. 31
( 39)


(b) Now project “New” becomes available. It will pay o¬ either +$50 or -$60. Show that it is in
the interest of shareholders for this project to be taken if the bond is not convertible.

(c) Now presume that the bond is also convertible into 75% equity. Show that it is no longer
necessarily in the interest of shareholders to take the bad project “New.”
¬le=caprest.tex: LP
606 Chapter 23. Other Capital Structure Considerations.

23·5. Inside Information

Our next important determinant of capital structure is inside information. Typically, ¬rm
New potential
partners/shareholders managers (acting on behalf of the old owners) have better information than new investors.
have less information
New investors should be careful that they are not exploited. As the old adage says, “Never bet
than current managers
with someone better informed than yourself.”
and owners.

Consider this scenario: you are a potential investor in an oil well, and you know that the current
If owners want partners
rather than lenders, the owner/manager (who has to raise new capital) already knows whether there is oil or not. You
project may not be as
do not know. You have to ask yourself:

• What will you believe about the oil well if the present owner o¬ers to make you a full
partner sharing in all future pro¬ts?

• What will you believe about the oil well if the present owner asks you for a loan to be paid
back that she is willing to collateralize with her present assets?

If you are o¬ered partnership, you should be reluctant to believe that there is oil. If, however,
the present owners want to keep the pro¬ts and simply borrow, she probably knows that the
project is pro¬table. This is sometimes called the winner™s curse or adverse selection. If you
receive the o¬er to become partner, it does not help you very much. If you do not receive the
o¬er to become partner, you would be better o¬ if you had indeed received it.

Table 23.6. Inside Calculations

Bad Luck Good Luck Future Ex- Today™s
Prob: 1/2 1/2 pected Value Present Value

Project FM $60 $160 $110 $100

Capital Structure LD: Bond with Face Value FV=$55
Bond(FV=$55) DT $55 $55 $55 $50
Equity EQ $5 $105 $55 $50

To raise $50, if:

Future Payo¬s
Project Creditors Percent of equity If equity ¬nanced, If debt ¬nanced,
is Believe to raise $50 owners keep owners keep
$55 + 65.6% · $160 = $160.00
Good Good $50/$145.45 = 34% $160
$55 + 65.6% · $60 = $94.36
Bad Good $50/$145.45 = 34% $60
$55 + 50% · $160 = $135.00
Good Unknown $50/$100 = 50% $160
$55 + 50% · $60 = $85.00
Bad Unknown $50/$100 = 50% $60
$55 + 8% · $160 = $68.33
Good Bad $50/$54.55 = 92% $160
$55 + 8% · $60 = $60.00
Bad Bad $50/$54.55 = 92% $60

The cost of capital in this example is 10% for all securities, which is equivalent to assuming risk-neutrality.
¬le=caprest.tex: RP
Section 23·5. Inside Information.

This analogy is directly transferable to capital structure. Sharing in the ¬rm™s equity is the A numerical example of
the inside information
equivalent of becoming a partner. Table 23.6 again considers our example, but adds the knowl-
edge of owners and your beliefs as a potential investor.

If you also know project quality: Not surprisingly, if the project is good and you believe this,
the owners end up with $160 next year. Similarly, if the project is bad, the owners end
up with $60 next year. Unfortunately, you do not know this.

If you believe either project quality is equally likely: This implies that you are willing to pur-
chase equity based on the expected project payo¬ of $110. Thus, you would lend $50 in
exchange for half the ¬rm.
Now consider what current owners would do. If they knew the project was good and
¬nanced through debt, they would be better o¬ ($160) than if they ¬nanced through
equity ($135). Conversely, if they knew the project was bad and they ¬nanced through
equity, they would be better o¬ ($85) than if they ¬nanced through equity ($60).
It follows that owners that prefer to ¬nance with equity know that their projects are
worse than average, and owners that prefer to ¬nance with debt know that their projects
are better than average. Therefore, it would not be reasonable for you to believe that the
project was equally likely to be good or bad. It would be irrational for you to maintain
such beliefs.

