Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Optimal Designation of Hedging Relationships Under FASB Statement 133

Download as ps, pdf, or txt
Download as ps, pdf, or txt
You are on page 1of 13

Optimal Designation of Hedging Relationships

Under FASB Statement 133


Katherine Wyatt
Logic Based Systems Lab
Brooklyn College
City University of New York

June 13, 1999

Abstract
New accounting standards for financial derivatives mandate on-balance sheet
reporting and the designation of hedged item - hedging derivative pairs
for reporting offsetting gain or loss on financial statements. A program
that uses disjunctive, or logical, constraints to model the criteria for hedge
accounting across all types of financial instruments and against allowable
risks is presented. The program returns an optimal designation that satisfies
the hedge-accounting rules, and uses standard risk management reports as
input. A linear program models compliance for most hedges against market
risk.

Introduction
In Accounting for Derivative Instruments and Hedging Activities (Statement
of Financial Accounting Standards No. 133), the Financial Accounting Stan-
dards Board outlines the reporting requirements for derivative instruments
on financial statements released after June 15, 2000. All derivative instru-
ments will be reported at fair value on the balance sheet and hedge ac-
counting will be allowed only for designated item - derivative pairs. Choos-
ing hedge relationships that satisfy the requirements for designated hedge
pairs is now an important issue for companies that hold derivatives, and we

1
2

present a mathematical program that a company can implement to select


the optimal hedge assignment that complies with the new standards.
The Statement outlines the conditions under which the change in value
of a derivative can be offset on the balance sheet by the change in value
of a financial item, in an adjustment to the carrying amount of the hedged
item. Derivative - item pairs that satisfy the hedge requirements must be
designated at the start of the period and companies must demonstrate ex-
pected as well as ongoing hedge effectiveness. If this offset is perfect, i.e.
the derivative exactly hedges the instrument, then the amount of gain or
loss on the derivative that is reported in earnings is equal to zero.

SFAS 133 notes that the hedge designation must be


reported at the inception of the hedge. At that time,

. . . there must be formal documentation of the hedg-


ing relationship and the entity’s risk management ob-
jective and strategy for undertaking the hedge, includ-
ing identification of the hedging instrument, the hedged
transaction, the nature of the risk being hedged, and
how the hedging instrument’s effectiveness in hedging the
exposure to the hedged transaction’s variability in cash
flows attributable to the hedged risk will be assessed. . . .
Both at inception of the hedge and on an ongoing basis,
the hedging relationship is expected to be highly effec-
tive in achieving offsetting cash flows attributable to the
hedged risk during the hedge. . . . an assessment of ef-
fectiveness is required whenever financial statements or
earnings are reported, and at least every 3 months.

These hedge accounting standards necessitate a method of determining


the optimal item - derivative pairs and of demonstrating expected and cur-
rent hedge effectiveness. Since it is a common practice to institute hedges
for total positions, which might, for example, be expressed as equivalents of
five- or ten-year Treasurys, and since assets and liabilities are often basketed
and hedged as portfolios, entities in many cases may not be able to assume
their hedges are effective. SFAS 133 strongly suggests that items can be
grouped together and hedged as a portfolio only if the percentage gain or
loss on an individual item is within 1% or 2% of the percentage gain or loss
on the whole portfolio over the period. This means that items whose gain
or loss is not within these bounds for any portfolio the company holds can
3

only be hedged individually. Further, the necessity of designating a hedged


item - hedging derivative pair at the beginning of the period argues against
reporting a dynamic hedging portfolio.
However, the FASB restrictions on hedging pairs can be expressed as log-
ical requirements in a mathematical program, and selecting optimal hedging
designations can be modeled as an objective function which is the minimum
of the total gain or loss on all derivative instruments that is not offset.
Assessing hedge effectiveness and assigning hedge relationships could
involve front and back office departments of a financial institution, as well
as accounting, treasury, and internal audit functions. However, managing
hedge effectiveness fits most squarely as a function of risk management since
the effect of the new accounting standards is entity-wide and since the risk
management group in most organizations is already equipped to measure the
sensitivities of the company’s holdings to different market and credit factors.
A risk manager can use the programs outlined below, first, to demonstrate
the historical effectiveness of a proposed hedging relationship, and, second,
to evaluate the performance of the hedge designations during a reporting
period.

