurve contains VBA code that implements the bootstrapping
technique described in previous sections of this chapter. The declaration
section of this module contains the following code:
1 Option Explicit
2
3 ' input
4 Dim aTenor () As String
5 Dim aRate () As Double
6 Dim sFixFreq As String
7 Dim sFixDCC As String
8 Dim sFltFreq As String
9 Dim sFltDCC As String
10 Dim sCalendar As String
11 Dim sBDC As String
12 Dim datCurveDate As Date
13 Dim datSet As Date
14 Dim iSetDays As Integer
15 Dim sYCDCC As String
16 Dim sOutputType As String
17
18 ' yield curve
19 Dim aDate () As Date ' tenor of the rates
20 Dim aDF () As Double ' discount factor

In the above code, we defined quite a few module-level variables that are
used to store the input information and the output yield curve. The swap
rates and their tenors are saved in the arrays aRate and aTenor, respectively.
The resulting yield curve is stored in the arrays aDate and aDF, respectively.
 Implementation 169

The array aDate contains the maturity dates of the swap rates, and the array
aDF contains the discount factors at these maturity dates.
The first procedure of this module is the sub procedure Initialize,
which is used to initialize the module-level variables based on the input
information. It also calculates the settlement date based on the number of
settlement days and the business calendar. We implement this sub proce-
dure as follows:
1 Sub Initialize ( ByRef rngRate As Range , ByRef rngParam
As Range )
2 Dim iNumRates As Integer
3 Dim i As Integer
4 iNumRates = rngRate . Rows . Count
5
6 ReDim aTenor (1 To iNumRates )
7 ReDim aRate (1 To iNumRates )
8
9 For i = 1 To iNumRates
10 aTenor ( i ) = rngRate . Cells (i , 1) . Value
11 aRate ( i ) = rngRate . Cells (i , 2) . Value / 100
12 Next i
13
14 sFixFreq = rngParam . Cells (1 , 1) . Value
15 sFixDCC = rngParam . Cells (2 , 1) . Value
16 sFltFreq = rngParam . Cells (3 , 1) . Value
17 sFltDCC = rngParam . Cells (4 , 1) . Value
18 sCalendar = rngParam . Cells (5 , 1) . Value
19 sBDC = rngParam . Cells (6 , 1) . Value
20 datCurveDate = rngParam . Cells (7 , 1) . Value
21 iSetDays = rngParam . Cells (8 , 1) . Value
22 sYCDCC = rngParam . Cells (9 , 1) . Value
23 sOutputType = rngParam . Cells (10 , 1) . Value
24
25 datSet = datCurveDate
26 For i = 1 To iSetDays
27 datSet = RollDate ( AddDays ( datSet , 1) , " Follow
" , sCalendar )
28 Next i
29
30 ReDim aDate (0 To 0)
31 ReDim aDF (0 To 0)
32 aDate (0) = datCurveDate
33 aDF (0) = 1
34 End Sub
 170 Bootstrapping Yield Curves

In this sub procedure, the settlement date is calculated, and the yield curve
is initialized. The output yield curve is stored in two array-type variables:
aDate and aDF. At the beginning, the variable aDate is initialized to contain
the curve date, and the variable aDF is initialized to contain 1, which is the
discount factor at the curve date.
For testing purposes, we also create the following sub procedure:
1 Sub PrintInfo ()
2 Dim i As Integer
3 For i = LBound ( aTenor ) To UBound ( aTenor )
4 Debug . Print " Swap Rate " & i & " : " & aTenor (
i ) & " -- " & aRate ( i )
5 Next i
6
7 Debug . Print " Fixed Leg : " & sFixFreq & " -- " &
sFixDCC
8 Debug . Print " Floating Leg : " & sFltFreq & " -- "
& sFltDCC
9 Debug . Print " Calendar : " & sCalendar
10 Debug . Print " Business Day Convention : " & sBDC
11 Debug . Print " Curve Date : " & datCurveDate
12 Debug . Print " Settlement Days : " & iSetDays
13 Debug . Print " Yield Curve Day Count Convention : "
& sYCDCC
14 Debug . Print " Settlement Date : " & datSet
15
16 For i = LBound ( aDate ) To UBound ( aDate )
17 Debug . Print " Discount Factor " & i & " : " &
aDate ( i ) & " -- " & aDF ( i )
18 Next i
19
20 End Sub

This sub procedure just displays the content of the module-level variables
in the Immediate Window.
To test the sub procedure Initialize, we use the following code:
1 Sub TestInitialize ()
2 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
3
4 PrintInfo
5 End Sub
 Implementation 171

Executing the above sub procedure produces the following output:

1 Swap Rate 1: 1 Y -- 0.0069
2 Swap Rate 2: 2 Y -- 0.0077
3 Swap Rate 3: 3 Y -- 0.0088
4 Swap Rate 4: 4 Y -- 0.0101
5 Swap Rate 5: 5 Y -- 0.0114
6 Swap Rate 6: 7 Y -- 0.0138
7 Swap Rate 7: 10 Y -- 0.0166
8 Swap Rate 8: 30 Y -- 0.0215
9 Fixed Leg : 6 M -- Thirty360
10 Floating Leg : 6 M -- ACT360
11 Calendar : NY
12 Business Day Convention : Modified
13 Curve Date : 2/8/2016
14 Settlement Days : 2
15 Yield Curve Day Count Convention : Thirty360
16 Settlement Date : 2/10/2016
17 Discount Factor 0: 2/8/2016 -- 1

If we compare the above output to the inputs shown in Figure 8.1, we see
that the sub procedure Initialize works correctly.
We implement the raw interpolation method as follows:
1 Function LogLinear ( ByVal datD As Date ) As Double
2 If datD < datCurveDate Then
3 Err . Raise Number :=2001 , Source := " MCurve .
LogLinear " , Description := " Date is ealier
than the curve date . "
4 End If
5 Dim i As Integer
6 Dim iL As Integer
7 Dim iH As Integer
8 iL = LBound ( aDate )
9 iH = UBound ( aDate )
10 For i = iL To iH
11 If datD = aDate ( i ) Then
12 LogLinear = aDF ( i )
13 Exit Function
14 End If
15 Next i
16
17 Dim dat1 As Date
18 Dim dat2 As Date
19 Dim dDF1 As Double
 172 Bootstrapping Yield Curves

20 Dim dDF2 As Double

21 Dim bFound As Boolean
22 bFound = False
23 For i = iL To iH - 1
24 If datD > aDate ( i ) And datD < aDate ( i + 1)
Then
25 dat1 = aDate ( i )
26 dat2 = aDate ( i + 1)
27 dDF1 = aDF ( i )
28 dDF2 = aDF ( i + 1)
29 bFound = True
30 End If
31 Next i
32
33 If Not bFound Then
34 dat1 = aDate ( iH - 1)
35 dat2 = aDate ( iH )
36 dDF1 = aDF ( iH - 1)
37 dDF2 = aDF ( iH )
38 End If
39
40 Dim dDCF1 As Double
41 Dim dDCF2 As Double
42 Dim dDCF As Double
43 dDCF1 = TFrac ( datCurveDate , dat1 , sYCDCC )
44 dDCF2 = TFrac ( datCurveDate , dat2 , sYCDCC )
45 dDCF = TFrac ( datCurveDate , datD , sYCDCC )
46
47 Dim dTemp As Double
48 dTemp = Work sheetFun ction . Ln ( dDF2 ) * ( dDCF -
dDCF1 ) + _
49 Works heetFunc tion . Ln ( dDF1 ) * ( dDCF2 - dDCF )
50 LogLinear = Exp ( dTemp / ( dDCF2 - dDCF1 ) )
51 End Function

The function procedure LogLinear has one argument, which is a date. The
function will raise an error if the input date is older than the curve date. If
the input date is equal to one of the dates stored in the variable aDate, the
function returns the discount factor corresponding to this date. If the date
is not equal to any dates stored in the variable aDate, the function returns
the discount factor interpolated by the raw interpolation method.
To test the function LogLinear, we use the following sub procedure:
1 Sub TestLogLinear ()
 Implementation 173

2 ReDim aDate (0 To 2)
3 ReDim aDF (0 To 2)
4
5 aDate (0) = #4/5/2016#
6 aDate (1) = #4/5/2017#
7 aDate (2) = #4/5/2026#
8 aDF (0) = 1
9 aDF (1) = 0.95
10 aDF (2) = 0.7
11
12 sYCDCC = " Thirty360 "
13
14 Debug . Print LogLinear (#4/5/2016#)
15 Debug . Print LogLinear (#10/5/2016#)
16 Debug . Print LogLinear (#4/5/2041#)
17 End Sub

Executing the above code gives the following output:

1 1
2 0.974 67943448 0896
3 0.420 78033130 0377

The output indicates that the function works appropriately.

We calculate the present value of an IRS as follows:
1 Dim iNum As Integer
2 Dim sPeriod As String
3 Dim i As Integer
4 Dim dDF As Double
5
6 Dim vFixDates As Variant
7 Dim dFixPV As Double
8 ExtractPeriod sIn := sTenor , iNum := iNum , sPeriod :=
sPeriod
9
10 ' calculate fixed leg cash flows and present
value
11 vFixDates = GenerateSchedule ( datSet , sFixFreq ,
sTenor , sCalendar , sBDC )
12 For i = LBound ( vFixDates ) + 1 To UBound ( vFixDates
)
13 dDF = LogLinear ( vFixDates ( i ) )
14 dFixPV = dFixPV + TFrac ( vFixDates ( i - 1) ,
vFixDates ( i ) , sFixDCC ) * dRate * dDF
 174 Bootstrapping Yield Curves

15 Next i
16
17 ' calculate floating leg cash flows and present
value
18 Dim vFltDates As Variant
19 Dim dFltPV As Double
20 Dim dFR As Double
21 Dim dT As Double
22 ' calculate fixed leg cash flows and present
value
23 vFltDates = GenerateSchedule ( datSet , sFltFreq ,
sTenor , sCalendar , sBDC )
24 For i = LBound ( vFltDates ) + 1 To UBound ( vFltDates
)
25 dDF = LogLinear ( vFltDates ( i ) )
26 dT = TFrac ( vFltDates ( i - 1) , vFltDates ( i ) ,
sFltDCC )
27 dFR = ( LogLinear ( vFltDates ( i - 1) ) /
LogLinear ( vFltDates ( i ) ) - 1) / dT
28 dFltPV = dFltPV + dT * dFR * dDF
29 Next i
30
31 PvSwap = dFixPV - dFltPV
32 End Function

The function procedure PvSwap has two arguments: the tenor and the rate
of the swap. This function uses some module-level variables, such as these
used to specify the payment frequency and the business calendar. It first
calculates the present value of the fixed leg, and then calculates the present
value of the floating leg.
To test the function PvSwap, we use the following sub procedure:
1 Sub TestPvSwap ()
2 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
3
4 ReDim Preserve aDate (0 To 1)
5 ReDim Preserve aDF (0 To 1)
6
7 aDate (1) = OffsetDate ( datSet , 1 , " Y " )
8 aDF (1) = 0.95
9
10 Debug . Print PvSwap ( aTenor (1) , aRate (1) )
11 End Sub
 Implementation 175

If we use the inputs shown in Figure 8.1, then executing the above code
gives
1 -4.30769841682971 E -02

The output shows that the present value of the fixed leg is lower than that of
the floating leg.
We implement Newton’s method as follows:
1 Sub SolveRate ( ByVal iInd As Integer )
2 ReDim Preserve aDate (0 To iInd )
3 ReDim Preserve aDF (0 To iInd )
4
5 Dim iNum As Integer
6 Dim sPeriod As String
7
8 ExtractPeriod sIn := aTenor ( iInd ) , iNum := iNum ,
sPeriod := sPeriod
9 aDate ( iInd ) = OffsetDate ( datSet , iNum ,
sPeriod )
10
11 Dim dx As Double
12 Dim dx0 As Double
13 Dim dx1 As Double
14 Dim df As Double
15 Dim df0 As Double
16 Dim df1 As Double
17
18 dx0 = Exp ( - aRate ( iInd ) * TFrac ( datCurveDate ,
aDate ( iInd ) , sYCDCC ) )
19 aDF ( iInd ) = dx0
20 df0 = PvSwap ( aTenor ( iInd ) , aRate ( iInd ) )
21 dx1 = dx0 + 0.001
22 aDF ( iInd ) = dx1
23 df1 = PvSwap ( aTenor ( iInd ) , aRate ( iInd ) )
24 Do While Abs ( dx1 - dx0 ) > 0.0000000001
25 dx = dx0 - df0 * ( dx1 - dx0 ) / ( df1 - df0
)
26 aDF ( iInd ) = dx
27 df = PvSwap ( aTenor ( iInd ) , aRate ( iInd ) )
28
29 dx0 = dx1
30 df0 = df1
31 dx1 = dx
32 df1 = df
 176 Bootstrapping Yield Curves

33 Loop
34
35 aDF ( iInd ) = dx
36 End Sub

The sub procedure SolveRate has only one argument, which is an integer
that represents the index of the input swap rates. This sub procedure also
modifies the variables used to store the yield curve. At the beginning of this
sub procedure, the variables aDate and aDF are redimensioned to contain
one more value. The existing values are preserved.
To test the sub procedure SolveRate, we use the following sub proce-
dure:
1 Sub TestSolveRate ()
2 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
3
4 SolveRate iInd :=1
5
6 Debug . Print aDate (1) & " -- " & aDF (1)
7 End Sub

If we use the inputs shown in Figure 8.1, then executing the above code
gives
1 2/10/2017 -- 0.993097540002 669

The output shows that the bootstrapped discount factor for 1 year is about
0.9931.
Now, we can implement a sub procedure to bootstrap all the discount
factors as follows:
1 Sub BuildCurve ()
2 Dim iL As Integer
3 Dim iH As Integer
4 Dim i As Integer
5 iL = LBound ( aTenor )
6 iH = UBound ( aTenor )
7
8 For i = iL To iH
9 SolveRate iInd := i
10 Next i
11 End Sub
 Implementation 177

The sub procedure BuildCurve just loops through the input swap rates
and calls the sub procedure SolveRate to bootstrap the discount factors.
To test this sub procedure, we use the following sub procedure:
1 Sub TestBuildCurve ()
2 Dim i As Integer
3 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
4
5 BuildCurve
6 PrintInfo
7 End Sub

Executing the above code gives the following output:

1 Swap Rate 1: 1 Y -- 0.0069
2 Swap Rate 2: 2 Y -- 0.0077
3 Swap Rate 3: 3 Y -- 0.0088
4 Swap Rate 4: 4 Y -- 0.0101
5 Swap Rate 5: 5 Y -- 0.0114
6 Swap Rate 6: 7 Y -- 0.0138
7 Swap Rate 7: 10 Y -- 0.0166
8 Swap Rate 8: 30 Y -- 0.0215
9 Fixed Leg : 6 M -- Thirty360
10 Floating Leg : 6 M -- ACT360
11 Calendar : NY
12 Business Day Convention : Modified
13 Curve Date : 2/8/2016
14 Settlement Days : 2
15 Yield Curve Day Count Convention : Thirty360
16 Settlement Date : 2/10/2016
17 Discount Factor 0: 2/8/2016 -- 1
18 Discount Factor 1: 2/10/2017 -- 0.9930 97540002 669
19 Discount Factor 2: 2/10/2018 -- 0.9847 07793704 918
20 Discount Factor 3: 2/10/2019 -- 0.9739 35374151 509
21 Discount Factor 4: 2/10/2020 -- 0.9603 49877939 723
22 Discount Factor 5: 2/10/2021 -- 0.9444 39556311 098
23 Discount Factor 6: 2/10/2023 -- 0.9073 27535413 443
24 Discount Factor 7: 2/10/2026 -- 0.8451 70756713 933
25 Discount Factor 8: 2/10/2046 -- 0.5147 31101479 042

From the output, we see that the yield curve was constructed.
Finally, we implement a function to create output based on a specified
output type. We implement this function as follows:
 178 Bootstrapping Yield Curves

1 Function CreateOutput () As Variant

2 Dim aOutDates () As Date
3 Dim i As Integer
4 Dim iL As Integer
5 Dim iH As Integer
6 Dim dT As Double
7 If StrComp ( sOutputType , " Input " , vbTextCompare ) =
0 Then
8 iL = LBound ( aDate )
9 iH = UBound ( aDate )
10 ReDim aOutDates ( iL To iH )
11 For i = iL To iH
12 aOutDates ( i ) = aDate ( i )
13 Next i
14 Else
15 aOutDates = GenerateSchedule ( datCurveDate ,
sOutputType , " 30 Y " , sCalendar , sBDC )
16 End If
17
18 iL = LBound ( aOutDates )
19 iH = UBound ( aOutDates )
20
21 ReDim aOut ( iL To iH , 0 To 4) As Double
22 For i = iL To iH
23 aOut (i , 0) = CDbl ( aOutDates ( i ) )
24 aOut (i , 1) = LogLinear ( aOutDates ( i ) )
25 dT = TFrac ( datCurveDate , aOutDates ( i ) , sYCDCC
)
26 If dT = 0 Then
27 aOut (i , 2) = 0
28 Else
29 aOut (i , 2) = - Wor ksheetFu nction . Ln ( aOut (i
, 1) ) / dT
30 End If
31 If i < iH Then
32 aOut (i , 3) = ( LogLinear ( aOutDates ( i ) ) /
LogLinear ( aOutDates ( i + 1) ) - 1) /
TFrac ( aOutDates ( i ) , aOutDates ( i + 1) ,
sYCDCC )
33 Else
34 aOut (i , 3) = aOut ( i - 1 , 3)
35 End If
36 aOut (i , 4) = TFrac ( datCurveDate , aOutDates ( i )
, sYCDCC )
37 Next i
 Implementation 179

38 CreateOutput = aOut
39 End Function

The sub procedure CreateOutput calculates other relevant quantities (e.g.,

forward rates, zero rates, and day count factors) based on the yield curve
and returns a two-dimensional array as a variant to the caller. If the output
type is “Input,” this sub procedure calculates these quantities only at the
maturities of the input swap rates. If the output is “xM” (e.g., 1M, 3M,
12M, etc.), this sub procedure calculates these quantities at a sequence of
dates generated by the given frequency. We will test this function using the
interface.

8.5.2 MInterface Module

The module MInterface contains VBA code for the button. We implement
this sub procedure as follows:
1 Sub Button1_Click ()
2 Dim vRes As Variant
3 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
4 BuildCurve
5 vRes = CreateOutput
6
7 Dim iRows As Integer
8 Dim i As Integer
9 Dim j As Integer
10 iRows = Sheet1 . Range ( " H3 " ) . End ( xlDown ) . Row
11 Sheet1 . Range ( Cells (3 , 8) , Cells ( iRows , 12) ) .
ClearContents
12 For i = LBound ( vRes , 1) To UBound ( vRes , 1)
13 For j = 0 To 4
14 Sheet1 . Cells ( i + 3 , 8 + j ) . Value = vRes (i
, j)
15 Next j
16 Next i
17 End Sub

This sub procedure is responsible for preparing the input and displaying
the output. The sub procedure first calls other three functions: Initialize,
BuildCurve, and CreateOutput. The three functions are defined in the
module MCurve. Then, the sub procedure writes the output to columns 8
180 Bootstrapping Yield Curves

to 12. If we want to rearrange the input or change the output to a different

location, we only need to change the VBA code in the module MInterface.
To test the program, we use the inputs as shown in Figure 8.1. If we click
the button, we see that outputs as shown in Figure 8.1.

Exercise 8.1. The bootstrapping procedure first bootstraps the discount

factor for the 1Y swap rate, then bootstraps the discount factor for the 2Y
swap rate, and so on. However, the VBA program does not sort the input
swap rates in terms of tenors. Instead, the VBA program bootstraps the
discount factors based on the input order of the swap rates. Write a sub
procedure named CheckInput to verify that the input order of the swap
rates is an increasing order of tenors. This sub procedure is included in the
module MCurve and should be called after the procedure Initialize and
before the procedure BuildCurve. The sub procedure CheckInput should
raise an error if the input order of the swap rates is not an increasing order
of tenors.

Exercise 8.2. The function procedure PvSwap in the module MCurve cal-
culates the present value of the floating leg by projecting and discounting
the individual cash flows. In fact, we do not have to do this. From Equation
(8.8), the net present value of a swap is given by
n
τ(Ti −1 , Ti )DF (Ti ) − DF (T0 ) + DF (Tn ).
X
r
i =1

Rewrite the function procedure PvSwap by using the above formula to cal-
culate the net present value.

Exercise 8.3. Newton’s method may not converge if the initial guesses are
not good. Another method for finding a root of an equation is the bisection
method. Suppose that we want to find a root of the equation f (x) = 0, where
f (x) is a continuous function of x. We first find two values x L < xU , such that
f (x L ) f (xU ) < 0. Then we know that there is a root in the interval (x L , xU ).
We can use the bisection method to find the root. To do that, we let
xL + x H
xM = .
2
If f (x L ) f (x M ) < 0, then we update x H = x M . If f (x M ) f (x H ) < 0, then we
update x L = x M . We continues updating x L and x H in the above way until
|x L − x H | < 10−10 .
Summary 181

Rewrite the sub procedure SolveRate by using the bisection method

described above. You can use the following initial values x L = exp(−10r T )
and x H = 1, where r denotes the swap rate under consideration and T
denotes its maturity.