If you believe equity-¬nanced projects are bad: If current owners knew the project was good
and ¬nanced through debt, they would be better o¬ ($160) than if they ¬nanced through
equity ($68.33). Conversely, if current owners knew the project was bad, no matter how
they ¬nanced the project, they would end up with $60. For convenience, we can assume
that they then prefer to ¬nance with equity. (Think about them getting 1 cent extra.) Thus,
your beliefs are con¬rmed”¬rms that ask you for equity ¬nancing are bad. You would
assume that debt-¬nanced ¬rms are better than average, and that equity-¬nanced ¬rms
are worse than average.

New equity investors are not inclined to assume that the project is good. They will assume that The “Pecking Order”:
More equity-like
their new claims are on a project that will eventually develop problems. Thus, when existing
(partner-like) shares are
owners announce a new equity o¬ering, it releases information that the ¬rm™s projects are bad news, and can only
worse than generally believed, and the new equity can only be sold for a very low price. This is be sold at a discount.
again an example of adverse selection”only companies fearing the future would want to share
their prospects. In real life, we indeed observe that when ¬rms announce that they plan to
raise about $1 by issuing new equity, their old public equity value declines by about 10 cents.
But this argument extends not only to equity, but to other claims as well. The more risky the
securities are that insiders want to sell rather than keep, the worse are their beliefs. Sharing in
more junior (risky) bonds is the equivalent of the present owners making you a “little partner,”
when they are not willing to collateralize their loan. Consequently, the announcement of a
risky junior security releases information that the ¬rm™s projects are not too great, but not too
bad, either. In contrast, the new issue of a collateralized loan (or a risk-free senior bond) will
indicate that the ¬rm™s projects are better than expected. The outcome is that the better the
¬rm™s projects are, the more senior the security the managers will o¬er for sale. This leads to
a pecking order view of capital structure: the best projects are ¬nanced by the most senior
debt, worse projects by junior debt, and the worst projects by equity.
What does this imply for the optimal capital structure? Consider a ¬rm that cannot issue debt Firms may want to avoid
issuing more shares of
easily because it has little collateral or because additional debt would unduly increase expected
anything, and equity in
bankruptcy costs. If it cannot issue equity because of these insider concerns, such a ¬rm may particular.
have to pass up on some good (but perhaps not stellar) projects, simply because owners do
not want to sell their projects at the price of the worst possible scenario. A publicly trading
¬rm thus may take on too much debt (incurring ¬nancial distress costs) or ration its projects,
failing to take at least some of its positive NPV projects.
¬le=caprest.tex: LP
608 Chapter 23. Other Capital Structure Considerations.

Important: Inside information concerns favor debt over equity as the cheaper
¬nancing vehicle.

When could a ¬rm issue equity without an insider penalty?
There is a way out!

• If there is a mechanism”e.g., a detailed audit”by which insiders with good projects can
credibly convey the true quality of the project, it would be in their interest to do so. Indeed,
if such a mechanism is known to exist, and owners do not undertake it, potential investors
should immediately assume that current owners are not doing so because they know that
the outcome will be bad.

• If current owners can convince potential investors that they have invested all of their own
money, that they have maxed out their personal credit cards, and they just cannot put any
more personal capital at risk than they already have, then there is no information in the
fact that they are asking to raise equity capital. In this case, external investors can assume
that the project is not necessarily bad. Indeed, no venture capitalist will ever invest in a
startup in which the current owners do not have most of their personal wealth at stake.

The pecking order (inside information) and free cash ¬‚ow (agency) theories have a very close
Agency and inside
information are closely family relationship. The former says that when ¬rms issue equity, managers signal that they
believe that the future will be worse. The latter says that when ¬rms issue equity, managers
will make the future worse”they will waste the money. Both send information signals to the
public about a worse future, although the latter is more causal than the former.
Solve Now!
Q 23.13 A house up for auction can be worth either $500 or $1,000 with 50-50 probability. The
other bidder knows the true value; you do not. If you bid for the house in an auction, what should
you bid? If you bid $750, what is your expected rate of return?
¬le=caprest.tex: RP
Section 23·6. Transaction Costs and Behavioral Explanations.

