Convertible Securities An Investors Guide. 2nd Edition October 2001 Jeremy Howard Michael O Connor

Convertible Securities: An Investors’ Guide
Convertible Securities
October 2001
An Investors’ Guide
2nd Edition October 2001

Jeremy Howard
Michael O’Connor
www.dbconvertibles.com
Note on the Authors: in answering our many market specific questions.
Jeremy Howard Lastly, this new Guide would not have been possible
co-Head of Global Convertible Research without the tireless and good-humoured editorial support of
Business Director - dbconvertibles.com Clodagh Muldoon, one of our junior European Analysts.
Jeremy Howard joined Deutsche Bank from Goldman Sachs

in August 1996 as Global Head of Convertible Research.
Prior to joining Goldman Sachs, Jeremy was an economic
advisor to the British Conservative Party. Jeremy holds a
Phd in Political Science from Oxford University. In addition
to co-heading the global research team, Jeremy is also
the business director of Deutsche Bank’s award-winning
convertible website: www.dbconvertibles.com.
Michael O'Connor
co-Head of Global Convertible Research
Michael O'Connor joined the convertible research team

at Deutsche Bank from ABN Amro in August 1999.
Michael spent three years on the ABN convertible desk
working in a combined sales and research role. Prior to this
Michael spent two years at Charterhouse Tilney selling and
researching UK convertibles. Michael holds an honours
degree from Hertford College, Oxford. Michael, together
with his European research team, were ranked Number 3 in
the Institutional Investor Analyst Poll in 2001.
Editorial Note:
We would like to thank Peter Darrell, Julian Moore and

Jonathan Poole for their expert technical advice on this
publication.
We also thank the European trading team of Simon Garcia,

Andy McDonnell and Steven Roth for their patience and help
2
Contents
Chapter Page
1. Convertible Securities - An Asset Class Grows Up 4
2. Convertibles for Beginners 6
3. Convertibles for Fixed Income Investors 18
4. Credit Protection for Convertibles (Asset Swaps and Credit Default Swaps) 22
5. Convertibles for Equity Investors 35
6. Convertibles for Hedge Funds 40
7. What Drives Convertible Prices Today? 56
8. Documentation 61
9. Convertibles for Issuers 66
10. Convertible Structures 70
© 2001. Publisher: Deutsche Bank AG London, 1 Great Winchester Street, London, EC2N 2EQ.
Authors: as referred to on the front cover. All rights reserved. When quoting, please cite Deutsche Bank AG as the source.
This publication is derived from selected public sources we believe to be reliable but neither its accuracy nor completeness is guaranteed. Opinions expressed herein reflect the
current judgement of the authors: they do not necessarily reflect the opinions of Deutsche Bank AG or any subsidiary or affiliate of the Deutsche Bank Group (together "the DB
Group") and may change without notice. Neither the author, nor the DB Group accepts any liability, whatsoever, with respect to the use of this publication, except as provided for
under applicable regulations. At the date hereof, the author, and/or the DB Group may be buying, selling, or holding significant long or short positions; acting as investment and/or
commercial bankers; be represented on the board of the issuer; and/or engaging in market making in securities mentioned herein. This publication should not be
construed as an offer, invitation or solicitation to subscribe for, purchase or sell any investment and neither it nor anything it contains shall form the basis of any contract
whatsoever. This report may not be copied, transmitted or distributed to any other party without the express written permission of the DB Group. Deutsche Banc Alex.Brown Inc.
have accepted responsibility for the contents and distribution of this publication in the United States under and to the extent permitted by applicable US requirements. Deutsche
Bank AG London (regulated by the SFA) has accepted responsibility for this report in the UK under section 57 of the Financial Services Act 1986. Interested parties are advised to
contact the Deutsche Bank entity they currently deal with, or the Deutsche Bank entity that has distributed this report to them. Transactions should be executed through a Deutsche
Bank entity in the client's home jurisdiction unless otherwise permitted by law. This publication is intended to provide information to assist investors in making their own
investment decisions, not to provide investment advice to any specific investor. Investments discussed and recommendations made herein may not be suitable for all investors:
readers must exercise their own independent judgement as to the suitability of such investments and recommendations in light of their own investment objectives, experience,
taxation status and financial position. Past performance is not necessarily indicative of future performance; the value, price or income from investments may fall as well as rise.
You should not deal in derivatives unless you understand the nature of the contract that you entering into and the extent of your exposure to risk.
Additional information available upon request.
Convertible Securities 2001 3

1. Convertible Securities - An Asset Class Grows Up
We published our first Investors' Guide to Convertibles in Fig 2: Global Market Capitalisation in USD
February 1997. Since then, a great deal has happened
in the asset class. Fortunately for everyone involved in the
product, the developments have been generally positive,
and convertibles are now firmly established as a
mainstream financing vehicle for corporates in every region.
We believe they have become among the most exciting
securities for dedicated investors, fixed income funds and
hedge funds. Market developments in 2000/2001 have also
drawn attention to the attractions of convertibles for equity
investors, who were previously only active in the Asia Pacific
markets.
Source: Deutsche Bank Convertible Research
The purpose of this updated Guide is to bring everyone

interested in convertibles up to date with the asset class as billion represents a very substantial global asset class. The
it currently stands. growth in convertibles has been driven by a number of
different factors. In the US, the world's largest market,
It is clear from Figures 1 and 2 that global convertible convertibles have survived the 'tech wreck' by attracting
issuance has been very strong over the past six years, with new higher credit quality companies through large, 0%
difficult market conditions in years such as 1998 and 2001 coupon, long dated securities. Investment grade issuers
failing to make a dent in the upwards march. have replaced the TMT corporates that fuelled the growth of
the late 1990s.
Fig 1: Global Market Issuance in USD
Europe has also maintained its upward trajectory.
Convertible issuance has come amidst corporate restructuring,
as shareholder value considerations have forced a major
programme of non-core holding divestments.
Asia has stayed on the map thanks to the Taiwanese

technology sector and to some high profile 'jumbo' issues in
Hong Kong and Korea.
The real underperformer since we last wrote in 1997 has

been Japan, which has failed to keep pace with the global
Source: Deutsche Bank Convertible Research growth trend. Despite occasional moments of hope, the
domestic market remains moribund, and the exchangeable
The general level of equity prices has affected the market revolution that has long been promised in Japan has yet to
cap of the product, but the total as of 2H 2001 of USD 451.9 get off the ground.
4
The major difference between this updated Guide
and the original is that we draw heavily on our
www.dbconvertibles.com website. We also use real
examples to illustrate our points as often as we can, rather
than relying on hypothetical securities.
We have expanded our coverage of Asset Swaps and Credit

Default Swaps, included whole new sections on how
convertibles trade in the market today, on convertible
documentation and also commentary on the new
structures that have appeared since 1997.

2. Convertibles for Beginners
Convertibles have an unfortunate (and totally unjustified, in So what do these terms mean?
our opinion) reputation amongst some investors for being
complex instruments. In its standard form, a convertible is Coupon (4%)
simply a corporate bond which gives the holder an additional
right to give up the bond in exchange for a fixed number of Coupons are the interest payments on the bond (convertible
ordinary shares. At maturity, convertibles are worth either preferreds pay fixed dividends). This can be paid either
their cash redemption value or the market value of the annually, semi-annually or even quarterly. This amount is
shares into which they are convertible, whichever is the (almost always) fixed for the life of the bond. A 4% annual
greater. Simple as that! coupon on a EUR 1,000 nominal value bond indicates that on
the coupon payment date each year the holder will receive
Although there are some exceptions to the plain-vanilla EUR 40 in cash. As with normal corporate bonds, interest
convertible structure, we believe it is important that all accrues between coupon payment dates according to the
investors are familiar with the basic model. Throughout this relevant market convention. These conventions are generally:
section we will use the real example of the Dutch food
UK 30 / 360 days
company Ahold 4% 2005 EUR. In order to understand how
Continental Europe Actual / Actual or 30 / 360 days
this convertible works, we will split our analysis into two
sections: A) the fixed terms and B) the market valuation. Asia Pacific 30 / 360 days
US/Latin America 30 / 360 days
A) The Fixed Terms Japan (Domestic) Actual / 365 days (in most cases)
1
Ahold 4% 2005 EUR An investor buying Ahold 4% 2005 EUR for settlement 16
July 2001 for 100 (clean price) would pay the seller 100.633
Coupon 4% (annual) or EUR 1,006.33 (ignoring any withholding tax). Accrued
Issue Amount EUR 920 million interest is paid up to (and including) settlement date (trade
First Settlement 19 May 2000 date + 3 days in all the Euroconvertible markets).
Issue Price 100.00%
Redemption Amount 100.00% The EUR 1,006.33 is calculated as follows:
Maturity 19 May 2005
Nominal Value EUR 1,000.00 Settlement (16 July 2001) - Last Coupon (19 May 2001)
Conversion Ratio 31.0463 (Nominal Value / Accrued Interest = (57 days / 360 days) x EUR 40
Conversion Price) = EUR 6.3333
Conversion Price EUR 32.21 (Nominal Value /
Conversion Ratio) Maturity (19 May 2005)
Call Protection Hard Call 3 years
The date on which the issuer must offer to redeem the
ISIN XS0111182597
securities for their redemption amount. If the value of the
1
Ahold 4% 2005 EUR compounds at 30 / 360
6
shares underlying the bond at maturity exceeds the 'Provisional' protection is when the bonds can be called
redemption amount the holder will convert the bond into subject to the share price being above a certain level (see
shares. below). Call protection is important for investors as it
guarantees the optionality of the convertible and whatever
Conversion Ratio (31.0463 shares per bond) yield advantage it has over the underlying shares for a fixed
period of time. The longer the call protection, the greater
The number of shares into which each EUR 1,000 bond is the benefit for investors.
convertible. A conversion ratio of 31.0463 means that each
bond can be converted into 31.0463 ordinary Ahold shares. 'Provisional' (or ‘Soft’) Call Protection
The conversion ratio remains fixed throughout the life of the
instrument, although it will usually be adjusted to account for Ahold 4% 2005 EUR does not have provisional call protection,
stock splits, special dividends or other dilutive events, and but provisional call protection means that the bond cannot
‘reset’ clauses. In fact , Ahold 4% 2005 EUR had an initial be called unless the stock trades above a specified level for
conversion price of EUR 33.02 that was adjusted as a result a certain period of time. Usually, this level is expressed as
of a dividend paid (partially) in shares. a percentage of the conversion price (i.e. bond is callable
if stock > 140% of conversion price). To prevent abnormal
Conversion Price (EUR 32.21) trading patterns triggering a provisional call, most convertible
prospectuses state that the stock must trade above the
The price at which shares are effectively 'bought' upon trigger price for 20 or 30 days before the provisional call is
conversion. Ahold 4% 2005 EUR is convertible into 31.0463 activated.
shares per bond. As the nominal value of the bond is EUR
1,000, the holder has paid EUR 1,000 for the right to convert The effect of call protection on convertible valuation is
the bond into 31.0463 shares. At maturity, if parity (share discussed more fully in Chapter 3.
price x conversion ratio) is higher than EUR 1,000 the
holder will convert into shares. As parity will only be greater B) The Market Valuation
than redemption amount if the stock is above EUR 32.21,
EUR 32.21 is effectively the 'strike price', or conversion This section examines the characteristics of the bond on a
price. A convertible is ‘in the money’ if: specific day (trade date 11 July 2001).
Share Price > Conversion Price Secondary Market Valuation
'Hard' Call Protection (3 Years Hard Call to 19 May 2003) Convertible Price 121.75
Parity 113.78
Most convertible bonds are issued with a period of call Conversion Premium 7.00%
protection during which the issuer may not redeem (call) the Current/Running Yield 3.285%
bonds early. 'Hard' call protection is the period during which Yield to Maturity -1.462%
the bonds cannot be called under any circumstances.

Underlying Stock Parity = Number of shares per bond x share price
Parity = 31.0463 x EUR 36.65 = EUR 1,137.8
Stock Price EUR 36.65
Stock Volatility 33% Parity is normally quoted as a percentage of par amount:
Stock Yield 1.5%
Breakeven 3.66 years = EUR 1,137.8 / EUR 1,000
= 113.78% (or just 113.78)
Fixed Income Analysis
Parity is crucial because it will influence not only the price of
Credit Spread LIBOR + 160 bps (4 Year the bond (convertible price should not normally be less than
LIBOR = 4.65%) parity), but also whether the bond will be converted at
Bond Value 91.91 maturity. Table 1 shows how this works.
Risk Premium 32.47%
Table 1: What Happens to Ahold 4% 2005 at Maturity?
Let's look at what each of these numbers means:
Stock Price Parity Redemption Bondholders’ Action
at Maturity Amount
Convertible Price (121.75)
1. EUR 10.00 31.05 100 Redeem for cash
In most markets, convertible bonds are quoted using the
standard fixed income percentage of par methodology. Indifferent between
2. EUR 32.21 100.00 100
Therefore, 121.75 means 121.75% of (in this case) EUR cash and shares
1,000 or EUR 1,217.50 per bond (given a nominal value of
EUR 1,000). France is an exception, where prices are quoted 3. EUR 40.00 124.19 100 Convert into shares
in absolute Euro / Franc amounts. Source: Deutsche Bank Estimates
Most prices (outside France and US Preferreds) are 'clean', 1. In Scenario 1 the bondholder will not take shares worth
in that they do not include accrued interest. An investor EUR 310.5 (per bond) when EUR 1,000 is the cash
purchasing the bond would actually pay more than EUR redemption value.
1,217.50 to reflect the accrued interest in the bond.
2. In Scenario 2 the bondholder is indifferent between EUR
Parity (113.78) 1,000 cash redemption value and EUR 1,000 in shares.2
Parity is a very important term for understanding how

3. In Scenario 3 the share price at maturity has risen to EUR
convertibles perform. Parity is the market value of the
40 and parity is 124.19. An investor will convert into
shares into which the bond can be converted at that time.
Like convertible price, it is also expressed on a percentage 2
This analysis is simplified to exclude the final coupon payment. In
of par basis (except in France). 113.78 is actually a market reality, the redemption value is EUR 1,040 due to the final annual coupon
abbreviation for 113.78% of EUR 1,000. payment, which will often be lost on conversion. An investor would usually
take the cash in Scenario 2.
8
shares worth EUR 1,242 at maturity rather than redeem Current Yield (3.285%) is:
the bond for EUR 1,000.
Coupon / Current Price of Bond
Premium (7.00%) = 4/121.75
= 3.285%
Premium is related to parity. Premium is the difference
between parity and the convertible bond price, expressed as
Yield to Maturity (-1.462%)
a percentage of parity. Premium expresses how much more
an investor has to pay to control the same number of shares The yield to maturity (YTM) is based on the price of the
via a convertible, rather than buying them outright. bond, the coupon payments to maturity and the final
redemption amount. YTM takes account of the capital
Premium = (Bond Price - Parity) / Parity gain/loss on the bond at maturity.
7.97
= x 100% = 7%
113.78 When a convertible has a put before maturity, investors will
also look at yield to put (YTP).
An investor buying Ahold 4% 2005 EUR for EUR 1,217.5
controls 31.0463 shares per bond. But an investor buying Breakeven (3.66 Years)
31.0463 shares in the market would only pay EUR 1,137.84
(31.0463 x EUR 36.65). The extra EUR 79.7 represents the Breakeven is the length of time needed to amortise the
convertible's premium. premium paid to buy the convertible via its yield advantage
over the underlying stock. We have explained above that
Premium also gives a guide as to how a convertible will investors in Ahold 4% 2005 EUR pay a EUR 79.66 'premium'
perform in relation to the underlying shares. Convertibles to own the convertible rather than the shares themselves.
with low premiums should be more sensitive to movements But convertible investors receive extra yield on their
in the underlying share price than convertibles where investment. Every year, the convertible will pay EUR 40 in
premium is high. cash. Assuming a 1.5% dividend yield, the same cash
outlay on the shares would have returned only EUR 18.26
Very ‘in the money’ (see page 11 for definition) convertibles (1.5% x EUR 1,217.5). We now have enough information to
will sometimes be quoted in terms of their ‘points’ premium calculate 'convertible breakeven'.
over parity. In the above example, points premium would be
121.75 - 113.78 = 7.97 points. Convertible investors receive EUR 21.74 more income than
investors in the same cash amount of stock each year (EUR
Current Yield (3.285%) 40.00 - EUR 18.26). Breakeven calculates how long it takes
for the EUR 79.66 premium to be amortised by the annual
Investors should think of current yield in exactly the same EUR 21.74 in extra yield.
way as dividend yield on the equity.