Exercise 8.4. Modify the sub procedure Button1_Click to do the following:

Create a chart sheet named Plot, if it does not exist, and plot the zero rates
and the forward rates against the day count factors in the chart sheet. (Hint:
you can use the macro recorder to get started.)

8.6 Summary
Bootstrapping a yield curve allows us to produce swap prices that are con-
sistent with market prices. In this chapter, we introduced how to construct a
yield curve by bootstrapping discount factors from swap rates, which cover
a wide range of maturities. We also introduced an interpolation method
and Newton’s method for solving a nonlinear equation that is required in
the bootstrapping process. For more information about interpolation meth-
ods for constructing yield curves, readers are referred to (Hagan and West,
2006, 2008). Other methods for solving nonlinear equations can be found
in (Press et al., 2002; Garrett, 2015).
9
Generating Risk-Neutral Scenarios

In this chapter, we implement a simple economic scenario generator for

generating risk-neutral scenarios of an equity security (e.g., a stock). The
interface of this risk-neutral scenario generator is shown in Figure 9.1.

Figure 9.1: Interface of the risk-neutral scenario generator

The inputs of the scenario generator include a list of forward rates and
six parameters: the volatility, the horizon of the scenarios, the time step, the
number of paths, a seed, and the precision of the scenarios. The seed allows
us to repeat the scenarios. In other words, if we use the same seed and the
same other parameters, then we get the same set of scenarios.

183
184 Generating Risk-Neutral Scenarios

9.1 Introduction
Economic scenario generators have been used by insurance companies
to price insurance products (e.g., variable annuities) embedding guaran-
tees and calculate capital adequacy. An economic scenario generator is
simply a computer program that produces random sequences of market
variables (e.g., equity returns) by assuming that those market variables
follow some stochastic process. For a particular market variable, we call a
random sequence of the market variable a path or a scenario.
There are two types of economic scenarios that are commonly used by
insurance companies: risk-neutral scenarios and real-world scenarios. Risk-
neutral scenarios, also referred to as arbitrage-free or market-consistent
scenarios, are economic scenarios that are consistent with market prices.
Risk-neutral scenarios are generally used for pricing purpose. Real-world
scenarios are created based on the historical experience of market variables
and are used for other purposes, such as calculating capital adequacy and
profitability of a product.

9.2 Black–Scholes Model

There are numerous approaches to generate risk-neutral scenarios for an
equity security. One common approach is to use the Black–Scholes model
(Black and Scholes, 1973). Under the Black–Scholes model, the equity se-
curity is assumed to follow the following stochastic process (Carmona and
Durrleman, 2005):
·µZ t
t
¶ ¸
S t = S 0 exp r s ds − σ2 + σB t , t ≥ 0, (9.1)
0 2

where r s is the short rate at time s, σ is the volatility of the equity security,
and {B t : t ≥ 0} is a standard Brownian motion (Gan et al., 2014).
The stochastic process given in Equation (9.1) is a continuous-time
stochastic process. Since a path is a discrete sequence, we need to discretize
the continuous-time stochastic process. Let ∆ be a time step in years. For
1
example, if we want to generate monthly scenarios, then ∆ = 12 . If we
want to generate scenarios at the annual time step, then ∆ = 1. Let m be
the number of time steps of a path. For example, if we want to generate
Black–Scholes Model 185

scenarios at the monthly time step for a horizon of 30 years, then m = 360.
For i = 0, 1, . . . , m, let t i = ∆i . Then, by Equation (9.1), we have

∆
·µZ t
S ti
¶ ¸
i
= exp r s ds − σ2 + σ(B ti − B ti −1 ) (9.2)
S ti −1 t i −1 2

for i = 1, 2, . . . , m.
Let
S ti
Ai = , i = 1, 2, . . . , m. (9.3)
S ti −1
We call A i the accumulation factor for the i th period. In other words, A i is
the growth of the equity security from time t i −1 = (i − 1)∆ to time t i = i ∆.
Then, by Equation (9.2) and Equation (9.3), we have

∆
µZ t ¶
i
ln A i = r s ds − σ2 + σ(B ti − B ti −1 ), (9.4)
t i −1 2

for i = 1, 2, . . . , m. The logarithm of A i , ln A i , is the continuous return of the

equity security for the i th period.
Let
Bt − Bt
Zi = i p i −1 , i = 1, 2, . . . , m. (9.5)
∆
Then, by the property of Brownian motions, we know that Zi is a standard
normal random variable (i.e., a normal random variable with a mean of 0
and a standard deviation of 1), and that Z1 , Z2 , . . ., Zm are independent.
We assume that the interest rate is a deterministic function of time, and
the forward rate is constant within each period. For i = 1, 2, . . . , m, let f i be
the annualized simply compounded forward rate for the period [t i −1 , t i ].
Then, by Equation (8.3), we have

1 P (t i −1 )
µ ¶
fi = −1 , (9.6)
∆ P (t i )
where P (t i ) is the price at time zero of the zero-coupon bond maturing at
time t i . Since the interest rate is deterministic rather than stochastic, we
have (Gan et al., 2014, Chapter 45)
µ Z t ¶
i
P (t i ) = exp − r s ds , i = 0, 1, . . . , m. (9.7)
0

Combining Equation (9.6) and Equation (9.7), we get

· µZ t ¶ ¸
1 i
fi = exp r s ds − 1 , (9.8)
∆ t i −1
186 Generating Risk-Neutral Scenarios

which gives Z ti
r s ds = ln(1 + ∆ f i ). (9.9)
t i −1
Now, plugging Equation (9.5) and Equation (9.9) into Equation (9.4), we
get
∆ 2 p
ln A i = ln(1 + ∆ f i ) − σ + σ ∆Zi (9.10)
2
for i = 1, 2, . . . , m. We use Equation (9.10) to generate risk-neutral scenarios.
Let n be the number of paths or scenarios we want to generate. Then,
the scenarios can be represented by an n × m matrix defined as
 
A 11 A 12 · · · A 1m
 A 21 A 22 · · · A 2m 
 
, (9.11)
 . .. .. 
 . ..
 . . . . 
A n1 A n2 · · · A nm

where A i j , called the accumulation factor for the j th period of the i th

scenario, is generated as
∆ 2 p
ln A i j = ln(1 + ∆ f j ) − σ + σ ∆z i j , (9.12)
2
where z i j is a random number generated from the standard normal distri-
bution.

9.3 Generating Random Normal Numbers

The probability density function of a normal distribution is defined as
(Evans et al., 2000, Chapter 30)

(x − µ)2
µ ¶
1
f (x, µ, σ) = p exp − , x ∈ (−∞, ∞), (9.13)
2πσ 2σ2
where µ is the mean, σ is the standard deviation, and exp(·) is the exponen-
tial function. The cumulative density function of the normal distribution is
defined as
Z y
F (y, µ, σ) = f (x, µ, σ)dx
−∞
(x − µ)2
Z y µ ¶
1
= p exp − dx, y ∈ (−∞, ∞). (9.14)
−∞ 2πσ 2σ2
 Implementation 187

The cumulative density function of the standard normal distribution is

F (y, 0, 1).
There are several ways to generate random normal numbers. One way is
to use the inversion method (Devroye, 1986). To use the inversion method
to generate a standard normal number, we first generate a uniform random
number u from [0,1], and then we calculate the number z by solving the
following equation:
u = F (z, 0, 1).
The number z is a standard random normal number.

9.4 Implementation
From Figure 9.1, we see that the input and the output are organized into the
“Main” sheet and the “Scenarios” sheet, respectively. When the button is
clicked, the scenarios are written to the “Scenarios” sheet. We organize the
Visual Basic for Applications (VBA) code into two modules: the MGenerator
module and the MInterface module.

9.4.1 MGenerator Module

This module contains VBA code for generating risk-neutral scenarios based
on the Black–Scholes models described before. The declaration section of
this module contains the following code:
1 Option Explicit
2
3 Private dSigma As Double
4 Private sHorizon As String
5 Private sTimeStep As String
6 Private lNumPath As Long
7 Private lSeed As Long
8 Private iPrecision As Integer
9 Private aFR () As Double

In the above code, we defined variables to store the parameters and the
forward rates. These variables are private variables that can only be used by
procedures in this module.
The following sub procedure, named Initialize, is used to initialize
the variables declared above:
 188 Generating Risk-Neutral Scenarios

1 Sub Initialize ( _
2 ByVal Sigma As Double , _
3 ByVal Horizon As String , _
4 ByVal TimeStep As String , _
5 ByVal NumPath As Long , _
6 ByVal Seed As Long , _
7 ByVal Precision As Integer , _
8 ByRef FR () As Double )
9 dSigma = Sigma
10 sHorizon = Horizon
11 sTimeStep = TimeStep
12 lNumPath = NumPath
13 lSeed = Seed
14 iPrecision = Precision
15
16 Dim lL As Long
17 Dim lU As Long
18 lL = LBound ( FR )
19 lU = UBound ( FR )
20
21 ReDim aFR (1 To lU - lL + 1)
22 Dim i As Long
23 For i = lL To lU
24 aFR ( i - lL + 1) = FR ( i )
25 Next i
26
27 End Sub

In the above sub procedure, the forward rates are passed by reference.
Other parameters are passed to the procedure by value. In Line 21, we re-
dimension the module-level array aFR to hold the forward rates. In Lines
23–25, we copy the forward rates to the array.
To make sure the sub procedure Initialize works as expected, we
need to test it. Before testing it, we need a sub procedure to display the
contents of the module-level variables. We can write such a sub procedure
as follows:
1 Sub PrintInfo ()
2 Debug . Print " Volatility : " & CStr ( dSigma )
3 Debug . Print " Horizong : " & sHorizon
4 Debug . Print " TimeStep : " & sTimeStep
5 Debug . Print " NumPath : " & CStr ( lNumPath )
6 Debug . Print " Seed : " & CStr ( lSeed )
 Implementation 189

7 Debug . Print " Precision : " & CStr ( iPrecision )

8 Debug . Print " Number of Forward Rates : " & CStr (
UBound ( aFR ) - LBound ( aFR ) + 1)
9 Dim i As Long
10 Dim iCount As Integer
11 For i = LBound ( aFR ) To UBound ( aFR )
12 Debug . Print Round ( aFR ( i ) , 6) ;
13 iCount = iCount + 1
14 If iCount >= 6 Then
15 Exit For
16 End If
17 Next i
18
19 End Sub

The sub procedure PrintInfo displays the values of the module-level vari-
ables in the Immediate window. Instead of displaying all the forward rates
in the array, we only show the number of forward rates and the first six
forward rates.
Now, we can use the following sub procedure to test the sub procedure
Initialize:
1 Sub TestInitialize ()
2 Dim rngParam As Range
3 Dim rngFirstFR As Range
4
5 Set rngParam = Sheet1 . Range ( " Param " )
6 Set rngFirstFR = Sheet1 . Range ( " FirstFR " )
7
8 Dim aFR () As Double
9 Dim lFirstRow As Long
10 Dim lLastRow As Long
11
12 lFirstRow = rngFirstFR . Cells (1 , 1) . Row
13 lLastRow = rngFirstFR . Cells (1 , 1) . End ( xlDown ) . Row
14
15 ReDim aFR (1 To lLastRow - lFirstRow + 1)
16 Dim i As Long
17 For i = 1 To lLastRow - lFirstRow + 1
18 aFR ( i ) = rngFirstFR . Cells (1 , 1) . Offset ( i - 1 ,
0) . Value
19 Next i
20
21 Initialize Sigma := rngParam . Cells (1 , 1) . Value , _
 190 Generating Risk-Neutral Scenarios

22 Horizon := rngParam . Cells (2 , 1) . Value , _

23 TimeStep := rngParam . Cells (3 , 1) . Value , _
24 NumPath := rngParam . Cells (4 , 1) . Value , _
25 Seed := rngParam . Cells (5 , 1) . Value , _
26 Precision := rngParam . Cells (6 , 1) . Value , _
27 FR := aFR
28
29 PrintInfo
30 End Sub

The test sub procedure requires two named ranges, which are shown in
Table 9.1. Since the number of forward rates varies, we only refer to the first
forward rate. We assume that the forward rates are organized continuously
in a column. In Lines 12–13, we calculate the first row and the last row
of the forward rates. In Lines 17–19, we copy the forward rates from the
worksheet to a procedure-level array named aFR. In Lines 21–27, we call
the sub procedure Initialize with appropriate arguments. Line 29 calls
the sub procedure PrintInfo to display the contents of the module-level
variables. Executing the above code, we see the following output in the
Immediate window:
1 Horizong : 30 Y
2 TimeStep : 1 M
3 NumPath : 100
4 Seed : 2016
5 Precision : 6
6 Number of Forward Rates : 360
7 0.00689 0.00689 0.00689 0.00689 0.00689 0.00689

Comparing the above output and Figure 9.1, we see that the sub procedure
Initialize works well.
Table 9.1: Range Names Defined in the Interface of the Risk-Neutral Scenario
Generator

Name Refers To
FirstFR =Main!$A$2
Param =Main!$D$2:$D$7

Note that the forward rates depend on the time step. In the above test,
we used a monthly time step. The forward rates are monthly forward rates
obtained from the yield curve bootstrapper (see Chapter 8). If the time step
 Implementation 191

is “1Y,” we need to input annual forward rates, which can also be obtained
from the yield curve bootstrapper by specifying the “Output Type” to “1Y.”
To generate scenarios, we need first to figure out the number of time
steps and the ∆ required by Equation (9.10). To do that, we need a procedure
to extract the number of periods and the period string from input strings,
such as “30Y” and “1M.” For this purpose, we can use the following sub
procedure:
1 Sub ExtractPeriod ( ByVal sIn As String , ByRef iNum As
Integer , ByRef sPeriod As String )
2 Dim iLen As Integer
3 Dim sNum As String
4 iLen = Len ( sIn )
5
6 If iLen < 2 Then
7 Err . Raise Number :=1003 , Source := " MHoliday .
ExtractPeriod " , Description := " Invalid
period string "
8 End If
9 sPeriod = Right ( sIn , 1)
10 sNum = Left ( sIn , iLen - 1)
11 If IsNumeric ( sNum ) Then
12 iNum = CInt ( sNum )
13 Else
14 Err . Raise Number :=1003 , Source := " MHoliday .
ExtractPeriod " , Description := " Invalid
period string "
15 End If
16 End Sub

The above sub procedure was first implemented in the yield curve boot-
strapper. For an explanation of this sub procedure, readers are referred to
Chapter 8.
Now, we can calculate the number of time steps using the following
function:
1 Private Function CalNumTimeStep ( ByVal Horizon As
String , ByVal TimeStep As String ) As Long
2 Dim lRet As Long
3
4 Dim iHorizonNum As Integer
5 Dim sHorizonPeriod As String
6 Dim iStepNum As Integer
7 Dim sStepPeriod As String
 192 Generating Risk-Neutral Scenarios

8
9 ExtractPeriod sIn := Horizon , iNum := iHorizonNum ,
sPeriod := sHorizonPeriod
10 ExtractPeriod sIn := TimeStep , iNum := iStepNum ,
sPeriod := sStepPeriod
11
12 If Not StrComp ( sHorizonPeriod , " Y " , vbTextCompare
) = 0 Then
13 Err . Raise Number :=1001 , Source := " MGenerator .
CalNumTimeStep " , _
14 Description := " Horizon should be specified
in years . "
15 End If
16
17 If StrComp ( sStepPeriod , " Y " , vbTextCompare ) = 0
Then
18 lRet = iHorizonNum / iStepNum
19 ElseIf StrComp ( sStepPeriod , " M " , vbTextCompare ) =
0 Then
20 lRet = 12 * iHorizonNum / iStepNum
21 Else
22 Err . Raise Number :=1002 , Source := " MGenerator .
CalNumTimeStep " , _
23 Description := " Unknown combination of
period strings . "
24 End If
25
26 CalNumTimeStep = lRet
27 End Function

The function CalNumTimeStep is a private function and requires two argu-

ments: a horizon string and a time step string. In the current version, the
horizon can only be specified in years. The time step can be specified in
months or years. The function will raise an error in other cases.
We can test the function CalNumTimeStep as follows:
1 Sub T es tC alN um Ti meS te p ()
2 Debug . Print CalNumTimeStep ( " 30 Y " , "1Y")
3 Debug . Print CalNumTimeStep ( " 30 Y " , "1M")
4 Debug . Print CalNumTimeStep ( " 10 Y " , "3M")
5 Debug . Print CalNumTimeStep ( " 30 Y " , "7M")
6 End Sub

Executing the above sub procedure gives

 Implementation 193

1 30
2 360
3 40
4 51

From the output, we see that the function calculates the number of time
steps correctly.
To calculate the time step ∆, we use the following function:
1 Private Function CalDelta ( ByVal TimeStep As String )
As Double
2 Dim dRet As Double
3
4 Dim iStepNum As Integer
5 Dim sStepPeriod As String
6
7 ExtractPeriod sIn := TimeStep , iNum := iStepNum ,
sPeriod := sStepPeriod
8
9 If StrComp ( sStepPeriod , " Y " , vbTextCompare ) = 0
Then
10 dRet = iStepNum
11 ElseIf StrComp ( sStepPeriod , " M " , vbTextCompare ) =
0 Then
12 dRet = iStepNum / 12#
13 Else
14 Err . Raise Number :=1004 , Source := " MGenerator .
CalDelta " , _
15 Description := " Unknown time step string . "
16 End If
17
18 CalDelta = dRet
19 End Function

The function CalDelta is also a private function that can only be used
within this module. This version can only handle time steps specified in
months or years. In Line 12, the sign # after 12 indicates that 12 is treated as
a double-type number rather than an integer.
To test the function CalDelta, we use the following sub procedure:
1 Sub TestCalDelta ()
2 Debug . Print CalDelta ( " 1 Y " )
3 Debug . Print CalDelta ( " 1 M " )
4 Debug . Print CalDelta ( " 3 M " )
5 End Sub
 194 Generating Risk-Neutral Scenarios

Executing the above test sub procedure gives the following output:
1 1
2 8.33333333333333 E -02
3 0.25

The output indicates that the function works well.

Now, we can implement the function for generating risk-neutral scenar-
ios. We implement this function as follows:
1 Function Generate () As Variant
2 Dim lNumTimeStep As Long
3 Dim dDelta As Double
4
5 lNumTimeStep = CalNumTimeStep ( sHorizon , sTimeStep
)
6
7 If UBound ( aFR ) < lNumTimeStep Then
8 Err . Raise Number :=1005 , Source := " MGenerator .
Generate " , _
9 Description := " There are not enough
forward rates . "
10 End If
11
12 dDelta = CalDelta ( sTimeStep )
13
14 Dim aSce () As Double
15 ReDim aSce (1 To lNumPath , 1 To lNumTimeStep )
16
17 Rnd -1
18 Randomize lSeed
19
20 Dim i As Long
21 Dim j As Long
22 Dim z As Double
23 Dim dTmp As Double
24 For i = 1 To lNumPath
25 For j = 1 To lNumTimeStep
26 z = Wo rksheetFu nction . NormInv ( Rnd , 0 , 1)
27 dTmp = Workshee tFunctio n . Ln (1 + dDelta *
aFR ( j ) ) - 0.5 * dDelta * dSigma *
dSigma + _
28 dSigma * Sqr ( dDelta ) * z
 Implementation 195

29 aSce (i , j ) = Round ( Exp ( dTmp ) , iPrecision )

30 Next j
31 Next i
32
33 Generate = aSce
34 End Function

The function Generate creates a matrix of scenarios and returns it to the

caller as a variant. In Lines 7–10, the function checks that there are enough
forward rates. In Line 15, the function redimension a two-dimensional array
to store the matrix of scenarios. In Lines 24–31, the function generates the
scenarios according to the model given in Equation (9.10). We do not need
to write a sub procedure to test this function, because we can test it using
the button, which will be implemented in the module MInterface.