23·6. Transaction Costs and Behavioral Explanations

Transaction costs have played an important role in all capital structure examples above: if Transaction costs are
everywhere. They
transaction costs had been zero, external pressures would force management to choose the
de¬nitely can prevent
best capital structure. But if transaction costs are high, managerial mistakes are di¬cult or optimal adjustment.
impossible to correct for outsiders. It is not just enough for an outsider to purchase shares
and then sell them. The appropriate corrective action requires accumulating enough shares and
pressuring management to improve the situation. Without the discipline of external pressure,
managers and investors can commit mistakes. They may take too much debt or too much
equity, and the market may not be able to correct it.
Section 19·2 has already described the link between high transaction costs and behavioral Transaction costs
“cause” behavioral
¬nance. In the corporate ¬nance context, the presence of high transaction costs”and, with
¬nance concerns.
it, the rarity of correcting mechanisms”means that behavioral ¬nance plays an important
role. It can rationalize a lot of managerial behavior, which is otherwise di¬cult to explain.
Unfortunately, on so vague a level, without a further description of what the mistakes are, it is
less prescriptive than the above theories. That is, it seems to o¬er little guidance as to how a
smart manager should act di¬erently (i.e., what the best debt/equity ratio is).
Then again, behavioral ¬nance is the most promising new direction in corporate ¬nance. It Speci¬c behavioral
errors can have speci¬c
is still too early to tell where and how it will help us better understand the world. Some
early insights suggest that there are certain behavioral mistakes that are more common than
others. For example, we now believe that overcon¬dence and overoptimism are common
traits among both managers and investors. If managers are overoptimistic, it may aggravate
agency concerns (they may take some negative NPV projects) and no-liquidation concerns, but
alleviate underinvestment problems. If investors are overoptimistic, issuing equity may not be
so disadvantageous as the inside information argument suggests. Investors may not necessarily
believe the worst”and there is some evidence that such was the case during the Internet bubble
at the turn of the millennium. Although it is less likely that markets rather than managers are
committing mistakes, there is good evidence that ¬nancial markets may be imperfect, too. If
markets indeed misvalue securities”either because they are irrational or imperfect”it would
be rational for managers to seek to time equity issuing activity.
Another domain where behavioral e¬ects seem particularly important are dividends. There are Dividend policy is
probably a behavioral
at least some investors out there who seem to prefer dividends, even if this is not a wise choice
managerial error.
from a tax perspective. In response, rational managers may want to pay dividends rather than
use share repurchases to send cash to their equity holders, even if this incurs a higher tax
obligation. In any case, there is little that ¬nancial markets can do about the wrong payout
policy. The world is how it is.

A more traditional view is that transaction costs also play a direct role. For example, the report- And then there are also
direct transaction
ing requirements and liabilities imposed by the 1933 Securities Acts for publicly traded equity
securities can be much larger than those for private borrowing. For many small companies,
these costs may be large enough to warrant a capital structure consisting exclusively of private
securities and bank debt.

Important: Behavioral considerations could favor either debt or equity.

Solve Now!
Q 23.14 What are behavioral explanations for dividend payout?
¬le=caprest.tex: LP
610 Chapter 23. Other Capital Structure Considerations.