Hedge Accounting Criteria


The FASB statement distinguishes, first, between types of hedging relation-
ships and, second, between risks (or attributable causes of change in value)
that can be hedged against. Companies may elect to designate a derivative
as an hedging instrument as follows:

• A hedge of the change in fair value of an asset or liability (fair value


hedge)

• A hedge of the variability in cash flows of an asset or liability, an


available-for-sale security, or of a forecasted transaction, or purchase
or sale of an asset or liability (cash flow hedge)

• A hedge of the foreign currency exposure of an asset or liability (foreign


currency fair value hedge), an available-for-sale-security or a forecasted
transaction (foreign currency cash flow hedge), or a net investment in
a foreign operation
4

The distinctions between fair value, cash flow, and foreign currency
hedges are reporting distinctions: the offset amount of gains and losses for
derivatives are reported in earnings, or other accumulated income, or as a
translation adjustment, depending on the type of hedge.
Gain or loss for a derivative in a fair value hedge is reported in earnings,
and the gain or loss due to the hedged risk adjusts the carrying amount for
the hedged item; the gain or loss for the derivative that is not offset remains
in earnings. In cash flow hedges, the effective part of the hedge, i.e., the
amount of gain or loss on the derivative that is offset by the gain or loss
on the hedged item due to the hedged risk, and that offset, is reported in
accumulated other income. This effective portion is transferred to earnings
in the period the cash flow or transaction occurs; the ineffective part of the
hedge is currently reported in earnings. Foreign currency fair value hedges
follow the former fair value-reporting rule, and foreign currency cash flow
hedges follow the latter reporting rule. The effective portion of the gain or
loss on a financial instrument that is designated as an economic hedge of the
net investment in a foreign operation is reported as a translation adjustment,
i.e., in a separate component of consolidated equity.
The important point to remember, however, is that the ineffective part of
the hedge, the amount of gain or loss that is not offset, is reported currently
in earnings. The optimal hedge assignment will be the one that minimizes
this amount.
The statement next delineates the restrictions for certain items on the
risks that can be hedged against. The four risks, or changes in value, that
the statement allows hedging for are:

Market risk: gains or losses attributed to changes in the market price of


the entire hedged item

Market interest rate risk: gains or losses attributed to changes in mar-


ket interest rates

Foreign exchange risk: gains or losses attributed to changes in foreign


exchange risks

Credit (default) risk: gains or losses attributed to changes in an obligor’s


creditworthiness
5

According to the statement, there are restrictions for certain items on


the risks that can be hedged against:

1. Financial assets and liabilities, the variable cash flows of financial as-
sets and liabilities, and the forecasted purchases and sales of financial
assets and liabilities can be hedged against either market risk, or mar-
ket interest rate risk, foreign exchange risk, or credit (default) risk.
Two or more of the latter group can be hedged simultaneously if de-
sired.

2. All or a portion of held-to-maturity debt securities, and the cash flows


related to a held-to-maturity security can be hedged against credit
(default) risk only.

3. Nonfinancial assets or liabilities can be hedged against market risk


only; the forecasted purchase or sale of a nonfinancial asset or liability
can be hedged against either market risk or foreign exchange risk.

4. All or a portion of the prepayment option of a held-to-maturity debt


security can be hedged against market risk only.

5. The foreign currency exposure of an unrecognized firm commitment or


available-for-sale security can be hedged against foreign exchange risk.
The foreign currency exposure to variability in the functional currency
equivalent cash flows associated with a forecasted foreign currency
denominated transaction (either intercompany or extra-entity) can be
hedged against foreign exchange risk. And, finally, the foreign currency
exposure of a net investment in a foreign operation can be hedged
against foreign currency gain or loss.

These restrictions have to be included in a program that models maximizing


hedge effectiveness; the table below lists possible allowable risks for a sample
portfolio.
The first requirement under the statement is to demonstrate that the
entity expects the gains or losses of the hedging derivative to be highly
effective in offsetting the gains or losses in the hedged item.
In many institutions, the Market Risk Management group analyzes and
reports the sensitivity of the institution’s holdings to a set of risk indicators
6

Items Allowable Risks


1. Natural gas inventory Market risk
2. Equity shares of XYZ Market risk
3. Forecasted purchase Market risk
of coffee Foreign exchange risk
4. Corporate bond (USD) with Market risk
quarterly coupon Interest rate risk
Credit risk
5. DM-denominated bond Market risk
Interest rate risk
Foreign exchange risk