Breakeven: Risk Premium (32.47%)
Cash Premium Paid

= Just as premium expresses the premium investors pay
Convertible Coupon - (Convertible Price x Stock Yield)
to own a fixed number of shares via a convertible, risk
EUR 79.66 premium refers to the premium of the convertible price over
=
EUR 40.00 - (EUR 1,217.50 x 1.5%) its bond floor. It could be thought of as the premium fixed
EUR 79.66 income investors pay for the equity call option.
=
(EUR 40.00 - EUR 18.26)
Risk Premium is:
EUR 79.66
=
EUR 21.74 (Convertible Price - Bond Floor)
= 3.66 Years Bond Floor
= (121.75 - 91.91) / 91.91
Some investors look at a simple measure of breakeven: = 32.47%
current yield / dividend yield. But this is not really an accurate
statement of the cash premium amortisation. In fact, How Convertibles Perform in the Market
breakeven analysis is today only applied to mandatory or
other very ‘equity-like’ convertibles. The best diagram for understanding the performance of
convertibles in the secondary market is shown in Figure 3
Bond Floor (91.91) (below). Newcomers to the convertible market are strongly
advised to familiarise themselves with this diagram, as it
Bond floor is the value of the convertible as a straight illustrates how convertibles perform in virtually all scenarios.
corporate (i.e. non-convertible) bond. To calculate bond floor
it is only necessary to know: (1) coupon, (2) maturity date, We break the diagram up into five main sections:
(3) redemption amount, (4) puts and (5) discount rates (the
risk-free rate + credit spread). Working from left to right, we find:
At 91.91, the accretion to redemption at par and coupons i. Distressed Debt ('Junk') Convertibles
would give a yield of 6.25%. This is exactly 160 bps over the ii. 'Out of the Money' Convertibles
risk-free rate of 4.65%. 6.25% is the YTM we would expect
iii. 'At the Money' Convertibles
to find on comparable straight Ahold bonds.
iv. 'In the Money' Convertibles
v. 'Discount to Parity' Convertibles
The bond floor of a convertible should provide a price floor if
interest rates and credit perceptions of the issuer remain
The expressions 'out of', 'at' and 'in the money' are loosely
unchanged.
defined in the convertible market. Technically, they should
be defined with reference only to the stock price and the
conversion price, such that:
10
Fig 3: How Convertibles Perform in the Secondary Market
Source: Deutsche Bank Estimates
'Out of the Money' = stock price < conversion price Delta* N/A
'At the Money' = stock price = conversion price Risk Premium < 5%
'In the Money' = stock price > conversion price
(*Delta is explained fully in Chapter 6 - it is a measure of the convertible's
sensitivity to stock price movements).
However, the 'At the Money' category is usually broadened
to include instruments whose stock prices are reasonably
close to the conversion price. When a convertible issuer's share price falls dramatically, or
its ability to finance its debt obligations is called
i) Distressed Debt / 'Junk' Convertibles into question, the convertible enters the distressed debt
category. The overriding concern for convertible investors,
and other bondholders, becomes the creditworthiness of
Parity = 0 - 40
the company. Distressed debt or 'junk' convertibles usually
Premium > 100%
trade with a large premium to parity, at the fixed income

value for the equivalent high yield instrument. The credit Fig 4: 'Out of the Money' Ahold 4% 2005 EUR
spread over the risk-free rate will be high to reflect the high
probability of default. Unlike convertibles which are just 'out
of the money', distressed bonds are often more sensitive to
credit perceptions of the guarantor than to interest rates.
These convertibles' prices can sometimes be correlated

almost 100% with the share price.
As with all high yielding debt, we believe 'junk' convertibles

should be approached with caution and avoided altogether
by investors without reliable credit research on the guarantor /
issuer. The last couple of years have seen an increase in the
number of high yield funds prepared to switch in and out of
convertibles as high yield alternatives. This has brought
greater efficiency into distressed convertible pricing.
Other considerations for assessing distressed debt Source: Deutsche Bank Estimates
convertibles include:
taking a position without a thorough grounding in the
1. Subordination - The level of subordination of the bond will bankruptcy laws of the applicable legal jurisdiction.
affect its 'recovery rate' (amount eventually paid back
to bondholders on default). Watch out for non-quoted 4. Structured Subordination - This can occur when
bank debt, assets and revenue streams that have convertibles are issued from holding companies which
been securitised, as well as the claims of holders of own geared operating assets.
more senior bonds.
ii) 'Out of the Money' Convertibles
2. Accountancy and Transparency - As an understanding of
the balance sheet will probably be the most important Parity = 40 - 70
consideration, an investor needs to be sure that they Premium > 35%
have all the relevant items. Off balance sheet liabilities Delta = 10% - 40%
can be hidden quite easily in some markets. Hidden Risk Premium = 5% - 20%
guarantees of related company liabilities (in Asia and
Japan especially) can also be an issue. Unlike 'distressed debt' convertibles, 'out of the money'
convertibles can (and often do) have perfectly creditworthy
3. Bankruptcy Laws - These vary hugely across the world. issuers. Their 'out of the money' status simply reflects the
No distressed debt convertible investor should think of fact that the underlying share price has not performed well
12
since issue, and is still well below the conversion price. This company improves (the credit spread contracts and the
means that parity will be well below 100. Remember that bond floor rises); and/or
parity is:
4. If the convertible's premium to its bond floor increases
Stock Price x Conversion Ratio as a result of a general richening of convertible valuations.
Nominal Value
‘Out of the money’ convertibles have become increasingly
So therefore parity must also be: attractive to fixed income funds who switch straight
corporate debt positions into these securities to take
Stock Price advantage of any sizeable rally in the underlying share price
x 100
Conversion Price (see Section 3).
Our sample bond Ahold 4% 2005 EUR was issued with iii) 'At the Money' Convertibles
parity at 80.37 (EUR 803.7 per bond) and a premium of
around 24.4%. If the share price were to fall to EUR 16.10 Parity = 70 - 130
(see Figure 4) parity would fall to around 50.0. So where Equity Premium = 10% - 35%
would the convertible trade in the secondary market? Delta = 40% - 80%
Risk Premium = 20% - 40%
We recall that the fixed income value, or bond floor, of the
convertible using the issuer's cost of debt (LIBOR + 160bps)
Fig 5: 'At the Money' Ahold 4% 2005 EUR
is 91.91. This provides a floor below which the convertible
should not trade. In fact, as Figure 4 shows, we expect the
convertible to trade at a small premium to its bond value
to reflect the time value of the (albeit out of the money)
equity call option contained in the convertible. The exact
convertible price is calculated using a binomial model, which
is explained in Chapter 6. The actual theoretical value is 92.78.
Investors purchasing out of the money convertibles will

profit under four different scenarios:
1. If the underlying share price rises significantly;
2. If interest rates fall (i.e. the bond market rallies) and bond
floor rises;
3. If the market's perception of the credit quality of the


'At the money' convertibles are often regarded as 'true' or money' profile, with parity usually somewhere between 70
'balanced' convertibles, as they tend to have the asymmetric and 90.
risk/return profile (upside participation with downside
protection versus stock price movements) that appeals to iv) 'In the Money' Convertibles
dedicated convertible investors. They are also the most
popular bonds with convertible hedge funds (see Chapter 6). Parity = 130 +
Premium = 0% - 10%
'At the money' convertibles have a higher delta (equity Delta > 80%
sensitivity) when the underlying shares rise than when they Risk Premium > 40%
fall. This means that the convertible will participate in more
equity upside than downside. It is clear from Figure 5 that As the stock price rises above the conversion price (EUR
as parity goes up, the convertible price line increases in 32.21 for Ahold 4% 2005 EUR), the convertible moves into
steepness, whereas it starts to flatten as parity falls. the 'in the money' section of the main 'ski jump'
diagram (Figure 3).
We can demonstrate the risk-adjusted profile of an 'at the
money' convertible by examining how Ahold 4% 2005 EUR Very 'in the money' convertibles are almost certain to
performs under different stock price scenarios when parity be converted at maturity. As they move further and
starts at 100. The values calculated in Table 2 are derived further from their bond floor, the premium investors will
from a binomial tree convertible valuation model, which will
be discussed fully in Chapter 6.
Fig 6: 'In the Money' Convertible Performance
Table 2: Asymmetric Risk/Return Profile of Ahold 4%
2005 EUR when 'At the Money'
Stock Parity Theo. Participation
Price Convertible with Stock
Value move
EUR 38.65 120 129.0 66.3%

EUR 32.21 100 113.9 0.00%
EUR 25.76 80 101.9 - 52.7%
When starting from an 'at the money' profile with parity at

100.00, Ahold 4% 2005 EUR participates in 66% of a stock
price rise of 20%, but only 53% of a similar downwards
move. This is the kind of asymmetric risk/return profile that
dedicated convertible investors and hedge funds look for.
Convertibles are almost always issued with an 'at the Source: Deutsche Bank Estimates
14
be prepared to pay over parity decreases. In the money Call Protection
convertibles will not trade at a significant premium to parity
unless either: (1) the bond has a very large income advantage Almost all convertibles allow the issuer to call (i.e. to
over the underlying shares and/or (2) there is a long time redeem) the bond before the maturity date. But of course if
until the first call date or maturity (giving significant time the convertible is ‘in the money’ investors will convert rather
value to the put inherent in the convertible). than accept the cash redemption. This call can sometimes
be subject to a trigger, such as parity needing to be 120% or
The value of the convertible's inherent put is the cash 130% before the call is activated.
redemption of 100 (or other redemption amount) at maturity.
The value of this put obviously declines as parity rises. The The following diagrams demonstrate the effect of call
put at 100 is worth much less when parity is 140 than when protection on convertible valuations. Remember that call
parity is 100. When the convertible is a long way in the protection is 'hard' if the issuer cannot call the bonds under
money the put is worth so little that an investor should pay any circumstances, and 'provisional' if the issuer can only
only the present value of the expected income advantage call the bonds if the stock / parity has risen above a trigger
over the underlying shares as premium. level.
‘In the money’ convertibles are far more sensitive to Figure 7 below shows how premium is retained if long call
movements in the underlying share price than to protection exists, whereas it collapses in Figure 8 where call
movements in interest rates. Because they trade at such a protection has expired.
low premium to parity, equity sensitivity (delta) is high. And
because they trade at such a high premium to bond floor, Fig 7: Call Protection Remains: Premium is Retained
interest rate sensitivity (rho) is low. This is demonstrated in
Table 3:
Table 3: As Parity Rises: Delta Rises
Parity Theo. Value Bond Floor Delta
60.00 93.94 91.91 21.87%

100.00 113.9 91.91 70.04%
140.00 146.0 91.91 92.04%
Valuation parameters for Ahold 4% 2005 EUR

Fig 8: Call Protection Elapsed: Premium Collapses upside participation with downside protection and yield
enhancement.
Fig 9: Theoretical Value Rises as Call Protection is

Extended
Source: Deutsche Bank Estimates Source: Deutsche Bank Estimates
Long call protection is good for the holder of a convertible v) 'Discount' Convertibles
because:
Parity High
1. The optionality of the bond lasts longer; and
Premium < 0%
2. The income advantage of the convertible over the Delta = 100% (can be > 100%)
underlying shares lasts longer. Risk Premium > 40%
Points 1 and 2 mean that investors will pay more for a bond A ‘discount’ convertible is one in which parity is greater than
with longer hard call protection. the convertible price. Sticking with the example of Ahold
4% 2005 EUR, if parity rose to 140 and the convertible was
The effect of hard call protection on the theoretical trading at 138, the bond would be trading at a -1.45%
value of a convertible is very significant. Figure 9 discount to parity.
demonstrates how the theoretical value of Ahold 4% 2005
EUR rises as hard call protection is extended. Why would a convertible trade at a discount to parity? It
should be obvious that as convertibles are usually American
Convertibles with reasonable call protection remaining and a style options (convertible at any time), a risk free profit
significant yield advantage over the underlying shares are could be made by buying the convertible, converting it
extremely attractive to equity investors, because they offer immediately and selling the shares. The word 'immediately'
16
gives a clue as to why convertibles occasionally do trade at parity (parity-accrued interest) to reflect the loss of accrued
a discount. upon conversion. A full explanation of this so called ‘screw
clause’ appears in Section 8.
In some convertible markets (especially in Asia) conversion
into the underlying shares is not straightforward. In Taiwan,
to take an extreme case, convertible securities are not
directly convertible into shares at all (as of 2001). Investors
must first convert into 'Entitlement Certificates', which are
then convertible into the underlying shares only on set dates
throughout the year (four times a year is usual).
In other markets, such as the Philippines, there is a delay (up

to one month) between lodging bonds for conversion and
receiving shares. Even this impediment would not cause
bonds to trade at a discount if sufficient stock borrow were
available. Investors in Ahold 4% 2005 EUR in the above
scenario would simply buy the bond for 138, sell short 100%
of the conversion property and lodge the bonds for
conversion. Even though there is a time delay in receiving
shares the investor is protected through his short position.
In many developing convertible markets, however, stock

borrow is either inadequate or non-existent, and investors
facing lengthy conversion procedures are exposed to
adverse movements in the share price with no means of
hedging themselves. This can result in a bond trading at a
slight discount. Alert readers will realise that for a natural
holder of the underlying shares, a ‘discount’ convertible
does indeed represent a free lunch!
In other situations, lack of liquidity in the underlying shares

can give the appearance of a discount when none really
exists. Investors need to be aware of the time and size of
price quotations.
In reality, however, discounts are very rare in the developed

markets except where a callable convertible trades at ‘net’

3. Convertibles for Fixed Income Investors
Convertibles can boost fixed income performance by for LIBOR-linked floating payments. The yield is
recreating straight corporate bond positions with additional enhanced because fixed income buyers of asset
exposure to rising equity markets. Specifically, fixed income swapped paper allow convertible arbitrage funds to
investors can buy 'out of the money' convertibles trading retain the optionality. Fixed income fund buying of 'asset
with yields very close to, or even at, their straight bond swapped' convertible paper has increased a great deal in
equivalent. Convertibles have the added attraction of recent years, and we will examine the mechanics of an
containing a long dated call option on the underlying asset swap transaction involving convertibles in Chapter 4.
stock, and will improve fixed income performance, if used
appropriately. 3. Write Credit Default Swaps against Convertibles
The newest area for fixed income funds to take

After the huge market declines of 2001, there is now
advantage of convertibles is by writing Credit Default
(October 2001) an unprecedented opportunity for fixed
Swaps referencing convertibles as the default obligation.
income investors to buy convertibles trading flat on their
CDSs are explained fully in Chapter 4, but they are
bond floors. Indeed, in Europe in October 2001, we
basically an insurance policy that pays par in the event of
estimated that 74% of the convertibles market was within
a default, in return for an agreed periodic fee. They are
5% of its bond floor. We believe the opportunity for fixed
becoming popular with high yield fixed income funds
income investors to enter the convertibles market has never
because they are tradeable instruments in which the
been so attractive.3
writer of the protection retains the credit upside (unlike
an asset swap transaction).
There are three front line strategies which allow fixed
income investors to use convertibles successfully:
Fixed income investors can assess the attractiveness of
a convertible very easily. Returning to the example of the
1. Buy Convertibles as Straight Corporate Bond Alternatives
Dutch Ahold 4% 2005 EUR convertible bond, we assume
Fixed income funds can buy convertibles trading close to that Ahold stock has fallen to EUR 20 and that the
(or even below) their bond floor (when the convertible's convertible is trading at 93 (at a 1.19% ‘risk’ premium to
yield is comparable to pari passu straights). If the shares its bond floor of 91.91). This means that fixed income
into which the bond is convertible rally, the bond will investors can switch into the convertible for a premium
outperform equivalent straight bonds (explained below). of 1.09 points (assuming they hold straight bonds of
similar profile).
2. Buy Asset Swapped Convertible Paper
In yield terms, the convertible’s YTM of 6.161% is only
Fixed income funds can also buy 'asset swapped' 36 bps below the 6.518% YTM on the equivalent straight
convertible paper. The transaction typically sees them debt.
entering into a swap agreement with a bank, passing the
fixed payments from the convertible to the bank in return The precise value of the embedded call in the convertible
3
We have even noticed instances, at times of heavy convertible credit can be calculated using a binomial tree model (see
hedging, when the convertible offers a yield pick-up over pari passu straights Chapter 6). But the simpler point is that if the yield 'give
(i.e. fixed income investors are actually paid to take an equity option!).
18
up' is very low, it ought to be attractive to move a straight Table 4: Swiss Re / ING 1.25% 2003 NLG -
bond position into the convertible alternative, if the exchangeable outperforms straight bonds
equity environment looks favourable. Swiss Re / ING Straight Bond
1.25% 2003 NLG Value
Convertibles as Straight Corporate Alternatives
5 Oct 1998 89.25 87.93
Considerable interest in convertibles as fixed income 1 Sept 2001 100.88 94.81
portfolio diversification instruments has been notable in
Income Adj. Return 13.0% 7.8%
Europe and the US from 1998/9 onwards. This interest has Source: Deutsche Bank Estimates
been sparked by theoretical analysis that shows that, at
least in reasonable equity markets or over the longer term, By switching from the straight bond to the exchangeable, a
adding convertibles to a fixed income portfolio can have the fixed income investor would have enjoyed a very significant
effect of increasing returns and lowering volatility. outperformance over holders of straight Swiss Re bonds.
The theory is simple. Convertibles contain an embedded But the case for a certain weighting in convertibles as an
call option on the stock, and therefore will outperform asset class at all times is more subtle, and is based on the
straight corporate bonds in a rising equity market. negative correlation between bonds and equities.
This is clearly indicated in Figure 10, where the rising This negative correlation between bonds and equities is
share price of Dutch bancassurance company ING leads to easy to demonstrate:
a huge outperformance of the exchangeable Swiss Re / ING
1.25% 2003 NLG exchangeable bond over the equivalent Table 5: Negative Correlation between Bonds and Equities
Swiss Re straight bonds.
Correlation S&P 500 versus US DAX / 10 Year
Treasury 10 year Bund
Fig 10: Swiss Re / ING 1.25% 2003 NLG
1991 - 2001 -0.603 -0.783
But what does this mean for convertible investors?
The secret strength of convertibles is that the negative

correlation between bonds and equities means that
convertibles will produce their performance enhancement
exactly when the fixed income investor needs it: when
equity prices are falling, bond prices will rise, and vice versa.

Before we go on to demonstrate this with statistics, it is therefore not suitable for inclusion in a general fixed income
important to note that the argument only works if the fund. NTL 6.75 2008 USD sees the bond price tracking the
bond and underlying equity component of the convertible share price downwards in Figure 13. These bonds may be
or exchangeable are not excessively positively correlated.
The bonds that are most suitable for fixed income Fig 12: National Bank of Greece 2% 2003 EUR
substitution are:
1. Exchangeable Bonds (where the issuer / guarantor and

underlying stock are different companies). Stock price
weakness should not affect perceptions of the credit of
the guarantor, as we see with Bell Atlantic (Verizon)
exchangeable into Cable and Wireless / NTL in Figure 11.
Fig 11: Bell / C&W - NTL 4.25% 2005 USD
Fig 13: NTL 6.75% 2008 USD
2. High Credit Quality Bonds. Generally, higher quality

issuers will not suffer credit spread expansion even if the
share price declines substantially (Figure 12 shows
National Bank of Greece 2% 2003 EUR).
However, sub-investment grade issues may exhibit a

positive correlation between stock and bond price, and are Source: Deutsche Bank Convertible Research
20
attractive for high yield investors, but are not suitable for Fig 14: Adding Convertibles Lowers Volatility and
fixed income substitution. Increases Returns
Having established which bonds are suitable for inclusion in

a fixed income fund, we can look at the effect of adding
convertibles to a portfolio of fixed income instruments in a
reasonable equity market. If we go back to a period of
reasonable equity returns (1997/1998), we can see the
performance enhancing characteristics of convertibles very
clearly.
An analysis of this period shows that, beyond the obvious

point that convertibles will outperform straight bonds if the
underlying stocks go up, the negative correlation between
the two asset classes (bonds and equities) causes a
significant initial improvement in the Sharpe Ratio (defined
below) of the fixed income portfolio as convertibles are
added to it.
But fixed income investors are certainly not restricted to
simply buying and holding convertibles in a mixed portfolio.
To demonstrate this effect, we added weightings in a
They can also buy Asset Swapped convertible paper or write
balanced global convertible portfolio to the JP Morgan
Credit Default Swaps against convertibles. These two
Global Bond Index and calculated the returns and volatility
strategies are the subject of the next chapter.
for progressively greater weightings in convertibles over the
time period (Figure 14).
Because the period studied covered a bullish equity

environment, the absolute return goes up with a greater
weighting in convertibles. No surprise there. But what we
find much more interesting is that the negative correlation
between bonds and stocks causes the volatility of the mixed
portfolio to fall as convertibles are added (up to a weighting
of around 15%).
When this is plotted in terms of Sharpe Ratio (the amount of

return above the risk free rate per unit of volatility), it
becomes clear just how beneficial a certain percentage
weighting in convertibles can have on a fixed income fund in
a positive equity market.