23·7. Corporate Payout Policy: Dividends and Share Repur-

You should realize that all these considerations apply to our capital structure, and therefore
apply both to the appropriate debt-equity level, as they apply to changes therein”¬rms may
issue debt or equity to send cash into the company, or they may pay out coupon and principal to
send cash back to owners. Dividends and equity share repurchases are particularly interesting,
because they are at the discretion of management.
When companies decide to return cash to shareholders, they do so either by dividends or share
Institutional details and
“the need for cash”?! repurchases. Many public companies pay dividends every quarter, others pay only once a
year. Sometimes, ¬rms declare “special dividends” to indicate that they will not repeat. The
process to pay dividend is that the board of director votes to pay a dividend (on the declaration
date), and then collects the names of all shareholders (on the record date). For shareholders,
shares that are owned before the ex-dividend day are called cum dividend and will receive the
dividend payment; shares owned after the ex-dividend date are without the dividend payment.
Naturally, shares are worth more with an extra dividend, so the price will fall after the ex-
dividend date”usually by just about the value of the dividend itself. Otherwise, a tax-exempt
investor can earn close-to-arbitrage pro¬ts. For example, if a $40 stock pays an annual dividend
of $2 (a dividend yield of 5%), this stock will trade at $38 after the ex-dividend date. Finally,
the dividend checks are mailed (on the payment date). Note that it makes no sense to argue
that dividends are paid because particular investors “need” money”if you hold 100 shares for
$4,000 and receive $200 worth of dividends, you would have $3,800 shares left; but you could
just as well have sold 5 shares for $200 on the stock exchange, which would similarly have left
you with $3,800 and $200 cash.
Share repurchases are direct substitutes for dividend payments. One naïve misconception is
Share repurchases and
dividends are perfect that share repurchases are di¬erent because they bene¬t only shareholders tendering their
substitutes in a perfect
shares into the repurchase”unlike dividends, which bene¬t all shareholders. This argument is
false. A ¬rm with 100 shareholders, each owning $10 worth of shares, could pay $50 worth of
dividends (50 cents to each shareholder), and the ¬rm would be worth $950. Each shareholder
would have a share worth $9.50 and $0.50 in dividends. If the ¬rm repurchased $50 worth
of shares, the ¬rm would now have 95 shareholders, each owning $10 worth of shares. In the
perfect M&M world, without taxes, all shareholders are equally well o¬ with either a repurchase
or a dividend payment.
Like the debt/equity choice, dividends and repurchases matter only if we leave the M&M world.
In the real world,
dividends are For example, investors who need money may now save on transaction costs of having to sell
tax-disadvantaged on the
shares, or the dividend payout may reduce agency con¬‚icts and money-wasting by managers,
investor side, relative to
or there may be tax di¬erences. We will focus primarily on the role of personal taxes. If taxes
share repurchases.
mean that there is less money left over for the dividend payments than there is for the share
repurchases, it must follow that share repurchases should be preferred from a tax perspective.
And, indeed, we know that dividends are more di¬cult to shelter than share repurchases, if
only because the investors that tender into the share repurchase will be those who ¬nd it
easiest to shelter the capital gains. The e¬ect is that e¬ective personal income taxes are higher
on dividends than on share repurchases. Again, the tax clienteles in the economy will evolve
to reduce the e¬ective dividend taxes paid. That is, clientele arguments cannot only explain
why some investors prefer holding equity over debt (to receive capital gains instead of taxable
interest payments), they can also explain why dividend taxes are not as punishing as they
would otherwise seem to be. (Some low-tax pension fund investors can avoid the personal
tax on dividend receipts.) But the clienteles can just reduce the penalty of dividends relative
to share repurchases”they cannot eliminate it. They can therefore not explain why ¬rms pay
dividends”share repurchases remain uniformly better, because they can de-facto avoid almost
all personal income taxes. Share repurchases are both a better and a perfect substitute for
¬le=caprest.tex: RP
Section 23·7. Corporate Payout Policy: Dividends and Share Repurchases.