Table 1: Allowable risks for sample portfolio

and calculates the risks inherent in these exposures. An example of risk


indicators is the RiskM etricsT M documentation list of indices and factors
against which a portfolio can be analyzed. One approach to demonstrating
that a company expects a derivative to be ”effective” is to require that, in
addition to having had offsetting gains or losses in the past, the hedged item
and hedging derivative share sensitivity to at least one of the risk indicators.
The advantage of this approach is that companies already have this informa-
tion about their holdings, and it does not require intensive computer time
as, for example, developing correlation matrices would. The shared sensitiv-
ities can be easily determined from standard risk reports. Companies can
determine how closely related the designated items and hedges should be
by specifying how many risk indicators must be shared for a designation
to qualify. The Market Risk Management Group is best situated, because
of their study of market risk factors, to review the effectiveness of hedge
designations, and to weigh the possible benefits from hedge ineffectiveness.
Table 2 presents risk indicators for the sample portfolio.

Formulation of Hedge Effectiveness


The elements of our model are

• the inputs of gains or losses for items, over the last financial period,
broken down according to the four allowable risks. Igk,a = gain for
item k due to risk a, and Ilk,a = the loss for item k due to risk a. Table
3 below shows the gains and losses for the sample portfolio; Ig1,1 = .55
and Il5,2 = .24.
7

Risk Indicators
Market Risk S&P500 Index
Natural gas 6M forward price
Coffee 6M futures price
Market Interest USD LIBOR spot
Rate Risk USD LIBOR 6M
USD LIBOR 12M
DM LIBOR spot
DM LIBOR 6M
FX Risk USD/Brazilian Real exchange rate
USD/DM exchange rate
Credit Risk Industry sector index

Table 2: Sample portfolio risk indicators

• the inputs of gains or losses, over the last financial period, for the
derivatives. Dgj = gain from derivative j and Dl j = loss from deriva-
tive j. Dg3 = .17 and Dl 2 = .38.

• for convenience, we also specify Iglk,a = Igk,a + Ilk,a and Dgl j =


Dgj + Dl j .
• the inputs of allowable risk sensitivities for the hedge designations:
j
qk,a = 0 if derivative j and item k are both sensitive to an indicator
j
for risk a, and qk,a = −1 if they do not share any sensitivity. For the
1 4 2 3 2 3 5
sample portfolio, q2,1 = q3,1 = q4,1 = q5,1 = q4,2 = q5,2 = q5,3 = 0.

j
• variable Yk,a for the change in value in item k due to allowable risk a
that is hedged by derivative j.

j
• decision variable Zk,a is a 0 − 1 variable that indicates whether the
hedge of item k by derivative j against risk a is a permissible designa-
tion.

• decision variable W j is a 0−1 variable that indicates whether derivative


Dj can be designated as hedging any item.
8

Market Interest Credit


Item Risk Rate Risk FX Risk Risk
1. Natural gas 0.55 (L) 0 0 0
inventory
2. Equity XYZ 0.21 (G) 0 0 0
3. Forecasted purchase 0.31 (L) 0 0.35 (L)
of coffee
4. USD Debt 0.24 (G) 0.46 (G) 0 0.25 (L)
5. German Bond (DM) 0.07 (L) 0.24 (L) 0.19 (G) 0

Table 3: Gains or losses on items due to allowable risks

Derivative Gain Loss


1. XYZ option 0 0.01
2. IRS USD 0 0.38
3. IRS DM 0.17 0
4. Coffee 6M futures 0.27 0
5. DM 6M forward 0 .05

Table 4: Gains or losses on derivatives

The optimal hedge in our model is the one that is least ineffective, i.e., the
hedge which minimizes the amount of gain or loss on a derivative which
remains after offset. Therefore, the objective function for the model is the
minimum of the sum over all derivatives of the absolute value of the change
in value of the derivative minus any permissible offsetting change in an item.
There is no explicit recommendation for a particular measure of hedge
effectiveness in SFAS 133. However, a traditional and simple test of hedge
effectiveness is that the ratio of gains or losses on the derivative to the losses
or gains on the hedged item be between 80% and 125% over the financial
period. This test can be applied to historical returns to demonstrate an
expectation that a proposed hedge will be effective.