4. Credit Protection for Convertibles (Asset Swaps and Credit Default Swaps)
In this chapter we look at how convertibles can be bifurcated dramatically since the mid 1990s. As of the middle of 2001,
into a credit and an equity component through the use of it is generally believed that at least 30% of the European
Asset Swaps and Credit Default Swaps. convertible universe had been asset swapped, with this
figure rising even higher in Japan. Even in the US, a market
Stripping the credit (or bond) part of a convertible from the whose credit quality previously precluded extensive asset
equity option enables the credit risk to be held by investors swapping, the emergence of the investment grade
who understand credit, but who have no knowledge of (or CoCo/CoPay issuer (see Chapter 10) has brought a sizeable
interest in) equity volatility. pick-up in the level of asset swapping in this market too.
In fact, many of the fixed income investors who buy The spreads on convertible asset swaps will always be
convertible credit are prohibited from owning any higher than on comparable straight bonds because of their
equity-related exposure at all, and so products that give callability (discussed below). The higher spread can make
them exposure to the fixed income component of the swapped convertible paper attractive to fixed income
convertible without the equity risk are very attractive. buyers, as can the fact that convertible asset swaps open up
Credit protection products allow convertible investors to a whole new universe in corporate credit. There are many
concentrate on the equity component of the security, examples in Europe and the US where a convertible issuer
although if the credit moves in their favour, they may also be has no publicly traded straight debt outstanding, and the
able to realise positive P/L by recalling the bonds and convertible often represents the only vehicle for credit
re-swapping them. investment available to a fixed income buyer.
Fig 15: Who Buys and Sells Convertible Credit ? For the convertible investor, the use of asset swaps has
transformed the convertibles market. It has led to much
larger issue sizes (in Europe and the US particularly) and has
enabled convertible investors to focus on what they really
understand, which is either the underlying equity (in the
case of outright convertible funds) or equity volatility (in the
case of hedge funds).
So how do convertible asset swaps work?

The Classic Par / Par Asset Swap
We will look first at Asset Swaps, before turning to the The majority of asset swaps are now conducted as Par / Par
newer market for convertible Credit Default Swaps. swaps. We will examine the mechanics of the classic Par /
Par swap and then discuss exceptions to this standard
model below.
Convertible Asset Swaps
The market in convertible asset swapping has grown Three players are generally involved in the full asset swap
22
Fig 16: Full Asset Swap Transaction
transaction (which is shown in Figure 16): Fig 17: Leg 1 - ‘Selling the Bond Floor’
1. Convertible Investor - who sells the bond floor of the

convertible to remove credit risk and free up balance
sheet.
2. Credit Investor - who takes on the convertible credit risk

for an attractive spread. Source: Deutsche Bank Estimates
3. Swap Intermediary - who facilitates the transaction. The

intermediary is usually an investment bank. The first point to understand about this transaction is that it
is not a physical interest rate swap. After the initial
The flows in the overall transaction are shown in Figure 16. exchange, in which the convertible investor sells the
We will split them into two legs: Leg 1 - between the actual convertible to the swap intermediary for the (asset
convertible investor and the swap intermediary, and Leg swap) bond floor, (retaining also the right to recall the bonds)
2 - between the credit investor and swap intermediary. no other cash flows are exchanged. However, asset swap
methodology is used to calculate the price of the asset
The Convertible Investor swap bond floor, and this is why it is crucial to understand
asset swap conventions.
Leg 1 - Convertible Investor to Swap Intermediary - 'Selling
the Bond Floor' The reason asset swap conventions are used to calculate
the bond floor in a convertible asset swap is that, even
We turn first to the 'swap' between the convertible investor though no actual asset swap cash flow payments are
and the swap intermediary, which we isolate in Figure 17. exchanged in Leg 1, they will be exchanged between the

swap intermediary and the end credit investor in the Swap Frequency Quarterly
second leg of the transaction. To avoid a mismatch of Bond Day-count actual / 360
payments, asset swap conventions are used on both legs. Convertible Price 102.00
Determining an 'Asset Swap' Bond Floor What happens in Table 6 is fairly straightforward. But before
we can look at the swap notional payment amounts
As stated above, an asset swap bond floor is calculated to (columns A & B), we need to know how to calculate implied
determine how much is paid to the convertible investor in LIBOR (column D), because this is the rate we will use to
return for selling the bonds to the swap intermediary. determine the amounts of the floating leg payments.
The asset swap bond floor in a Par / Par swap is:

Calculating Implied LIBOR:
PV of an interest rate swap where LIBOR + credit spread To calculate implied LIBOR, the relevant zero coupon yield
payments are exchanged for the convertible coupons curve is boot strapped from the offered side of the swap
curve. We need a zero coupon curve in order to find the
(remember that this 'swap' is notional only - no exchange of discount factors that we will use to calculate implied LIBOR.
coupons and LIBOR + credit swap payments actually occurs As the above swap is being calculated on a quarterly basis,
on Leg 1). we have four legs a year on the floating side, and therefore
need 20 3M zero coupon points (for a five year bond).
This notional swap is valued using asset swap conventions
and is not simply the PV of the convertible cash flows From the zero curve, we calculate discount factors using:
discounted using a credit spread. The asset swap bond floor
is not the same at the traditional convertible bond floor, and 1/(1 + zero coupon rate / 100)
number of days / 360
is not calculated in the same way as yield to maturity. In our

view, this is very important. From these discount factors (some of which appear in
4
column C) we get the implied LIBOR by applying:
Asset Swap Bond Floor = Traditional Bond Floor
((DFT-1 / DFT ) - 1) x 36000 / number of days
To show how an asset swap bond floor is calculated, we will
use an example:
Therefore to calculate implied LIBOR for 24 September
2006 we have
ABC 5% 2006 USD
0.798079682
Swap Settlement Date 24 September 2001 -1 x (36,000 / 92 ) = 6.25%
0.785539738
Maturity Date (convertible) 24 September 2006
Coupon Pay Dates (annual) 24 September annually
4
Redemption 100% Note: You cannot calculate the exact implied LIBOR in column D for all
the points from the DFs in column C as we use many more zero coupon
Credit Spread 100 bps
points in actual calculation of implied LIBOR.
24
Table 6: Calculating the Asset Swap Bond Floor of ABC 5% 2006 USD
Column A Column B Column C Column D Column E

Date Fixed Income LIBOR Floating Zero Coupon Implied NPV
(T) Flows (USD) Flows (USD) Curve Discount LIBOR (%)
Factors
24 Sep 01 10,000,000.00 -10,000,000.00 1 0.00

24 Dec 01 89,315.01 0.993464241 2.53 88,731.27
24 Mar 02 90,012.97 0.987047154 2.60 88,847.04
24 Jun 02 96,580.46 0.980086102 2.78 94,657.16
24 Sep 02 -500,000.00 99,735.36 0.972869376 2.90 -389,405.21
24 Dec 02 122,149.12 0.963535478 3.83 117,695.01
24 Mar 03 125,747.49 0.953924924 4.03 119,953.67
24 Jun 03 139,318.81 0.943194832 4.45 131,404.78
24 Sep 03 -500,000.00 148,863.67 0.93170614 4.83 -327,155.88
24 Dec 03 146,316.43 0.92056376 4.79 134,693.60
24 Mar 04 153,724.48 0.908889375 5.08 139,718.55
24 Jun 04 163,934.78 0.896483901 5.41 146,964.89
24 Sep 04 -500,000.00 170,906.83 0.88364008 5.69 -290,799.92
24 Dec 04 161,501.90 0.871764543 5.39 140,791.63
24 Mar 05 163,676.82 0.859840548 5.55 140,735.97
24 Jun 05 174,418.80 0.847228431 5.83 147,772.57
24 Sep 05 -500,000.00 180,087.32 0.8343353 6.05 -266,914.44
24 Dec 05 170,875.03 0.822361936 5.76 140,521.12
24 Mar 06 170,839.07 0.810541081 5.83 138,472.08
24 Jun 06 181,697.84 0.798079682 6.11 145,009.35
24 Sep 06 -10,500,000.00 10,185,193.31 0.785539738 6.25 -247,295.52
NPV 394,397.75
Notional 10,000,000.00
Net Bond Floor 9,605,602.25
Asset Swap Bond Floor = 96.056


Column D shows the implied LIBOR points. We then add Fig 18: Zero Coupon Curve and Implied LIBOR Curve
the spread (100 bps) to the implied LIBOR rate and apply
that to the principal (USD 10 million) to calculate the floating
leg payments:
Notional x (implied LIBOR + credit spread) x (number of days / 360)
In the example in Table 6, we have rounded the implied

LIBOR rates, but in order to calculate the exact floating
payment amount on 24 December 2001 the equation is:
USD (10,000,000 x 0.03533341) x (91 / 360) = EUR 89,315.01
This amount is then discounted using the zero coupon curve

discount factor (column C) to find its present value:
Source: Reuters
(USD 89,315.01 x 0.993464241) = EUR 88,731.27

What does the Convertible Investor Own?
We then build up our twenty floating legs (together with the
Physical ownership of the bonds passes from the convertible
exchange of principal) as in column B. The fixed flows are
investor to the swap intermediary (and then probably on to
known in advance and are entered in column A. All the
the credit investor), at the time of the 'swap'.
flows on both sides of the swap are then PV'd using the
original zero coupon discount rates (column C) and then
But the convertible investor retains the right to call the
simply summed to get the asset swap bond floor.
bonds back at any time (subject to conditions explained
below) at the recall spread. This gives the convertible
This asset swap bond floor is the amount the convertible
investor an option on both the underlying stock and the credit
investor will be paid in exchange for the convertible at the
spread of the issuer. The recall spread is agreed at the time
time of the swap.
of the initial swap. The recall spread will always be tighter
than the original swap spread (i.e. a higher price). The asset
Note on the Implied LIBOR Curve
swap will therefore usually be quoted LIBOR + 100 to
The implied LIBOR curve is well above the zero coupon LIBOR + 90 bps, with 100 bps the bid and 90 bps the recall
curve, as is demonstrated in Figure 18. The large difference spread.
between the two curves is one of the reasons why the
asset swap bond floor is different from the traditional bond Why Exercise the Option to Recall the Swap?
value.
There are four reasons why the convertible investor might
call the bonds back, either at maturity or before:
26
1. To Convert the Bonds - If the convertible is 'in the money' 1. Auto Exercise - If the issuer calls the convertible, normal
at maturity, the bonds will be recalled so the convertible procedure would see the swap intermediary call the
investor can convert them. bonds back from the credit investor on behalf of the
convertible investor, on the basis that the convertible
2. If the Spread has Tightened - If the spread tightens investor will want to convert them (if in the money).
significantly it may be profitable for the convertible
investor to recall the existing swap and re-swap the 2. Contingent Mandatory Exchange (the so-called 'Death
bonds at a higher bond floor. Spiral') - Many issues in the US (and a few in Europe)
allow the issuer to redeem the convertible in shares
3. Corporate Action - It may occasionally be necessary rather than cash. This can have many unfortunate
to recall the bonds to participate in some kind of consequences, and one of them is that the credit
corporate action. investor is usually unable to take delivery of physical
shares on redemption. In this instance, normal procedure
4. To Sell the Convertible - If the convertible investor wants is for the swap intermediary to have the right to sell the
to sell the convertible for any reason, they must first shares delivered in lieu of cash back to the convertible
recall the bonds. investor in return for the redemption amount in cash.
The amount that the convertible investor must pay to recall 3. Event of Default - In the event of default, the convertible
the bonds is calculated using exactly the same methodology investor usually has a limited time period (often 5 days)
as was used to calculate the original asset swap bond floor. to decide whether or not to call the bonds. It is hard to
The only differences are: foresee a situation in which this call would be exercised,
and so the objective is to give the credit investor the
1. The Recall Spread is the spread used. maximum amount of time to enter into negotiations with
the issuer about recovery, restructuring and administration.
2. The date for calculating the recall bond floor will
be the day of the recall, with the current swap curve Recall Penalties
being applied.
Unwinding the swap early often causes problems for the
3. The bid side of the swap curve will be used - as end buyer of the credit. The credit investor will usually have
opposed to the ask side which is used to calculate the set up liabilities of its own to match the inflow of floating
original bond floor when the swap is first calculated. payments, and so will face significant disruption if it is
forced to unravel these transactions early.
Other Recall Scenarios:
In order to compensate for this, small penalties are usually
The right to recall the bonds is held by the convertible payable in the event of early recall.
investor and is generally only exercised when it is in the
convertible investor's favour. However, a couple of extra 'Make Whole' Provisions - The most common remedy is the
points need to be made. 'make whole' payment, which is usually a payment of the

PV of the spread that would have been payable on the swap When asset swapping these bonds, there are two possible
for the specified 'make whole' period - usually six months. methodologies:
This should give the credit investor time to untangle its own
obligations and replace the lost swap with another one. 1. Swap Using Issue Price as Notional - This method uses
the issue price as the notional and uses the difference
Occasionally the ‘make whole’ penalty is an absolute between the issue price and the first put as a balloon
amount (e.g. 0.5 - 1.0 point). Absolute amount make wholes payment exactly as in the premium redemption case
are widely used in Asia. In the high yield market, this amount above. Swap intermediaries tend to avoid this method as
is usually punitive (perhaps 5 points), meaning that convertible it results in considerable credit investor counterparty
investors entering into these swaps are effectively locked in risk. This is because the balloon payment, that is the
until maturity. difference between the issue price and the first put
price, is back end loaded and very high.
In very rare instances there can be a restrictive non-call
period, but this is not common due to the necessity of 2. Swap to First Put Date - Swapping to the first put date
bonds being recalled to participate in unforeseen corporate means that the first put price is the notional amount. The
actions. credit investor will pay the full amount of the put up
front and this lessens the counterparty risk significantly
Variations on the Par / Par Theme versus the issue price notional swap. This method is
preferred by swap intermediaries.
Although the Par / Par structure is now the established
standard for convertible asset swapping, there are other Fig 19: Discount Swaps: Verizon 0% 2021 USD
methodologies:
Premium Redemption
Many bonds in Europe are issued at par (100%) but redeem

at a premium (say 125%).
These bonds are easily handled by treating the difference

between par (or more correctly par + final coupon) and the
redemption amount as a balloon final coupon.
OID (Original Issue Discount)
Many US convertibles in 2000 and 2001 were structured Method 1: 60.841 is notional, with 8.048 as balloon payment in year 5
as deeply discounted 0% coupon bonds with 20 year Method 2: 68.889 is notional, with no other payment passing from credit
investor to swap intermediary
maturities and rolling 5 year puts.
28
Interestingly, the bond floors on the two methods will not be isolates the equity call option with a floating strike
exactly the same, with the issue price swap generally giving (struck at the recall spread) on the underlying shares.
a marginally higher asset swap bond floor. This is because The implied volatility of this option can be calculated, and
the swap notional is smaller using just the issue price swap, whether or not it is attractive versus comparable options
and therefore the PV of the floating payments is lower so the and the volatility of underlying stock, will be key in
PV of the swap is lower, resulting in a higher bond floor. determining whether the trade is put on.
The two methods are illustrated in Figure 19, where we The Credit Investor
show two alternative methods for asset swapping Verizon
0% 2021 USD. Leg 2 - Credit Investor to Swap Intermediary - 'Buying the
Convertible Credit'
Why Asset Swap?
Having examined the rationale and methodology behind the
The majority of asset swap transactions involve convertible convertible investor's sale of the credit, we now turn to the
arbitrage funds. Why do they asset swap convertibles? position of the end buyer of the credit (the credit investor).
There are four main reasons: It is important to stress that the sale of the convertible
credit to the credit investor is a full interest rate asset swap,
1. Credit Protection. The hedge fund is no longer exposed involving a full schedule of cash flow exchanges.
to the credit of the issuer on the downside if the position
is swapped. This is the primary reason for asset swapping. As Figure 20 reminds us, in the classic Par / Par swap the
credit investor:
2. Leverage. Once a position is swapped the hedge fund
receives the value of the bond floor in cash. An 'out of 1. Buys the convertible for par;
the money' position such as ABC Corp 5% 2006 USD
might be put on for only a few points (5.94 in this case). 2. Passes the coupons on the convertible to the swap
If the delta of the remaining 'stub' option is high enough, intermediary; and
once the delta (hedge ratio x stock price) is sold, the
hedge fund might receive a net cash credit to put on the 3. Receives LIBOR + credit spread on the full par amount
trade! Asset swapping can also help with investor risk based on the frequency of the contract (but usually
limits. quarterly referenced to 3M LIBOR).
3. Cross Margining. The asset swap is recognised by more It is important to understand that the credit investor
sophisticated prime brokers (e.g. Deutsche Bank), which receives LIBOR + credit spread on the full par amount of the
enables cross margining of the bond option against the bonds, in exchange for assuming the full credit risk of the
stock short. convertible. This is because the credit investor physically
owns the bonds, and will only receive whatever is paid on
4. Floating Strike Call Option. The sale of the bond floor the bonds at maturity (recovery rate in the event of default).

Fig 20: Leg 2 - ‘ Buying the Convertible Credit’ It is usual for the credit investor to physically own the bonds
during the life of the swap, but the bonds might be placed
in an SPV (Special Purpose Vehicle) if specific counterparty
risks are identified, or if the credit investor is precluded from
holding convertible instruments.
Based on our example of ABC Corp 5% 2006 USD above,

it appears that 3.944 points have 'gone missing' in the
It is important also to note that the swap with the credit transaction. The swap intermediary has paid the bond floor
investor is callable by the swap intermediary, to mirror the of 96.056 to the convertible investor but has received 100
call the convertible investor has. If the swap is called early from the credit investor.
the credit investor will simply receive the par amount back
early, plus whatever early redemption penalties were So what happened to the missing 3.944 points?
agreed.
Figure 21 demonstrates that 3.944 is exactly the PV of the
It is also a feature of Leg 2, between the credit investor and swap that the swap intermediary must conduct internally
the swap intermediary, that it is not contingent on Leg 1 in with its own swap desk in order to convert the fixed income
any way. This means that in the event of default, or any flows from the convertible into the LIBOR + credit spread
other event, the credit investor will still be required to payments on the full par amount to give to the credit investor.
honour the swap payments - paying the fixed coupons
based on the convertible and receiving floating as per the
original agreement.
Fig 21: ABC 5% 2006 USD - Fully Asset Swapped
30
Other Variations which case the protection seller is obligated to make a
contingent settlement (i.e. to compensate the convertible
Fixed / Fixed Swaps investor for the credit event).
In Japan, where the credit buying community is more The rise of CDSs has given investors a new alternative to
accustomed to receiving fixed interest rather than floating asset swaps for hedging the default risk of an issuer.
LIBOR, most asset swaps are done as fixed / fixed
transactions. There is no material difference with the Credit events are determined by the 1999 ISDA Credit
floating swap described above, except that the credit Derivative definitions. Investors are given full protection
investor passes the fixed coupons from the convertible and against:
receives fixed payments as determined by a pre-agreed
schedule. 1. Failure-to-Pay;
2. Bankruptcy;
Credit Default Swaps (CDS) 3. Obligation Acceleration/Obligation Default;
4. Repudiation / Moratorium; and
A credit default swap is very similar to an insurance policy.
5. Restructuring*
A holder of a security, a loan, or any other debt obligation
pays a fixed premium (usually annual) to ‘insure’ the risk of
*(can be ISDA 1999 full restructuring events or ‘Modified Restructuring’).
the obligation not being fulfiled.
Credit default swaps provide an efficient hedge for