Historically, it was an important puzzle why corporations would ever pay dividends. Both the Share repurchases are
becoming relatively
tax advantage of share repurchases over dividends and the average corporate dividend yield
more important over
in the economy were high. Firms eventually ¬gured this out, too. Bagwell and Shoven (1989) time.
found that big ¬rms paid out about 7 times more through dividends than through share re-
purchases in 1977”but by 1987, this ratio had dropped to about 1.5. In any case, by today,
this payout puzzle has become fairly unimportant. As of 2003, typical dividend yields (the
amount of dividends divided by the share price) have declined to about two percent per year. Of
three tax advantages of capital gains over repurchases”a lower statutory rate, taxation only at
realization, and an o¬set against capital losses”only the latter two remain. The Bush tax cuts
(in e¬ect at least until 2008) have much reduced statutory rate di¬erences. Therefore, the ef-
fective remaining advantage of capital gains vs. dividends is now fairly small. (A good number
of ¬rms responded to the Bush tax cuts in the logical way, most prominently Microsoft”they
started paying dividends!) Add to this that some dividend payments are received by tax-exempt
institutions (who often purchase shares just before the dividends are paid and sell them just
thereafter)”and the question of whether to pay out cash in dividends or share repurchases is
no longer a very important for the typical company. Dividends may still not be the smartest
method for managers to pay out, but the wasted money is too small to o¬er large arbitrage
opportunities (buying the ¬rm, ¬ring the CFO, changing the payout policy, and selling the ¬rm
for a higher price). The presence of dividends is less of a mystery than it once was.
So, with the prime historical di¬erence between dividends and share repurchases (personal Remaining differences:
changes in inside
income taxes) much reduced, what are the remaining important di¬erences? Not much, but
ownership, an informal
here they are: understanding about
persistence of future
payments, and
1. Executives often receive stock options in the company, whose value depends on the share “behavioral ¬nance” type
price. Because share prices drop when companies pay dividends, managers with many investor preferences.
options prefer repurchases to dividend payments.

2. Executives and insiders are often not permitted to tender their shares into share repur-
chase o¬ers, and thus will own relatively more of the company after a share repurchase
than after an equivalent dividend payment.

3. Dividends tend to be regular and are expected by shareholders to continue. An ordinary
dividend payment therefore informally obliges management to continue, thereby signal-
ing more optimism about the future. In contrast, many share repurchases are done in
one-time chunks. The distinction is not perfect, though. Many companies have semi-
regular share repurchase programs, which make repurchases almost as regular as divi-
dend payments. And many other companies pay “special dividends,” which immediately
signal their one-time nature to investors. Such special dividends are as much “one-time”
as share repurchases.

4. There is behavioral ¬nance empirical evidence that many retail investors just “like” divi-
dends better than share repurchases.
Today, this matters less, but it used to paradoxical, when dividend payouts were badly
tax-disadvantaged. The argument that investors like dividends “because they need cash”
does not hold water. Selling a fraction of the shares in stocks that pay zero dividends
provides physical cash, too”except that the investor would not have had to pay as much
in personal income tax. Still, it seems that many investors wrongly think only of share
sales but not of dividend receipts as reduction in their “investment substance.”
The empirical evidence is still under academic investigation. My guess is that the answer
will likely be that these individual investor preference e¬ects are real, but that they are
not universal. For example, we also know that many tax-exempt institutions are obliged
by their charters to hold only dividend paying stocks”just as the clientele theory would
have predicted.
¬le=caprest.tex: LP
612 Chapter 23. Other Capital Structure Considerations.

So what payout policy should a company choose? The most important recommendation is
that a company should pay out cash when the alternative uses are not positive NPV projects.
Of course, many managers will not like to hear this advice or they will assert that all of their
projects are high NPV (even if they are not). Compared to the question of whether the ¬rm
should pay out or not pay out, the question of whether the form of payout should be dividends
or share repurchases is of secondary importance today. In cases of doubt, share repurchases
seem better than corporate dividends. If the Bush dividend tax reductions will disappear in
2009, share repurchases would seem much better than dividends.
Before we close, you may sometimes hear of stock dividends and stock splits”these are really
Stock dividends and
splits. not payouts, at all. For example, a $1 million company whose shares are trading for $100/share
may issue a stock dividend of 1 share for every 10 outstanding shares. Its 10,000 shares would
become 11,000 shares, each worth $90.91. Or it may splits its shares 2-for-1, which means that
its 10,000 shares would become 20,000 shares, each worth $50. Another strange creature is
the dividend reinvestment plan (DRIP), in which shareholders can volunteer to automatically
reinvest dividend receipts in the company. They therefore never receive any cash. All that
they receive is a tax obligation at the end of the year for the dividends that they presumably
received. If the company had just kept all the money, it would have saved its investors the tax
obligations. (To complicate this further, if set up with the company rather than a brokerage
¬rm, many DRIPs o¬er the shares at a discount or at a rate that is not the current market value
but the average value over the most recent quarter.)