Representing Requirements
Hedge accounting requirements differ according to the risk being hedged,
whether entire items and derivatives or portions of them are being used,
and whether basis swaps or written options appear as hedging instruments.
We will first consider the case where there are no basis swaps or written
options in the set of hedging derivatives.
The gains and losses over the period should be easily extracted from
9

available P&L and DV01 reports. Since we have excluded basis swaps from
the set of derivatives, we assume here that at most one of {Igk,a , Ilk,a} and
{Dgj , Dl j } is greater than 0, for all a, j, k. Assume an entity holds the
sample portfolio of items and derivatives and has collected the input data
for this portfolio.

Hedges against market risk for fair value and cash flow hedges
Hedge designation is most straightforward when market risk is being hedged
(since if market risk is hedged, it is the only allowable risk) and the items
and derivatives are such that either the entire instrument or a portion (per-
centage) of it can be designated in a hedge. In this case, if we restrict our
test of effectiveness to an upper bound on the item-derivative offset, we
have a formulation with linear constraints and a piecewise linear objective
function.

Constraints
To ensure that designated pairs have similar risk profiles, we require
j j
∀j ∀k qk,1 ∗ Yk,1 ≥0

If item k and derivative j are not sensitive to the same indicator, then
j j
qk,1 = −1 and the constraint can be satisfied only by setting Yk,1 = 0.
Item k is hedged effectively by derivative j if the following inequalities
hold, since these constraints require that the hedged gain for the item is less
than 125% of the hedging loss on the derivative and the hedged loss on the
item is less than 125% of the hedging gain on the derivative.
m
j
X
∀j Igk,1 ∗ Yk,1 ≤ 1.25Dl j
k=1

m
j
X
∀j Ilk,1 ∗ Yk,1 ≤ 1.25Dgj
k=1

j
If Dgj = 0 and if Ilk,1 > 0, then Yk,1 is forced to 0; similarly, if Dl j = 0 and
j
if Igk,1 > 0, Yk,1 must be 0. This ensures that the gain or loss on a hedged
item offsets the loss or gain on a hedging derivative.
10

Objective function: Maximizing hedge effectiveness is equivalent to min-


imizing gain or loss from derivatives that is not offset:
n
m


j
X X
j
minimize Dgl − Igl ∗ Y

k,1 k,1

j=1 k=1

Hedges against all four allowable risks


As noted above, if market risk is being hedged in fair value and cash flow
hedges, then that is the only risk being hedged, and a = 1 for all j and k.
However, if market risk is not being hedged, then one or more of interest rate
risk, foreign exchange risk, and credit risk can be hedged simultaneously in
cash flow and fair value hedges. Therefore, to select the allowable designa-
j
tions that will provide the greatest offset a 0 − 1 variable Zk,a is introduced,
where a ∈ 1, . . . , 4 and
j
Zk,a = 1 ⇒ derivative j and item change k, a are a designated hedge
j
Zk,a = 0 ⇒ derivative j and item change k, a can not be a designated hedge

Then the requirement that a designated item and derivative are sensitive to
the same indicators is expressed by
j j
∀a ∀j ∀k qk,a ∗ Zk,a ≥0

j j
This is true because Zk,a is either 0 or 1, and if qk,a = −1, indicating that
j j
item k and derivative j share no sensitivities, then the product qk,a ∗Zk,a ≥0
j
implies that Zk,a = 0, or that the hedge is not permitted.
Inequalities modeling the requirement that market risk not be hedged in
conjunction with any other risk are added to the preceding program:
j j
∀j ∀k Zk,1 + Zk,2 ≤ 1
j j
Zk,1 + Zk,3 ≤ 1
j j
Zk,1 + Zk,4 ≤ 1
j j
∀a ∀j ∀k Yk,a ≤ Zk,a

These inequalities are satisfied only if at most one of market risk and
market interest rate risk, or market risk and foreign exchange risk, or market
j j
risk and credit risk is hedged. Then Yk,a and Zk,a are linked by requiring
j j
that if Yk,a = 1, then Zk,a must also equal 1.
11