Credit Default Swaps transfer the credit risk of a corporate
longer-term credit risk (credits of over one year to maturity).
from the buyer of the swap to the seller, and as such allow
The credit spread is locked in for the term of the swap, with
investors to short the credit of an issuer. Two parties enter
the credit seller (convertible investor) retaining exposure to
into a swap agreement where the protection buyer pays a
changes in interest rates. Maturities can be customized to
fixed periodic payment, usually expressed in basis points
suit convertible bond idiosyncrasies, reflecting put and / or
per annum on the notional amount, for the life of the
call dates if necessary.
agreement. The protection seller (the credit investor) makes
no payments unless a specified credit event occurs, in
From the perspective of the CDS seller (credit investor), the
credit default swap provides a synthetic long bond position
Fig 22: A Credit Default Swap
giving a fixed return for the term of the default swap (in
excess of that which would be available from buying the
bond directly). CDS writers can also gain exposure to
specific credits in sizes and maturities not always available
in the corporate bond market.
The wider CDS market is highly liquid and currently (2001)

accounts for over 50% of the total credit derivatives market.
Source: Deutsche Bank Estimates The market has grown with phenomenal speed. Starting

from just USD 50 billion in 1996, the market had increased Fig 23: CDS Gives Credit Tightening Upside to CDS Writer
to USD 350 billion by 1998. The USD 1 trillion mark was
reached in 2000.
With more and more players and solid demand from both
sides of the market, the bid to offer spread for high-grade
corporate names can be as low as 10 bps or less for
maturities between one and five years (in normal market
conditions). The normal range for maturities is one to ten
years, and the market for investment grade names tends to
be very liquid, with transaction sizes ranging from USD 5
million - USD 10 million, but occasionally up to USD 500
million for portfolio trades. Contract transfer and assignability
has been greatly enhanced by the introduction of standardised
documentation and terms by ISDA (Deutsche Bank's ISDA
documentation is available on www.dbconvertibles.com).
The key difference between CDSs and asset swaps, is that The pay-off profile is completely different in a CDS, as
the writer of the CDS retains the credit upside. This means compared to an asset swap. Figure 23 illustrates that the
that CDSs will usually provide protection at a lower rate than CDS writer will make a positive capital gain if the spread
an asset swap (in which the credit investor actually sells the tightens, whereas the buyer of asset swapped convertible
credit tightening option through the callability of the asset paper (also a credit investor) does not benefit from a spread
swap). tightening.
Figure 24: Credit Default Swaps - The Mechanics
32
How are CDSs priced? original level. But how is the 'value' of a CDS determined?
Because swaps are priced in terms of spreads over LIBOR,
Although CDS spreads have a strong correlation to spreads the 'price' (which may be a positive or negative for the
in the cash markets, they are tradeable securities in their convertible investor) is actually the net present value of
own right and subject to unique supply and demand factors. the difference between the initial spread and the spread
The 'basis' between corporate bonds and CDS spreads is at the time of closing the swap.
influenced by a number of factors, including mismatches in
documentation, although the repo costs of the underlying We have approached the pricing of a CDS using the ‘no
securities is probably the most significant influence on any arbitrage’ method above (deriving the correct CDS level
apparent arbitrage. from the difference between current credit spreads and
financing costs). But there is another, more predictive,
Arbitrage Approach to Pricing a CDS method to pricing. This second approach grew from option
5
pricing, and treats the default swap as a credit option.
The repo rate is central to the bond / credit default swap
arbitrage. This is best illustrated by looking at a sample The predictive model requires both probabilities of default
trade: an investor buys default protection, paying x basis and the most likely loss in the event of default. The rate of
points per annum. This is used to hedge bonds which earn recovery (i.e. the percentage recovery value per 100 of
LIBOR + y basis points. The bonds are financed at LIBOR + exposure) may be observed from historical data for different
z basis points. For 'arbitrage-free' pricing the rate earned on industry sectors and ratings.
the bonds should exactly equal the insurance cost plus the
cost of financing (i.e. x = y - z). Consequently the CDS If we make the assumption that existing credit spreads in
spread should approximate to the spread over LIBOR the cash market contain all known information about the
minus the repo spread. This is more easily explained likelihood of default, then probability of default is:
diagrammatically (see Figure 24).
Probability of Default = Spread / (1 - Recovery Rate / 100)
The potential for arbitrage means that the mark to market for
CDSs should be closely related to changes in the credit This is a crude yet reasonably accurate approximation of the
spread of the issuer in the corporate bond market. Thus probability of default.
CDSs should hedge specific changes in the credit spread of
the issuer, as well as the outright risk of default, though this But the credit curve will have a term structure, requiring
can be limited by liquidity in the repo market. different values at different time intervals. Where gaps exist
Unwinding Credit Default Swaps 5

Academics have constructed an option pricing model for default swap
pricing. But although the CDS transaction does involve the payment of a fee
CDSs allow investors to short bonds synthetically. CDSs are in return for a payoff should a default occur, CDSs should not be confused
not 'callable' by either side, and so there is no optionality. with genuine default options. The key characteristic of options is their
asymmetrical payoff and price performance as the price (or credit spread) of
Consequently, if a convertible investor wants to break a CDS
the underlying obligation changes. With CDSs, by contrast, the price varies
early they will have to 'buy' the credit back from the credit directly with changes in the credit spread of the underlying, in much the
investor, and the price may well have moved from the same way as the price of an interest rate swap.

in the observable maturity spectrum, assumptions have to Table 7: Differences Between Asset Swaps and CDSs
be made. In an ideal world, there would be a complete set of
zero coupon bonds across every maturity, for every corporate Credit Default Swap Asset Swap
name, from which to extract the default probabilities. But Standard contracts Non-standard contracts
unfortunately the world is not that simple! So in practice the
Not callable Callable
credit curve is built up from a matrix of observable bonds
and existing CDS contracts. Hedges spread risk Hedges spread and rate risk
No leverage Leverage
When calculating the present value of a default swap, the
Easier to assign Difficult to assign
discount factors are derived from the credit curve, (in the
same way as they are derived from the zero curve in the Credit spread option struck
No fixed income optionality
swap market). at the recall spread
Lower spread over LIBOR Higher spread over LIBOR
The above is a very brief summary of a much more complex Source: Deutsche Bank
subject, but it gives the basic outline of how the present

value of a CDS is calculated. CDSs give credit buyer upside from spread
improvement
The main differences between CDSs and asset
For sub-investment grade bonds, CDSs may well prove
swaps
more attractive than asset swaps from the credit investor’s
perspective. High yield investors, in particular, look as
To summarise, there are two main differences between
much for the capital gain as for actual income in realising
asset swaps and CDSs. Investors who hedge a convertible
their investment objectives. Under an asset swap, the
via an asset swap receive an up front payment for the
convertible investor retains the option to recall the
fixed income portion of the bond. This effectively gives the
bonds, and therefore retains the upside from any credit
convertible investor off-balance sheet gearing. With CDSs
improvement. This is clearly against the interests of the high
this does not occur; the investor simply makes a periodic
yield investor. Consequently, with high yield asset swaps,
'insurance' payment and there is no gearing benefit.
credit buyers are seeking harsher and harsher penalties for
early recall.
Secondly, asset swaps give optionality; the bond can be
repurchased at the recall spread, giving the convertible
CDSs give the credit buyer the upside from any spread
investor upside in any improvement in the credit. CDSs
improvement, and so are more suitable for high yield funds
create a synthetic short, and therefore result in full exposure
as an investment vehicle. From the point of view of the
to any changes in the credit spread. This means that they
convertible investor, giving up the credit upside is often a
eliminate any exposure when used as a hedging instrument.
small price to pay for the ability to hedge more risky credits,
CDSs give no protection against rate changes. These
which often have more interesting (and volatile) underlying
differences are reflected in the cost of the two instruments.
equities.
34
5. Convertibles for Equity Investors
Studies of convertible performance have suggested that,

over a long time period, convertibles can replicate virtually all
of the upside performance of equities with substantially
lower volatility. 2. Put option on equity - The 'put option' on the underlying
shares is the convertible's cash redemption at maturity.
For equity investors, convertibles offer the following three An equity investor buying Ahold 4% 2005 EUR for 121.75
components: when parity is 113.78 pays a small premium (7.00%) to
own the same number of shares, but gains a lot of
Fig 25: Convertibles for an Equity Investor potential downside protection. If, at maturity, the shares
have fallen from EUR 36.65 to EUR 18 (i.e. parity
drops from 113.78 to 55.88), the bond will still be worth
100 (redemption value). The shares have fallen by
50.89% but the bond is only down 17.86% (even before
the convertible's yield advantage is taken into account).
The put value of the convertible can be roughly modelled

using the Deutsche Bank Option Calculator (located on
the www.dbconvertibles.com home page). An investor
can model the put by setting up a European style option
We will deal with each of these three components in turn, with parity as the equity spot price and 100 (or whatever
and in doing so will explain how an equity investor can the redemption value is) as the strike price.
determine whether to switch an equity position into a
convertible position. For our calculation we use 4.65% as the risk free rate
and 27% for the volatility. For the dividend yield we
1. Convertible holder still owns equity - The most important used 1.5%, and maturity for Ahold 4% 2005 EUR is 19
point to stress is that a convertible investor still May 2005. Valuing the put to maturity in 2005, we get:
controls the underlying shares. With most convertibles,
conversion can occur at any time. The number of shares Put Value = EUR 10.20 (EUR 102.00 per bond)
owned via a convertible is simply:
Obviously there is a degree of credit risk attached to
Conversion Ratio x Number of Bonds the put (as with all puts). Therefore investors need to
be more wary of lower credit quality convertibles
Once again turning to the example of the Dutch food in this respect.
group Ahold 4% 2005 EUR, we will assume that Ahold
stock is trading at EUR 36.65 and the convertible at 121.75.
Equity Value = 31.0463 x EUR 36.65

= 113.78 (EUR 1,137.80 per bond) 3. Income Advantage - An equity investor switching out of

shares and into a convertible gives up equity dividends This 131.75 total is well above the 121.75 theoretical value
and receives convertible coupons. This is likely to calculated using a binomial tree. Unfortunately, the equity
enhance the current yield profile. In order to calculate investor has not discovered a free lunch here. The disparity
the present value of the yield advantage it is necessary is caused by the issuer's option to call the bond early and
to project the stock dividends and coupons into the force conversion if it is ‘in the money’. An early call will
future.The difference between the cash received is then reduce the value of both the put and the income advantage.
PV'd to calculate the value of the income advantage. To value the bond properly, a binomial tree is used to
account for the issuer's call option and to model its exact
Table 8 below shows how this calculation is arrived at. effect on the bond’s theoretical value.
PV of Income Advantage = 7.77 Breakeven Horizon
The total economic value of the convertible to an equity Discounted cash flow analysis also allows equity investors
investor is therefore: to work out what premium they should pay for a convertible
in order to 'breakeven' by a certain time.
Equity Value = 113.78
+ Table 8 shows how discounted income advantage can be
Put Value = 10.20 used to calculate how long it takes to amortise different
+ levels of premium. From Table 8 it should be clear that an
Income Swap = 7.77 investor paying EUR 59.60 per bond (premium of 4.95%)
will 'breakeven' in exactly three years.
Total = 131.75 (per bond)
The ringed box shows that in 2004, the PV of the income
Table 8: Ahold 4% 2005 EUR - Value of Income Advantage over Shares
Year Convertible Coupon Stock Dividend Yield PV Running

(4% on EUR 1,000) (1.5% x EUR 1,217.50) Advantage (4.65%) Total
2002 EUR 40.00 EUR 18.26 EUR 21.74 EUR 20.78 EUR 20.78
Total Income Advantage per Bond = EUR 77.73
We use dividend yield x market price of convertible (1.5% x EUR 1,217.50) for this table to calculate the opportunity cost of owning the same cash value of
shares via the convertible.
36
advantage is EUR 59.60, giving a 3 year ‘breakeven’ to an convertible. Because the convertible’s premium is so low, if
equity investor switching into the convertible in 2001. the shares rally the bond will rise almost 1 for 1 with the
underlying shares.
Equity Switching
If they fall significantly, however, the convertible should fall
Convertibles can be highly effective in equity portfolios by much less. After a decline in the stock, premium will
when they form part of a 'switching' strategy. Investors have opened out and the investor can switch back into the
'switch' between the underlying shares and the convertible shares in anticipation of the next rally.
to protect profits and retain equity exposure in difficult
market conditions. Equity investors switching into In order to emphasise the success of this strategy, we have
convertibles at low premiums often give themselves a made our 'switches' at the most advantageous moments,
highly asymmetric risk/return profile. If the shares continue but even a simple predetermined strategy such as:
to rally, the convertible will rise on virtually a 100% delta (i.e.
1 for 1) with the shares. If the shares fall, the convertible Switch into convertible when premium < 5%
will outperform as premium expands on the way down. Switch back into shares when premium > 15%
And all the time the convertible is likely to be yielding more
than the underlying shares. would have outperformed simple 'buy and hold' strategies
of either Seacor stock or the convertible.
The main objection to using 'in the money' convertibles
from equity fund managers is that they only buy stocks Fig 26: Switching Strategy at Work - Seacor 5.375% 2006 USD
which they believe are going up, and therefore do not need
the protection inherent in convertibles. This may be so for
active managers (although the markets of 2000 and 2001
rather undermine this argument), but funds benchmarked
against major indices will usually be compelled to retain
exposure to certain large cap names even though they may
have a cautious stance on the stock (Vodafone in 2001 is a
good example). For index-related funds, convertible and
exchangeable bonds can be liquid, inexpensive and obviously
defensive alternatives to OTC put strategies.
We chose a benchmark US convertible, Seacor 5.375%

2006 USD, to demonstrate the 'switching' strategy in action
(Figure 26).
The Switching Strategy in Action Having identified the switching points we can look at the P/L
of the strategy (Table 9). The P/L is simple to calculate. At
When the convertible's premium is very low an equity the first point (12/01/01), we switch from the stock into the
investor sells the underlying shares and switches into the convertible. The convertible falls -7.42% to the next

Table 9: P/L of the Switching Strategy
Date Switch CB Price Parity CB Gain (Loss) Parity Gain (Loss)
12/01/01 Switch into CB 123.00 122.159
Switch out of CB
22/01/01 113.875 109.94 -7.42% -10.00%
back into Stock
Switch out of Stock
08/02/01 119.00 1117.72 4.5% 7.08%
back into CB
Switch out of CB
30/03/01 111.75 101.13 -6.09% -14.09%
back into Stock
Switch out of Stock
06/06/01 114.25 111.36 2.24% 10.12%
back into CB
Switch out of CB
19/07/01 106.00 94.56 -7.22% -15.09%
finally back into Stock
switching point (22/01/01), but parity is down -10.00%. fell almost 23%. But all the time the investor retained
Premium has expanded and so we switch back into the exposure to the underlying shares.
stock at 22/01/01. The switching then goes on until 19/07/01
when we switch back into the stock for the final time. Convertibles also offer a number of other attractions for
equity investors:
The results of this strategy (although exaggerated by the
effect of hindsight) are impressive. 1. Reasonable Liquidity - Particularly in Europe, convertible
and exchangeable issue sizes have increased a great deal
The investor who employed the switching strategy achieved in the last couple of years. This has resulted in much
an almost flat performance, even though Seacor stock better liquidity in convertible bonds than was the case in
the mid 1990s. Equity 'switchers' will generally get good
Table 10: 'Switching' - Total Return Analysis prices from convertible market makers if they sell (buy)
the shares and buy (sell) the convertible as part of the
Strategy Price Gain Yield Total
same transaction, because the market maker will be
/ Loss Return
naturally hedged. Convertibles can even offer better
1 Hold Stock -22.59% 0.0% -22.59% liquidity than the underlying shares in certain cases. In
some markets, most notably the Asia Pacific region, a
2 Hold Convertible -13.82% 2.76% -11.06%
large Euroconvertible issue can be the most liquid way to
3 Switching -3.53% 1.50% -2.03% gain exposure to the underlying shares, especially if they
Source: Deutsche Bank Estimates do not have ADRs or GDRs outstanding.
38
Convertibles can also offer exposure to local shares A convertible position also buys an equity investor time
when foreign share limits have been exhausted. Many to exit a market more gradually if it experiences sudden
convertibles in the Asia Pacific region are convertible into weakness. This is because the convertible will fall much
local shares, so investors have exposure to the local more slowly than the underlying shares and a large
securities without having to own them physically. position can be unwound over a longer period.
2. Broadens Investment Horizons - Convertibles can have 3. Currency Protection - Cross currency convertibles (bonds
the effect of turning more speculative shares / markets denominated in a currency other than that of the underlying
into much safer investments. An example might be a shares) give equity investors inexpensive currency
stock like Pacific Century Cyberworks in Hong Kong. In protection in volatile markets. US investors very often
October 2000, an investor who thought the story looked buy US dollar denominated convertibles on European
interesting but who was nervous of an outright stocks. These convertibles participate in European
investment in the stock could have bought the PCCW currency appreciation through an appreciation in parity,
3.5% 2005 USD convertible instead. As it turned out, but are protected from large falls in the local currency by
October 2000 was too early to buy the stock, and it more the redemption value in USD.
than halved in the next nine months. But the convertible
limited the loss to -6% only, including the interest. The This feature of convertibles is also what drove the US
investor was wrong but lived to fight another day, whereas dollar denominated Asian Euroconvertible market in the
a big outright position in the stock would have been a mid 1990s, and is still driving issuance from that region
painful experience indeed! today (2001).
Fig 27: Convertibles Offer a Safer Investment in Difficult

Markets - PCCW 3.5% 2005 USD

6. Convertibles for Hedge Funds
Since we wrote our first Investors' Guide in 1997 perhaps Convertibles, like other derivative instruments, are good for
the most important development in the market has been the arbitrage investing as two assets (the convertible and the
explosive growth of the convertible hedge fund industry. It underlying shares) trade with predictable correlations (at
is guestimated that arbitrage funds now own as much as least theoretically). This allows a (theoretically) risk-free
50% of the world's USD 450 billion stock of convertible portfolio to be constructed with a long convertible position
securities! On any given day in Europe or the US, hedge hedged with a short position in stock. Ignoring, for the
fund trading can represent up to 75% of convertible moment, the problem of credit exposure, the portfolio
volumes. Any investor or issuer involved with convertibles should display all the characteristics of a hedged equity call
should have at least some idea what these funds do. option, with the call struck at the conversion price and
maturing at the maturity of the convertible.
Hedge funds differ from outright investors in that they do
not generally take directional views, and instead try to We will deal with more of the complexities later, but first
exploit relative valuation anomalies in the market through let's examine a simple long convertible/short stock delta
the use of leverage. Within convertibles, hedge funds buy hedging scenario.
convertibles and short shares against them to try to isolate
undervalued equity options. The concept of delta is the first thing we need to
understand. Convertible delta is the bond's price sensitivity
In our opinion, a lot of ill-informed comment has been to movements in the underlying share price. It is the
written about the operation of convertible arbitrage funds, 'equity sensitivity' of the convertible.
but we believe that such funds have had a demonstrably
beneficial effect in four crucial areas: Convertible Price
Delta =
Stock Price x Conversion Ratio
1. Efficiency of Pricing - ensuring that convertible securities
trade closer to their theoretical or 'fair' value, ensuring Given that it is a hedge ratio, it must also follow that:
that the full economic benefit of these instruments
(Delta x Stock Price x Conversion Ratio) = Convertible Bond
accrues to all investors and issuers.
for any small movement in the share price.