9.4.2 MInterface Module

The MInterface module contains VBA code for passing the parameters to
the module MGenerator and writing the resulting scenarios to the “Scenar-
ios” sheet.
In this module, we only need to implement a sub procedure for the
button. We can implement the sub procedure as follows:
1 Sub Button1_Click ()
2 Dim rngParam As Range
3 Dim rngFirstFR As Range
4
5 Set rngParam = Sheet1 . Range ( " Param " )
6 Set rngFirstFR = Sheet1 . Range ( " FirstFR " )
7
8 Dim aFR () As Double
9 Dim lFirstRow As Long
10 Dim lLastRow As Long
11
12 lFirstRow = rngFirstFR . Cells (1 , 1) . Row
13 lLastRow = rngFirstFR . Cells (1 , 1) . End ( xlDown ) . Row
14
15 ReDim aFR (1 To lLastRow - lFirstRow + 1)
16 Dim i As Long
17 For i = 1 To lLastRow - lFirstRow + 1
18 aFR ( i ) = rngFirstFR . Cells (1 , 1) . Offset ( i - 1 ,
0) . Value
19 Next i
20
 196 Generating Risk-Neutral Scenarios

21 MGenerator . Initialize Sigma := rngParam . Cells (1 , 1)

. Value , _
22 Horizon := rngParam . Cells (2 , 1) . Value , _
23 TimeStep := rngParam . Cells (3 , 1) . Value , _
24 NumPath := rngParam . Cells (4 , 1) . Value , _
25 Seed := rngParam . Cells (5 , 1) . Value , _
26 Precision := rngParam . Cells (6 , 1) . Value , _
27 FR := aFR
28
29 Dim vSce As Variant
30 vSce = MGenerator . Generate
31
32 ' write scenarios to Sheet2
33 Sheet2 . Cells . ClearContents
34 Dim j As Long
35 For i = 1 To UBound ( vSce , 1)
36 For j = 1 To UBound ( vSce , 2)
37 Sheet2 . Cells (i , j ) . Value = vSce (i , j )
38 Next j
39 Next i
40
41 End Sub

The code in the first part of this sub procedure is similar to that of the sub
procedure TestInitialize implemented in the module MGenerator. In
Lines 21–30, the sub procedure calls Initialize and Generate to generate
scenarios. In Lines 35–39, it writes the scenarios to Sheet2.
We assign the sub procedure Button1_Click to the button in the inter-
face. If we use the parameters and forward rates shown in Figure 9.1, we
will get the scenarios shown in Figure 9.2.

g
Exercise 9.1. The volatility is a positive real number. However, we did not
check the positivity of the volatility parameter in our implementation. Add
code to the sub procedure Initialize to check the positivity of the volatil-
ity.
Exercise 9.2. The scenarios are saved to the worksheet as accumulation
factors A i j for i = 1, 2, . . . , n and j = 1, 2, . . . , m, where n is the number of
paths, and m is the number of time steps. Add code to the sub procedure
Button1_Click to do the following:
• For each j = 1, 2, . . . , m, calculate the mean µ j and the standard devia-
tion σ j of ln A 1 j , ln A 2 j , . . ., ln A n j , that is, calculate the mean and the
Summary 197

Figure 9.2: First few rows and columns of the risk-neutral scenarios

standard deviation of the continuous returns for each column of the

scenarios.
µ µ µ
• Write the annualized means ∆1 , ∆2 , . . . , ∆m to the row that is two rows
below the last row of the scenarios, and write the annualized volatilities
σ σ σ
p 1 , p 2 , . . . , pm to the row that is three rows below the last row of the
∆ ∆ ∆
scenarios.
Exercise 9.3. Generating a large number of paths (e.g., n = 5000) takes
some time. Add code to the function Generate in the module MGenerator
to display the following message, “Generating path i of n” in the status bar,
where i is the index of the path and n is the total number of paths. (Hint:
use Application.StatusBar and DoEvents.)
h

9.5 Summary
In this chapter, we implemented a simple risk-neutral scenario generator
based on the Black–Scholes model. The scenario generator generates risk-
198 Generating Risk-Neutral Scenarios

neutral scenarios for a single equity security. To generate scenarios for

multiple equity securities with a correlation structure, readers are referred
to (Carmona and Durrleman, 2005) and (Gan, 2015b). For more information
about risk-neutral scenarios, readers are referred to (Sheldon and Smith,
2004).
10
Valuing a GMDB

In this chapter, we implement a simple program to calculate the fair market

value and Greeks of a guaranteed minimum death benefit (GMDB) embed-
ded in a variable annuity contract. The interface of the program is shown in
Figure 10.1.

Figure 10.1: Interface of the GMDB valuation program

The inputs to the program include contract information, the mortality

table, forward rates, and risk-neutral scenarios. The outputs include the
fair market value, the dollar delta, and the dollar rho. The forward rates are
obtained from the yield curve bootstrapper developed in Chapter 8. The
risk-neutral scenarios are generated by the scenario generator developed
in Chapter 9. We generate three sets of risk-neutral scenarios based on the
base yield curve, the yield curve shifted 10bps 1 up, and the yield curve
1 Here, “bps” means “basis points.” A basis point is equal to 0.0001 or 0.01%.

199
200 Valuing a GMDB

shifted 10bps down, respectively. The three sets of scenarios are saved in
the worksheets Base, AllUp, and AllDn, respectively.

10.1 Introduction
A variable annuity is a long-term life insurance contract between a policy-
holder and an insurance company under which the policyholder agrees to
make a lump-sum purchase payment or a series of purchase payments to
the insurance company, and, in return, the insurance company agrees to
make benefit payments to the policyholder, beginning immediately or at a
future date (The Geneva Association Report, 2013). Variable annuities are
known as unit-linked products in Europe and segregated funds in Canada.
Under a variable annuity contract, the policyholder can allocate the
account value to several subaccounts of the insurance company’s separate
account. The separate account of an insurance company is managed sepa-
rately from the insurance company’s general account and is protected from
claims against the insurance company in the event of insolvency. Subac-
counts are similar to mutual funds and allow the policyholder to participate
in the stock market.
One main feature of variable annuities is that they contain guarantees,
which include GMDB and guaranteed minimum living benefits (GMLB).
The death benefit is paid to the beneficiary of a policyholder upon the death
of the policyholder. There are several types of GMLB: guaranteed minimum
accumulation benefits (GMAB), guaranteed minimum maturity benefits
(GMMB), guaranteed minimum income benefits (GMIB), and guaranteed
minimum withdrawal benefits (GMWB). The GMAB rider provides a guaran-
teed minimum survival benefit during a specified window after a specified
waiting period, which is usually 10 years. The specified widows typically
begins on an anniversary date and remains open for 30 days (Brown et al.,
2002). The GMMB rider is similar to the GMAB rider, but the guaranteed
amount is only available at the maturity of the contract. The GMIB rider
gives a policyholder annuitized payments during a specified window after a
specified waiting period. The annuitized payments are equal to the greater
of payments determined by applying current fixed-annuitization rates to
the payments determined by applying guaranteed fixed-annuitization rates,
which are set at issue. The GMWB rider gives a policyholder the right to
Life Table Construction 201

withdrawal a specified amount during the life of the contract until the initial
investment is recovered.
Such guarantees are financial guarantees and cannot be addressed
adequately by traditional actuarial approaches (Boyle and Hardy, 1997).
Dynamic hedging is a popular approach to managing the financial risk as-
sociated with the guarantees and is adopted by some insurance companies.
However, dynamic hedging requires calculating the Greeks (i.e., sensitivities
of the value of guarantees on market factors) of the guarantees. Due to the
complexity of guarantees, the Monte Carlo simulation method is usually
used to value the guarantees. In the Monte Carlo simulation method, the
cash flows of the guarantees are projected along many risk-neutral paths
(see Chapter 9), and the value of the guarantees is calculated as the average
present value of the cash flows.
The cash flows of different types of guarantees are projected differently,
depending on the contract specifications. In this chapter, we introduce how
to value a GMDB using the Monte Carlo simulation method.

10.2 Life Table Construction

Valuing a GMDB embedded in a variable annuity requires calculating
mortalities. We use the US life tables 1999–2001 downloaded from http:
//mort.soa.org/. In these life tables, the mortality rates q x are given for
integer ages x from 0 to 109. Since valuing a GMDB involves fraction ages,
we need to interpolate the mortalities of fraction ages from integer ages.
We use a linear method to interpolate the mortalities (Gupta and
Varga, 2002). Let `n be the number of persons surviving to age n, where
n = 0, 1, . . . , 109. Then,
n
`n = (1 − q i ).
Y
(10.1)
i =0

Let x be an arbitrary age. Then, we interpolate `x as follows:

`x = (1 − {x})`[x] + {x}`[x]+1 , (10.2)

where [x] and {x} represent the integer part and the fractional part of x,
respectively.
Under the linear interpolation method, t p x , the probability that a per-
202 Valuing a GMDB

son aged x will survive to the age of x + t , is calculated as

`x+t (1 − {x + t })`[x+t ] + {x + t }`[x+t ]+1

t px = = . (10.3)
`x (1 − {x})`[x] + {x}`[x]+1

The probability that a person aged x will die before the age of x + t is calcu-
lated as
t qx = 1 − t p x . (10.4)

10.3 GMDB Valuation

Let F i j (i = 1, 2, . . . , n, j = 1, 2, . . . , m) denote the risk-neutral scenarios for the
underlying fund that are generated by the risk-neutral scenario generator
from Chapter 9. Here, n and m denote the number of paths and the number
of time steps, respectively. In general, the larger the n and the m, the more
accurate the results. However, the runtime increases when n and m increase.
When using the Monte Carlo simulation method, there is a tradeoff between
accuracy and runtime. In our example, we use n = 1, 000 and m = 360
months.
For a variable annuity contract with a GMDB, we use the following
notation to denote the cash flows of the GMDB that occur within the period
(t j −1 , t j ] along the i th risk-neutral path:

DB i j denotes the claim of death benefit when the policyholder dies within
the period (t j −1 , t j ].

RC i j denotes the risk charge for providing the guarantees.

In addition to the above notation, we use the following notation to denote

the real and paper account values:

GD i j denotes the guaranteed death benefit.

A i j denotes the account value.

We use the following notation to denote various fees:

φM E denotes the annualized mortality and expense (M&E) fee of the con-
tract. This fee is charged as a percentage of the fund value at the begin-
ning of each anniversary.
GMDB Valuation 203

φI E denotes the initial expense used to cover the sales commission. This
fee is charged once at the inception of the policy.

φOE denotes the ongoing expense that is used to cover the administrative
cost. This expense is a fixed dollar amount charged at the beginning of
each anniversary.

φF denotes the annualized fund management fee of the underlying fund.

Usually this fee is charged at the end of each period and goes to the fund
manager rather than the insurance company.

The M&E fee includes the guarantee fee of the GMDB.

We can project the cash flows in a way that is similar to the way used by
Bauer et al. (2008). For the sake of simplicity, we assume there is one type
of event that will occur during the term of the contract: the policyholder
passes away. We consider two GMDB riders:

DBRP contains a return-of-premium GMDB rider.

DBSU contains a step-up GMDB rider. The death benefit is reset to the
account value annually if the latter is higher.

Suppose that the fees are charged from the account values at the end
of every period. For j = 0, 1, . . . , m − 1, the cash flows from t j to t j +1 are
projected as follows:

• The account value evolves as follows:

A i j F i , j +1 1 − ∆φF 1 − φM E − φOE , if t j +1 is an anniversary
½ ¡ ¢¡ ¢
A i , j +1 =
A i j F i , j +1 1 − ∆φF ,
¡ ¢
if otherwise,
(10.5)
where ∆ = t j +1 − t j . Here, we assume that the fund management fees
are deducted at the end of each period. Here, we consider only single-
premium products. The fund value at inception is calculated as follows:

A i ,0 = Premium × (1 − φI E )(1 − φM E ) − φOE .

• The risk charges are projected as

A i j F i , j +1 1 − ∆φF φM E , if t j +1 is an anniversary
½ ¡ ¢
RC i , j +1 =
0, if otherwise.
(10.6)
The risk charge at inception is calculated as

RC i ,0 = Premium × (1 − φI E )φM E .
204 Valuing a GMDB

• If the contract contains a return-of-premium GMDB rider, the death

benefit is equal to the initial premium:

GD i , j +1 = GD i , j , GD i ,0 = Premium. (10.7)

• If the contract contains a step-up GMDB rider, we have

½
max{GD i , j , A i , j +1 }, if t j +1 is an anniversary.
GD i , j +1 =
GD i , j , if t j +1 is not an anniversary,
(10.8)
with GD i ,0 = Premium.
• If the policyholder dies within the period (t j , t j +1 ], then the claim of
death benefit is projected as

DB i , j +1 = max{0,GD i , j +1 − A i , j +1 }. (10.9)

• After the maturity of the contract, all the state variables are set to zero.

Once we have all the cash flows, we can calculate the fair market values
of the riders as follows:
1X n X m
V0 = ( j −1)∆ p x 0 · ∆ q x 0 +( j −1)∆ L j DB i , j d j , (10.10)
n i =1 j =1

where x 0 is the age of the policyholder, p is the survival probability, q is

the probability of death, r L is the annual lapse rate, L j is the lapse rate
calculated as
(1 − ∆r L ) j , if t j < T ,
½
Lj =
0, if t j ≥ T ,
and d j is the discount factor defined as
à !
j
X
d j = exp −∆ fl .
l =1

Here, r L is the annual lapse rate, T is the maturity of the policy, and
f 1 , f 2 , . . . , f m are forward rates. The risk charge value is calculated as

1X n X m
RC 0 = j ∆ p x 0 RC i , j L j d j . (10.11)
n i =1 j =1

The fair market value is then calculated as

F MV0 = V0 − RC 0 . (10.12)
Greek Calculation 205

10.4 Greek Calculation

Hedging the financial risks associated with the guarantees usually requires
calculating the Greeks of the guarantees. The Greeks are the sensitivities of
the fair market value of the guarantees to market variables, such as equity
and the interest rate. The sensitivity to equity is referred to as the delta,
while the sensitivity to the interest rate is referred to as rho.
There are several ways to calculate the Greeks (Cathcart et al., 2015; Fu,
2015). Here, we use the finite difference method to calculate the Greeks. To
calculate the delta of the GMDB, we shock the initial account values A i ,0 up
1% and calculate the fair market value F MVu of the GMDB. Then we shock
the initial account values A i ,0 down 1% and calculate the fair market value
F MVd of the GMDB. Then, we calculate the dollar delta as
F MVu − F MVd
$Delta = .
0.02
The dollar delta approximates the dollar value change of the fair market
value of the guarantee when the equity increases 1%.
Calculating the rho is a little bit tedious. First, we shock the swap rates up
10bps (i.e., 0.001) and bootstrap the shocked swap rates to get the forward
rates, which are used to generate risk-neutral scenarios. Then, we calculate
the fair market value F MVu of the guarantee based on the risk-neutral
scenarios. Second, we shock the swap rates down 10bps (i.e., −0.001) and
bootstrap the shocked swap rates to get the forward rates. We generate risk-
neutral scenarios and calculate the fair market value F MVd of the guarantee
again based on the new scenarios. The dollar rho is calculated as
F MVu − F MVd
$Rho = .
20
The dollar rho approximates the dollar value change of the fair market value
of the guarantee when the yield curve shifts up 1bps.

10.5 Implementation
We organize the Visual Basic for Applications (VBA) code of the valuation
program into three modules: the MLifeTable module, the MGMDB module,
and the MInterface module. A list of range names used by the program is
given in Table 10.1.
 206 Valuing a GMDB

Table 10.1: Range Names Used by the GMDB Valuation Program

Name Refers To
FRAllDn =Main!$L$3:$L$362
FRAllUp =Main!$K$3:$K$362
FRBase =Main!$J$3:$J$362
MTFemale =Main!$G$3:$G$112
MTMale =Main!$F$3:$F$112
Param =Main!$B$3:$B$13

10.5.1 MLifeTable Moduel

The MLifeTable module contains code for interpolating the life table. The
declarations section of this module contains the following code:
1 Option Explicit
2
3 Private aLT () As Double
4 Private iMaxAge As Integer

In this module, we declared two private module-level variables. The first

variable is used to store the mortality rates shown in Figure 10.1. The second
variable is used to store the maximum age available in the mortality table.
We need a procedure to create the life table from the mortality rates. We
implement the procedure as follows:
1 Sub Initialize ( ByRef rngMTMale As Range , ByRef
rngMTFemale As Range )
2 Dim nRow As Integer
3
4 nRow = rngMTMale . Rows . Count
5 If nRow <> rngMTFemale . Rows . Count Then
6 Err . Raise Number :=1001 , Source := " MLifeTable .
Initialize " , Description := " Input ranges
have different size . "
7 End If
8
9 iMaxAge = nRow - 1
10 ReDim aLT (0 To nRow - 1 , 0 To 1)
11
12 ' convert mortality rates to survival
probabilities
13 Dim i As Integer
 Implementation 207

14 aLT (0 , 0) = 1 - rngMTMale . Cells ( i + 1 , 1) . Value

15 aLT (0 , 1) = 1 - rngMTFemale . Cells ( i + 1 , 1) . Value
16 For i = 1 To nRow - 1
17 aLT (i , 0) = aLT ( i - 1 , 0) * (1 - rngMTMale .
Cells ( i + 1 , 1) . Value )
18 aLT (i , 1) = aLT ( i - 1 , 1) * (1 - rngMTFemale .
Cells ( i + 1 , 1) . Value )
19 Next i
20 End Sub

The sub procedure requires two arguments: a range containing male mor-
tality rates, and a range containing female mortality rates. In Lines 4–7, we
check whether the two input ranges have the same number of rows. If the
two ranges do not have the same number of rows, an error will be raised.
The maximum available age in the mortality rates is the number of rows
minus one, as the mortality rates start from age 0. In Line 10, we redimen-
sion the array aLT so that it can hold all the survival probabilities, which are
calculated in Lines 14–19.

g
Exercise 10.1. Write a sub procedure named PrintInfo to display the
values of the module-level variables in the Immediate window.
Exercise 10.2. Write a sub procedure named TestInitialize to test the
sub procedure Initialize. In particular, you first call the sub procedure
Initialize and then call PrintInfo (see Exercise 10.1) to show the values
of the module-level variables.
h

We need to implement some functions to calculate the survival prob-

ability for fractional ages. We first implement the interpolation method
described in Section 10.2 as follows:
1 Function Interpolate ( ByVal x As Double , ByVal g As
Integer ) As Double
2 Dim iL As Integer
3 Dim iU As Integer
4 Dim i As Integer
5
6 If x < 0 Or x > iMaxAge Then
7 Err . Raise Number :=1002 , Source := " MLifeTable .
Interpolate " , Description := " Age x is out
of range . "
 208 Valuing a GMDB

8 End If
9
10 iL = WorksheetFuncti on . Floor (x , 1)
11 iU = iL + 1
12
13 Interpolate = ( iU - x ) * aLT ( iL , g ) + ( x - iL ) *
aLT ( iU , g )
14 End Function

In Lines 6–8 of the above code, we first check that the input age is not
less than zero or larger than the maximum age. Since we use the linear
interpolation method, we cannot extrapolate the survival probabilities.
In Line 10, we get the integer part of the input age. The code in Line 13
implements the linear interpolation method based on Equation (10.2).
To test the function Interpolate, we use the following sub procedure:
1 Sub TestInterpolate ()
2 Dim rngMTMale As Range
3 Dim rngMTFemale As Range
4
5 Set rngMTMale = Range ( " MTMale " )
6 Set rngMTFemale = Range ( " MTFemale " )
7
8 Initialize rngMTMale := rngMTMale , rngMTFemale :=
rngMTFemale
9
10 Debug . Print Interpolate (0 , 0)
11 Debug . Print Interpolate (35.5 , 0)
12 Debug . Print Interpolate (35.5 , 1)
13 Debug . Print Interpolate (99.1 , 0)
14 End Sub

In the above test procedure, we used two named ranges defined in Table
10.1. We calculated the survival probabilities for a male aged 0, 35.5, and
99.1. We also calculated the survival probability for a female aged 35.5.
Executing the above sub procedure gives the following output:
1 0.99239
2 0.961165379370226
3 0.979048981352752
4 6.21200647276638 E -03

From the output, we see that a female aged 35.5 has a higher survival prob-
ability than a male of the same age.
 Implementation 209

Now, we can implement the functions for calculating t p x and t q x for

arbitrary t and x. We implement the two functions as follows:
1 Function tpx ( ByVal x As Double , ByVal t As Double ,
ByVal g As Integer ) As Double
2 tpx = Interpolate ( x + t , g ) / Interpolate (x , g )
3 End Function
4
5 Function tqx ( ByVal x As Double , ByVal t As Double ,
ByVal g As Integer ) As Double
6 tqx = 1 - tpx (x , t , g )
7 End Function

In the above code, we implemented Equations (10.3) and (10.4) straightfor-

wardly.

Exercise 10.3. Write a sub procedure named Testtpx to test the function
tpx by calculating the following probabilities: 1 p 35 and 3.6 p 35.2 for a male,
and 3.6 p 35.2 for a female.