Tax schemes also apply to investors If personal investors want to hold dividend-paying stocks
Side Note:
without paying dividend taxes, they can do so for most of the year”as long as they sell them to institutions prior
to the ex-dividend day, which is the day on which the owner of the shares is determined for the distribution
of the dividend.
In real life, this “tax-arbitrage” indeed happens; tax-exempt funds compete to purchase these shares, driving up
the share prices before the ex-dividend date. The Financial Times reports that such transactions are known as
“bed-and-breakfast” deals for equity, and “bond-washing” for bonds”even though both the IRS and the Bank
of England have speci¬cally prohibited such tax arbitrage. The latter has imposed a 1-week holding period for
tax-free institutions purchasing around dividend dates.
Of course, there are limits to such dividend tax arbitrage. Some high-tax investors would have to pay capital
gains taxes when they sell their shares, and thus prefer paying taxes on the dividends instead of on their capital
gains. In addition, the round-trip transaction costs limit the possible pro¬ts from the avoidance of dividend
Units are combination securities that consist of a debt security and an equity security. If the ¬rm pays interest,
it shifts its tax burden to the unit owners. If the ¬rm pays dividends, it shifts its unit owners™ tax burdens to

Anecdote: Ralph Nader and Microsoft
On January 4, 2002, Ralph Nader wrote an open letter to William H. Gates, III, Chairman of Microsoft, that begins
as follows:

We are writing to ask Microsoft to change its practice of not paying dividends to shareholders. Our
reasons are as follows.
1. The quantitative failure to pay dividends year after year is an inappropriate and we believe
unlawful device to shelter Microsoft earnings from federal income taxes.
By not paying dividends, wealthy Microsoft shareholders such as yourself avoid paying the top
marginal tax rate of 39.6 percent that would apply to income distributed as dividends. By taking
earnings entirely through stock sales, wealthy shareholders lower their tax rate to the maximum
20 percent that applies to capital gains. According to the most recent SEC reports on insider trades,
you personally sold more than $2.9 billion in Microsoft stock last year, bene¬ting enormously from
the lower tax rate that applies to stock sales.

This letter does not even point out that 20% is an overstatement: Bill is taxed only on realized capital gains! If
Bill does not sell his shares, he su¬ers zero taxes on increases in his wealth over the year. And, with the GOP
elimination of the estate tax, neither may his heirs. The Bush tax reforms of 2003 have further signi¬cantly
reduced the taxes on dividend payments. Microsoft promptly started paying dividends in 2003”many billions
Here is an interesting question: Is it the fault of Bill Gates (who is also a proli¬c political campaign donor) or
is it the fault of the U.S. government that Gates has su¬ered only minimal tax obligations on his wealth gains
over the last 20 years?
¬le=caprest.tex: RP
Section 23·8. Further Considerations.

Solve Now!
Q 23.15 If a normal investor cannot participate in a share repurchase program, is she better o¬
with a dividend payout than with a share repurchase?

Q 23.16 What companies should pay dividends?

23·8. Further Considerations

Of course, we have not covered everything possible about capital structure in our chapter. The
real world is considerably more complex than the surgical scenarios we have explored. But,
you now have a very good grasp of the most important issues. Let me close our chapter with a
number of important issues that did not ¬t earlier.

23·8.A. Interactions

Typically, ¬rms do not face each of the above problems in isolation, but all at the same time. Executing the
The presence of one problem”or attempts to reduce it”may worsen another. That is, there
strategy may not be
can be signi¬cant costs to move from a suboptimal to an optimal capital structure. possible.