2. Promoting Liquidity - frequent trading by arbitrage funds
has dramatically improved liquidity, in Europe particularly.
It is clear from this equation that in order to calculate delta
we have to understand the theoretical price change of the
3. Product Innovation - the convertible asset swap and
convertible. In fact, it is only possible to do this accurately
credit default swap markets have developed largely as a
by using a convertible model such as those described below,
result of their activities, to name but two innovations.
from which delta is derived from observations of a
lattice-based (binomial or trinomial) tree.
4. Issue Size Increases - the new demand for convertible
securities has resulted in a significant increase in issue
We assume that our model gives us 60.00% as a delta for
sizes in Europe and the US.
Ahold 4% 2005 EUR, and that we are hedging EUR
40
1,000,000 of bonds. The following facts are relevant: We can see the hedge ratio of 60% was correct, given that
the net P/L of the entire position was almost unchanged.
Nominal Value = EUR 1,000
Conversion Ratio = 31.0463 shares per bond So, a convertible's delta is a key number. It determines how
Share Price = EUR 29.60 many shares must be shorted to isolate the arbitrage
Parity = 91.90 portfolio (long convertible / short stock) from small
movements in the underlying stock.
Convertible Price = 100
So far, so good. But we haven't actually made any money

The hedge calculation is:
yet! In fact, as a hedge fund, we might even have lost money
on the position above because our prime broker may be
Delta x (Conversion Ratio x No. of Bonds) = shares to be shorted
charging us interest on the (probably) leveraged position, and
60% x (31.0463 x 1,000) = 18,628 shares
we will almost certainly be paying a fee to borrow the
shares we shorted. Clearly, small stock moves are not going
We can check that this is the correct hedge by running a to be much help. But what about larger ones?
scenario. If we increase the share price by 2%, parity will be
93.74, and our model tells us that the bond should trade at Figure 28 shows that convertible delta does not remain
101.10. Let's look at the P/L of our hedged portfolio to check constant, but changes as the convertible's profile changes.
that we were correctly hedged: The change is attributable to gamma, which we shall explain
more fully overleaf.
Table 11: Initial Position
Long
Fig 28: Delta is the Slope of a Tangent Drawn on the
1,000 bonds (EUR 1,000 nominal) @100 = EUR 1,000,000
Convertible Price Line
Short
18,628 shares @ E 29.60 = EUR 551,389
We now increase the share price by 2% to EUR 30.19
Table 12: Position after 2% Share Price Rise

Long
1,000 bonds @ 101.10 = EUR 1,011,000
+ EUR 11,000
Short
18,628 shares @ EUR 30.19 = - EUR 562,379
= - EUR 10,990
Source: Deutsche Bank Estimates Source: Deutsche Bank Estimates

Table 13: A Greater Move in the Underlying Shares Produces a Larger Gain for the Hedged Portfolio
Date Convertible Stock Hedge Ratio New P/L on

Price (EUR) at Outset Theo. Delta Portfolio (EUR)
14/08/01 118.81 35.52 60.00% 70.00% +10,422.24

14/08/01 106.74 29.60 60.00% 60.00% 0.00
14/08/01 97.54 23.68 60.00% 50.00% +18,277.76
Note: The P/L on the portfolio is based on long EUR 1,000,000 bonds / short 18,628 shares
If we move Ahold's share price up and down by 20% for 1 percentage move in the stock. This methodology
instead of 2%, the hedge portfolio returns become much is attractive because it means that gammas are readily
more interesting, as Table 13 illustrates. comparable across different securities. But convertible
models will often express gamma in terms of points change
Table 13 demonstrates that the initial hedge of 60% only in delta for a one point move in stock price, and we revert to
keeps the P/L constant for small movements in the equity this latter methodology below.
price. For greater moves, the convertible is affected by
gamma (the rate of change of delta). Like delta, gamma has to be derived from the binomial (or
other) tree, and once again it is derived from observations in
Gamma the tree itself.
Gamma is the rate of change of delta for movements in the Table 14 shows that delta alone is a very accurate measure
underlying share price. of the future value of Ahold 4% 2005 EUR for rise in parity
of one point.
Delta
Gamma =
Parity However, for a larger move of 10 points in parity, only using
the initial delta results in an unacceptably large error, as
Investors need to be very careful with gamma units. Our Table 15 demonstrates.
favourite expression of gamma is of points change in delta
Table 14: Ahold 4% 2005 EUR - Parity Rises by 1 point
Convertible Parity Delta Parity Theo. Price using Theo. Price Difference
Price (+ 1 point) initial delta using full model
106.74 91.90 60.00% 92.90 107.34 107.35 0.01%

42
Table 15: Ahold 4% 2005 EUR - Parity Rises by 10 points
Price Parity Delta Parity Theo. price using Theo. Price Difference
(+ 10 points) initial delta using full model
106.74 91.90 60.00 % 101.90 112.74 113.94 1.06%

To predict the convertible price for larger movements in Fig 29: Ahold 4% 2005 EUR - Delta Profile
parity we apply the gamma equation:
(Delta x expected move in parity in points) + ½ (Gamma

in points) x (Expected move in parity in points)2
Table 16 (below) shows that the application of the ‘gamma

correlation’ to delta results in a much more accurate result.
Convertible gamma is clearly another key variable for hedge

fund investors.
A look at the delta profile for Ahold 4% 2005 EUR (Figure

29) shows that delta changes significantly at different stock Source: Deutsche Bank Estimates
prices.
Capturing Volatility
The point of maximum gamma is often the most attractive
point at which to 'set-up' a convertible with a risk neutral In order to make money with hedged 'risk neutral' portfolios
hedge. For Ahold 4% 2005 EUR, this point would be when of long convertible and short stock, it is necessary for
parity is around 80.0 (Figure 30). investors to 'capture' stock volatility by trading the portfolio.
When parity rises (falls) and delta rises (falls) more stock is
sold short (bought back) to keep the portfolio 'delta neutral'.
Table 16: Ahold 4% 2005 EUR - Parity Rises by 10 points - Application of ‘Gamma Correlation’
Price Parity Delta New Parity Theo. price Theo. price Theoretical
(+10 points) using initial using Price from
delta only Delta and Model
Gamma
106.74 91.90 60.00 101.92 112.74 113.67 113.94


Fig 30: Ahold 4% 2005 EUR - Gamma Profile 2. To capture a short term volatility ‘event’; and/or
3. To capture a Vega 'richening' of the convertible (explained

a little later).
1. Capturing Excess Volatility
Convertibles are issued, and often trade, at an implied

volatility below the level of volatility arbitraguers would
expect to capture by gamma trading the stock. Indeed no
arbitrageur would buy a convertible if this were not the case,
as they would almost be guaranteed to lose money!
So the most basic strategy is to buy bonds where the

arbitrageur expects to be able to capture more volatility than
The P/L of selling stock high and buying it back low is what that implied by the bonds.
makes up ‘gamma trading profits’.
2. Capturing a Volatility ’Event’
A convertible arbitrageur hedging the portfolio continuously
will return exactly the risk free rate if he ‘captures’ exactly Convertible arbitrageurs often 'set up' portfolios where they
the volatility ‘implied’ by the convertible. This is the definition expect some short-term volatility event.
of 'implied' volatility.
An example of a profitable (and realistic) short term 'set up'
High gamma is good because the large changes in delta, of Ahold 4% 2005 EUR might have been over year end
even for small movements in the shares, give an arbitrageur 2000/01. We assume that a EUR 10 million portfolio was
the opportunity to capture more P/L because they will be established on 29 December 2000 (Table 17):
rehedging more shares, more frequently (i.e. selling when
they go up and buying when they fall). This will result in a Table 17: Ahold 4% 2005 EUR - ‘Millennium Set Up’
higher ‘gamma trading’ P/L than would have been the case
29 Dec 2000 15 Jan 2001
for a lower gamma name.
Ask Price 117.00 109.00
Returning the risk free rate is better than making no money Parity 106.68 94.38
at all, but again, the arbitrageur is hoping to do rather better.
Stock Price EUR 34.36 EUR 30.40
In fact the convertible hedge fund is hoping for one (or all) of
three things: Delta 74.50% 66.43%
1. To capture more volatility than that 'implied' by the

convertible over a longer time period; and / or
44
Table 18: Portfolio: In assessing the margining requirement the prime broker
will assess three things:
29 Dec 2000 15 Jan 2001
Long EUR 10million EUR 11,700,000 EUR 10,900,000 (i) Counterparty risk
Short 231,295 shares - EUR 7,947,296 - EUR 7,031,368 (ii) Profile of the convertible security
Source: Deutsche Bank Estimates (iii) Effectiveness of the convertible hedge
So now we can look at the overall profitability of the position:

(i) Counterparty Risk
Table 19: Portfolio: P/L
The counterparty will be tiered according to the level of risk
- EUR 800,000 long bonds associated with the fund. The fund's size, strategies
employed, level of diversification and the experience of the
+ EUR 915,928 short stock
+ EUR 115,928 positive P/L managers will all be taken into account.

(ii) Convertible Portfolio
In this example, we have ignored the issue of funding. How an individual position is assessed depends on whether
Adding in the funding reduces the profit somewhat, but this it forms part of a diversified strategy or not. A theoretically
would still have been a very profitable strategy and the sort hedged position in a good quality name will probably be
of position that hedge funds routinely look for. assessed according to simple profile rules such as those in
Table 20.
Funding and Margining
Table 20: Margin Rules for Convertible in Diversified Portfolio
It is important at this stage to look more carefully at how
convertible arbitrage positions are funded, because this is Low Delta Convertible x% of notional (usually below 10%)
very relevant for positions that are going to be run for any High Delta Convertible Cash Premium Only
length of time. Arbitrage positions such as the ones Source: Deutsche Bank Estimates
described above will almost always be held with an

However, a single convertible position, that does not form
investment bank or 'prime broker', who will calculate the
part of a diversified portfolio, will be assessed more
required margin on the position and the funding (interest
harshly:
charge).
Table 21: Margin Rules for Convertible in Isolation
The prime broker should understand the way convertible
hedges work and should require only the correct margin for Low Delta 3 x % notional
the risk of the position being funded. Mid Delta 2 x % notional
High Delta Cash Premium Only

Table 22: Deviating from Theoretical Hedge Requires Higher Margin Charge
Theoretically Hedged Portfolio Non-Theoretically Hedged Portfolio
1. Long EUR 10 million Ahold 4% 2005 EUR 2. Long EUR 10 million Ahold 4% 2005 EUR
Theo Delta = 74.50% Theo Delta = 74.50%
Actual Hedge (74.50%) = 231,295 shares short Actual Hedge (60%) = 186,278 shares short
Naked Stock Position = 0 Naked Stock Position = 45,017 shares (long)
Margin Required if Convertible Price is 117: Margin Required if Convertible Price is 117:
(EUR 10 million x 117%) x 5%1 = EUR 585,000 (EUR 10 million x 117%) x 5% = EUR 585,000
(45,017 shares x EUR 34.36) x 20%2 = EUR 309,357
Margin = EUR 585,000 Margin = EUR 894,357

1 2
5% is an assumption and not necessarily the rate a prime broker would charge. Assumed margin. For a very volatile stock this could be higher.
(iii)Effectiveness of Convertible Hedges Fig 31: Axa #2 3.75% 2017 EUR - Implied Volatility and
Stock Volatility
Using its own database, the prime broker will also calculate
the correct delta for the position and will judge the client's
actual position against this. Any deviation from the theoretical
hedge will generally be treated as a naked equity long /
short. Naked equity positions will usually attract a higher
margin penalty. Consider two different Ahold 4% 2005 EUR
positions in a diversified portfolio of convertibles (Table 22).
3. Capturing Vega Richening
The rule that convertibles (and other options) always trade at

theoretical value at maturity is always worth remembering.
But often a convertible arbitrageur will not have to wait until Source: Deutsche Bank Convertible Research
maturity for the market to re-evaluate an undervalued
convertible. The implied volatility graph (with 260 day richen of their own accord over time (Figure 31).
stock volatility overlaid) for Axa #2 3.75% 2017 EUR shows
clearly that bonds that have been artificially depressed by Figure 31 shows that arbitrageurs should not be too
some event (in this case excess issuance from Axa) will concerned with short-term market sentiment. Providing the
46
potential for arbitrage exists, bonds will richen/cheapen to Table 24: Vega Effect on Ahold 4% 2005 EUR
their correct level over time.
Old New Move Vega New P/L
Ivol Ivol (in IVol points) (per 1%) Price (EUR)
The richening (cheapening) of convertibles with changes in
underlying stock volatility is estimated by Vega. 28.1% 31.1% +3 0.49625 118.49 +149,000
Vega Conversely (in risk management terms) if the whole market

were to cheapen and the bonds were to lose 3 implied
Vega is the sensitivity of the convertible price to changes in
volatility points, the effect would be as in Table 25.
implied volatility. Vega is usually expressed as the change in
convertible price in points for a one percent point increase in
6 Table 25: Vega Effect on Ahold 4% 2005 EUR
implied volatility.
Old New Move Vega New P/L
Ivol Ivol (in IVol points) (per 1%) Price (EUR)
Convertible Price
Vega =
Volatility Stock 28.1% 25.1% -3 0.49625 115.51 -149,000
Vega is used in both P/L projection and in risk
management. Let's return once again to our Ahold example:
Rho
Long EUR 10 million Ahold 4% 2005 EUR We have already dealt with credit protection for convertibles
in Chapter 4. However, we should also cover simple Rho
Table 23: Vega Effect on Ahold 4% 2005 EUR hedging for convertible positions.
IVol% of Vega Convertible
convertible [per 1% move in IVol] Price A convertible's Rho is its sensitivity to movements in
interest rates.
28.1% 0.49625 117.00
Convertible Price
Rho =
Interest Rate
An investor has bought the bond on an implied volatility of
28.1% on the assumption that it will richen by 3 IVol (implied For fixed income investors, Rho will be familiar as Adjusted
volatility) points to 31.1%. Ignoring the stock short, funding Modified Duration. Rho is usually expressed as the points
etc., in this case the investor would make: move in a convertible price for a 1 basis point move in
interest rates (uniform curve shift is usually assumed).
6
Vega can also be applied to interest rate volatility in exactly the same
way. This is useful in two factor convertible models where the volatility of Rho indicates what hedge is required for changes in interest
interest rates is taken into account when constructing the model:
rates. For example, the Rho of Ahold 4% 2005 EUR in all
Convertible Price the above examples is 0.012, which means that if interest
Vega =
Volatility Interest Rate

rates rise by 1 basis point, the bond price should fall by Theta
0.012 points.
All options have 'time value'. The longer the option exists for,
Investors with long convertible positions will hedge their the more it is worth.
interest rate exposure with reference to this Rho number.
For convertible arbitrage funds and market makers, Rho Theta is the rate of change in the convertible price over time.
hedging is usually an approximate affair. Usually a liquid Theta is usually expressed as the change in theoretical
futures contract in the currency of the bond is used as the convertible price over one day.
hedging vehicle. A long position of EUR 10 million nominal
in Ahold 4% 2005 EUR, on the 3 September 2001, would Price
Theta =
require a hedge as shown in Figure 33 below: Time
Fig 32: Rho Declines as Delta Increases - Ahold 4% 2005 Convertible theta is more complex than a simple option
theta because the time decay of the call option can
be offset by the accretion of the bond to its redemption
value. Accretion can mean that deep OID (Original Issue
Discount) structures can have positive theta, as the capital
appreciation of the bond to maturity is greater than the level
of time decay in the call option.
In fact, in September 2001, over 75% of the European

convertible universe had positive theta!
Again, we can see the impact of theta on the Ahold 4%

2005 EUR bond in Tables 26 and 27.
Rho becomes progressively less important as parity and
delta rise (Figure 32).
Fig 33: Rho Hedging of Ahold 4% 2005 EUR
Nominal Bonds Price of Bond Rho of Bond

x x x Conversion Factor of Future
Contract Size Price of CTD Modified Duration
Of Future Of Future Of Future
EUR 10 m 114.125 0.012

x x x 0.939069 = 30 contracts
100,000 101.363 4.28%
48
Table 26: Theta in Ahold 4% 2005 EUR Lattice Based Modelling
Date Theo Price Parity Theta/day Note: The following section draws on theoretical calculus and therefore
maybe beyond the interest of some investors. The basic thread of
11 July 2001 121.75 113.78 0.009504
our analysis can be picked up again by jumping straight to Chapter 7.
For readers who are interested in our modelling discussion, we use
The theta in Table 26 of 0.0095 per day would imply that one the notation set out by John C. Hull in his excellent Futures, Options
year later in time the bond would have 'decayed' to the level and other Derivatives, 4th Edition, 2000.
in Table 27. Table 27 implies that theta is a linear function,
though in fact the rate of theta decay is a complex function The lattice approach allows an investor to build up a grid of
of time to maturity, convertible profile and other factors. possible stock (or other price) possibilities between now and
the maturity of the convertible. This lattice is then used to
Table 27: Theta in Ahold 4% 2005 EUR find the present value of the convertible by calculating its
value at different 'nodes'.
Date (+365 days) New Theoretical Value
11 July 2002 (-0.0095 x 365) = -3.46 A 'one factor' model uses only stock price / parity as a
121.75 - 3.46 = 118.29 variable. To build a lattice of expected parity values we start
Source: Deutsche Bank Estimates with the current share price. We decide in advance how
many branches or 'nodes' the tree will have; the more
The above table will only accurately predict the new nodes, the more accurate will be the model.
theoretical value if no other variables change but time, it is
certainly an important metric, and one that is constantly Between one node and the next, parity can either rise to
assessed by arbitrageurs. P(u), or fall to P(d).
Convertible Modelling The amount the stock is assumed to rise or fall depends on
the length of time to the next node (T- t ) and the volatility of
Most of the analysis above has relied on the calculation of the underlying stock price ( ).
theoretical values, implied volatilities and other metrics such
as delta, gamma, vega, rho and theta. The notation is:
These numbers are all derived from convertible models. We e = base of the natural logarithm (used to continuously
now turn our attention to how these models are constructed. compound interest rates). The continuously
The shortcomings of simple bond plus options models have compounded 5% interest rate approximates to
0.05
been exposed (in Chapter 2), where we saw that they cannot 5.127% (e = 1.05127)
accurately take account of issuer call options and investor = stock price volatility = 25%
puts. In reality, nobody today uses this kind of analysis. T = time at next node
t = current time (so T- t = 1 in our one year, one
step example below)

An upward move (u) is defined as: than falling to P(d) in a positive yield curve environment. The
probability (p) of an upwards move is:
(T-t)
u=e
r(T-t)
e -d
p(u) =
u=e
0.25x1
= 1.2840 u-d
and the probability of a downwards move is:

Similarly, a downwards (d) move is:
p(d) = 1 - p(u)
- (T-t)
d=e
We will assume the 1 year zero coupon risk free rate is 5%.
It is also a condition that: Continuously compounded this approximates to 5.127%.
1 The probability of an upwards move is therefore:

u =
d
1.05127 - 0.7788
p (u) =
1.284 - 0.7788
Therefore, if a convertible's parity is 80 currently and stock
volatility is 25%, the expected upward / downward move
p (u) = 0.53933
over a 1 year period is:
and
P(u) = 1.2840 x 80
p (d) = 0.46067
The downwards move is:
So a one year, one node model looks like Figure 34.
P(d) = 0.7788 x 80
Fig 34: One Node Binomial Tree
Therefore, after one year, parity either rises or falls: Now Year 1
Parity = 102.72
P(u) = 80 x 1.2840 = 102.72 (53.93% probability)
Parity
= 80
P(d) = 80 x 0.7788 = 62.30 Parity = 62.30
(46.07% probability)
The tree must also have a 'drift component' because it is a Source: Deutsche Bank Estimates
central argument of risk neutral valuation that a risk-less
portfolio of long bond / short stock can only return the Summing these values and probabilities
continuously compounded risk-free rate. This means that
there must be a greater probability of parity rising to P(u) (102.72 x 0.53933) + (62.30 x 0.46067) = 84.10
50
r(T-t)
This is exactly what we expect, as 80 e = 84.10 worth at each node to get to the present theoretical value.
For simplicity, we will assume that the bond has no coupons,
Once we have the values for parity and the probabilities, the puts or calls, that the stock pays no dividends and that there
tree can be populated as shown in Figure 35. are no borrow fees (we explain how these are dealt with later).
Once the tree is populated with parity values, we work We start with the convertible value at maturity, which is
backwards from maturity calculating what the convertible is known with certainty; because all derivatives are worth
Fig 35: The Parity 'Binomial Tree'
Now Year 1 Year 2 Year 3 Year 4 Year 5

exactly their intrinsic value at maturity. At maturity, the As the CV is greater than parity, CV is the theoretical value
convertible is worth the greater of: (1) Par + Final Coupon or of the convertible at node 4.
(2) Parity.
Working back through the tree, applying the above
We can illustrate the way the tree works by doing a very methodology, we arrive at a current theoretical value of
crude valuation of XYZ Corp 0% 2006 USD. We will build a 95.478. Our analysis reveals that this very crude five node
very basic tree of just five nodes: analysis actually gives a theoretical value that is remarkably
close to the value calculated using 250 nodes in a full
Node 5 (Year 5) - Maturity binomial tree model.
We know that all parity values below 100 will return 100 as Partial Derivatives From the Lattice
the convertible value (redemption amount). And all parity
levels above 100 will return parity. First and second derivatives (i.e. delta and gamma) can be
obtained directly from empirical observations of the lattice.
Node 4 (Year 4) - One Node Before Maturity We will demonstrate how.
To get the values at Node 4 (one year before maturity), the Starting from the ‘Now’ node and moving to Year 1, delta is
convertible value will the greater of: computed as:
1. Parity; or 111.523 - 87.32

Delta =
102.72 - 62.304
2. 'Continuation Value' (CV)
= 59.9%
The 'continuation value' of the convertible (CV) is calculated
using the next two node values multiplied by their
probability and discounted at the risk free rate. A glance at
the tree itself will make this clear (Table 28).
In order to calculate the theoretical value of the convertible

at the circled node in year 4, we take the maximum of
parity and the CV:
CV:
Discounted Expected Value:
= (102.716 x 0.53933) + (100.00 x 0.46067)

1.05127
= 96.516
52
Table 28: Simple 5 Node Binomial Tree Model of XYZ Corp 0% 2006 USD
Now Year 1 Year 2 Year 3 Year 4 Year 5
279.200
279.199
217.445
217.445
169.350 169.346
169.349 169.346
131.892 131.890
135.064 131.890
102.720 102.718 102.716
111.523 109.957 102.716
80.000 79.998 79.997
95.478 96.374 96.516
62.304 62.303 62.301
87.320 91.199 100.000
48.522 48.521
86.438 95.123
37.789 37.788
90.484 100.000
29.430
95.123
Parity 22.920
Convertible Theoretical Value 100.000

Gamma Absolute callability is approached in the same way. At any
node where the bond is absolutely callable, the value at
Gamma can be calculated using the change of delta each node will be:
between Year 1 and Year 2
The greater of:
135.064 - 96.374
Delta 1 =
131.892 - 79.998 a) Parity
= 74.56% b) The lower of [CV and Call Price (including accrued

interest)]
96.374 - 86.438
Delta 2 =
79.998 - 48.522 This is because if the CV is greater than the call price it will
be in the issuer's interest to call the bonds; and conversion
= 31.57% will be forced if parity is higher than the call price.
Change in Parity: Call Cushions
131.892 - 48.522 Some models employ a 'call cushion', such that the
= 83.37 convertible issuer will not call a convertible unless it
is trading at a comfortable level above the conversion
price (10% is usual). Although there is evidence that
Therefore, as parity moves 83.37 points, delta moves
this is what issuers do in practice, it is difficult to arrive
74.56% - 31.57% = 42.99 points, and so parity gamma (i.e.
at an approximation of what this 'call cushion' should be.
the move in delta for a one point move in parity) is:
The addition of a cushion will lead to a higher, and less
conservative, valuation.
42.99
= 0.516
83.37 Puts
Puts are added into the tree at the relevant node and a put
Calls and Puts is assumed to be exercised if the put value (plus accrued) is
higher than both parity and CV.
Having built the basic tree, we can add boundary conditions
very easily. If we want to assume that the bond is Discount Rate
provisionally callable in Year 4 and 5, with a 140% trigger, we
simply require that any parity value in those years that is The question of which discount rate to use in the tree is still
greater than 140 (the provisional call trigger) will be the the subject of some controversy. 'Blended discount rate'
theoretical value at this node (i.e. for the model to assume models keep track of the bond and equity components of
conversion). the convertible value at each node and discount the bond
54
component at the risk free rate plus credit spread and the Adding Second and Third Factors
parity component at the risk free rate only.
The preceding analysis of a 'one factor' model assumes a
However, other models use the company's risky discount constant interest rate and credit spread. Two and three
rate for the whole convertible value, using the argument that factor models allow investors to model changes in interest
credit risk cannot be eliminated by hedging the bond rates and credit spreads.
with stock. The Black Scholes replicating portfolio of long
bond / short stock would not accurately replicate the Interest Rate as Second Factor
performance of the convertible unless the portfolio
contained the same level of credit risk as the convertible. An interest rate tree is built up in the same way as
the parity tree but with an equal probability of an upwards
In practice, the different approaches do not make much or downwards move. The up and down moves must
difference except when valuing Original Issue Discount conform to the present term structure of interest rates. The
(OID) structures (see below). relationship between an 'up' and a 'down' move is given by:
2 (T-t)
Coupons Ru = Rde
Coupons are added to the appropriate nodes in the tree (i.e. At each node the model generates a quadrinomial tree with
7
at nodes where the convertible is assumed to be held - not four branches SuRu, SuRd, SdRu and SdRd.
converted - over a coupon pay date). The coupon is
discounted back from the node at which it was added using Because of the generally low volatility of interest rates, this
the appropriate discount rate. Care needs to be taken factor is often omitted from convertible models.
where interest is lost on conversion to ensure that the
correct decision is assumed. Credit Spread as a Second / Third Factor
Of much more interest to investors, especially in higher risk

Dividends
bonds, is the correlation between stock price and credit
Stock dividends can easily be accommodated by growing spread.
parity at a rate equal to the risk-free rate minus the
continuous dividend yield. If the dividends themselves are Any convertible investor understands this relationship
known, they are present valued and subtracted from parity instinctively. But from a technical perspective it is fairly easy
before it is grown at the risk-free rate through the tree. In to specify a condition that, below a certain share price, as
practice, trying to use actual dividends complicates the tree the stock drops lower the credit spread widens. This is far
considerably, although it can be (and often is) done. more useful as a second factor, particularly outside the
investment grade universe.
7
Modelling interest rate volatility is a complex area. Interested readers are
referred to Hull, Chapter 16.

7. What Drives Convertible Valuations Today?
The global convertible markets have travelled a huge Arbitrage Funds Now Determine Valuations
distance since the mid 1990s. Until fairly recently, analysis
of convertibles was still dominated by the concept of Our first conclusion is that for the larger and more liquid
'theoretical value'. Theoretical value was an easy concept issues, it is convertible arbitrage funds that now determine
to understand. It posited that a convertible was 'worth' the pricing of convertibles.
whatever came out of a binomial tree model when the
stock's historical volatility (and certain other assumptions) This is not to say that outright accounts have no influence,
were keyed in. and less liquid, or more ‘out of the money’, bonds are still
primarily driven either by outright convertible investors or by
To calculate ‘theoretical value’, two assumptions were buyers from other asset classes (such as high yield).
generally made:
But for the balanced and liquid universe, wherever we look
1. The underlying stock's 260 day historical volatility was in the world, it is arbitrage funds that are able to mobilise
used; and the greatest liquidity to buy undervalued issues or to sell
expensive ones. In the previous two chapters we looked at
2. Volatility was capped at 40%. how credit protection could be used to separate the equity
derivative component from the bond. We also looked at how
The problem with these assumptions was that they were convertible arbitrage works.
completely arbitrary:
But what really makes an arbitrage fund think that a
1. Why use 260 day stock volatility? Why not a shorter time convertible is rich or cheap?
period (100 day), or even a longer time period. The
majority of convertibles are at least three year options. We believe a number of observations can be made about
So why not use 780 day volatility? how valuations are driven in modern markets:
2. Why cap the volatility at 40%? The traditional answer to 1. Bonds do not Stay Mispriced Forever
this question was that convertibles 'would not trade richer
Our analysis suggests that convertibles do not remain
than this'. But this is completely disproved now.
undervalued for long periods if the potential for arbitrage
Siemens / Infineon 1% 2005 EUR, amongst other bonds,
exists. Even if bonds are 'tainted' following a weak debut or
has traded with an implied volatility well above 50% for
some disadvantageous corporate activity, they will recover
long periods (quite justifiably).
over time. Axa #2 3.75% 2017 EUR is a good example of a
bond which suffered a difficult birth and a certain amount of
So it is clear that we need to look well beyond simplistic
stigma in its early stages. But Figure 36 demonstrates
conceptions of theoretical value if we are looking for what
that an implied volatility far below the historical volatility
drives valuations in today's convertibles markets.
eventually leads to the bond richening to a more realistic
valuation.
56
Fig 36: Axa #2 3.75% 2017 EUR 2. Historic Volatility of a Stock is Not Always a Good
Guide to Valuations
Having emphasised the importance of historical volatility in

convertible analysis, it is necessary to temper this by
highlighting the drawbacks of historic volatility analysis in
isolation. The major drawback, of course, is that historic
volatility is a backwards looking indicator.
Fig 38: Munich Re / Allianz 1% 2005 EUR
Similarly, after a period of dislocation, Allianz / Siemens 2%

2005 EUR has once again become more closely correlated
with changes in the historical volatility of Siemens.
Fig 37: Allianz / Siemens 2% 2005 EUR
Investors trading the Munich Re / Allianz 1% 2005

EUR bond were not fooled by the artificially high historic
volatility of Allianz in late 2000, caused by speculation about
potential corporate activity in the German financial sector.
And sure enough, it was the historic volatility that came

crashing down to vindicate the more cautious stance of the
convertible market.
The fact that historical volatility can often be a poor guide

Source: Deutsche Bank Convertible Research to future volatility is one of the reasons why investors are
increasingly turning to the (forward looking) OTC and listed

volatility markets for clues on where convertible valuations exchangeable, but it is clearly not as close or as useful to a
are going. trader as what is going on in the OTC / listed markets.
3. OTC / Listed Volatility is the Dominant Factor in Fig 40: Allianz / Siemens 2% 2005 EUR
Efficient Markets
Where continuous and stable credit protection exists,

convertible valuations are driven to a greater and greater
extent by pricing in OTC and listed volatility markets.
In Germany, an exchangeable such as Allianz / Siemens 2%

2005 EUR illustrates the point very well.
Fig 39: Allianz / Siemens 2% 2005 EUR
When using the OTC and listed volatility markets, it is

important to isolate the correct volatility contract. We have
used simple ‘at the money’ contracts (ATM) in our
examples above, but it is very important not to apply an ATM
volatility valuation to a deeply ‘out of the money’ or ‘in the
money’ convertible.
Loews / Diamond Offshore 3.125% 2007 USD illustrates

the point (Figure 41). The bond looks very cheap against
ATM volatility at both 6M and 18M maturities. However,
Figure 39 shows a close relationship between the short the problem is that the premium of 103.84% means a
term ups and downs in the OTC volatility markets and the conversion price of USD 65.04, which is a long way above
IVol of the exchangeable. It seems very clear that there is the stock price of USD 27.0.
arbitrage activity between the two markets.
A look at the volatility surface for Diamond Offshore for the
Figure 40 also demonstrates that there is a correlation 18M contract gives a clue as to why the exchangeable is
between the historic volatility of Siemens and the IVol of the trading where it is.
58
Fig 41: Loews / Diamond Offshore 3.125% 2007 USD Options a long way out of the money, particularly to the
upside, often trade much more cheaply than ATM options.
Figure 42 shows that at 150% of the current stock

price, Diamond Offshore implied volatility in the options
market falls to 38.5%. Interpolating this out to the almost
250% ‘out of the money’ exchangeable makes an implied
volatility of 30% look reasonable.
The OTC / listed volatility markets are not just of use to the
convertible analyst. The increasing availability of credit
protection means that it is more and more common for
hedge funds to sell OTC volatility against convertibles where
significant disparities exist. While this can be a profitable
strategy, it is fraught with difficulties due to:
1. Difficulty of matching strike and maturity (particularly the

Figure 42 introduces the concept of volatility ‘skew’. Skew latter given the lack of liquidity in longer dated volatility
reflects the fact that in the options markets, options with markets); and
different strikes trade at different implied volatilities.
2. Documentation mis-match. The main problem here is
Fig 42: Diamond Offshore - Option Implied Volatility Skew a lack of standardization in what will happen to
convertibles in corporate action scenarios versus what
will happen to OTC / listed contracts (see Chapter 8:
Documentation).
4. Hedge Funds and 'Busted' Convertibles
In our view, it is a misconception that arbitrage funds shun

convertibles on which there is not credit protection available.
But what they do will certainly have a bearing on how the
convertible trades.
Consider a UK / US issue such as NTL 6.75% 2008 USD at

issue. The convertible is rated B-/B3, and there is neither an
asset swap nor a CDS offer available. A hedge fund has two
Source: Deutsche Bank Equity Derivatives Research choices: a) ignore the issue completely, or b) employ an
'overhedging' strategy.

Using comparisons with similar companies, we come up
with a 'guesstimate' credit spread of about US LIBOR +
1,200 bps. Putting this spread into the model gives a
theoretical delta of 68.74%. One strategy for the hedge
fund might be to overhedge the position to protect against
credit spread expansion. In the above scenario, if the stock
were to fall by 50% the credit spread is likely to increase
significantly. In this case, the convertible's realised delta
will be much greater than that predicted by a one factor
model.
So a hedge fund might overhedge by perhaps ten delta

percentage points, perhaps 80% in the above NTL example,
in order to allow for some credit spread expansion should
the stock fall significantly.
60
8. Documentation
A great deal of progress has been made in convertible Fig 43: KDIC / KEPCO 2.25% 2005 USD - Prospectus
documentation in recent years (particularly in Europe). But it
is still important for investors to understand the legal
framework in which they are investing.
Contract Based Securities
Unlike most OTC derivatives, convertibles are not covered

by generic documentation. Investors buy convertibles on
the basis of a trust deed / indenture or other documentation
which lays out the contract in detail. Before the trust deed
is produced investors may have to work with a preliminary
prospectus. But where a trust deed is produced it will
always be the legally binding document, and differences
between it and a preliminary prospectus will always be
settled by reference to the trust deed (and we have seen
many instances of this occurring).
An example of a prospectus (for KDIC / KEPCO 2.25% 2005

USD in Korea) is shown in Figure 43 opposite.
Trustees
When a trust deed has been issued, a trustee will normally

act (in theory at least) in the convertible holders’ interests in
any dispute involving the issuer of the securities. While
the trustee can make tiny adjustments to a trust deed,
a full convertible holders' meeting is required to make Source: Korea Deposit Insurance Corporation, Korea Electric Power Corporation
any substantive change. The trust deed will set out the
mechanics of how meetings are called and conducted. Terms and Conditions
Governing Law The 'Ts & Cs' are the real bones of the documentation, and
are where investors need to be especially focused. The
It is very important to be aware of which governing law has basic information is usually straightforward, but problems
been specified in the prospectus. If a dispute goes to can arise in the following areas:
litigation, this will govern where the case will be heard and
the legal framework that will be applied.