10.5.2 MGMDB Moduel

The MGMDB module implements the GMDB valuation model described in
Section 10.3. The declarations section of this module contains the following
code:
1 Option Explicit
2
3 Private iGender As Integer
4 Private datDOB As Date
5 Private dPremium As Double
6 Private dMEFee As Double
7 Private dFundFee As Double
8 Private dIE As Double
9 Private dOE As Double
10 Private iTerm As Integer
11 Private sFeature As String
12 Private dLapseRate As Double
13 Private datValDate As Date
14 Private iNumSce As Integer
 210 Valuing a GMDB

15 Private rngFR As Range

16 Private shtSce As Worksheet
17
18 Private Const dTimeStep As Double = 1# / 12
19 Private ap (0 To 360) As Double ' ap ( i ) = x survives
to ( i * dTimeStep )
20 Private aq (1 To 360) As Double ' aq ( i ) = x dies in ((
i -1) * dTimeStep , i * dTimeStep )
21 Private ad (0 To 360) As Double
22 Private aL (0 To 360) As Double
23
24 Private aA (0 To 360) As Double
25 Private aRC (0 To 360) As Double
26 Private aGD (0 To 360) As Double
27 Private aDB (0 To 360) As Double

In Lines 3–14 of the above code, we declared some variables to hold the
parameters of the contract. In Lines 15–16, we declared two variables to hold
the references to the range of forward rates and the worksheet of scenarios.
In Line 18, we defined a constant for the time step. In this program, we only
consider monthly time steps. In Lines 19–22, we defined some arrays to
hold the survival probabilities, the probabilities of dying within a period,
the discount factors, and the survivorship. In Lines 24–27, we declared some
variables to hold temporary values. The names of the variables match those
used in the formulas described in Section 10.3.
Valuing a GMDB requires two sets of information: the contract and the
scenarios. Hence, we implement two procedures to initialize the module-
level variables. We first implement the contract initialization procedure as
follows:
1 Sub InitializeContr ac t ( _
2 ByVal Gender As Integer , _
3 ByVal DOB As Date , _
4 ByVal Premium As Double , _
5 ByVal MEFee As Double , _
6 ByVal FundFee As Double , _
7 ByVal IE As Double , _
8 ByVal OE As Double , _
9 ByVal Term As Integer , _
10 ByVal Feature As String , _
11 ByVal LapseRate As Double , _
12 ByVal ValDate As Date , _
13 ByVal NumSce As Integer )
14 iGender = Gender
 Implementation 211

15 datDOB = DOB
16 dPremium = Premium
17 dMEFee = MEFee
18 dFundFee = FundFee
19 dIE = IE
20 dOE = OE
21 iTerm = Term
22 sFeature = Feature
23 dLapseRate = LapseRate
24 datValDate = ValDate
25 iNumSce = NumSce
26
27 Dim dx As Double
28 dx = DateDiff ( " d " , datDOB , datValDate ) / 365#
29 Dim i As Integer
30 ap (0) = 1
31 aL (0) = 1
32 For i = 1 To 360
33 ap ( i ) = MLifeTable . tpx ( dx , i * dTimeStep ,
iGender )
34 aq ( i ) = ap ( i - 1) * MLifeTable . tqx ( dx + ( i -
1) * dTimeStep , dTimeStep , iGender )
35 If i < iTerm * 12 Then
36 aL ( i ) = aL ( i - 1) * (1 - dTimeStep *
dLapseRate )
37 Else
38 aL ( i ) = 0
39 End If
40 Next i
41 End Sub

The sub procedure InitializeContract requires quite a few arguments,

which correspond to the parameters of the contract and the valuation
model. In Lines 14–25, we initialize the variables using the input values. In
Line 28, we calculate the age of the contract holder by dividing the number
of days between the date of birth and the valuation date by 365. In Lines
32-40, we calculate the survival probabilities, the probability of dying within
each period, and the survivorship.
The second initialization procedure is simple and is implemented as
follows:
1 Sub I ni ti ali ze Sc ena ri o ( _
2 ByRef Sce As Worksheet , _
3 ByRef FR As Range )
 212 Valuing a GMDB

4 Set shtSce = Sce

5 Set rngFR = FR
6
7 Dim i As Integer
8 ad (0) = 1
9 For i = 1 To 360
10 ad ( i ) = ad ( i - 1) * Math . Exp ( - dTimeStep *
rngFR . Cells (i , 1) )
11 Next i
12 End Sub

This sub procedure requires only two arguments: the range of forward rates
and the worksheet of scenarios. Since the arguments are objects, we pass
the arguments by reference. In this sub procedure, we also calculate the
discount factors of all periods that can be used later.
Once we have all the contract information and the scenarios, we can
project the cash flows of the guarantee and calculate the present values. We
first implement a function to project the cash flows along a single path. We
implement this function as follows:
1 Function Project ( ByVal dShock As Double , ByVal ind As
Integer ) As Double
2 Dim i As Integer
3
4 aA (0) = (1 + dShock ) * ( dPremium * (1 - dIE ) * (1
- dMEFee ) - dOE )
5 aRC (0) = dPremium * (1 - dIE ) * dMEFee
6 aGD (0) = dPremium
7
8 For i = 1 To 360
9 If i Mod 12 = 0 Then
10 aA ( i ) = aA ( i - 1) * shtSce . Cells ( ind , i ) .
Value * _
11 (1 - dTimeStep * dFundFee ) * (1 -
dMEFee ) - dOE
12 aRC ( i ) = aA ( i - 1) * shtSce . Cells ( ind , i )
. Value * _
13 (1 - dTimeStep * dFundFee ) * dMEFee
14 If StrComp ( sFeature , " RP " , vbTextCompare )
= 0 Then
15 aGD ( i ) = aGD ( i - 1)
16 Else
17 aGD ( i ) = Worksh eetFunct ion . Max ( aGD ( i
- 1) , aA ( i ) )
 Implementation 213

18 End If
19 Else
20 aA ( i ) = aA ( i - 1) * shtSce . Cells ( ind , i ) .
Value * _
21 (1 - dTimeStep * dFundFee )
22 aRC ( i ) = 0
23 aGD ( i ) = aGD ( i - 1)
24 End If
25 aDB ( i ) = Worksh eetFunct ion . Max (0 , aGD ( i ) - aA
(i))
26 Next i
27
28 Dim dPVBenefit As Double
29 Dim dPVCharge As Double
30 For i = 1 To 360
31 dPVBenefit = dPVBenefit + aq ( i ) * aL ( i ) * aDB
( i ) * ad ( i )
32 dPVCharge = dPVCharge + ap ( i ) * aL ( i ) * aRC ( i
) * ad ( i )
33 Next i
34 Project = dPVBenefit - dPVCharge
35 End Function

The function Project requires two arguments: an equity shock amount

and an index of scenario. The equity shock argument is used to calculate the
dollar delta described in Section 10.4. In Lines 4–6, we calculate the account
value, the risk charge, and the guaranteed benefit at time 0. In Lines 8–26,
we project the cash flows for 360 months. In Lines 10–18, we project the
cash flows when the time is an anniversary. Lines 20–23 project the cash
flows when the time is not an anniversary. The death benefit for each month
is calculated in Line 25. In Lines 30–34, we calculate the present value of the
cash flows.

Exercise 10.4. Write a sub procedure named TestProject to test the func-
tion Project by doing the following:

• Add a new worksheet named Temp if it does not exist.

• Initialize the modules MLifeTable and MGMDB.
• Call the function Project with dShock=0 and ind=1.
• Write all the module-level arrays to the worksheet Temp, such that each
 214 Valuing a GMDB

row contains an array, and the name of the array appears in the first
column.

Now, we implement the function to calculate the fair market value of

the GMDB as follows:
1 Function FMV ( ByVal dShock As Double ) As Double
2 Dim dSum As Double
3 Dim i As Integer
4
5 For i = 1 To iNumSce
6 dSum = dSum + Project ( dShock , i )
7 Next i
8
9 FMV = dSum / iNumSce
10 End Function

In the above function, we calculate the average present value by calling the
function Project for the specified number of scenarios.

10.5.3 MInterface Module

The MInterface module contains the VBA macro for calling various func-
tions to calculate the fair market value, the dollar delta, and dollar rho of a
GMDB embedded in a variable annuity contract. This module also contains
code to write the results to the worksheet. We implement the module as
follows:
1 Option Explicit
2
3 Sub Button1_Click ()
4 Dim dFMVBase As Double
5 Dim dFMVEU As Double
6 Dim dFMVED As Double
7 Dim dFMVIU As Double
8 Dim DFMVID As Double
9
10 MLifeTable . Initialize rngMTMale := Range ( " MTMale " ) ,
rngMTFemale := Range ( " MTFemale " )
11 With Range ( " Param " )
12 MGMDB . In iti al iz eCo nt ra ct _
13 Gender :=. Cells (1 , 1) . Value , _
14 DOB :=. Cells (2 , 1) . Value , _
 Implementation 215

15 Premium :=. Cells (3 , 1) . Value , _

16 MEFee :=. Cells (4 , 1) . Value , _
17 FundFee :=. Cells (5 , 1) . Value , _
18 IE :=. Cells (6 , 1) . Value , _
19 OE :=. Cells (7 , 1) . Value , _
20 Term :=. Cells (8 , 1) . Value , _
21 Feature :=. Cells (9 , 1) . Value , _
22 LapseRate :=. Cells (10 , 1) . Value , _
23 ValDate :=. Cells (11 , 1) . Value , _
24 NumSce :=. Cells (12 , 1) . Value
25 End With
26
27 MGMDB . Initializ eSc en ar io Sce := Sheets ( " Base " ) , FR
:= Range ( " FRBase " )
28 dFMVBase = MGMDB . FMV (0)
29 dFMVEU = MGMDB . FMV (0.01)
30 dFMVED = MGMDB . FMV ( -0.01)
31
32 MGMDB . Initializ eSc en ar io Sce := Sheets ( " AllUp " ) , FR
:= Range ( " FRAllUp " )
33 dFMVIU = MGMDB . FMV (0)
34
35 MGMDB . Initializ eSc en ar io Sce := Sheets ( " AllDn " ) , FR
:= Range ( " FRAllDn " )
36 DFMVID = MGMDB . FMV (0)
37
38 Sheet1 . Range ( " O3 " ) . Value = dFMVBase
39 Sheet1 . Range ( " O4 " ) . Value = ( dFMVEU - dFMVED ) /
0.02
40 Sheet1 . Range ( " O5 " ) . Value = ( dFMVIU - DFMVID ) / 20
41 End Sub

In the above function, we first initialize the module MLifeTable in Line 10.
Then, we initialize the contract information in the module MGMDB in Lines
11–25. In Line 27, we pass the worksheet containing the base scenarios and
the range containing the corresponding forward rates to the module MGMDB.
In Lines 28–30, we calculate the fair market values for different equity shocks.
In Lines 32–36, we calculate the fair market values for different interest rate
shocks. Finally we write the results to the worksheets in Lines 38–40.
If we use the inputs shown in Figure 10.1, we see that outputs shown
in the same figure. The fair market value of this GMDB rider is –$5368.11.
Since it is negative, the present value of the benefits is lower than that of
the fees. The dollar delta is –$8256.88, meaning that when the market goes
216 Valuing a GMDB

up 1%, the fair market value of the guarantee will decrease about –$82.5688.
The dollar rho is close to zero, meaning that the fair market value of the
GMDB guarantee is not sensitive to the yield curve changes.

Exercise 10.5. Recalculate the output by changing the guarantee feature

from Return of Premium (RP) to Step-Up (SU). Which guarantee is more
expensive between a GMDB with RP and a GMDB with SU, holding all other
parameters constant?

Exercise 10.6. Modify the sub procedure Button1_Click by adding code

to calculate the runtime of the valuation program. Write the runtime to the
cell “O6” of the main worksheet. (Hint: Use the Timer function.)

10.6 Summary
Variable annuities are insurance products that contain guarantees. Since
these guarantees are complicated, closed-form formulas to calculate the
fair market values of these guarantees do not exist. Insurance companies
rely heavily on Monte Carlo simulation to calculate the fair market values of
guarantees. In this chapter, we implemented a simple Monte Carlo method
to calculate the fair market value of a GMDB.
An advantage of the Monte Carlo simulation method is that it is flexible
to value various guarantees. However, Monte Carlos simulation is time-
consuming, especially when used to value a large portfolio of variable an-
nuities. Metamodeling approaches have been proposed to speed up the
Monte Carlo simulation method. See (Gan, 2013), (Gan, 2015a), (Gan and
Lin, 2015), and (Gan and Lin, 2016) for details.
11
Connecting to Databases

In this chapter, we implement a simple tool in Visual Basic for Applications

(VBA) to extract variable annuity contracts from an Access® database that
are issued in a particular year. This tool also allows us to save extracted
contracts into a new table in the database. The interface of this inforce tool
is shown in Figure 11.1.

Figure 11.1: Interface of the variable annuity inforce tool

Table 11.1: Range Names Used by the Variable Annuity Inforce Tool

Name Refers To
FileName =Main!$C$4
IssueYear =Main!$C$3
TableName =Main!$C$5

217
218 Connecting to Databases

Table 11.1 gives a list of range names used by the simple tool. Figure
11.2 shows the variable annuity inforce table in an Access® database named
“va.accdb.” The name of the table is “Inforce10k,” and the table contains
10,000 variable annuity contracts, which are issued in different time. These
variable annuity contracts are generated randomly by a Java program (Gan,
2015b).

Figure 11.2: Inforce10k table in an Access® database

11.1 ActiveX® Data Objects

Excel® is a powerful tool for analyzing data and producing customized
reports. However, Excel may fall short as a data storage application when
dealing with extremely large volumes of data. Databases are usually used to
store such large volumes of data.
A database is an organized collection of data and is created and man-
aged by a database management system (DBMS), which is a computer soft-
ware application that provides users and programmers with a systematic
way to create, retrieve, update, and manage data. Popular database man-
agement systems include Microsoft Access, Microsoft SQL Server, Oracle® ,
and MySQL® .
Excel VBA provides some methods to access databases. One method
to access databases in Excel VBA is through Microsoft’s ActiveX Data Ob-
ActiveX® Data Objects 219

jects (ADO). ADO consists a set of objects for accessing data sources. A
main advantage of ADO is that it allows application developers to write pro-
grams to access data without knowing how the database is implemented.
In particular, ADO allows us to do the following:

• Connect to various external databases.

• Add, delete, and edit records from a database.
• Query data from a database to get a recordset.

Figure 11.3: Checking the reference for Microsoft ActiveX Data Objects 2.8
Library

To use ADO to access databases in Excel VBA, we need to select “Mi-

crosoft ActiveX Data Objects 2.8 Library” in the References window (see
Figure 11.3). To show the References window, you can click “Tools/Refer-
ences” in the Visual Basic Editor.
In this chapter, we will use the following objects of ADO: the
Connection object, the RecordSet object, and the Command object. The
Connection object is used to establish a connection between Excel and a
 220 Connecting to Databases

database. The RecordSet object is used to retrieve a single record or multi-

ple records from a database table. The RecordSet object can also be used
to save records from Excel to a database table. The Command object is used
to hold information related to queries to be executed.

11.2 SQL Queries

SQL stands for Structured Query Language, which is a database language
used for querying, updating, and managing relational databases. SQL is a
complex language. In this section, we introduce some commonly used SQL
statements.
To retrieve data from one ore more tables, we use the SELECT statement.
The SQL expression to retrieve specified columns and all rows from a table
is as follows:
1 SELECT [ ColumnName1 ] , [ ColumnName2 ] FROM TableName

The column names are separated by commas. If a column name contains

spaces or is an SQL keyword, we need to enclose the column name with
square brackets “[].”
If we want to retrieve all columns and all rows from a table, we use the
following simplified expression:
1 SELECT * FROM TableName

We use * to replace all the column names.

To retrieve specified columns and specified rows from a table, we use
the following SQL expression:
1 SELECT [ ColumnName1 ] , [ ColumnName2 ] FROM TableName
WHERE Condition

To retrieve specified rows, we add a WHERE clause to specify the condition

for the rows. To sort the retrieved data, we can add an ORDER BY clause at
the end:
1 SELECT [ ColumnName1 ] , [ ColumnName2 ] FROM TableName
WHERE Condition ORDER BY [ ColumnName1 ]

To delete a table in a database, we use the following SQL expression:

 Implementation 221

1 DROP TABLE TableName

The SQL expression used to create a table in a database is given by

1 CREATE TABLE TableName ( ColumnName1 DataType ,
ColumnName2 DataType , ColumnName3 DataType ) ;

There are many other SQL statements (Forta, 2012; Tale, 2016). The SQL
statements mentioned above will be used in this chapter.

11.3 Implementation
We organize the VBA code of the inforce tool into two modules: the
MDatabase module and the MInterface module.

11.3.1 MDatabase Module

The MDatabase module contains VBA code to access databases. The decla-
ration section of this module contains the following code:
1 Option Explicit
2
3 Dim dbConnection As ADODB . Connection

In the declaration section, we define only one module-level variable, which

is an ADODB.Connection object. This variable is used to hold the connec-
tion to a database.
The module-level variable is initialized by the following sub procedure:
1 Sub Connect ( ByVal FileName As String )
2 Dim sConn As String
3 sConn = " Provider = Microsoft . ACE . OLEDB .12.0; Data
Source = " & _
4 FileName & " ; Persist Security Info = False ; "
5
6 Set dbConnection = New ADODB . Connection
7 dbConnection . ConnectionString = sConn
8 dbConnection . Open
9 End Sub
 222 Connecting to Databases

This sub procedure requires one argument, which is the name of the Access
database. In Line 3, we construct a connection string by specifying the
Provider, the Data Source, and the Persist Security Info. The code in Line
3 is a common syntax for version 2007 and later versions of Excel. We set
Persist Security Info to false so that security-sensitive information (e.g., the
password) will not be returned as part of the connection. In Line 6, we
create a new ADODB.Connection object and assign it to the module-level
variable. In Line 7, we assign the connection string to the object. In Line 8,
we open the connection.
To close the connection, we use the following sub procedure:
1 Sub Disconnect ()
2 dbConnection . Close
3 Set dbConnection = Nothing
4 End Sub

In Line 2 of the above code, we call the Close method to close the connec-
tion. In Line 3, we release the memory of the variable by setting the variable
to Nothing.
The following function is used to retrieve a list of tables stored in an
Access database:
1 Function ListTables ()
2 Dim dbRecordSet As ADODB . Recordset
3
4 Set dbRecordSet = dbConnection . OpenSchema (
adSchemaTables )
5
6 Dim vTmp As Variant
7 vTmp = dbRecordSet . GetRows
8
9 Dim nRow As Integer
10 nRow = UBound ( vTmp , 2)
11
12 Dim aTable () As String
13 ReDim aTable (0 To nRow )
14
15 Dim i As Integer
16 Dim j As Integer
17 For i = 0 To UBound ( vTmp , 1)
18 If StrComp ( dbRecordSet . Fields ( i ) . Name , "
TABLE_NAME " , vbTextCompare ) = 0 Then
19 For j = 0 To UBound ( vTmp , 2)
20 aTable ( j ) = vTmp (i , j )
 Implementation 223

21 Next j
22 Exit For
23 End If
24 Next i
25
26 ListTables = aTable
27 End Function

In Line 2 of the above code, we declare an ADODB.Recordset object variable.

In Line 4, we call the OpenSchema method to get schema of the database
and assign it to the ADODB.Recordset object variable. The schema of the
database contains the names of the tables stored in the database. In Line
7, we convert the record set into a two-dimensional array by calling the
GetRows method. In Lines 17–24, we extract the table names and put them
into a one-dimensional array.
To test the function ListTables, we use the following sub procedure:
1 Sub TestListTables ()
2 Dim sFileName As String
3
4 sFileName = ThisWorkbook . Path & " \ va . accdb "
5 Connect ( sFileName )
6
7 Dim vRes As Variant
8 vRes = ListTables
9
10 Disconnect
11
12 Dim i As Integer
13 For i = 0 To UBound ( vRes )
14 Debug . Print vRes ( i )
15 Next i
16 End Sub

In the above test procedure, we assume that the Access database va.accdb
is saved in the same folder as the workbook. In Line 5, we open the con-
nection by calling the procedure Connect. In Line 8, we call the function
ListTables to get the table names. In Line 10, we close the connection
by calling the procedure Disconnect. Executing the above test procedure
gives the following output:
1 Inforce10k
2 MSysA ccessSto rage
3 MSysACEs
 224 Connecting to Databases

4 MS ys Co mpl ex Co lum ns
5 MSysIMEXColumns
6 MSysIMEXSpecs
7 MSysNameMap
8 MSysNavPaneGroupCategories
9 MSysN avPaneGr oups
10 MSysNavPaneGroupToObjects

From the output, we see that many names start with “MSys.” These tables
are Microsoft Access system tables, which contain general information
about various specifications. The first name “Inforce10k” is the name of the
variable annuity inforce table.
Before creating a table in a database, we need to make sure that the
table does not exist in the database. The following function can be used to
detect whether a table exists or not in the database:
1 Function HasTable ( ByVal TableName As String )
2 Dim vTables As Variant
3 Dim bRes As Boolean
4
5 vTables = ListTables
6 bRes = False
7 Dim i As Integer
8 For i = 0 To UBound ( vTables )
9 If StrComp ( vTables ( i ) , TableName ,
vbTextCompare ) = 0 Then
10 bRes = True
11 Exit For
12 End If
13 Next i
14
15 HasTable = bRes
16 End Function

In this function, we first call the function ListTables to get the names of
all tables in the database. Then we use a For-Next loop to check whether
the given table name matches one of the names in the database. To test this
function, we use the following sub procedure:
1 Sub TestHasTable ()
2 Dim sFileName As String
3
4 sFileName = ThisWorkbook . Path & " \ va . accdb "
5 Connect ( sFileName )
 Implementation 225

6
7 Dim vRes As Variant
8 vRes = ListTables
9
10 Debug . Print HasTable ( " Inforce10k " )
11 Debug . Print HasTable ( " 2008 " )
12
13 Disconnect
14 End Sub

Executing the above test code gives:

1 True
2 False

The output indicates that table “2008” does not exist in the database.
The following sub procedure is used to delete a table in the database:
1 Sub DeleteTable ( ByVal TableName As String )
2 Dim sSQL As String
3 Dim dbCmd As ADODB . Command
4
5 Set dbCmd = New ADODB . Command
6 sSQL = " DROP TABLE " & TableName
7
8 dbCmd . ActiveConnection = dbConnection
9 dbCmd . CommandText = sSQL
10
11 dbCmd . Execute
12
13 Set dbCmd = Nothing
14 End Sub

In this function, we use the ADODB.Command object to execute the SQL

query. In Lines 8–9, we specify the properties ActiveConnection and
CommandText for the object. In Line 11, we call the method Execute to
run the SQL query to delete the table.