For example, we already know that if a ¬rm is close to bankruptcy, issuing equity could avoid Con¬‚ict among different
interest groups can
or reduce bankruptcy costs, which in turn would increase ¬rm value. But the infusion of more
prevent optimal
equity may mostly bene¬t bondholders, so equity holders may not agree to put in more equity. solutions.
So, although a reorganization (i.e., a new start) could install a capital structure to increase ¬rm
value, there are problems to be resolved to get there, given the current capital structure.

Important: Interaction e¬ects can make it di¬cult to optimally adjust capital
structure in the future. This can favor a more ¬‚exible capital structure (more
equity and slack) today.

Of course, equity infusions to stave o¬ bankruptcy are not always good, either. For example, it Equity Infusions are not
always good, either,
could be that it would be more e¬cient to liquidate the ¬rm, and the equity infusion allows the
¬rm to continue to operate ine¬ciently. Raising more equity might thus facilitate the wrong
managerial action.

23·8.B. Reputation and Capital Structure Recommendations

We have already mentioned another important factor in the real world”reputation. It can lower Sometimes, owners are
best off building a
¬nancing costs, improve incentives, and increase ¬rm value. Remember an earlier example, in
corporate reputation,
which the presence of the ex post ability to expropriate bondholders hurts the ¬rm. If managers which can help alleviate
had a reputation not to take such bad projects, perhaps overly restrictive covenants could be investor worries.
avoided, in e¬ect lowering ¬nancing costs ex ante. More importantly, the example assumed that
everyone knew exactly what expropriation opportunities existed, and what their probabilities
were. But, despite restrictive covenants, bondholders will always have the nagging suspicion
that they may be expropriated after all when unforeseen opportunities appear. Only the build-
ing of trust and reputation can overcome such suspicions, with their associated increase in
¬nancing costs.
Do investors trust managers? Can investors trust managers? Should investors trust managers?
When is it worthwhile for a manager/¬rm to build such a reputation? How can this e¬ectively
be accomplished? These are di¬cult questions to answer empirically, but they are important
in the real world.
¬le=caprest.tex: LP
614 Chapter 23. Other Capital Structure Considerations.

Ultimately, the trick in being a good manager is to judge and weigh the plethora of marginal
Choosing the best
Capital Structure is a costs and marginal bene¬ts of projects, of debt, and of equity, and to have sound judgment in
combination of art and
deciding on a good combination thereof. Choosing a good capital structure remains as much
an “art” as it is a “science.” This is good news for today™s business students: capital structure
choices are unlikely to be taken over by a computer program anytime soon.

23·8.C. Final Note: Cost of Capital Calculations

As with personal income taxes, managers need not compute what expected value consequences
WACC still works for
managers. These factors their decisions will have: if they choose a capital structure leading to inferior future payo¬s,
impact changes in the
they have to give away a larger percentage of the ¬rm today, which manifests itself in the WACC
cost of capital, r, that
interest rate or APV cash ¬‚ows that they experience.
managers face.

Figure 23.1. Capital Structure E¬ects and Formulas

Project Value

= “
Perfect Project Value All Distortions

Project Value Under
= +
Distortions: Taxes, Dis-
All Remedies
tress Costs, Transaction
Costs, Opinions, etc.
¨ d

Debt E¬ects
Equity E¬ects
e.g., signals higher quality Tax Deductibility
e.g., lowers
lower issuing costs, etc. of Interest
distress costs
c c c
NPV of Cash Flows
• ignoring corporate income tax NPV of Corporate Income Tax Shelter
APV = +
mitigation due to debt; Arising From Presence of Debt
• but taking into account all other
mitigating factors.

“ignoring corporate income tax mitigation due to debt” means that cash ¬‚ows are computed as if no interest payments
are tax deductible.