Constraints on the decision variables W j can be used to enforce the lower


bound for effectiveness:
j
X
∀j Iglk,a ∗ Yk,a − .8 ∗ Dgl j ∗ W j ≥ 0
k,a

j j
Using this inequality forces W j = 0 if all the Yk,a = 0, and if any Yk,a >0
j
then the gain or loss on all the Yk,a for that j must be at least 80% of the
loss or gain on derivative j.
The constraints modeling hedge effectiveness have to be expanded to
include gains or losses due to all four allowable risks:
m
4 X
j
X
∀j Igk,a ∗ Yk,a ≤ 1.25Dl j
a=1 k=1

m
4 X
j
X
∀j Ilk,a ∗ Yk,a ≤ 1.25Dgj
a=1 k=1

and the objective function is now:



n m
4 X
j
X X
j
minimize Dgl − Igl ∗ Y

k,a k,a

j=1 a=1 k=1

If we apply this formulation to the sample portfolio of items and deriva-


tives, we find an optimal designation of:
• the loss from derivative 1 is offset by the gain from a portion (3.2%) of
item 2
• the loss from derivative 2 is offset by the gain due to interest rate risk of
a portion 82% of item 4
• the gain from derivative 3 is offset by the loss due to interest rate risk of
70% of item 5
• the gain from derivative 4 is offset by the loss from a portion (88%) of
item 3
• the loss from derivative 5 is offset by the gain due to foreign exchange
risk of a portion (29%) of item 5
These designations yield perfect hedges: the offset for this example is 0.

Designating hedged pairs with entire items


In the case of items of the following types, the entire item, not a portion,
must be hedged:
12

(a) one or more selected contractual cash flows

(b) a put option, a call option, an interest rate cap, or an interest rate
floor embedded in an existing asset or liability and clearly and closely
related to the host instrument

(c) the residual value in a lessor’s net investment in a direct-financing or


sales-type lease
j j
For these items, Yk,a is replaced by the 0 − 1 variable Zk,a in the con-
straints and the objective function.

Hedging with basis swaps or written options


Using basis swaps and written options as hedging instruments carries addi-
tional restrictions. Basis swaps can be designated as hedging instruments
only if they link an identified asset and liability, with the basis of one leg of
the swap identical to the basis of the asset cash flows and the basis of the
other leg identical to the basis of the liability cash flows. This requirement
can be modeled by considering the basis swap as two derivatives; if the basis
for each swap leg is included in the list of risk indicators for the portfolio,
then we can require that the item hedged by each leg share sensitivity to
the basis of that swap leg’s cash flows.
A written option can only be designated as a hedge of an item with
an embedded, closely related purchased option. A company can implement
this requirement by including a ”written option” indicator in the list of risk
indicators, and specifying that an item will be recorded as sensitive to this
indicator only if it has an embedded purchased option. A derivative will be
sensitive to this indicator only if it is a written option.

Conclusions
We have presented a suite of mathematical programs that risk managers
can use to assign optimal hedge relationships among the items and deriva-
tives held by a company. These programs match items and derivatives so
that the least amount of gain or loss on derivatives is reported while satis-
fying the SFAS 133 requirements. The program for designating most items
and derivatives in a hedge of market risk has a piecewise linear objective
function and linear constraints and can be solved using standard linear pro-
gramming methods. Programs for designating all possible types of items and
derivatives, in hedges of any allowable risk, include 0 − 1 decision variables
13

and linear and disjunctive constraints. These latter programs can be solved
using branching search methods combined with linear programs. Experi-
ments with programs modelling hedge effectiveness for portfolios of items
and derivatives are being carried out at the Logic Based Systems Lab at
Brooklyn College of the City University of New York.

Selected Readings
Financial Accounting Standards Board, Statement of Financial Accounting
Standards No. 133, Financial Accounting Series, No. 186-B, June
1998.

J.P. Morgan/Reuters, RiskM etricsT M Technical Document, 4th Edition,


Morgan Guaranty Trust Company of New York, 1996.

Ken McAloon and Carol Tretkoff, Optimization and Computational Logic,


Wiley, New York, 1996.

Alexander Schrijver, Theory of Linear and Integer Programming, Wiley,


Chichester, 1986.

H.P. Williams, Model Building in Mathematical Programming, 3rd Ed.,


John Wiley & Sons, Chichester, 1990.

Katherine Wyatt, Decomposition Techniques and Disjunctive Linear Pro-


gramming for Fixed-Income Portfolio Selection, doctoral dissertation,
Mathematics, City University of New York (1997).

Financial Optimization, Stavros A. Zenios, ed. Cambridge University Press,


Cambridge, 1993.

You might also like