Exercise 11.1. Open the database va.accdb and manually create a ta-
ble named Table1 in the database. Then write a sub procedure named
TestDeleteTable to delete the table Table1 from the database.
h
 226 Connecting to Databases

To create a table in the database, we use the following sub procedure:

1 Sub CreateTable ( ByRef Data As Range , ByVal TableName
As String , ByVal PrimaryKey As String )
2 If HasTable ( TableName ) Then
3 DeleteTable ( TableName )
4 End If
5
6 Dim sCmdText As String
7
8 sCmdText = " CREATE TABLE " & TableName & " ( "
9
10 Dim i As Integer
11 For i = 1 To Data . Columns . Count
12 Select Case VarType ( Data . Cells (2 , i ) . Value )
13 Case vbDouble
14 sCmdText = sCmdText & " [ " & Data .
Cells (1 , i ) . Value & " ] DOUBLE "
15 Case vbString
16 sCmdText = sCmdText & " [ " & Data .
Cells (1 , i ) . Value & " ] VARCHAR "
17 Case vbDate
18 sCmdText = sCmdText & " [ " & Data .
Cells (1 , i ) . Value & " ] DATETIME "
19 End Select
20 If StrComp ( Data . Cells (1 , i ) . Value , PrimaryKey
, vbTextCompare ) = 0 Then
21 sCmdText = sCmdText & " PRIMARY KEY "
22 End If
23 If i < Data . Columns . Count Then
24 sCmdText = sCmdText & " , "
25 End If
26 Next i
27 sCmdText = sCmdText & " ) "
28
29 Dim dbCmd As ADODB . Command
30 Set dbCmd = New ADODB . Command
31
32 dbCmd . ActiveConnection = dbConnection
33 dbCmd . CommandText = sCmdText
34
35 Debug . Print sCmdText
36 dbCmd . Execute
37 Set dbCmd = Nothing
38 End Sub
 Implementation 227

This sub procedure requires three arguments: a range of data, a table name,
and the name of the primary key. We assume that the first row of the range
contains the column names of the table. In Lines 8–26, we construct the SQL
query for creating the table. The data types are derived from the data in the
second row of the range. Here, we only consider three types of data: Double,
String, and Date. We also assume that the first column of the range is the
primary key, which is a special column designated to uniquely identify
all table records. In Lines 28–34, we create an ADODB.Command object to
execute query.

Exercise 11.2. Suppose that the range "O2:Q5" contains the following data:

First Name Last Name Grade

Courtney Salazar A
Carroll Paul B
Jane Taylor C

Write a sub procedure named TestCreateTable to create a table named

Grade in the database based on the range "O2:Q5".

The following sub procedure can be used to save a table in a worksheet

to the database:
1 Sub ExcelToAccess ( ByRef Data As Range , ByVal
TableName As String )
2 CreateTable Data := Data , TableName := TableName
3
4 Dim dbRecordSet As ADODB . Recordset
5 Set dbRecordSet = New ADODB . Recordset
6 dbRecordSet . CursorLocation = adUseServer
7 dbRecordSet . ActiveConnection = dbConnection
8 dbRecordSet . CursorType = adOpenDynamic
9 dbRecordSet . LockType = adLockOptimistic
10 dbRecordSet . Source = TableName
11
12 dbRecordSet . Open
13
14 Dim i As Integer
15 Dim j As Integer
 228 Connecting to Databases

16 For i = 2 To Data . Rows . Count

17 dbRecordSet . AddNew
18 For j = 1 To Data . Columns . Count
19 dbRecordSet ( Data . Cells (1 , j ) . Value ) . Value
= Data . Cells (i , j ) . Value
20 Next j
21 dbRecordSet . Update
22 Next i
23
24 dbRecordSet . Close
25 Set dbRecordSet = Nothing
26 End Sub

In Line 2, we create the table in the database by calling the sub procedure
CreateTable. In Lines 4–10, we create and configure an ADODB.Recordset
object. In Line 12, we open the record set for appending new records. In
Lines 16–22, we use a loop to append the rows in the range to the record set.

Exercise 11.3. Continue from Exercise 11.2 and write a sub procedure
named TestExcelToAccess to save the range "O2:Q5" to a table named
Grade in the database.
h

The following function is used to retrieve data from the database:

1 Function ExtractRecord ( ByVal SQL As String ) As
Variant
2 Dim dbRecordSet As ADODB . Recordset
3 Dim dbCmd As ADODB . Command
4
5 Set dbCmd = New ADODB . Command
6 dbCmd . ActiveConnection = dbConnection
7 dbCmd . CommandText = SQL
8
9 Set dbRecordSet = dbCmd . Execute
10
11 Dim vTmp As Variant
12 vTmp = dbRecordSet . GetRows
13
14 Dim aRes () As String
15 Dim iU1 As Integer
16 Dim iU2 As Integer
 Implementation 229

17 Dim i As Integer
18 Dim j As Integer
19 iU1 = UBound ( vTmp , 1)
20 iU2 = UBound ( vTmp , 2)
21 ReDim aRes (0 To iU1 , 0 To iU2 + 1)
22 For j = 0 To iU1 - 1
23 aRes (j , 0) = dbRecordSet . Fields ( j ) . Name
24 Next j
25 For i = 0 To iU2 - 1
26 For j = 0 To iU1 - 1
27 aRes (j , i + 1) = vTmp (j , i )
28 Next j
29 Next i
30
31 dbRecordSet . Close
32 Set dbRecordSet = Nothing
33 Set dbCmd = Nothing
34
35 ExtractRecord = aRes
36 End Function

The function requires one argument, which is an SQL query. In Lines 2–12,
we create an ADODB.Recordset object and an ADODB.Command object to
retrieve data from the database. In Lines 22–29, we convert the record set to
a two-dimensional array.

Exercise 11.4. Suppose that the database va.accdb contains a table named
Grade with the data given in Exercise 11.2. Write a sub procedure named
TestExtractRecord to retrieve the students whose grade is B.
h

11.3.2 MInterface Module

The MInterface module contains VBA macros for the two buttons shown
in Figure 11.1. The macro assigned to the first button is as follows:
1 Sub Button1_Click ()
2 Dim sFileName As String
3 Dim sSQL As String
4 Dim vRes As Variant
5
 230 Connecting to Databases

6 sFileName = ThisWorkbook . Path & " \" & Range ( "

FileName " ) . Value
7 Connect ( sFileName )
8
9 sSQL = " SELECT * FROM " & Range ( " TableName " ) .
Value & _
10 " WHERE year ([ issueDate ]) = " & Range ( "
IssueYear " ) . Value & " ; "
11 vRes = ExtractRecord ( sSQL )
12
13 Disconnect
14
15 Dim ws As Worksheet
16 Set ws = Sheets ( " Cohort " )
17 ws . Cells . ClearContents
18
19 Dim i As Integer
20 Dim j As Integer
21 For i = 0 To UBound ( vRes , 2)
22 For j = 0 To UBound ( vRes , 1)
23 ws . Cells ( i + 1 , j + 1) . Value = vRes (j , i )
24 Next j
25 Next i
26 End Sub

The sub procedure uses three named ranges defined in Table 11.1. In Line
10, the SQL function year is used to get the year part from a date. In Lines
21–25, we write the retrieved contracts to the worksheet “Cohort.”
If we click the first button, we see that output as shown in Figure 11.4.
From the output, we see that all the contracts have an issue date in 2009.
The macro assigned to the second button is given below:
1 Sub Button2_Click ()
2 Dim ws As Worksheet
3
4 Set ws = Sheets ( " Cohort " )
5
6 Dim nCol As Integer
7 Dim nRow As Integer
8 nCol = ws . Cells (1 , 1) . End ( xlToRight ) . Column
9 nRow = ws . Cells (1 , 1) . End ( xlDown ) . Row
10
11 Dim sTableName As String
12 sTableName = " Cohort " & Range ( " IssueYear " )
Implementation 231

Figure 11.4: Output in the worksheet when the first button is clicked

Figure 11.5: New table in the database when the second button is clicked
 232 Connecting to Databases

13 Dim rngData As Range

14 Set rngData = ws . Range ( ws . Cells (1 , 1) , ws . Cells (
nRow , nCol ) )
15
16 Dim sFileName As String
17 sFileName = ThisWorkbook . Path & " \" & Range ( "
FileName " ) . Value
18 Connect ( sFileName )
19
20 ExcelToAccess Data := rngData , TableName :=
sTableName , PrimaryKey := " recordID "
21
22 Disconnect
23 End Sub

In this sub procedure, we just write the table in the “Cohort” worksheet to
the database. The name of the table in the database is “Cohort” plus the
issue year. If we click the second button, the data in the worksheet “Cohort”
will be saved to a new table, as shown in Figure 11.5.

Exercise 11.5. Modify the variable annuity inforce tool as follows:

• Add a new parameter named Product Type to the input area of the
interface shown in Figure 11.1.
• Modify the macro assigned to the first button to retrieve the contracts
with specified issue year and product type.

How many contracts with “MB” guarantee have an issue date in 2009?

11.4 Summary
In this chapter, we introduced how to connect to a Microsoft Access
database in Excel VBA. In particular, we introduced how to retrieve data
from an Access database to Excel, how to create tables in Access, and how
to transport data in Excel to an Access database. To learn more about the
SQL, readers are referred to (Forta, 2012; Tale, 2016).
12
Object-Oriented Programming

In this chapter, we use object-oriented programming techniques to imple-

ment the bootstrapping tool introduced in Chapter 8. The interface of the
bootstrapping tool to be implemented in this chapter is identical to the one
shown in Figure 8.1. However, we implement the bootstrapping tool in an
object-oriented way by using classes.

12.1 Introduction
Object-oriented programming (OOP) is a programming paradigm intro-
duced in the 1980s (Urtis, 2015). Since its introduction, OOP has grown
in popularity and been supported in many other programming languages,
such as C++ (Gan, 2011). It is not surprising that today’s Visual Basic for
Applications (VBA) programming language also supports OPP.
A computer program developed using the OOP method consists of
individual objects, each of which has its own properties and methods. For
example, Excel® is an application developed using the OOP method. In
Section 2.1, we introduced the Excel® object model. For example, a range
is an object in Excel® and contains properties (e.g., value and text) and
methods (e.g., select and copy), as described in Section 2.5.
In OOP, the objects of a computer program interact with each other.
Each object models an aspect of the problem one tries to solve. In OOP, an
object is called an instance of a class, which can be thought as a template of
objects. For example, Range is a class in Excel® VBA, and Range("A1") is a
Range object. In Section 1.3, we introduced how to define a VBA class.
An object-oriented (OO) VBA program has the following characteristics
(Sintes, 2001; Gan, 2011):

(a) Natural. An OO program solves a problem by modeling the problem at a

functional level.
(b) Reliable. An OO program consists of well-designed objects. Changing

233
234 Object-Oriented Programming

an object will not affect other objects as long as we do not change the
interface of the object.
(c) Reusable. The classes created for an OO program can be used in other
programs.
(d) Maintainable. Since changing an object does not affect other objects,
fixing one part of an OO program will not cause changes in other parts.

Unlike other OOP languages, such as C++ and Java, VBA does not support in-
heritance. In other words, we cannot define a class by inheriting properties
and methods from another class.

12.2 Objects
As we mentioned before, an OO program consists of objects, each of which
models an aspect of the problem. To develop an OO bootstrapping tool, we
need the following objects: dates, yield curves, holiday calendars, swaps,
and a bootstrapping procedure.
Since an object is created from a class, which serves as a template for
objects. According to our analysis above, we need classes for dates, yield
curves, holiday calendars, swaps, and the bootstrapping procedure. Since
the VBA has the Date data type, we do not need to implement our date class.
We also do not need to implement a holiday calendar class, because we do
not need to keep copies of a calendar object. A holiday calendar is global in
the sense that we only need one copy of the holiday calendar. As a result, we
can just implement the functions of a holiday calendar in a regular module
rather than in a class module. Note that the VBA does not keep separate
copies of data in a regular module. However, VBA keeps separate copies of
data stored in a class module.

12.3 Implementation
To develop the OO bootstrapping tool, we decide to implement three classes:
a yield curve class, a swap class, and a bootstrapper class. These classes al-
low us to create yield curve objects, swap objects, and bootstrapper objects,
respectively.
 Implementation 235

12.3.1 CZeroCurve Module

We implement a yield curve class in a class module named CZeroCurve.
The declarations section of this class module contains the following code:
1 Option Explicit
2
3 Dim datCurveDate As Date
4 Dim sYCDCC As String
5 Dim aDate () As Date
6 Dim aDF () As Double

In the above code, we declare four module-level variables, which are prop-
erties used to store the curve date, the yield curve day count convention,
dates, and discount factors at these dates.
The sub procedure used to initialize the module-level variables is im-
plemented as follows:
1 Sub Initialize ( ByVal CurveDate As Date , ByVal YCDCC
As String )
2 datCurveDate = CurveDate
3 sYCDCC = YCDCC
4
5 ReDim aDate (0 To 0)
6 ReDim aDF (0 To 0)
7
8 aDate (0) = CurveDate
9 aDF (0) = 1
10 End Sub

This sub procedure requires two arguments: the curve date and the yield
curve day count convention. The array aDate is initialized to contain one
element, which is the curve date. The array aDF is also initialized to contain
one element, which is the discount factor at the curve date.
Sometimes we need to make a copy of a CZeroCurve object. To do that,
we can use the Clone method implemented as follows:
1 Function Clone () As CZeroCurve
2 Dim cZC As CZeroCurve
3 Dim iU As Integer
4 Dim i As Integer
5
6 Set cZC = New CZeroCurve
7 cZC . Initialize CurveDate := datCurveDate , YCDCC :=
sYCDCC
 236 Object-Oriented Programming

8 For i = LBound ( aDate ) + 1 To UBound ( aDate )

9 cZC . PushBack datIn := aDate ( i ) , dDF := aDF ( i )
10 Next i
11
12 Set Clone = cZC
13 End Function

In Line 6 of the above code, we use the New function to create a new object.
In Lines 7–10, we copy the data of this object to the new object. This function
calls another function called PushBack to copy the dates and the discount
factors from this object to a new object called cZC. Note that we use Set
to assign the new object to the function name at the end of this function.
We have to use Set here because cZC is an object. This method is useful,
as it allows us to experiment with a copy of a CZeroCurve object without
affecting the object.
The PushBack method is implemented as follows:
1 Sub PushBack ( ByVal datIn As Date , ByVal dDF As Double
)
2 Dim iU As Integer
3 iU = UBound ( aDate )
4
5 If datIn <= aDate ( iU ) Then
6 Err . Raise Number :=2002 , Source := " CZeroCurve .
PushBack " , Description := " The input date is
ealier than the last date . "
7 End If
8
9 ReDim Preserve aDate (0 To iU + 1)
10 ReDim Preserve aDF (0 To iU + 1)
11 aDate ( iU + 1) = datIn
12 aDF ( iU + 1) = dDF
13 End Sub

The method PushBack is a sub procedure that inserts a date and a discount
factor to the end of the arrays aDate and aDF. This method takes two argu-
ments: a date and a real number. The method checks whether the input
date is equal to or before the last date in the array aDate. If the input date is
equal to or before the last date in the array, the method will raise an error.
This method ensures that the dates in the array are sorted in an increasing
order.
We implement the raw interpolation method described in Section 8.2
in the following method:
 Implementation 237

1 Function LogLinear ( ByVal datD As Date ) As Double

2 If datD < datCurveDate Then
3 Err . Raise Number :=2001 , Source := " CZeroCurve .
LogLinear " , Description := " Date is ealier
than the curve date . "
4 End If
5 Dim i As Integer
6 Dim iL As Integer
7 Dim iH As Integer
8 iL = LBound ( aDate )
9 iH = UBound ( aDate )
10 For i = iL To iH
11 If datD = aDate ( i ) Then
12 LogLinear = aDF ( i )
13 Exit Function
14 End If
15 Next i
16
17 Dim dat1 As Date
18 Dim dat2 As Date
19 Dim dDF1 As Double
20 Dim dDF2 As Double
21 Dim bFound As Boolean
22 bFound = False
23 For i = iL To iH - 1
24 If datD > aDate ( i ) And datD < aDate ( i + 1)
Then
25 dat1 = aDate ( i )
26 dat2 = aDate ( i + 1)
27 dDF1 = aDF ( i )
28 dDF2 = aDF ( i + 1)
29 bFound = True
30 End If
31 Next i
32
33 If Not bFound Then
34 dat1 = aDate ( iH - 1)
35 dat2 = aDate ( iH )
36 dDF1 = aDF ( iH - 1)
37 dDF2 = aDF ( iH )
38 End If
39
40 Dim dDCF1 As Double
41 Dim dDCF2 As Double
 238 Object-Oriented Programming

42 Dim dDCF As Double

43 dDCF1 = TFrac ( datCurveDate , dat1 , sYCDCC )
44 dDCF2 = TFrac ( datCurveDate , dat2 , sYCDCC )
45 dDCF = TFrac ( datCurveDate , datD , sYCDCC )
46
47 Dim dTemp As Double
48 dTemp = Work sheetFun ction . Ln ( dDF2 ) * ( dDCF -
dDCF1 ) + _
49 Works heetFunc tion . Ln ( dDF1 ) * ( dDCF2 - dDCF )
50 LogLinear = Exp ( dTemp / ( dDCF2 - dDCF1 ) )
51 End Function

This function takes one argument, which is a date, and returns the discount
factor at the input date. If the input date is older than the curve date, the
function will raise an error. If the input date matches one of the dates saved
in the array aDate, the function just returns the corresponding discount
factor saved in the array aDF. Otherwise, the function searches for two dates
in aDate that enclose the input date. If the input date is not enclosed in any
two dates in aDate, the function will use the last two dates in aDate. Then,
the function performs the raw interpolation based on the two dates and the
two corresponding discount factors.
The following function allows us to get the dates corresponding to the
discount factors:
1 Function Dates () As Variant
2 Dates = aDate
3 End Function

The method Dates returns the array aDate as a variant. This method allows
us to extract the dates saved in the variable aDate, which is private to a
CZeroCurve object.
The methods mentioned above allow us to perform necessary actions
on a CZeroCurve object. If needed, we can add more methods to the class.