1. ‘Capital’ Distributions to Shareholders (including special capital distributions are made to shareholders, other than in
dividends); the form of 'normal' dividends. This compensation almost
2. Issuer Calls ('Screw' Clauses); always comes via an adjustment to the conversion ratio, so
3. Merger / Acquisition Activity; that the bond will be convertible into more shares to leave
4. Look Backs; and the ex-date parity value unchanged.
5. Status of Shares Received
Unfortunately, the definition of 'normal' dividends will often
We will examine each of the potentially problematic areas allow a company to distribute some amount of cash or
in turn: securities, above and beyond its normal annual dividend,
before convertible holders are compensated. In Europe and
1. ‘Capital’ Distributions Asia, this amount is becoming standardised at between
3% and 5% (depending on the nature of the company and
We believe the most hazardous situation for convertible its dividend policy). In the US, it is more of a 'free-for-all'.
holders is often when a company distributes ‘capital’ to The American Tower 5% 2010 USD, for example, rather
shareholders. If they are not compensated, convertible generously (for the shareholders!) allows the issuer to pay
holders are disadvantaged in that any cash distribution a special dividend of up to 15% without compensating
will cause the share price to fall on an ex-date by the amount bondholders at all!
of that distribution. Hedged investors will also find
themselves short the distribution without any corresponding Non Cash Distributions
compensation.
Where an issuer makes a distribution of non-cash securities
Normal dividends are the most common form of distribution (bonus shares, bonus warrants, debt securities, etc) the
(for which bond holders are not compensated), but for holder will be at the mercy of the indenture. There will
the most part these can be predicted and factored into usually be some compensation, but the way it will be
theoretical models. The increasing trend towards share calculated can vary considerably:
buy-backs (in the US at least), as opposed to special
dividend payments, as a means of returning excess capital (i) Actual Ex-Rights Price - (i.e. compensation is based on
to shareholders, is therefore a very welcome development where the stock trades when it goes ex-whatever the
for the convertible markets. Buy backs conducted ‘at distribution is).
market’ leave the value of the shares unchanged (in theory)
and do not disadvantage convertible bondholders. (ii) Theoretical Ex-Rights Price - (i.e. where the stock ought
to trade based on the level of dilution and the 'price' at
However, other distributions such a Special Dividends, which the distribution is made).
Bonus Share Issues, Bonus Warrants etc. will disadvantage
convertible bondholders if there is no compensation. The conversion ratio will always be adjusted to account for
stock splits (i.e. if the stock is split in half, the conversion
To protect investors, the majority of indentures contain ratio is doubled etc).
provisions stating that compensation must be offered if
62
2. Issuer Calls ('Screw Clauses') It is much less clear what happens to conversion rights
outside the (rare) full legal merger.
'Screw' clauses state that convertible holders converting
into shares will not be paid the accrued interest on their In some countries, notably Germany, it is not legally
bonds by the issuer. This means that if a bond is callable an possible to bind an acquirer to honour conversion rights into
issuer can wait until just before a coupon is payable before an outstanding bond.
calling the bond, forcing conversion, and 'screwing'
investors out of their accrued interest. Unlike unforeseen So in Europe a series of conventions have grown up to
and uncompensated distributions to shareholders, however, protect investors under change of control scenarios. In
the market will warn investors of a likely 'screw' because France and the Netherlands it is usual for an acquirer to
callable 'in the money' convertibles subject to a 'screw' allow continued conversion into the acquired shares. In the
clause will trade at 'net parity'. Net parity means parity UK, however, it is more usual for investors to rely on
minus accrued interest (i.e. the price reflects the probable remedies which offer an enhanced conversion window.
loss of accrued interest in the event of call and early
conversion). There are two main conventions for enhanced conversion
windows:
3. Merger / Acquisition Activity
(i) Step Conversion Clauses - They stipulate that, in a
In our view, probably the most treacherous area for convertible change of control, conversion is allowable at an
investors is M & A activity. When the conversion / exchange enhanced ratio. This enhanced ratio falls as time goes by
property (i.e. underlying equity) is acquired or merged with according to a schedule that is laid out at issue in the
another entity, the overriding concern for convertible holders trust deed. On the announcement of a change of
will be what the new conversion rights change to. control, convertible holders will have a window of
perhaps 60 days to convert at the prevailing enhanced
In virtually all jurisdictions, if a 'full legal merger' of two or ratio.
more joint stock companies occurs, conversion rights will be
protected and will be on-going into the new legal entity. But, (ii) Average Premium Enhancement - is also used in the UK.
in fact, full legal mergers in the strict sense are very rare, This clause states that under a change of control the
and the continuing legal existence of companies that have conversion ratio is enhanced according to the average
been 'merged' or 'acquired' can last indefinitely. premium of the bond over a given time period (often
one year). Again the enhanced conversion window only
Under this latter scenario, there will be no automatic right of applies for a specified time period (60 days is the norm).
on-going conversion into the dominant company (in many
countries), and investors will need to rely on remedies in the Change of Control Puts
indenture and / or the goodwill of the acquiring / dominant
company in order to determine their future rights. These state that the convertible will be puttable (usually at
par / accreted value or just above) in the event of a change

of control. These provisions are very common everywhere, comprise cash, debt securities or other non-equity
including the US. They obviously act as a poison put if parity securities. This is fairly disastrous for the holder in that
is very low and the bonds are trading well below par. non-equity securities will usually have little or no (in the
case of cash) volatility. Optionality is therefore worthless,
What to watch out for! and the exchangeable will collapse to intrinsic value.
(i) When is an Acquisition Not an Acquisition? The Also note that the issuer of an exchangeable does not
HypoVereins Bank / Bank Austria Precedent have to take account of the interests of exchangeable
holders when deciding whether to accept (or what to
We find it regrettable that some companies occasionally
accept) when an offer is made for the exchange property.
take steps that appear to avoid what was intended to
protect investors who have bought convertibles in good
In order to address these concerns, indenture language
faith.
has evolved (in Europe first of all) to require exchangeable
issuers to use any cash received in an acquisition to
In 2000, when the German HypoVereins Bank "acquired"
purchase shares (or even index futures) to top-up the
Bank Austria, (the word "acquired" appeared in a
exchange property, thus preserving optionality and
Tombstone in the Financial Times, 16 November 2000),
protecting the theoretical value of the exchangeable.
the deal was structured in such a way that the company
was not obligated to honour a clause in the Bank
4. Look Backs
Austria 0% 2004 EUR trust deed, which provided for
compensation in the event that Bank Austria was 'Look back' clauses are often overlooked when convertibles
acquired. are issued, but they can be nasty and virtually unhedgeable
short options when a holder eventually comes to convert.
It is clear from this episode that the structure of any
M & A activity is crucially important in determining what A 'look back' clause allows the issuer to decide whether to
compensation will accrue to convertible investors. give a converting bond holder shares or the cash value of
the shares. A problem arises where an issuer can choose to
(ii) Exchangeables deliver cash or shares after the calculation period.
The treatment of exchangeables can differ from standard

There is increasing evidence, certainly in the US, that issuers
convertibles in M & A situations.
are opportunistic in exploiting these clauses, choosing
between shares or cash on a 'cheapest to deliver' basis.
A particular problem can arise when non-equity
securities are included as part of an acquisition of the
If the calculation period is very short (one day is not
company that forms the exchange property.
uncommon in the US) and the delivery / notice delay is a
number of days later, investors can be badly disadvantaged
In this scenario, without express language to the
if the share price is volatile.
contrary in the indenture, the conversion property can
64
XYZ Corp 5% 2006 USD has a look back option with a one 5. Status of Shares Received
day averaging period. Converting holders are told what they
receive (cash or shares) two days after the conversion / Investors need to be aware of restrictions on stock delivered
reference date. XYZ Corp is a highly volatile stock, and on conversion. Any restrictions should be highlighted in the
as Table 29 shows, this disadvantages bond holders indenture, but if in doubt, it may be necessary to check with
significantly if the company acts opportunistically. the conversion agent (who is appointed by the issuer to
handle the issuance of shares to converting bondholders).
For a hedged investor, a negative gamma position is created:
The most common area of complication is dividend
Table 29: The ‘Look Back’ in action entitlement. In some countries in Europe, the shares
delivered on conversion may not rank for the next dividend.
Scenario Day 1 Day 2 Day 3 In the Netherlands, part of the last coupon paid may even be
(reference date) (notice date) repayable if the coupon accrual period overlaps a dividend
Stock Price Stock Price Stock Price
entitlement period. In France, it is also common for shares
Company delivered on conversion not to rank for dividends paid in the
A) 50.00 47.00 45.00
delivers stock same financial year.
Company
B) 50.00 52.00 55.00 delivers USD
50.00 per share
Scenario A: Hedge Bought Back
If the hedge (assumed 100%) is bought back at the time of

conversion, this creates: 1) a long position in the stock only
if the stock falls (i.e. stock will be delivered on Day 3 if it has
fallen) or 2) a flat position if the stock goes up (USD 50.00
per share will be delivered).
Scenario B: Hedge Not Bought Back
However, if the hedge is retained until Day 3, cash will be

delivered if the stock has risen to USD 55, and the investor
will lose USD 5 per share buying back the hedge. Whereas
if the stock falls, the position will only be flat as the delivered
stock is used to close the short.

9. Convertibles for Issuers
Many different explanations have been put forward for why Of course, if a company knows that its share price is
companies issue convertible securities. But what seems going up then it should not issue any kind of equity, and
clear is that there is no all-embracing reason. Different should raise debt instead. But very few companies can
structures can be used by different companies at different realistically be in this position, and for most issuers
stages in their development. convertibles offer a good way of selling equity forward at
a more attractive price whilst raising cash immediately.
As we demonstrate in Chapter 10, there are many different
potential convertible structures, but they almost all have one There is also evidence that, as with other equity products,
thing in common: the issuer sells equity at a premium to its companies will act opportunistically and issue equity
current price in return for giving an investor some kind of linked securities if they perceive their share price to be
downside protection and / or extra yield. too high.
The following corporate characteristics tend to tempt 2. Lack of adequate credit rating - For many small or fast
issuers into the convertibles market: growing companies a straight debt issue is impossible
because the lack of an investment grade rating leads to
1. Share price perceived to be too low for equity a prohibitive interest charge. By attaching an equity
issuance; option, via a convertible, the company can reduce its
interest expense dramatically.
2. Lack of good (any) credit rating makes straight debt
expensive; 3. Non-strategic equity stake held in another company -
Exchangeable bonds have been extremely useful in
3. Non-strategic equity stake held in another company; Europe in helping to unwind corporate cross holdings.
and / or
Issuing an exchangeable bond allows a stake-holding
4. Specific tax considerations. company to:
We will examine each in turn: 1) Dispose of the stake at a premium;
1. Share price perceived to be too low - A company that 2) Retain economic control over the stake (retain voting
perceives its share price to be undervalued will not be rights etc.);
keen to sell equity at prevailing levels. A convertible
issue is one way of raising cash today whilst (potentially) 3) Delay crystallisation of any capital gains taxes; and
selling equity at a price closer to the value ascribed to it
by the company in the future. If the shares rise as the 4) Finance at rates well below their normal funding cost.
company expects, convertible holders will convert at
maturity or before, effectively buying shares at the Particularly for reason three, exchangeables proved
higher price that was originally judged appropriate by the especially popular in Germany in the late 1990s, ahead of a
issuer. major reduction in disposal taxes.
66
4. Specific tax considerations - Tax considerations are growth rate of the company, and that the shares are
always important in any capital raising exercise. It is undervalued by around 25%.
probably not an exaggeration to say that convertible
structures have been the subject of more ingenious tax If the company were to issue straight debt in US Dollars, it
engineering than almost any other form of financial is advised that a YTM/P of USD LIBOR + 500 bps would be
security. Among the most important innovations in tax necessary. The company does not have enough free cash to
engineering are: meet such high interest payments.
a) Capital Gains Tax Deferral - Exchangeable bonds allow A convertible bond issue may be a good solution to XYZ
the (potential) forward sale of a stake without crystallising Corp's financing needs. The company could raise USD 100
the tax liability until the bonds are actually converted. In to USD 300m for five years at a YTM of perhaps USD LIBOR
fact, exchangeable bonds usually include the option for flat if the conversion premium was set at around 25%.
the company to deliver cash instead of shares, thus
giving the issuer even more flexibility to manage disposal If the shares rally as the company expects, it will have sold
tax liabilities. its equity forward for the price it originally wanted, whilst
delaying dilution for up to 5 years. If the shares do not
b) MIPS Structures - Although this structure has now been perform, the company will at least have saved 500 bps a
closed off, MIPS provided a very important funding year in yield versus a straight bond issue.
vehicle in the mid 1990s in the US. A special purpose
vehicle was set up between the issuer and the investor, Exercising Issuer Calls
allowing a company to issue mandatory securities
(which ranked as equity on the balance sheet), but then The most important decision a company will have to make
to claim tax deductibility on the interest payments. after it has issued a convertible is when, or whether, to
exercise its call option (if it has one). We discussed
c) Contingent Payment and Conversion - The innovation 'provisional' and 'hard' call protection in Chapter 2, and
for the early 2000s (in the US at least) is certainly we now revisit these call features from the issuer's point
contingent conversion and payment features being of view.
added to long dated bonds issued by investment grade
companies. These issues give the issuer significantly Provisionally Callable Convertibles - We assume that the
improved dilution / tax treatment. 'CoCo' and 'CoPay' are objective of all convertible issuers is to get their securities
explained fully in Chapter 10 'Convertible Structures'. converted into shares either at maturity or before. When a
bond is provisionally callable with a trigger at (say) 140% of
An Issuance Scenario conversion price, it is a fairly straightforward decision for the
issuer whether or not to call if the shares trade above this
XYZ Corp 5% 2006 USD is a fast growing technology level. If the dividend yield on the shares is less than the
company. The company needs to increase working capital in coupon on the convertible (after adjusting for tax), it will
order to increase productive capacity. Management usually be in the company's interests to call and force
believes that the market is not fully valuing the potential conversion.

The company will have to give notice of its intention to Launch Date: - 1 month (or more)
redeem the bond (normally at least 30 days), giving
investors the opportunity to convert. If parity > 140 there is Officials from the potential issuer listen to 'pitches'
little danger of parity slipping below 100 and the company from investment banks regarding the many different
having to find cash to pay bondholders the redemption structures / pricing that could meet their financing / disposal
amount. It will almost always be advantageous for a requirements. The basic size and structure of the issue
company to call a convertible just before a coupon payment takes shape. The company reaches tentative conclusions on
date if the bond has a 'screw' clause. the structure and target pricing for the premium and YTM/P.
Hard Calls - When a bond is absolutely callable the same Launch Date: - 2 weeks
considerations as above apply, but the difficulty occurs if
A leading convertibles investment bank is mandated to
parity is above 100 (redemption value), but not by a sufficient
manage the sale. The company decides the minimum it
margin to preclude the possibility that it might fall below 100
would like to raise is EUR 1 billion, but grants the lead
during the conversion period. In this situation the company
manager an option to sell another EUR 200 million of
has a problem. If parity is not sufficiently above redemption
the bonds if demand is strong (the 'greenshoe' option). A
value, parity might drop below redemption value, forcing the
target date is set to launch the transaction, but market
company to redeem the securities for cash instead. It is
conditions will play the most important role, and the issuer's
also in the investors' interest to wait until the end of the
stock price will be crucially important in determining the
conversion period before deciding what to do, thus
launch date.
maximising the 'put' value of the convertible.
Launch Date: - 1 week

In practice, issuers often build a 'call cushion' into their
call assessment. This is usually around 10% above the Legal work on documentation, drafting the preliminary
redemption amount, depending on the volatility of the stock. prospectus and discussions with relevant regulators or
exchanges is well advanced. Officials from the company and
Buying Convertibles Back the bank are now watching the company’s stock price, the
bond market and the level of valuations in the convertible
If a company judges its convertibles to be undervalued in
market very closely, searching for exactly the right moment
the market, it may choose to buy them back. But this
to launch the transaction.
scenario is dependent on the existing convertibles not being
fully valued by convertible investors.
Launch Date: - 1 day
The New Issue Process: A Typical Convertible Deal The issuer and bank meet after the close and agree to
launch the transaction at 7:00 am the next day. The exact
Opposite we sketch a slightly stylised, but broadly realistic,
pricing range is agreed, consisting of 50 bps spread on the
new issuance scenario for a book built transaction.
YTM and 5 percentage points of premium range. Relevant
regulators and / or exchange officials are notified, and senior
68
salespeople from the investment bank are briefed on the Importantly, the bank may elect to exercise the 'greenshoe'
forthcoming sale. option to create extra securities to meet excessive demand.
Launch Date: 7:00am
Deal launched!! The lead manager plus other members of

syndicate begin calling convertible investors and the book
building process starts. A three day book building process
is envisioned, with the company’s management touring
investor centres on a 'roadshow' of presentations.
Launch Date: + 2 days (pm)
The 'book' of demand for the bond now exceeds EUR 6

billion, consisting of mainly non-limited orders. The bonds
are quoted 101.00 - 101.50 in the 'grey' (pre-issue) market.
The stock came under a small amount of pressure on
Day 1 of the marketing period, but has since recovered.
Management and bank officials, who have been in constant
contact, meet and decide to close the book early and to
increase the issue size to EUR 1.25 billion. A 'pricing call' is
held and the final terms of the convertible are agreed
between the investment bank and the company. Pricing is
towards the low end of the yield range and the high end of
the premium range, to reflect the level of demand. The two
sides go through the 'book' and decide how to 'allocate' the
securities.
Launch Date: + 3 days (am)
Subscribers are informed of their allocations and, only once

this process is completed, bonds begin formal trading.
After the Issue
Many things happen after the first day of trading. Bonds

become listed on exchanges, 'season' to allow initially
exempt investors to purchase them and are included in the
market making universe of the lending convertible brokers.

10. Convertible Structures
Unfortunately for the newcomer, convertible structures the disadvantage (from an investor's point of view) inherent
continue to become more complex and innovative. In this in zeros.
section we discuss the main types of convertible structure
and consider the valuation issues associated with them. Fig 44: Zero Coupon Convertible - Defensive Security
Zero Coupon Convertibles
Zero Coupon convertibles pay no coupon and have a

current yield of 0%. Yield to maturity or put (YTM/P) is
achieved by means of an issue price below final redemption
value and / or put values (i.e. investors pay 60 today for a
bond maturing in five years at 100 - giving a YTM of around
10.5%). The bonds are sometimes called Original Issue
Discount (OID) bonds.
Roche 0% 2010 USD is a good example of this structure.

The bond was issued at 35.628 on 20 April 1995 and
redeems at par (100) in 2010, giving a YTM of 7% at issue.
The bond also has puts every five years at accreted value. Source: Deutsche Bank Estimates
Calls and puts on 'zeros' are calculated according to Zero coupon bonds received a huge new lease of life in the
'accreted value'. Accreted value is the price that keeps US in 2000 as a result of the ‘CoCo / ‘CoPay’ revolution,
YTM constant. Roche 0% 2010 USD is callable on 20 April which we discuss below.
2003 at 61.778 (accreted value). This is the value the bond
will have 'accreted' to in 2003 in order to keep YTM constant
Knock-Out Puts
at 7%.
Some issues employ knock-out puts, where the put lapses
Zero coupon bonds have a much higher effective conversion if the share price trades above a pre-determined level (the
price at maturity than couponed bonds, and as such have 'knock-out level') at any time before the put. The knock-out
lower deltas and more defensive characteristics. They tend level is usually set at 130% to 140% of the conversion price.
to be issued by higher credit quality companies.
Bond with Warrants
The 'payoff' diagram (Figure 44) shows that the return on
the bond remains constant unless the share price rises a Bonds are sometimes issued with attached warrants.
long way during the life of the bond. The difference Where the warrants are detachable from the bond and
between the conversion price based on the issue price and exercisable for cash, they are stripped in the market and
the conversion price based on the final redemption price is trade separately.
70
Bonds can also be issued with 'non-detachable warrants', 2) DECSTM
where the warrants can only be exercised by tendering the
whole bond. The 'bond with warrant' structure can be useful Note that almost all mandatory structures are European
in specific circumstances. It can help with pre-emption style options.
rights problems and can limit withholding tax liabilities
for investors. An example is the Daimler Benz 4.125% 1) PERCSTM are very simple securities. They are usually
2003 DEM N.E.W.S. ("Notes with Equity Warrant issued at the prevailing share price, convertible into one
Securities"). Such structures behave as if they were plain ordinary share, with an enhanced dividend yield. Above
vanilla convertibles. a certain stock price, however, the conversion ratio will
fall by the amount the stock rises, capping the upside at
Mandatory Structures that level. PERCSTM are convertible 1:1 below this cap
level, which means that they participate 100% on the
In the US in the 1990s, the falling dividend yield on ordinary downside. PERCSTM have a payoff structure as shown
shares led to the creation of a whole new class of in Figure 45.
mandatorily convertible instruments. The rationale behind
these structures was to give equity income investors Fig 45: PERCSTM are Fairly Defensive Instruments
greater yield in exchange for slightly muted upside
participation. Because these preference shares mandatorily
convert into ordinary shares at maturity, there is no danger
of the issuer having to find cash to redeem them. From this
simple idea has grown a bewildering array of mandatory
securities (with ever sillier associated acronyms). But all
share the same basic idea of the investor receiving
enhanced dividend income in return for slightly reduced
equity upside participation.
The key to understanding mandatory structures is that the

conversion ratio will change depending on the price of
the shares at maturity. This means that if the share price
rises, within certain bands, the conversion ratio will fall. This
reduces an investor's upside participation. Source: Deutsche Bank Estimates
From an issuer's point of view, the structure will be In option terms, a PERCSTM investor is:
attractive if the shares rise, because investors will be buying
equity at a premium. There are two main 'families' of 1. Long stock;
mandatory security:
2. Short 1 European call struck at 'cap level'; and
1) PERCSTM ; and