12.3.2 CSwap Module

The module CSwap is a class module implementing an interest rate swap.
The declarations section of this class module contains the following code:
1 Option Explicit
2
3 Dim sFixFreq As String
4 Dim sFixDCC As String
 Implementation 239

5 Dim sFltFreq As String

6 Dim sFltDCC As String
7 Dim sCalendar As String
8 Dim sBDC As String
9 Dim iSetDays As Integer
10 Dim datEffective As Date
11 Dim sTenor As String
12 Dim dRate As Double
13
14 Dim datSet As Date
15 Dim vFixPaymentDates As Variant
16 Dim vFltPaymentDates As Variant

As we discussed in Section 8.3, an interest rate swap consists of two legs: a

fixed leg and a floating leg. The properties defined in Lines 3–12 represent
various specifications of an interest rate swap contract. The properties
defined in Lines 14–16 hold some convenient values calculated from other
properties.
The properties sFixFreq and sFixDCC hold the payment frequency
and the day count convention of the fixed leg. Similarly, the properties
sFltFreq and sFltDCC hold the payment frequency and the day count
convention of the floating leg. In our implementation, we allow the fixed
leg and the floating leg of a swap to have different specifications.
The properties sCalendar and sBDC hold the business calendar and the
business day count convention of the yield curve. For example, the US yield
curve uses the US holiday calendar. The business day count convention
tells how to handle the cases when a payment date falls on a holiday or
weekend.
The property iSetDays is used to hold the number of settlement days of
a swap. For example, the number of settlement days in the US is 2 business
days, meaning that a swap transaction is settled 2 business days after the
transaction date. The transaction date of a swap is stored in the property
datEffective.
The properties sTenor and dRate hold the tenor and the fixed rate of a
swap. The value of sTenor can be “1Y,” “5Y,” etc.
The remaining properties are used to store some values calculated
from the specifications of a swap. For example, the property datSet holds
the settlement date of the swap. The properties vFixPaymentDates and
vFltPaymentDates hold the payment dates of the fixed leg and the floating
leg, respectively.
 240 Object-Oriented Programming

The class CSwap contains only two methods, which are given in the
following code:
1 Sub Initialize ( ByVal FixFreq As String , ByVal FixDCC
As String , ByVal FltFreq As String , _
2 ByVal FltDCC As String , ByVal Calendar As String ,
ByVal BDC As String , _
3 ByVal SetDays As Integer , ByVal EffectiveDate As
Date , ByVal Tenor As String , _
4 ByVal Rate As Double )
5
6 sFixFreq = FixFreq
7 sFixDCC = FixDCC
8 sFltFreq = FltFreq
9 sFltDCC = FltDCC
10 sCalendar = Calendar
11 sBDC = BDC
12 iSetDays = SetDays
13 datEffective = EffectiveDate
14 sTenor = Tenor
15 dRate = Rate
16
17 Dim i As Integer
18 datSet = datEffective
19 For i = 1 To iSetDays
20 datSet = RollDate ( AddDays ( datSet , 1) , " Follow
" , sCalendar )
21 Next i
22
23 vFixPaymentDates = GenerateSchedule ( datSet ,
sFixFreq , sTenor , sCalendar , sBDC )
24 vFltPaymentDates = GenerateSchedule ( datSet ,
sFltFreq , sTenor , sCalendar , sBDC )
25 End Sub
26
27 Function NPV ( ByRef cZC As CZeroCurve ) As Double
28 Dim dFixPV As Double
29 Dim dFltPV As Double
30 Dim dDF As Double
31 Dim i As Integer
32 Dim dT As Double
33 Dim dFR As Double
34
35 For i = LBound ( vFixPaymentDates ) + 1 To UBound (
vFixPaymentDates )
 Implementation 241

36 dDF = cZC . LogLinear ( vFixPaymentDates ( i ) )

37 dFixPV = dFixPV + TFrac ( vFixPaymentDates ( i -
1) , vFixPaymentDates ( i ) , sFixDCC ) * dRate
* dDF
38 Next i
39
40 For i = LBound ( vFltPaymentDates ) + 1 To UBound (
vFltPaymentDates )
41 dDF = cZC . LogLinear ( vFltPaymentDates ( i ) )
42 dT = TFrac ( vFltPaymentDates ( i - 1) ,
vFltPaymentDates ( i ) , sFltDCC )
43 dFR = ( cZC . LogLinear ( vFltPaymentDates ( i - 1) )
/ dDF - 1) / dT
44 dFltPV = dFltPV + dT * dFR * dDF
45 Next i
46
47 NPV = dFixPV - dFltPV
48 End Function

The method Initialize is used to initialize an object by setting values

of the properties. The payment dates of the fixed leg and the floating leg
are created by calling the function GenerateSchedule, which is a function
defined in the module MHoliday (see Section 7.6.2).
The method NPV calculates the net present value of a swap. This method
takes one argument, which is a reference to a yield curve object. This
method first calculates the present value of the fixed cash flows and then
calculates the present value of the floating cash flows.

12.3.3 CBootstrapper Module

The bootstrapper class is implemented in the class module CBootstrapper.
The declarations section of this class module contains the following code:
1 Option Explicit
2
3 ' input
4 Dim aTenor () As String
5 Dim aRate () As Double
6 Dim sFixFreq As String
7 Dim sFixDCC As String
8 Dim sFltFreq As String
9 Dim sFltDCC As String
10 Dim sCalendar As String
11 Dim sBDC As String
 242 Object-Oriented Programming

12 Dim datCurveDate As Date

13 Dim iSetDays As Integer
14 Dim sYCDCC As String
15
16 ' settlement date
17 Dim datSet As Date
18
19 ' yield curve
20 Dim cZC As CZeroCurve

The properties of the class CBootstrapper are used to store the input
and output data. For example, the properties aTenor and aRate stores
the tenors and rates of the input swap rates. The properties sFixFreq and
sFixDCC represent the payment frequency and the day count convention of
the fixed leg. The properties sFltFreq and sFltDCC represent the payment
frequency and the day count convention of the floating leg. The proper-
ties sCalendar and sBDC hold the business calendar and the business day
count convention of the swap. The properties datCurveDate, iSetDays,
and sYCDCC store the curve date, the number of settlement days, and the
curve’s day count convention, respectively.
The remaining two properties are used to store some immediate and
output values. For example, the property datSet stores the settlement date
of the swaps. For convenience purposes, we save the settlement date into a
variable, which can be used later. The property cZC stores the yield curve
that is bootstrapped from the input data.
To initialize the properties, we implement the following method:
1 Sub Initialize ( ByVal FixFreq As String , ByVal FixDCC
As String , ByVal FltFreq As String , _
2 ByVal FltDCC As String , ByVal Calendar As String ,
ByVal BDC As String , _
3 ByVal SetDays As Integer , ByVal EffectiveDate As
Date , ByVal YCDCC As String , _
4 ByRef Tenors () As String , ByRef Rates () As Double
)
5 Dim iL As Integer
6 Dim iU As Integer
7 Dim i As Integer
8
9 iL = LBound ( Tenors )
10 iU = UBound ( Tenors )
11
12 ReDim aTenor ( iL To iU )
 Implementation 243

13 ReDim aRate ( iL To iU )
14
15 For i = iL To iU
16 aTenor ( i ) = Tenors ( i )
17 aRate ( i ) = Rates ( i )
18 Next i
19
20 sFixFreq = FixFreq
21 sFixDCC = FixDCC
22 sFltFreq = FltFreq
23 sFltDCC = FltDCC
24 sCalendar = Calendar
25 sBDC = BDC
26 iSetDays = SetDays
27 datCurveDate = EffectiveDate
28 sYCDCC = YCDCC
29
30 datSet = EffectiveDate
31 For i = 1 To iSetDays
32 datSet = RollDate ( AddDays ( datSet , 1) , " Follow
" , sCalendar )
33 Next i
34
35 Set cZC = New CZeroCurve
36 cZC . Initialize CurveDate := datCurveDate , YCDCC :=
sYCDCC
37 End Sub

The method Initialize is used to initialize the properties of an object. The

values used to initialize these properties are passed from arguments. Most
arguments of this method are ByVal arguments except for the arguments
Tenors and Rates, which are ByRef arguments. We have to use the ByRef
modifier for Tenors and Rates because it is not possible to pass an array
using ByVal.
Newton’s method described in Section 8.4 is implemented in the follow-
ing method:
1 Private Sub SolveRate ( ByVal iInd As Integer )
2 Dim cIRS As CSwap
3 Dim cZCTmp As CZeroCurve
4 Dim datMaturity As Date
5 Dim dRate As Double
6
7 Dim iNum As Integer
 244 Object-Oriented Programming

8 Dim sPeriod As String

9
10 ExtractPeriod sIn := aTenor ( iInd ) , iNum := iNum ,
sPeriod := sPeriod
11 datMaturity = OffsetDate ( datSet , iNum , sPeriod )
12
13 Set cIRS = New CSwap
14 cIRS . Initialize FixFreq := sFixFreq , FixDCC :=
sFixDCC , FltFreq := sFltFreq , FltDCC := sFltDCC , _
15 Calendar := sCalendar , BDC := sBDC , SetDays :=
iSetDays , EffectiveDate := datCurveDate , _
16 Tenor := aTenor ( iInd ) , Rate := aRate ( iInd )
17
18 Dim dx As Double
19 Dim dx0 As Double
20 Dim dx1 As Double
21 Dim df As Double
22 Dim df0 As Double
23 Dim df1 As Double
24
25 dx0 = Exp ( - aRate ( iInd ) * TFrac ( datCurveDate ,
datMaturity , sYCDCC ) )
26 Set cZCTmp = cZC . Clone
27 cZCTmp . PushBack datIn := datMaturity , dDF := dx0
28 df0 = cIRS . NPV ( cZCTmp )
29
30 dx1 = dx0 + 0.001
31 Set cZCTmp = cZC . Clone
32 cZCTmp . PushBack datIn := datMaturity , dDF := dx1
33 df1 = cIRS . NPV ( cZCTmp )
34 Do While Abs ( dx1 - dx0 ) > 0.0000000001
35 dx = dx0 - df0 * ( dx1 - dx0 ) / ( df1 - df0 )
36 Set cZCTmp = cZC . Clone
37 cZCTmp . PushBack datIn := datMaturity , dDF := dx
38 df = cIRS . NPV ( cZCTmp )
39
40 dx0 = dx1
41 df0 = df1
42 dx1 = dx
43 df1 = df
44 Loop
45
46 cZC . PushBack datIn := datMaturity , dDF := dx
47 End Sub
 Implementation 245

The method SolveRate bootstraps a discount factor from a given swap rate.
The new discount factor is added to the yield curve.
The following method bootstraps all discount factors:
1 Sub BuildCurve ()
2 Dim iL As Integer
3 Dim iH As Integer
4 Dim i As Integer
5 iL = LBound ( aTenor )
6 iH = UBound ( aTenor )
7
8 For i = iL To iH
9 SolveRate iInd := i
10 Next i
11 End Sub

The method BuildCurve calls the method SolveRate repeatedly to boot-

strap all the discount factors based on the given swap rates.
To calculate the output specified by users, we implement the following
method:
1 Function CreateOutput ( ByVal OutputType As String ) As
Variant
2 Dim aOutDates () As Date
3 Dim i As Integer
4 Dim iL As Integer
5 Dim iH As Integer
6 Dim dT As Double
7 Dim aDate As Variant
8
9 If StrComp ( OutputType , " Input " , vbTextCompare ) =
0 Then
10 aDate = cZC . Dates
11 iL = LBound ( aDate )
12 iH = UBound ( aDate )
13 ReDim aOutDates ( iL To iH )
14 For i = iL To iH
15 aOutDates ( i ) = aDate ( i )
16 Next i
17 Else
18 aOutDates = GenerateSchedule ( datCurveDate ,
OutputType , " 30 Y " , sCalendar , sBDC )
19 End If
20
21 iL = LBound ( aOutDates )
 246 Object-Oriented Programming

22 iH = UBound ( aOutDates )
23
24 ReDim aOut ( iL To iH , 0 To 4) As Double
25 For i = iL To iH
26 aOut (i , 0) = CDbl ( aOutDates ( i ) )
27 aOut (i , 1) = cZC . LogLinear ( aOutDates ( i ) )
28 dT = TFrac ( datCurveDate , aOutDates ( i ) , sYCDCC
)
29 If dT = 0 Then
30 aOut (i , 2) = 0
31 Else
32 aOut (i , 2) = - Wor ksheetFu nction . Ln ( aOut (i
, 1) ) / dT
33 End If
34 If i < iH Then
35 aOut (i , 3) = ( cZC . LogLinear ( aOutDates ( i ) )
/ cZC . LogLinear ( aOutDates ( i + 1) ) -
1) / _
36 TFrac ( aOutDates ( i ) , aOutDates ( i + 1) ,
sYCDCC )
37 Else
38 aOut (i , 3) = aOut ( i - 1 , 3)
39 End If
40 aOut (i , 4) = TFrac ( datCurveDate , aOutDates ( i )
, sYCDCC )
41 Next i
42 CreateOutput = aOut
43 End Function

The method CreateOutput calculates the output quantities based on a

given output type. If the output type is “Input,” then this method calculates
the discount factors, forward rates, zero rates, and the year fractions at the
maturity dates of the input swap rates. Otherwise, this method calculates
these quantities based on a given frequency.

12.3.4 OO Bootstrapping Tool

The OO bootstrapping tool consists of three regular modules and three class
modules. The VBA code of the three class modules are shown in Sections
12.3.1, 12.3.2, and 12.3.3, respectively. The three regular modules include
MDate, MHoliday, and MInterface. The first two regular modules are im-
plemented in Chapter 7.
The VBA code of the module MInterface is given below:
 Implementation 247

1 Sub Button1_Click ()
2 Dim vRes As Variant
3 Dim aTenor () As String
4 Dim aRate () As Double
5 Dim sFixFreq As String
6 Dim sFixDCC As String
7 Dim sFltFreq As String
8 Dim sFltDCC As String
9 Dim sCalendar As String
10 Dim sBDC As String
11 Dim datCurveDate As Date
12 Dim iSetDays As Integer
13 Dim sYCDCC As String
14 Dim sOutputType As String
15
16 Dim iNumRates As Integer
17 Dim i As Integer
18 Dim iRows As Integer
19 Dim j As Integer
20
21 Dim cBT As CBootstrapper
22
23 iNumRates = Range ( " Rate " ) . Rows . Count
24 ReDim aTenor (1 To iNumRates )
25 ReDim aRate (1 To iNumRates )
26 With Range ( " Rate " )
27 For i = 1 To iNumRates
28 aTenor ( i ) = . Cells (i , 1) . Value
29 aRate ( i ) = . Cells (i , 2) . Value / 100
30 Next i
31 End With
32 With Range ( " Param " )
33 sFixFreq = . Cells (1 , 1) . Value
34 sFixDCC = . Cells (2 , 1) . Value
35 sFltFreq = . Cells (3 , 1) . Value
36 sFltDCC = . Cells (4 , 1) . Value
37 sCalendar = . Cells (5 , 1) . Value
38 sBDC = . Cells (6 , 1) . Value
39 datCurveDate = . Cells (7 , 1) . Value
40 iSetDays = . Cells (8 , 1) . Value
41 sYCDCC = . Cells (9 , 1) . Value
42 sOutputType = . Cells (10 , 1) . Value
43 End With
44
 248 Object-Oriented Programming

45 Set cBT = New CBootstrapper

46 cBT . Initialize FixFreq := sFixFreq , FixDCC := sFixDCC
, FltFreq := sFltFreq , _
47 FltDCC := sFltDCC , Calendar := sCalendar , BDC :=
sBDC , SetDays := iSetDays , _
48 EffectiveDate := datCurveDate , YCDCC := sYCDCC ,
Tenors := aTenor , Rates := aRate
49 cBT . BuildCurve
50 vRes = cBT . CreateOutput ( sOutputType )
51
52 iRows = Sheet1 . Range ( " H3 " ) . End ( xlDown ) . Row
53 Sheet1 . Range ( Cells (3 , 8) , Cells ( iRows , 12) ) .
ClearContents
54 For i = LBound ( vRes , 1) To UBound ( vRes , 1)
55 For j = 0 To 4
56 Sheet1 . Cells ( i + 3 , 8 + j ) . Value = vRes (i
, j)
57 Next j
58 Next i
59 End Sub

The VBA code in the module MInterface is the macro assigned to the
button. The sub procedure gets input data from the worksheet, calls other
functions to bootstrap the yield curve, and extracts the output data to the
worksheet. In Lines 45, a new object of the class CBootstrapper is created.
In Line 46, the object is initialized with data from the worksheet. In Lines 49–
50, the functions BuildCurve and CreateOutput are called to bootstrap
the yield curve and get the output, respectively.
The input and output of the OO bootstrapping tool are identical to those
of the bootstrapping tool implemented in Chapter 8. If we click the button
of this tool, we see the results shown in Figure 8.1.

Exercise 12.1. In the MInterface module, we see that the three functions
Initialize, BuildCurve, and CreateOutput are called in sequence. How-
ever, it is possible that some users may not call these functions in sequence.
Modify the VBA code of the class CBootstrapper by doing the following:

(a) Add two properties named bInitialized and bCurveBuilt and assign
false to them.
(b) Modify the function Initialize by assigning true to the variable
bInitalized at the end of this function
Summary 249

(c) Modify the function BuildCurve by checking whether bInitialized

is true. If bInitialized is false, raise an error. Also assign true to the
variable bCurveBuilt at the end of this function.
(d) Modify the function CreateOutput by checking whether bCurveBuilt
is true. If bCurveBuilt is false, raise an error.

Exercise 12.2. In Exercise 12.1, you are asked to modify the code of the
class CBootstrapper to address the issue that the three functions have to
be called in sequence. However, users still need to call the three functions
explicitly. We can modify the code such that users only need to call the func-
tions Initialize and CreateOutput, which calls the other function auto-
matically, if necessary. Modify the VBA code of the class CBootstrapper
and the module MInterface as follows:

(a) Add two properties named bInitialized and bCurveBuilt and assign
false to them.
(b) Modify the function Initialize by assigning true to the variable
bInitalized at the end of this function.
(c) Modify the function BuildCurve by checking whether bInitialized
is true. If bInitialized is false, raise an error. Also assign true to the
variable bCurveBuilt at the end of this function.
(d) Modify the function CreateOutput by checking whether bCurveBuilt
is true. If bCurveBuilt is false, call the function BuildCurve.
(e) Modify the function Button1_Click in the module MInterface by re-
moving the function call cBT.BuildCurve.

12.4 Summary
In this chapter, we implemented a bootstrapping tool using OO techniques.
We implemented three classes, each of which models an aspect of the
bootstrapping algorithm. In particular, we implemented a yield curve class,
a swap class, and a bootstrapper class. The advantage of using classes is that
VBA keeps a copy of the data contained in an object. By using classes, we
can create multiple objects of a class to contain different copies of data. For
250 Object-Oriented Programming

more information about class modules, readers are referred to Walkenbach

(2013a).
 A
Solutions to Selected Exercises

A.1 Introduction to VBA

Solution 1.2. We can create the module by right clicking the Modules
folder in the Project window. We can change the module name through the
Properties window.
Solution 1.4. The recorded macro looks like:
1 Sub Macro2 ()
2 '
3 ' Macro2 Macro
4 '
5
6 '
7 Range ( " A1 : B2 " ) . Select
8 Selection . Copy
9 Range ( " F1 " ) . Select
10 Selection . PasteSpecial Paste := xlPasteValues ,
Operation := xlNone , SkipBlanks _
11 := False , Transpose := False
12 Application . CutCopyMode = False
13 Selection . Style = " Percent "
14 Selection . NumberFormat = " 0.0% "
15 Selection . NumberFormat = " 0.00% "
16 End Sub

A.2 Excel® Objects

Solution 2.2. We can implement the sub procedure as follows:
1 Sub CountRowColumn ()
2 Debug . Print Rows . Count

251
 252 Solutions to Selected Exercises

3 Debug . Print Columns . Count

4 End Sub

Executing the above sub procedure gives the following output:

1 1048576
2 16384

Solution 2.4. We can write the sub procedure as follows:

1 Sub InputBoxDemo ()
2 Debug . Print Application . InputBox ( prompt := " Input a
number " , Type :=1) + 5
3 End Sub

Solution 2.6. The first and second statements are true. The third state-
ment is false. If the Visual Basic code is part of an add-in, ThisWorkbook
won’t return the active workbook. In this case, ActiveWorkbook is the work-
book calling the add-in, whereas ThisWorkbook returns the add-in work-
book.
Solution 2.8. The Index property is read-only. We cannot assign a value
to it.
Solution 2.10. We can find out the address using the following code:
1 Sub FindAddress2 ()
2 Debug . Print Cells (20 , 51) . Address
3 End Sub

Executing the sub procedure gives "AY 20."

Solution 2.12. The output is:
1 $D$4 : $F$5
2 $D$4 : $F$5

Solution 2.14. We can write the sub procedure as follows:

1 Sub CopyTable ()
2 Workbooks ( " USLifeTable1999 -2001 Male . xls " ) . Sheets
(1) . Range ( " A25 : B134 " ) . Copy
3 ThisWorkbook . Sheets (1) . Range ( " F1 " ) . PasteSpecial
xlPasteValues
4 End Sub

Solution 2.16. We can calculate these quantities as follows:

 Variables, Data Types, and Scopes 253

1 Sub W o r k s h e e t F u n c t i o n D e m o 3 ()
2 Debug . Print Works heetFunc tion . Pi * 2
3 Debug . Print Works heetFunc tion . NormDist (1.5 , 0 , 1 ,
True )
4 Debug . Print Works heetFunc tion . NormInv (0.025 , 0 ,
1)
5 End Sub

Executing the code produces the following output:

1 6.28318530717959
2 0.933 19279873 1142
3 -1.95996398454005

A.3 Variables, Data Types, and Scopes

Solution 3.2. We need to use the Set statement to assign an object to
an Object variable.
Solution 3.4. The code does not contain any errors.
Solution 3.6. We can write the sub procedure as follows:
1 Sub WorksheetDemo6 ()
2 Dim ws As Worksheet
3 Set ws = Worksheets . Add
4 ws . Name = " SheetB "
5 Debug . Print ActiveSheet . Name
6 End Sub

Solution 3.8. The base can only be 0 or 1.

Solution 3.10. We can write the sub procedure as follows:
1 Sub ArrayDemo7 ()
2 Dim arM (1 To 2 , 1 To 2 , 1 To 2) As Integer
3
4 arM (1 , 1 , 1) = 3
5 arM (1 , 1 , 2) = 4
6 arM (1 , 2 , 1) = 4
7 arM (1 , 2 , 2) = 5
8 arM (2 , 1 , 1) = 4
9 arM (2 , 1 , 2) = 5
 254 Solutions to Selected Exercises

10 arM (2 , 2 , 1) = 5
11 arM (2 , 2 , 2) = 6
12
13 Debug . Print arM (1 , 1, 2) , arM (1 , 2, 1)
14 Debug . Print arM (1 , 2, 1) , arM (1 , 2, 2)
15 Debug . Print arM (2 , 2, 1) , arM (2 , 2, 2)
16 Debug . Print arM (2 , 2, 1) , arM (2 , 2, 2)
17 End Sub

Executing the code gives the following output:

1 4 4
2 4 5
3 5 6
4 5 6

Solution 3.12. Objects cannot be defined as constants. Defining

constants also requires constant expression. In this procedure, conBook,
conCell, and conCellsInRow are not valid.
Solution 3.14. We cannot assign new values to constants.
Solution 3.16. We can write the sub procedure as follows:
1 Sub StringDemo6 ()
2 Dim arV As Variant
3 Dim arWord () As String
4 Dim arCount () As Integer
5 Dim i As Integer
6 Dim j As Integer
7 Dim intMax As Integer
8 Dim iL As Integer
9 Dim iU As Integer
10 Dim bFound As Boolean
11 Dim iInd As Integer
12
13 arV = RandomWords
14 iL = LBound ( arV )
15 iU = UBound ( arV )
16
17 ReDim arWord (1 To 1)
18 ReDim arCount (1 To 1)
19 arWord (1) = arV ( iL )
20 arCount (1) = 1
21
22 For i = iL + 1 To iU
23 bFound = False
 Variables, Data Types, and Scopes 255

24 For j = 1 To UBound ( arWord )

25 If StrComp ( arV ( i ) , arWord ( j ) ,
vbTextCompare ) = 0 Then
26 bFound = True
27 iInd = j
28 End If
29 Next j
30
31 If bFound Then
32 arCount ( iInd ) = arCount ( iInd ) + 1
33 Else
34 ReDim Preserve arWord (1 To UBound ( arWord )
+ 1)
35 ReDim Preserve arCount (1 To UBound (
arCount ) + 1)
36 arWord ( UBound ( arWord ) ) = arV ( i )
37 arCount ( UBound ( arCount ) ) = 1
38 End If
39 Next i
40
41 intMax = 0
42 For i = 1 To UBound ( arWord )
43 If arCount ( i ) > intMax Then
44 intMax = arCount ( i )
45 End If
46 Next i
47
48 Dim dSum As Double
49 For i = 1 To UBound ( arWord )
50 If arCount ( i ) = intMax Then
51 Debug . Print arWord ( i ) ; arCount ( i )
52 End If
53 dSum = dSum + arCount ( i )
54 Next i
55
56 Debug . Print dSum
57 End Sub

Executing the above code produces the following output:

1 bcbd 6
2 1000

Solution 3.18. We can write the sub procedure as follows:

 256 Solutions to Selected Exercises

1 Sub StringDemo11 ()
2 Dim strA As String
3 Dim strB As String
4 Dim intA As Integer
5 strA = " This car is beautiful . That car is small .
Your car is big . "
6
7 intA = InStr ( strA , " car " )
8 strB = Mid ( strA , 1 , intA ) & Replace ( strA , " car " ,
" house " , intA + 1 , 1)
9 Debug . Print strB
10 Debug . Print Len ( strA )
11 Debug . Print Len ( strB )
12 End Sub

Solution 3.20. We can write the sub procedure as follows:

1 Sub StringDemo14 ()
2 Dim strA As String
3 strA = " 1.1 ,0.2 ,3.14 ,4.2 ,0.5 ,4.8 ,1.3 ,6.2 ,1 "
4
5 Dim arB () As String
6 Dim arC () As Double
7
8 arB = Split ( strA , " ," )
9 Dim i As Integer
10
11 ReDim arC ( LBound ( arB ) To UBound ( arB ) )
12 For i = LBound ( arB ) To UBound ( arB )
13 arC ( i ) = CDbl ( arB ( i ) )
14 Debug . Print arC ( i ) ,
15 Next i
16 End Sub

Solution 3.22. We can write the sub procedure as follows:

1 Sub DateDemo6 ()
2 Dim datBeg As Date
3 Dim datEnd As Date
4 Dim arA () As Date
5 Dim intW As Integer
6 Dim i As Integer
7 datBeg = #1/1/2016#
8 datEnd = #5/31/2016#
9 intW = DateDiff ( " w " , datBeg , datEnd )
 Variables, Data Types, and Scopes 257

10 ReDim arA (1 To intW + 1)

11 For i = 1 To intW + 1
12 arA ( i ) = DateAdd ( " ww " , i - 1 , datBeg )
13 Debug . Print arA ( i )
14 Next i
15 End Sub

The output is:

1 1/1/2016
2 1/8/2016
3 1/15/2016
4 1/22/2016
5 1/29/2016
6 2/5/2016
7 2/12/2016
8 2/19/2016
9 2/26/2016
10 3/4/2016
11 3/11/2016
12 3/18/2016
13 3/25/2016
14 4/1/2016
15 4/8/2016
16 4/15/2016
17 4/22/2016
18 4/29/2016
19 5/6/2016
20 5/13/2016
21 5/20/2016
22 5/27/2016

Solution 3.24. We can write the function as follows:

1 Function IsMLK ( dat As Date )
2 Dim m As Integer
3 Dim d As Integer
4 Dim w As Integer
5 m = Month ( dat )
6 d = Day ( dat )
7 w = Weekday ( dat )
8
9 If m = 1 And w = 2 And ( d >= 15 And d <= 21) Then
10 IsMLK = True
11 Else
 258 Solutions to Selected Exercises

12 IsMLK = False
13 End If
14 End Function

The output is
1 True
2 True
3 False

A.4 Operators and Control Structures

Solution 4.2. The output is 2016.
Solution 4.4. We will get a type mismatch error. The string "abc" cannot
be converted to a number.
Solution 4.6. The output is
1 True
2 False

Solution 4.8. It is an infinite loop. Executing the code may freeze your
computer.
Solution 4.10. We can write the sub procedure as follows:
1 Sub IfThenDemo4 ()
2 Dim intDay As Integer
3 Dim intM As Integer
4 Dim intWeekDay As Integer
5
6 intM = Month ( Date )
7 intDay = Day ( Date )
8 intWeekDay = Weekday ( Date )
9
10 ' Labor Day is the first Monday in September
11 If intM = 9 And intDay < 8 And intWeekDay = 2
Then
12 Debug . Print " Today is Labor day "
13 Else
14 Debug . Print " Today is not Labor day "
15 End If
16 End Sub
 Operators and Control Structures 259

Solution 4.12. We can write the sub procedure as follows:

1 Sub SelectDemo4 ()
2 Select Case VarType ( Range ( " A1 " ) . Value )
3 Case vbDouble
4 Debug . Print " Double "
5 Case vbString
6 Debug . Print " String "
7 Case vbCurrency
8 Debug . Print " Currency "
9 Case vbBoolean
10 Debug . Print " Boolean "
11 Case Else
12 Debug . Print " Other "
13 End Select
14 End Sub

Solution 4.14. We can calculate the sum as follows:

1 Sub ForNextDemo3 ()
2 Dim dSum As Double
3 Dim n As Integer
4
5 For n = 10 To 100
6 dSum = dSum + n ^ 2 + 3 * n ^ 3
7 Next n
8
9 Debug . Print dSum
10 End Sub

Executing the code gives 76839490.

Solution 4.16. We can create the array as follows:
1 Sub ForNextDemo6 ()
2 Dim vA (1 To 30) As Double
3 Dim n As Integer
4
5 For n = LBound ( vA ) To UBound ( vA ) Step 3
6 vA ( n ) = 2
7 vA ( n + 1) = 1
8 vA ( n + 2) = 3
9 Next n
10
11 Debug . Print vA (1) ; vA (2) ; vA (3) ; vA (4) ; vA (5) ; vA
(6)
12 End Sub
 260 Solutions to Selected Exercises

Solution 4.18. We can write the sub procedure as follows:

1 Sub ForNextDemo9 ()
2 Dim mA (1 To 100 , 1 To 100) As Integer
3 Dim i As Integer
4 Dim j As Integer
5
6 For i = 1 To 100
7 For j = 1 To 100
8 mA (i , j ) = i + j - 2
9 ActiveSheet . Cells (i , j ) = mA (i , j )
10 Next j
11 Next i
12 End Sub

Solution 4.20. The For-Next statements are not nested properly.

Solution 4.22. The loop is infinite and will not stop.

A.5 Functions, Events, and File IO

Solution 5.2. We can write the VBA function as follows:
1 Function DNorm ( x As Double , mu As Double , sigma As
Double )
2 Dim dPi As Double
3 dPi = Works heetFunc tion . Pi
4 DNorm = Exp ( -( x - mu ) ^ 2 / (2 * sigma ^ 2) ) / (
Sqr (2 * dPi ) * sigma )
5 End Function

The output of the sub procedure is

1 0.398 94228040 1433
2 0.106 48266850 7451
3 0.398 94228040 1433
4 0.106 48266850 7451

Solution 5.4. We can write the function procedure as follows:

1 Function Factorial ( n As Long )
2 If n = 0 Then
 Functions, Events, and File IO 261

3 Factorial = 1
4 Else
5 Factorial = n * Factorial ( n - 1)
6 End If
7 End Function

The output of the sub procedure is

1 1
2 3628800
3 2.43290200817664 E +18

Solution 5.6. The first function always returns the same random num-
bers. However, the second function returns different random numbers for
different calls.
Solution 5.8. We can add event-handlers for the NewChart and New-
Sheet events as follows:
1 Private Sub Workbook _NewChar t ( ByVal Ch As Chart )
2 If ThisWorkbook . Sheets . Count >= 5 Then
3 Application . DisplayAlerts = False
4 Ch . Delete
5 Application . DisplayAlerts = True
6 End If
7 End Sub
8
9 Private Sub Workbook _NewShee t ( ByVal Sh As Object )
10 If Sheets . Count > 5 Then
11 Application . DisplayAlerts = False
12 Sh . Delete
13 Application . DisplayAlerts = True
14 End If
15 End Sub

Solution 5.10. We can add an event-handler for the Active event of the
worksheet as follows:
1 Private Sub Wo rk she et _A cti va te ()
2 MsgBox ActiveSheet . Name
3 End Sub

Solution 5.12. We can write the sub procedure as follows:

1 Sub FileDemo6 ()
2 Dim strFile As String
 262 Solutions to Selected Exercises

3 Dim strV () As String

4 Dim j As Integer
5 Dim n As Integer
6 strFile = ThisWorkbook . Path & " : words . txt "
7
8 n = 1
9 Open strFile For Input As #1
10 Do Until EOF (1)
11 Line Input #1 , Line
12 strV = Split ( Line , " " )
13 For j = 0 To UBound ( strV )
14 Cells (n , j + 1) . Value = strV ( j )
15 Next j
16 n = n + 1
17 Loop
18 Close #1
19 End Sub

A.6 Error Handling and Debugging

Solution 6.2. The code contains runtime errors. For example, when the
active sheet is a chart sheet or the active sheet is protected, the code will
raise runtime errors.
Solution 6.4. We use the watch window to find out the value at the
900th iteration. The value is 0.68491667509079.

A.7 Generating Payment Schedules

Solution 7.2. The adjusted dates are given by the following:

Payment Date Following Modified Following Proceeding

January 16, 2016 January 19 January 19 January 15
May 30, 2016 May 31 May 31 May 27
November 25, 2016 November 28 November 28 November 24

Solution 7.4. We can implement the function as follows:

 Generating Payment Schedules 263

1 Function GetHolidayNY ( ByVal lY As Long , ByVal

sHoliday As String ) As Variant
2 Dim adHoliday (1 To 2) As Date
3 Dim datHoliday As Date
4 Dim datObserved As Date
5 Dim iWD As Integer
6
7 Select Case sHoliday
8 Case " NewYear "
9 datHoliday = DateSerial ( lY , 1 , 1)
10 If Weekday ( datHoliday ) = vbSaturday Then
11 datObserved = DateSerial ( lY - 1 , 12 ,
31)
12 ElseIf Weekday ( datHoliday ) = vbSunday
Then
13 datObserved = DateSerial ( lY , 1 , 2)
14 Else
15 datObserved = datHoliday
16 End If
17 Case " MLK "
18 datHoliday = DateSerial ( lY , 1 , 1)
19 iWD = Weekday ( datHoliday )
20 If iWD <= vbMonday Then
21 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 14)
22 Else
23 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 21)
24 End If
25 datObserved = datHoliday
26 Case " President "
27 datHoliday = DateSerial ( lY , 2 , 1)
28 iWD = Weekday ( datHoliday )
29 If iWD <= vbMonday Then
30 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 14)
31 Else
32 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 21)
33 End If
34 datObserved = datHoliday
35 Case " Memorial "
36 datHoliday = DateSerial ( lY , 5 , 31)
37 iWD = Weekday ( datHoliday )
 264 Solutions to Selected Exercises

38 If iWD = vbSunday Then

39 datHoliday = AddDays ( datHoliday , -6)
40 Else
41 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD )
42 End If
43 datObserved = datHoliday
44 Case " Independence "
45 datHoliday = DateSerial ( lY , 7 , 4)
46 If Weekday ( datHoliday ) = vbSaturday Then
47 datObserved = DateSerial ( lY , 7 , 3)
48 ElseIf Weekday ( datHoliday ) = vbSunday
Then
49 datObserved = DateSerial ( lY , 7 , 5)
50 Else
51 datObserved = datHoliday
52 End If
53 Case " Labor "
54 datHoliday = DateSerial ( lY , 9 , 1)
55 iWD = Weekday ( datHoliday )
56 If iWD <= vbMonday Then
57 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD )
58 Else
59 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 7)
60 End If
61 datObserved = datHoliday
62 Case " Columbus "
63 datHoliday = DateSerial ( lY , 10 , 1)
64 iWD = Weekday ( datHoliday )
65 If iWD <= vbMonday Then
66 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 7)
67 Else
68 datHoliday = AddDays ( datHoliday ,
vbMonday - iWD + 14)
69 End If
70 datObserved = datHoliday
71 Case " Veterans "
72 datHoliday = DateSerial ( lY , 11 , 11)
73 If Weekday ( datHoliday ) = vbSaturday Then
74 datObserved = DateSerial ( lY , 11 , 10)
75 ElseIf Weekday ( datHoliday ) = vbSunday
Then
 Generating Payment Schedules 265

76 datObserved = DateSerial ( lY , 11 , 12)

77 Else
78 datObserved = datHoliday
79 End If
80 Case " Thanksgiving "
81 datHoliday = DateSerial ( lY , 11 , 1)
82 iWD = Weekday ( datHoliday )
83 If iWD <= vbThursday Then
84 datHoliday = AddDays ( datHoliday ,
vbThursday - iWD + 21)
85 Else
86 datHoliday = AddDays ( datHoliday ,
vbThursday - iWD + 28)
87 End If
88 datObserved = datHoliday
89 Case " Christmas "
90 datHoliday = DateSerial ( lY , 12 , 25)
91 If Weekday ( datHoliday ) = vbSaturday Then
92 datObserved = DateSerial ( lY , 12 , 24)
93 ElseIf Weekday ( datHoliday ) = vbSunday
Then
94 datObserved = DateSerial ( lY , 12 , 26)
95 Else
96 datObserved = datHoliday
97 End If
98 Case Else
99 Err . Raise Number :=1008 , Source := " MHoliday
. GetHoliday " , Description := " Unknown
holiday . "
100 End Select
101
102 adHoliday (1) = datHoliday
103 adHoliday (2) = datObserved
104 GetHolidayNY = adHoliday
105 End Function

Solution 7.6. We can implement the function as follows:

1 Function EasterSunday ( ByVal lY As Long ) As Date
2 Dim lN As Long
3 Dim lG As Long
4 Dim lE As Long
5 Dim lH As Long
6 Dim lSOL As Long
7 Dim lLUN As Long
 266 Solutions to Selected Exercises

8 Dim lV As Long
9 Dim lC As Long
10 Dim lR As Long
11 Dim lS As Long
12
13 lN = 7 - ( lY + lY \ 4 - lY \ 100 + lY \ 400 - 1)
Mod 7
14 lG = 1 + lY Mod 19
15 lE = (11 * lG - 10) Mod 30
16 lH = lY \ 100
17 lSOL = lH - lH \ 4 - 12
18 lLUN = ( lH - 15 - ( lH - 17) \ 25) \ 3
19 lV = lE \ 24 - lE \ 25 + ( lG \ 12) * ( lE \ 25 -
lE \ 26)
20 lE = (11 * lG - 10) Mod 30 - ( lSOL - lLUN ) Mod 30
+ lV
21 If lE < 24 Then
22 lR = 45 - lE
23 Else
24 lR = 75 - lE
25 End If
26 lC = 1 + ( lR + 2) Mod 7
27 lS = lR + (7 + lN - lC ) Mod 7
28
29 EasterSunday = DateSerial ( lY , 3 , lS )
30 End Function

A.8 Bootstrapping Yield Curves

Solution 8.2. We can write the function PvSwap as follows:
1 Function PvSwap ( ByVal sTenor As String , ByVal dRate
As Double ) As Double
2 Dim iNum As Integer
3 Dim sPeriod As String
4 Dim i As Integer
5 Dim dDF As Double
6
7 Dim vFixDates As Variant
8 Dim dFixPV As Double
 Bootstrapping Yield Curves 267

9 ExtractPeriod sIn := sTenor , iNum := iNum , sPeriod :=

sPeriod
10 ' calculate fixed leg cash flows and present
value
11 vFixDates = GenerateSchedule ( datSet , sFixFreq ,
sTenor , sCalendar , sBDC )
12 For i = LBound ( vFixDates ) + 1 To UBound ( vFixDates
)
13 dDF = LogLinear ( vFixDates ( i ) )
14 dFixPV = dFixPV + TFrac ( vFixDates ( i - 1) ,
vFixDates ( i ) , sFixDCC ) * dRate * dDF
15 Next i
16
17 PvSwap = dFixPV - LogLinear ( datSet ) + LogLinear (
vFixDates ( UBound ( vFixDates ) ) )
18 End Function

Solution 8.4. We can modify the sub procedure as follows:

1 Sub Button1_Click ()
2 Dim vRes As Variant
3 Initialize rngRate := Range ( " Rate " ) , rngParam :=
Range ( " Param " )
4 CheckInput
5 BuildCurve
6 vRes = CreateOutput
7
8 Dim iRows As Integer
9 Dim i As Integer
10 Dim j As Integer
11 iRows = Sheet1 . Range ( " H3 " ) . End ( xlDown ) . Row
12 Sheet1 . Range ( Cells (3 , 8) , Cells ( iRows , 12) ) .
ClearContents
13 For i = LBound ( vRes , 1) To UBound ( vRes , 1)
14 For j = 0 To 4
15 Sheet1 . Cells ( i + 3 , 8 + j ) . Value = vRes (i
, j)
16 Next j
17 Next i
18
19 ' plot the zero rates and forward rates
20 Dim bTag As Boolean
21 Dim csPlot As Chart
22 bTag = False
23 For i = 1 To Charts . Count
 268 Solutions to Selected Exercises

24 If StrComp ( Charts ( i ) . Name , " Plot " ) = 0 Then

25 Set csPlot = Charts ( i )
26 bTag = True
27 End If
28 Next i
29 If Not bTag Then
30 Set csPlot = Charts . Add ( after := ActiveSheet )
31 csPlot . Name = " Plot "
32 End If
33
34 Dim seriesA As Series
35 For Each seriesA In csPlot . SeriesCollection
36 seriesA . Delete
37 Next seriesA
38
39 Set seriesA = csPlot . SeriesCollection . NewSeries
40 seriesA . XValues = " = Curve ! $L$3 : $L$ " & CStr ( UBound
( vRes , 1) + 3)
41 seriesA . Values = " = Curve ! $J$3 : $J$ " & CStr ( UBound (
vRes , 1) + 3)
42 seriesA . ChartType = xlXYScatterLines
43 seriesA . Name = " Zero Rate "
44 Set seriesA = csPlot . SeriesCollection . NewSeries
45 seriesA . XValues = " = Curve ! $L$3 : $L$ " & CStr ( UBound
( vRes , 1) + 3)
46 seriesA . Values = " = Curve ! $K$3 : $K$ " & CStr ( UBound (
vRes , 1) + 3)
47 seriesA . ChartType = xlXYScatterLines
48 seriesA . Name = " Forward Rate "
49
50 End Sub

A.9 Generating Risk-Neutral Scenarios

Solution 9.2. We can revise the sub procedure as follows:
1 Sub Button1_Click ()
2 Dim rngParam As Range
3 Dim rngFirstFR As Range
4
5 Set rngParam = Sheet1 . Range ( " Param " )
 Generating Risk-Neutral Scenarios 269