3. Long income swap of PERCSTM dividends for ordinary constant between these stock prices. Above Trigger 2, the
dividends conversion ratio is again constant, but at a lower level than
Trigger 1, so (unlike the PERCSTM), the DECSTM participates
Investors modelling PERCSTM need only calculate the price again in further equity upside. Because of their greater
of the call they are short (struck at 'cap' level) and compare upside participation, DECSTM will typically have a yield
that with the present value of the PERCSTM income advantage below that of a similar PERCSTM.
advantage. PERCSTM are generally issued by companies that
are bullish on their own share price. Valuing a DECSTM
2) DECSTM are a little more complicated than PERCSTM, and Holding a DECSTM, an investor is:
are more equity-like. Instead of upside participation
ending completely above the cap level, DECSTM employ a 1. Long 0.xx European calls at upper trigger level
'dead zone' of no equity participation, but offer holders (Number of calls = Initial Stock Price / Trigger 2)
renewed upside above a higher trigger. DECSTM are also
usually issued at the prevailing share price with an initial 2. Short 1 European put at Trigger 1 level
conversion ratio of 1:1. The total return is shown in
3. Long risk-free 0% bond paying initial stock price at
Figure 46.
maturity
Fig 46: DECSTM Give Upside and Downside Participation
4. Long PV of risky dividend payments
- With More Yield
The key to valuing a DECSTM is to get the skew right

between the valuation of the call and the put.
Both options have the same maturity and European style,

but the different strike prices require the prevailing term
structure of OTC volatility to be applied.
Convertible Preferreds
In the US market, a large number of convertibles are

structured as convertible preference shares. These 'prefs'
often have a nominal value of USD 50.
Source: Deutsche Bank Estimates There are a number of differences between convertible
preference shares and standard convertibles bonds:
The DECSTM investor gives up upside in return for enhanced
yield. Between Trigger 1 and Trigger 2, the conversion 1. Redemption: 'Prefs' can be redeemable or irredeemable.
ratio falls as the share price rises to keep the total return If irredeemable, the market convention is to treat them
72
as 50 year redeemable structures. Fig 47: 'Parity Refix' Convertible Payoff
2. Denomination: Usually USD 50 par amount.
3. Listing: Often listed on an equity exchange (NYSE).
4. Dividends: Usually paid quarterly.
5. Ranking: Usually subordinate to all debt obligations.
6. Trading: Trade 'dirty' (inclusive of accrued interest).
Other than the above, there is no reason why convertible

preferreds will not trade as normal convertible instruments.
Parity Resets

In our view, one of the most interesting developments in
the convertible market in the mid 1990s was the introduction
They are then worth the sum of:
of 'parity resets' clauses. Resets mean that on certain
dates the conversion price is reset to a level at (or near) the
1. The present value of the reset amount (usually
prevailing share price (i.e. - if the shares fall you get more of
Nominal Value)
them per bond to make up for the fact that they are worth less).
Minus (-)
We believe parity resets are very attractive for investors.
They are effectively puts payable in stock at par. However, 2. Puts struck at the Lowest Conversion Price =
parity refixes always have a maximum limit on the amount
of the conversion ratio adjustment. Usually, this means that Nominal Value
the conversion price can be adjusted down to a certain Lowest Convertible Price
minimum, such as 60% of the initial conversion price at
issue. Plus (+)
The payoff for a parity reset convertible bond is shown in 3. Calls struck at the Initial Conversion Price =
Figure 47.
Nominal Value
Many of the Japanese resettable securities issued in the Initial Convertible Price
1990s were mandatory preferreds rather than bonds.
Investors can get a reasonable approximation of the value of
mandatory preferreds with resets by valuing them as
European options to the next reset date only.

Modelling a Parity Reset Bond is More Difficult Cross Currency Convertibles
Modelling a parity reset bond, as opposed to mandatory Cross currency convertibles are those in which the bond
preferred, is more difficult, because although the investor is is denominated in a different currency from that of the
still short the puts, downside is limited by the fixed income underlying stock. This means that investors have currency
value of the convertible. In effect, the investor is only short exposure because parity on a cross currency convertible is:
a put between the reset amount (par) and the bond floor.
(stock price x conversion ratio) / FX in local currency
Monte Carlo Simulation
This formula means that if the currency of the underlying
The most accurate method for valuing path dependent stock weakens, parity will fall even if the stock price remains
derivatives such as parity reset convertibles is Monte Carlo unchanged. When the underlying stock is an exporter this
simulation. This can be done in a spreadsheet using the effect may be mitigated somewhat by an appreciation in the
following methodology: share price accompanying a decline in the currency. Parity
volatility (as opposed to stock volatility) will be lower and the
1. Set up a spreadsheet using Excel and take a univariate theoretical value of the convertible will be depressed
standardised normal distribution as described in John C. accordingly.
Hull's Options, Futures and Other Derivatives (p. 218).
Investors can project this effect by regressing the share
The mean should be the expected annualised return on price against the local currency in terms of the bond
the stock and the standard deviation should be the currency. The correlation can be used to calculate parity
expected volatility over the full life of the security. volatility as follows:
2. Create all the days between the current (or valuation) Parity Volatility:
date and maturity and apply the normal distribution to
calculate the movements in the stock price. 2 2
(Stk. Vol. + FX Vol. ) - 2(Stk. Vol. x FX Vol.) (Corr.)
3. At the reset dates apply an "IF" function to reset the Alternatively, investors can calculate the stock’s volatility
conversion ratio to the appropriate number. in the currency of the bond. This will take into account the
correlation.
Step-Up Coupon Convertibles
This value is then used in the binomial model as the
A step-up coupon convertible has a coupon which is
volatility assumption for the cross currency convertible.
increased on set dates during the life of the instrument.
YTM can still be calculated without any difficulty at issue, as
Contingent Conversion Convertibles (‘CoCos’)
all fixed income flows are known in advance.
‘CoCo’ structures came to the fore amongst investment
grade issuers in the US in 2000. The innovation is very
74
simple. The bonds do not become convertible unless the significant. The D.R. Horton 0% 2021 USD convertible had
stock trades above a trigger level, usually expressed in an initial YTM/P of 3.25%, whereas the company stated in
relation to the conversion price (110% is common). the indenture that its comparable non-convertible yield
would be 8.88%. This provides a very sizeable current tax
In order to protect investors, the bonds will become shield.
automatically convertible under most M & A scenarios, or if
any qualifying capital distributions are made. Conversion is In addition, if the bonds are converted, the difference
automatically allowable if the bonds are called. between parity at the time of conversion and the accreted
value of the comparable non-convertible bond (less any
Because the zone of non-performance is small (between coupons) can also usually be used as as tax shield.
100% and 110%), projected far out into the future (many
CoCos are structured as 20 year obligations) and early The downside risk for investors is that the difference
exercise is rarely optimal, the loss of theoretical value is between the convertible and non-convertible yields may
small. result in 'phantom income' to non US Federal income tax
exempt investors. For this reason, these securities have
However, investors should be concerned about the need to been marketed more towards tax exempt offshore funds
convert a bond in order to participate in an unforeseen and (particularly hedge funds).
uncompensated corporate action event.
‘Phones’ and ‘Zones’
The CoCo structure (especially when combined with a
CoPay feature) significantly increases the attractiveness These securities also include contingent payment features,
of the convertible asset class to the issuer, because where the issuer can voluntarily make payments on the
although proceeds are taken in immediately, the dilution of bond during its life and have this amount deducted from the
the new shares does not have to be accounted for until the principal due at maturity.
contingent conversion threshold is satisfied.
Contingent Payment Convertibles (‘CoPays’)
‘CoPay’ features are often used alongside the CoCo

structure, although they are separate features and can be
used independently. Contingent payment features are
much more varied than CoCo structures, but all have
characteristics which allow them to be classified under the
US IRS's Contingent Payment Debt Instrument (CPDI)
regulations. Such classification allows the issuer to claim
deductions based on the non-convertible cost of debt (which
will usually be stated in the indenture). This can be very

Index
Accrued Interest, 6, 8, 17, 54, 63, 73, 76 Credit Investor, 23-24, 26-34
ADRs, 39, 76 Credit Spread, 8, 10, 12-13, 20, 24, 26, 29-31, 33-34, 55,
Ahold 4% 2005 EUR, 6-7, 9, 12-18, 35-36, 40, 42-44, 46-49 60, 76
Allianz, 57-58, 76 Cross Currency, 39, 74, 76
American Tower, 62 Cross Margining, 29
Asia Pacific, 4, 6, 38-39, 76 current yield, 9-10, 36, 70, 76
Asset Swap, 18, 22-27, 29, 32, 34, 40, 59, 76 death Spiral, 27
At the Money, 10-11, 13-14, 58, 76 TM
DECS , 71-72, 76
ATM, 58-59 Default, 3, 5, 12, 18, 21-22, 27, 29-34, 40
Axa, 46, 56-57 Delta, 11-16, 29, 37, 40-46, 48-49, 52, 54, 60, 76
Bank Austria, 64 Diamond Offshore, 58-59
Bell Atlantic, 20 Discount, 10, 16-17, 24-26, 28, 34, 48, 54-55, 70, 76
binomial model, 13, 74, 76 Distressed Debt, 10-12, 76
binomial tree, 14, 18, 36, 50-53, 56, 76 dividend yield, 9-10, 35-36, 55, 67, 71, 76
Black Scholes, 55, 76 dividends, 6-7, 36, 51, 55, 61-62, 65, 72-73, 76-77
Bond Floor, 10, 13-15, 18, 23-27, 29-30, 74, 76 Duration, 47-48, 76
Bond Value, 8, 13, 19, 26, 76 Entitlement Certificates, 17, 76
Bond with Warrants, 70 Equity, 3-4, 9-10, 13-16, 18-22, 29, 35-40, 42, 46, 54, 56,
Breakeven, 8-10, 36-37, 76 63, 66-67, 71-73, 76
Cable and Wireless, 20 Equity Switching, 37
call cushion, 54, 68, 76 Euroconvertible markets, 6, 76
Call Price, 54, 76 Europe, 4, 6, 18-19, 22, 27-28, 38, 40, 61-66, 76
Call Protection, 6-7, 15-16, 67, 76 Exchangeable Bonds, 20, 37, 66-67
Change of control, 63 France, 8, 63, 65, 76
CoCo, 22, 67, 70, 74-75 Gamma, 41-44, 49, 52, 54, 65, 76
Continuation Value, 52, 76 GDRs, 39, 76
conversion, 6-8, 10-11, 13-14, 17, 35-36, 40-41, 48, 54-55, Germany, 58, 63, 66, 76, 78
58, 62-65, 67-68, 70-76 hedge funds, 3-4, 14, 22, 40, 45, 59, 75-76
Conversion Price, 6-7, 10-11, 13-14, 40, 54, 58, 67, 70, 73, Hong kong, 4, 39
75-76 Implied LIBOR, 24-26
Conversion Ratio, 6-7, 13, 35, 40-41, 62-63, 71-74, 76 Income Swap, 36, 72, 76
convertible issuance, 4, 76 interest rates, 10, 12-13, 15, 31, 47, 49, 55, 76
convertible preferreds, 6, 72-73, 76 ISDA, 31-32
CoPay, 22, 67, 70, 75 ISIN, 6
counter-party risk, 28, 45, 76 issuer, 3, 6-7, 10-13, 15, 20, 22, 26-27, 29, 31, 33, 36, 40,
Coupon, 4, 6, 8-10, 24-26, 28, 34, 36, 50-51, 55, 63, 65, 49, 54, 61-68, 71, 75-76
67-68, 70, 74, 76-77 Japan, 4, 6, 12, 22, 31, 76, 78
Credit Default Swaps, 3, 5, 18, 21-22, 31-33 Junk, 10-12, 76
76
KEPCO, 61 Running Yield, 7, 77
Knock-Out Puts, 70 S&P 500, 19, 77
Korea, 4, 61, 76 screw, 17, 62-63, 68, 77
Latin America, 6, 76 Seacor, 37-38, 77
LIBOR, 8, 13, 24-26, 29-31, 33-34, 60, 67, 76 Sharpe Ratio, 21
Liquidity, 17, 33, 38, 40, 56, 59, 76 Siemens, 56-58, 77
Look Back, 64-65 special dividends, 7, 61-62, 77
M & A, 63-64, 75 Special Purpose Vehicle, 30, 67, 77
Make Whole, 27-28 SPV, 30
Mandatory, 10, 27, 67, 71, 73-74 Step-Up Coupon, 74, 77
Margining, 29, 45 stock borrow, 17, 77
MIPS, 67 stock splits, 7, 62, 77
Munich Re, 57 Stock Yield, 8, 10, 77
N.E.W.S., 71, 76 Stripping, 22, 77
National Bank of Greece 2% 2003 EUR, 20 Swapping, 22, 28-29, 77
Nominal Value, 6-8, 13, 41, 72-73, 76 Swiss Re / ING 1.25% 2003 NLG, 19
NTL 6.75% 2008 USD, 20, 59 Taiwan, 17, 77
OID, 28, 48, 55, 70 Terms and Conditions, 61
Original Issue Discount, 28, 48, 55, 70, 76 the Philippines, 17, 77
OTC, 37, 57-59, 61, 72 theoretical value, 13, 16, 36, 46, 49, 51-54, 56, 64, 74-75, 77
Par, 8, 10, 18, 22, 24, 28-30, 51, 63-64, 70, 73-74, 76 Theta, 48-49, 77
Par / Par Asset Swap, 22 Trustee, 61
Parity, 7-17, 35, 38-39, 41-44, 48-55, 62-64, 68, 73-76 UK, 2-3, 6, 59, 63, 77
PCCW, 39 US, 3-4, 6, 8, 19, 22, 27-29, 37, 39-41, 59-60, 62, 64, 67,
TM
PERCS , 71-72, 76 70-72, 74-75, 77
Phones, 75 Vega, 44, 46-47, 49, 77
pre-emption rights, 71, 76 Verizon, 20, 28-29
preference shares, 71-72, 76 Vodafone, 37
premium, 7-16, 18, 28, 31, 35-38, 45, 58, 63, 66-69, 71, 76-77 volatility, 8, 19, 21-22, 29, 35, 43-44, 46-47, 49-50, 55-59,
Prime broker, 41, 45-46 64, 68, 72, 74, 77
Provisional Call trigger, 54, 76 Withholding Tax, 6, 71, 77
put, 9, 15, 28-29, 31, 35-37, 54, 64, 66, 68, 70, 72, 74, 76 www.dbconvertibles.com, 2, 5, 32, 35
Recall penalties, 27 yield advantage, 7, 9, 16, 35-36, 72, 77
Redemption Value, 6, 8, 35, 39, 48, 68, 70, 76 Yield to Maturity, 7, 9, 24, 70, 77
Refixes, 73, 76 Zero Coupon, 24-26, 34, 50, 70, 77
Rho, 15, 47-49, 76 Zones, 75
Risk Premium, 8, 10-14, 16, 18, 77
Roche, 70, 77

Contacts & Office Locations
Deutsche Bank AG London Deutsche Bank AG Tokyo

Winchester House Sanro Park, 2-11-1 Nagatacho
1 Great Winchester Street Chiyoda-hu
London EC2N 2DB 100-6170 Tokyo
England Japan
Tel: +44 (20) 7545 8000 Tel: +81 3 5156 4000

Fax: +44 (20) 7545 4455 Fax: +81 3 5156 6001
Deutsche Bank AG Frankfurt Deutsche Bank AG Singapore

Grosse Gallusstrasse 10-14 8 Shenton Way
Frankfurt am Main # 20-01
60272 Frankfurt Treasury Building
Germany Singapore 068811
Tel: +49 69 910 00 Tel: +65 423 8001

Fax: +49 69 910 34227 Fax: +65 225 4911
Deutsche Bank New York Deutsche Bank AG Paris

31 West 52nd Street 3 Avenue, De Friedland
New York, NY 10019-6160 Paris, 75008
USA France
Tel: +1 212 474 8000 Tel: +33 1 44 95-64 00

Fax: +1 212 474 8560 Fax: +33 1 53 75-07 08
Deutsche Bank AG Sydney

Level 18
Grosvenor Place
225 George Street
NSW 2000 Sydney
Australia
Tel: +612 9258 1234

Fax: +612 9258 1400
78

Convertible Securities An Investors Guide. 2nd Edition October 2001 Jeremy Howard Michael O Connor

Uploaded by

Copyright:

Available Formats

Convertible Securities An Investors Guide. 2nd Edition October 2001 Jeremy Howard Michael O Connor

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Convertible Securities An Investors Guide. 2nd Edition October 2001 Jeremy Howard Michael O Connor

Uploaded by

Copyright:

Available Formats

What is the purpose of the document?

What is the purpose of the document?

What are convertible securities?

What are convertible securities?

Convertible Securities: An Investors’ Guide

2nd Edition October 2001

Jeremy Howard joined Deutsche Bank from Goldman Sachs

Michael O'Connor joined the convertible research team

We would like to thank Peter Darrell, Julian Moore and

We also thank the European trading team of Simon Garcia,

1. Convertible Securities - An Asset Class Grows Up 4

2. Convertibles for Beginners 6

3. Convertibles for Fixed Income Investors 18

5. Convertibles for Equity Investors 35

6. Convertibles for Hedge Funds 40

7. What Drives Convertible Prices Today? 56

9. Convertibles for Issuers 66

10. Convertible Structures 70

Convertible Securities 2001 3

The purpose of this updated Guide is to bring everyone

Asia has stayed on the map thanks to the Taiwanese

The real underperformer since we last wrote in 1997 has

We have expanded our coverage of Asset Swaps and Credit

Convertible Securities 2001 5

Share Price > Conversion Price Secondary Market Valuation

Convertible Securities 2001 7

Parity is a very important term for understanding how

Convertible Securities 2001 9

Cash Premium Paid

Source: Deutsche Bank Estimates

Convertible Securities 2001 11

These convertibles' prices can sometimes be correlated

As with all high yielding debt, we believe 'junk' convertibles

Investors purchasing out of the money convertibles will

1. If the underlying share price rises significantly;

3. If the market's perception of the credit quality of the

Convertible Securities 2001 13

EUR 38.65 120 129.0 66.3%

When starting from an 'at the money' profile with parity at

Table 3: As Parity Rises: Delta Rises

Parity Theo. Value Bond Floor Delta

60.00 93.94 91.91 21.87%

Source: Deutsche Bank Estimates

Convertible Securities 2001 15

Fig 9: Theoretical Value Rises as Call Protection is

Source: Deutsche Bank Estimates Source: Deutsche Bank Estimates

In other markets, such as the Philippines, there is a delay (up

In many developing convertible markets, however, stock

In other situations, lack of liquidity in the underlying shares

In reality, however, discounts are very rare in the developed

Convertible Securities 2001 17

The newest area for fixed income funds to take

But what does this mean for convertible investors?

The secret strength of convertibles is that the negative

Source: Deutsche Bank Convertible Research

Convertible Securities 2001 19

1. Exchangeable Bonds (where the issuer / guarantor and

Fig 11: Bell / C&W - NTL 4.25% 2005 USD

Source: Deutsche Bank Convertible Research

Fig 13: NTL 6.75% 2008 USD