6 Set rngFirstFR = Sheet1 . Range ( " FirstFR " )

7
8 Dim aFR () As Double
9 Dim lFirstRow As Long
10 Dim lLastRow As Long
11
12 lFirstRow = rngFirstFR . Cells (1 , 1) . Row
13 lLastRow = rngFirstFR . Cells (1 , 1) . End ( xlDown ) . Row
14
15 ReDim aFR (1 To lLastRow - lFirstRow + 1)
16 Dim i As Long
17 For i = 1 To lLastRow - lFirstRow + 1
18 aFR ( i ) = rngFirstFR . Cells (1 , 1) . Offset ( i - 1 ,
0) . Value
19 Next i
20
21 MGenerator . Initialize Sigma := rngParam . Cells (1 , 1)
. Value , _
22 Horizon := rngParam . Cells (2 , 1) . Value , _
23 TimeStep := rngParam . Cells (3 , 1) . Value , _
24 NumPath := rngParam . Cells (4 , 1) . Value , _
25 Seed := rngParam . Cells (5 , 1) . Value , _
26 Precision := rngParam . Cells (6 , 1) . Value , _
27 FR := aFR
28
29 Dim vSce As Variant
30 vSce = MGenerator . Generate
31
32 ' write scenarios to Sheet2
33 Sheet2 . Cells . ClearContents
34 Dim j As Long
35 For i = 1 To UBound ( vSce , 1)
36 For j = 1 To UBound ( vSce , 2)
37 Sheet2 . Cells (i , j ) . Value = vSce (i , j )
38 Next j
39 Next i
40
41 ' calculate summary statistics
42 Dim lRow As Long
43 Dim lCol As Long
44 lRow = UBound ( vSce , 1)
45 lCol = UBound ( vSce , 2)
46 Dim aTmp () As Double
47 ReDim aTmp (1 To lRow )
48
 270 Solutions to Selected Exercises

49 Dim dDelta As Double

50 dDelta = MGenerator . GetDelta
51 For j = 1 To lCol
52 For i = 1 To lRow
53 aTmp ( i ) = W orksheet Function . lN ( vSce (i , j )
)
54 Next i
55
56 Sheet2 . Cells ( lRow + 2 , j ) . Value =
Works heetFunc tion . Average ( aTmp ) / dDelta
57 Sheet2 . Cells ( lRow + 3 , j ) . Value =
Works heetFunc tion . StDev ( aTmp ) / Sqr ( dDelta
)
58
59 Next j
60 End Sub

The function GetDelta is a function of the module MGenerator and is

implemented as
1 Function GetDelta () As Double
2 GetDelta = CalDelta ( sTimeStep )
3 End Function

A.10 Valuing a GMDB

Solution 10.2. We can write the test procedure as follows:
1 Sub TestInitialize ()
2 Dim rngMTMale As Range
3 Dim rngMTFemale As Range
4
5 Set rngMTMale = Range ( " MTMale " )
6 Set rngMTFemale = Range ( " MTFemale " )
7
8 Initialize rngMTMale := rngMTMale , rngMTFemale :=
rngMTFemale
9
10 PrintInfo
11 End Sub
 Valuing a GMDB 271

Solution 10.4. We can implement the test procedure as follows:

1 Sub TestProject ()
2 Dim bFound As Boolean
3 bFound = False
4 Dim i As Integer
5 For i = 1 To Sheets . Count
6 If StrComp ( Sheets ( i ) . Name , " Temp " ) = 0 Then
7 bFound = True
8 Exit For
9 End If
10 Next i
11 Dim ws As Worksheet
12 If Not bFound Then
13 Set ws = Worksheets . Add
14 ws . Name = " Temp "
15 End If
16
17 MLifeTable . Initialize rngMTMale := Range ( " MTMale " ) ,
rngMTFemale := Range ( " MTFemale " )
18 With Range ( " Param " )
19 MGMDB . In iti al iz eCo nt ra ct _
20 Gender :=. Cells (1 , 1) . Value , _
21 DOB :=. Cells (2 , 1) . Value , _
22 Premium :=. Cells (3 , 1) . Value , _
23 MEFee :=. Cells (4 , 1) . Value , _
24 FundFee :=. Cells (5 , 1) . Value , _
25 IE :=. Cells (6 , 1) . Value , _
26 OE :=. Cells (7 , 1) . Value , _
27 Term :=. Cells (8 , 1) . Value , _
28 Feature :=. Cells (9 , 1) . Value , _
29 LapseRate :=. Cells (10 , 1) . Value , _
30 ValDate :=. Cells (11 , 1) . Value , _
31 NumSce :=. Cells (12 , 1) . Value
32 End With
33
34 MGMDB . In iti al iz eSc en ar io Sce := Sheets ( " Base " ) , FR
:= Range ( " FRBase " )
35
36 Project dShock :=0 , ind :=1
37 Set ws = Sheets ( " Temp " )
38 ws . Cells . ClearContents
39 ws . Cells (1 , 1) . Value = " ap "
40 ws . Cells (2 , 1) . Value = " aq "
41 ws . Cells (3 , 1) . Value = " ad "
 272 Solutions to Selected Exercises

42 ws . Cells (4 , 1) . Value = " aL "

43 ws . Cells (5 , 1) . Value = " aA "
44 ws . Cells (6 , 1) . Value = " aRC "
45 ws . Cells (7 , 1) . Value = " aGD "
46 ws . Cells (8 , 1) . Value = " aDB "
47 For i = 0 To 360
48 ws . Cells (1 , i + 2) . Value = ap ( i )
49 If i > 0 Then
50 ws . Cells (2 , i + 2) . Value = aq ( i )
51 End If
52 ws . Cells (3 , i + 2) . Value = ad ( i )
53 ws . Cells (4 , i + 2) . Value = aL ( i )
54 ws . Cells (5 , i + 2) . Value = aA ( i )
55 ws . Cells (6 , i + 2) . Value = aRC ( i )
56 ws . Cells (7 , i + 2) . Value = aGD ( i )
57 ws . Cells (8 , i + 2) . Value = aDB ( i )
58 Next i
59 End Sub

Solution 10.6. We can modify the sub procedure as follows:

1 Sub Button1_Click ()
2 Dim dFMVBase As Double
3 Dim dFMVEU As Double
4 Dim dFMVED As Double
5 Dim dFMVIU As Double
6 Dim DFMVID As Double
7
8 Dim dStartTime As Double
9 dStartTime = Timer
10
11 MLifeTable . Initialize rngMTMale := Range ( " MTMale " ) ,
rngMTFemale := Range ( " MTFemale " )
12 With Range ( " Param " )
13 MGMDB . In iti al iz eCo nt ra ct _
14 Gender :=. Cells (1 , 1) . Value , _
15 DOB :=. Cells (2 , 1) . Value , _
16 Premium :=. Cells (3 , 1) . Value , _
17 MEFee :=. Cells (4 , 1) . Value , _
18 FundFee :=. Cells (5 , 1) . Value , _
19 IE :=. Cells (6 , 1) . Value , _
20 OE :=. Cells (7 , 1) . Value , _
21 Term :=. Cells (8 , 1) . Value , _
22 Feature :=. Cells (9 , 1) . Value , _
23 LapseRate :=. Cells (10 , 1) . Value , _
 Connecting to Databases 273

24 ValDate :=. Cells (11 , 1) . Value , _

25 NumSce :=. Cells (12 , 1) . Value
26 End With
27
28 MGMDB . In iti al iz eSc en ar io Sce := Sheets ( " Base " ) , FR
:= Range ( " FRBase " )
29 dFMVBase = MGMDB . FMV (0)
30 dFMVEU = MGMDB . FMV (0.01)
31 dFMVED = MGMDB . FMV ( -0.01)
32
33 MGMDB . In iti al iz eSc en ar io Sce := Sheets ( " AllUp " ) , FR
:= Range ( " FRAllUp " )
34 dFMVIU = MGMDB . FMV (0)
35
36 MGMDB . In iti al iz eSc en ar io Sce := Sheets ( " AllDn " ) , FR
:= Range ( " FRAllDn " )
37 DFMVID = MGMDB . FMV (0)
38
39 Sheet1 . Range ( " O3 " ) . Value = dFMVBase
40 Sheet1 . Range ( " O4 " ) . Value = ( dFMVEU - dFMVED ) /
0.02
41 Sheet1 . Range ( " O5 " ) . Value = ( dFMVIU - DFMVID ) / 20
42 Sheet1 . Range ( " O6 " ) . Value = Timer - dStartTime
43 End Sub

A.11 Connecting to Databases

Solution 11.2. We can write the sub procedure as follows:
1 Sub TestCreateTable ()
2 Dim sFileName As String
3
4 sFileName = ThisWorkbook . Path & " \ va . accdb "
5 Connect ( sFileName )
6
7 CreateTable Data := Sheets ( " Main " ) . Range ( " O2 : Q5 " ) ,
_
8 TableName := " Grade " , PrimaryKey := " Name "
9
10 Disconnect
11 End Sub
 274 Solutions to Selected Exercises

Solution 11.4. We can write the sub procedure as follows:

1 Sub Tes tExtract Record ()
2 Dim sFileName As String
3 Dim sSQL As String
4 Dim vRes As Variant
5
6 sFileName = ThisWorkbook . Path & " \ va . accdb "
7 Connect ( sFileName )
8
9 sSQL = " SELECT * FROM Grade WHERE [ Grade ]= 'B ' "
10 vRes = ExtractRecord ( sSQL )
11
12 Disconnect
13
14 Dim i As Integer
15 Dim j As Integer
16 For i = 0 To UBound ( vRes , 1)
17 For j = 0 To UBound ( vRes , 2)
18 Debug . Print vRes (i , j ) & " - " ;
19 Next j
20 Debug . Print " "
21 Next i
22 End Sub

Executing the sub procedure gives the following output:

1 First Name - Carroll -
2 Last Name - Paul -
3 Grade - B -
References

Bauer, D., Kling, A., and Russ, J. (2008). A universal pricing framework
for guaranteed minimum benefits in variable annuities. ASTIN Bulletin,
38(2):621–651.

Black, F. and Scholes, M. (1973). The pricing of options and corporate

liabilities. Journal of Political Economy, 81(3):637–54.

Bovey, R., Wallentin, D., Bullen, S., and Green, J. (2009). Professional Excel
Development: The Definitive Guide to Developing Applications Using Mi-
crosoft Excel, VBA, and .NET. Addison-Wesley Professional, 2nd edition.

Boyle, P. and Hardy, M. (1997). Reserving for maturity guarantees: Two

approaches. Insurance: Mathematics and Economics, 21(2):113–127.

Brown, R. A., Campbell, T. A., and Gorski, L. M. (2002). Valuation and capital
requirements for guaranteed benefits in variable annuities. Record, 28(3).

Carmona, R. and Durrleman, V. (2005). Generalizing the Black-Scholes

formula to multivariate contingent claims. Journal of Computational
Finance, 9(2):43–67.

Cathcart, M. J., Lok, H. Y., McNeil, A. J., and Morrison, S. (2015). Calculating
variable annuity liability “greeks” using Monte Carlo simulation. ASTIN
Bulletin, 45(2):239–266.

Devroye, L. (1986). Non-Uniform Random Variate Generation. Springer,

New York, NY.

Evans, M., Hastings, N., and Peacock, B. (2000). Statistical Distributions.

John Wiley & Sons, Inc., Hoboken, New Jersey, 3rd edition.

Forta, B. (2012). SQL in 10 Minutes, Sams Teach Yourself. Sams Publishing,

4th edition.

Fu, M. C. (2015). Stochastic gradient estimation. In Handbook of Simulation

Optimization, volume 216 of International Series in Operations Research
& Management Science, pages 105–147. Springer, New York, NY.

275
276 References

Gan, G. (2011). Data Clustering in C++: An Object-Oriented Approach. Data

Mining and Knowledge Discovery Series. Chapman & Hall/CRC Press,
Boca Raton, FL, USA.

Gan, G. (2013). Application of data clustering and machine learning

in variable annuity valuation. Insurance: Mathematics and Economics,
53(3):795–801.

Gan, G. (2015a). Application of metamodeling to the valuation of large

variable annuity portfolios. In Proceedings of the Winter Simulation Con-
ference, pages 1103–1114.

Gan, G. (2015b). A multi-asset Monte Carlo simulation model for the val-
uation of variable annuities. In Proceedings of the Winter Simulation
Conference, pages 3162–3163.

Gan, G. and Lin, X. S. (2015). Valuation of large variable annuity portfo-

lios under nested simulation: A functional data approach. Insurance:
Mathematics and Economics, 62:138–150.

Gan, G. and Lin, X. S. (2016). Efficient Greek calculation of variable annuity

portfolios for dynamic hedging: A two-level metamodeling approach.
Revised and resubmitted to North American Actuarial Journal.

Gan, G., Ma, C., and Xie, H. (2014). Measure, Probability, and Mathematical
Finance: A Problem-Oriented Approach. John Wiley & Sons, Inc., Hoboken,
NJ.

Garrett, S. (2013). Introduction to the Mathematics of Finance. A Determin-

istic Approach. Butterworth-Heinemann, Waltham, MA, 2nd edition.

Garrett, S. (2015). Introduction to Actuarial and Financial Mathematical

Methods. Academic Press, Singapore.

Getz, K. and Gilbert, M. (2001). VBA Developer’s Handbook. Sybex, 2nd

edition.

Gupta, A. K. and Varga, T. (2002). An Introduction to Actuarial Mathematics.

Springer, New York, NY.

Hagan, P. S. and West, G. (2006). Interpolation methods for curve construc-

tion. Applied Mathematical Finance, 13(2):89–129.

Hagan, P. S. and West, G. (2008). Methods for constructing a yield curve.

WILMOTT magazine, May/June:70–81.
References 277

ISDA (2006). 2006 ISDA definitions. International Swaps and Derivatives

Association.

Lai, D. C. F., Tung, H. K. K., and Wong, M. C. S. (2010). Professional Financial

Computing Using Excel and VBA. John Wiley & Sons, Inc., Hoboken, NJ.

Martin, R. (2012). Excel Programming with VBA Starter. Packt Publishing,

Birmingham, UK.

Press, W., Teukolsky, S., Vetterling, W., and Flannery, B. (2002). Numerical
Recipes in C++: The Art of Scientific Computing. Cambridge University
Press, 2nd edition.

Richards, E. G. (2012). The Explanatory Supplement to The Astronomical

Almanac, chapter 15, pages 585–624. University Science Books.

Roman, S. (2002). Writing Excel Macros with VBA. O’Reilly Media, 2nd
edition.

Sheldon, T. J. and Smith, A. D. (2004). Market consistent valuation of life

assurance business. British Actuarial Journal, 10(3):543–626.

Sintes, A. (2001). Sams Teach Yourself Object Oriented Programming in 21

Days. Sams, Indianapolis, Indiana, 2nd edition.

Tale, S. (2016). SQL: The Ultimate Beginners Guide: Learn SQL Today. Cre-
ateSpace Independent Publishing Platform.

The Geneva Association Report (2013). Variable annuities —

an analysis of financial stability. Available online at: https:
//www.genevaassociation.org/media/618236/ga2013-variable_
annuities.pdf.

Urtis, T. (2015). Excel VBA 24-Hour Trainer. Wrox.

Walkenbach, J. (2013a). Excel 2013 Power Programming with VBA. John

Wiley & Sons, Inc., Hoboken, NJ.

Walkenbach, J. (2013b). Excel VBA Programming For Dummies. John Wiley

& Sons, Inc., Hoboken, NJ, 3rd edition.
Index

A database management system,

accumulation factor, 185 218
ACT/360, 134 day count convention, 133
ACT/365, 133 DBMS, 218
ACT/ACT, 133 delta, 205
ActiveX Data Objects, 218 derivative, 130
add-in, 24 discount factor, 163
ADO, 218 dollar delta, 205
arbitrage-free, 184 dynamic array, 48
array, 48 dynamic string, 57

B E
basis point, 199 Easter Sunday, 158
bisection method, 180 economic scenario generator, 184
bps, 205 Excel macro recorder, 13
bug, 115
business day convention, 134 F
fixed-length string, 57
C following, 134
calendar, 131 forward rate, 163
class, 11, 233 function procedure, 9, 95
class module, 10
collection, 17 G
comment, 15 GMAB, 200
compilation error, 109 GMDB, 200
connection string, 222 GMLB, 200
constant, 54 GMMB, 200
continous return, 185 GMWB, 200
cumulative density function, 186 Greek, 205
Gregorian calendar, 131
D guarantee, 200
database, 218

278
Index 279

I OOP, 17, 233

immediate window, 19
interest rate swap, 130, 165 P
inversion method, 187 path, 184
IRS, 130, 165 preceding, 134
primary key, 227
J probability density function, 186
Julian calendar, 131 property, 17
property procedures, 12
L
label, 82 R
leap year, 133 recursive function, 98
life table, 201 rho, 205
Locals windows, 115 risk-neutral, 184
logic error, 109 runtime error, 109
logical conjunction, 80
logical disjunction, 80 S
logical exclusion, 80 scenario, 184
logical inclusion, 80 schema, 223
logical opposite, 80 separate account, 200
loop, 86 solar calendar, 131
spot rate, 162
M SQL, 220
macro, 4 standard module, 10
market-consistent, 184 static variable, 70
method, 17 string, 57
modified following, 134 Structured Query Language, 220
module, 9 sub procedure, 9
modulus operator, 75 subaccount, 200
Monte Carlo simulation, 201
mortality, 201 T
30/360, 134
N
naming convention, 120 V
Newton’s method, 167 variable, 43, 73
normal density function, 40 variable annuity, 200
VBA, 3
O VBA constant, 36
object, 17 Visual Basic Editor, 4
object-oriented programming, 17
280 Index

W
Watch window, 115

Y
yield curve, 162

Z
zero-coupon bond, 162
zero-coupon rate, 162
Index of VBA Keywords

Symbols Byte, 45
*, 75 ByVal, 124
+, 75
-, 75 C
/, 75 Calculate, 30
;, 88 Case, 81, 85
=, 79 Case Else, 86
#, 77, 146 CDbl, 77
&, 78 Ceiling, 38
ˆ, 75 Cells, 28, 33
>, 79 Charts, 25, 28
>=, 79 Clear, 36
<, 79 ClearContents, 20
<=, 79 Close, 26, 27, 31
Column, 35
A Columns, 20, 21, 28
Activate, 20, 26, 30 Command, 219
ActiveCell, 21 CommandText, 225
ActiveChart, 25 Connection, 219
ActiveConnection, 225 Const, 54
ActiveSheet, 21, 25 Copy, 20, 31, 36
ActiveWorkbook, 21 Count, 27, 35, 38
Add, 27, 31 CountIf, 38
Address, 35 CREATE TABLE, 221
ADODB.Connection, 221 CStr, 78
ADODB.Recordset, 223 Currency, 45
And, 80
Application, 17, 21 D
Date, 45, 66, 84
B DateAdd, 66
Boolean, 45 DateDiff, 66
ByRef, 124, 153 DatePart, 66

281
282 Index of VBA Keywords

DateSerial, 66, 139 I

DateValue, 66 If, 81
Day, 66, 84 Index, 28
Debug.Print, 19 InputBox, 22
Delete, 36 InStr, 59, 62
Dim, 44 InStrRev, 59
DisplayAlerts, 21, 31 Integer, 45
Do, 87 Interior, 35
Double, 45 Is, 79
DROP TABLE, 220 IsAddin, 25
IsEmpty, 91
E IsNot, 79
EOF, 107 IsNumeric, 111
Err.Number, 113 Item, 27
Err.Raise, 151
Evaluate, 22 J
Execute, 225 Join, 59
Exit For, 90
Exit Sub, 110 L
EXP, 96 LBound, 49
Explicit, 9 LCase, 59
Left, 59
F Len, 58, 59
FileDialog, 21 Ln, 38
Filter, 59 Long, 45
FindFile, 22 Lookup, 38
Floor, 38 LSet, 59
Font, 35 LTrim, 59
For, 87
Format, 59, 66, 86 M
Formula, 35 Max, 38
FullName, 25 Mid, 59, 63
Min, 38
G Minute, 66
GetOpenFilename, 22 Mod, 75
GetRows, 223 Month, 66, 84
GoTo, 81 MonthName, 66

H N
Hour, 66 Name, 25
Index of VBA Keywords 283

New, 12 Replace, 59, 63

Next, 28, 87 Resume, 113
NormDist, 38 Right, 59
NormInv, 38 Rnd, 100
Not, 80 Row, 35
Nothing, 222 Rows, 20, 21
Now, 66 Run, 22
NumberFormat, 35
S
O Save, 26
Object, 45 SaveAs, 26, 30
Offset, 34 SELECT, 220
On Error, 113 Select, 30, 36, 81, 85
OnKey, 22 Set, 44
OnTime, 22 Sheets, 21, 25, 28
Open, 27, 31 Single, 45
OpenSchema, 223 Space, 59
Option, 9 Split, 59, 64
Option Base 1, 49 Static, 70
Option Compare, 60 StrComp, 59
Option Explicit, 47 String, 45
Option Private Module, 124 sub, 5
Or, 80
ORDER BY, 220 T
Text, 35
P Then, 81
Paste, 30 ThisWorkbook, 21
paste, 37 Time, 83
PasteSpecial, 30, 36 Trim, 59, 61
Path, 25 Type, 28
Pi, 38
Power, 38 U
Preserve, 52 UBound, 49
Print, 105 UCase, 59
Until, 87
R UsedRange, 28
RandBetween, 38
Range, 32 V
RecordSet, 219 Value, 35
ReDim, 52 Variant, 45
284 Index of VBA Keywords

VarType, 47, 86
vbBlue, 36
vbWhite, 36

W
WeekDay, 66, 84
WeekDayName, 66
WHERE, 220
While, 87
Workbook, 17, 23
Workbooks, 17
Worksheet, 17, 28, 32
WorkSheetFunction, 17
WorksheetFunction, 37
Worksheets, 17, 25

X
xlPasteAll, 36
Xor, 80

Y
Year, 66
year, 230