Figure 23.1 again illustrates that even though the tax shelter created by the tax deductibility of
interest plays a special role in the APV formula, other factors can play just as important a role.
However, they do not manifest themselves as their own algebraic terms in the APV formula;
instead, they manifest themselves in the cash ¬‚ow part of the formula. The reason why these
interest payments are broken out is

• the tax shelter can be very large, so some extra emphasis is not harmful;
• it is straightforward to compute the corporate cash ¬‚ows “as if interest payments were not
tax deductible,” which in turn makes it easy to add back and understand the quantitative
magnitude of the tax shelter.

You could also try to ¬rst compute the value of the corporation not taking into account other
bene¬ts of debt or equity, and then add back these e¬ects. For example, you could work with
cash ¬‚ows to owners as if they were fully taxed at the personal level, and then add back the
personal tax savings created by not paying them out but reinvesting them (thus creating capital
¬le=caprest.tex: RP
Section 23·9. Summary.

gains). (Indeed, the so-called Miller Debt and Taxes formulas have tried to do this.) But such
formulas are very di¬cult to apply in the real world, if only because managers do not know the
e¬ective marginal income tax rates of their investors”so few managers would ever do so.

Important: Only corporate income taxes ¬‚ow directly into the WACC or APV
formulas through their own algebraic terms. All other distortions, ranging from
personal investor taxes to bankruptcy distortions, manifest themselves in the cost
of capital (the appropriate interest rate) that investors demand, and/or the value of
the cash ¬‚ows of projects. The optimal capital structure allows managers to raise
¬nancing at the lowest ¬rm cost of capital, and thereby maximizes the wealth of
current owners.

23·9. Summary

The chapter covered the following major points:

• The managerial objective should be to minimize the overall tax burden”the sum of taxes
paid by the corporation and its investors.

• Investor clientele e¬ects arise because they reduce overall tax payments. They are

High Tax Investors
Low Tax Investors
Choice (e.g., Pension Funds) (e.g., High-Income Individuals)

Hold (low-dividend) stocks
Hold bonds
Better (or very high-dividend stocks) with high capital gains

Hold (low-dividend) stocks Hold bonds
Worse with high capital gains (or very high-dividend stocks)

High Tax Corporations Low Tax Corporations
Choice (e.g., “cash cows”) (e.g., “growth ¬rms”)

Finance With Stocks
Better Finance With Bonds (pay out with share repurchases
instead of dividends)

Worse Finance With Stocks Finance With Bonds

It is the market prices for the cost of capital that incentivice smart ¬rms and smart in-
vestors to arrange themselves in this clientele fashion to reduce taxes.

• Dividends used to be a great mystery, but nowadays they are almost as good as share
repurchases. Today, di¬erences between repurchases and dividends are minor.
One remaining di¬erence is “behavioral”: some individual investors just seem to like
dividends for their own sake”even when they were severely tax-disadvantaged prior to
¬le=caprest.tex: LP
616 Chapter 23. Other Capital Structure Considerations.

• There are numerous other tax reduction schemes that ¬rms can undertake. Some are
mentioned in the ¬nal section.

• Capital structure can in¬‚uence managerial behavior in good times and bad times, and
both positively or negatively.

• Equity has the advantage in that it reduces the likelihood of ¬nancial distress, and with it
deadweight bankruptcy costs in bad times. This includes both direct costs (such as legal
fees) and indirect costs (such as underinvestment, reluctance to liquidate, and excessive

• Debt has the advantage in that it imposes discipline on managers and thus reduces money
wasting in good times. Managers and employees will work harder if poor performance
can lead to bankruptcy.

• Equity has the advantage in that it does not tempt managers to expropriate creditors.
If bondholders fear expropriation from subsequent increases in corporate risk or from
the issuance of more debt with earlier payments or payments that are equal or higher in
priority, they demand a higher cost of capital.

• Debt has the advantage in that it signals con¬dence. If owners”or managers acting on
behalf of owners”prefer to sell partnership shares rather than debt, they probably believe
that the project™s true quality is worse. Thus, the cost of raising equity is high”new
partners will assume the worst.

• Managers can continue to use the WACC or APV formulas from Chapter 22, because the
issues described in this chapter ¬‚ow into the ¬rm™s cost of capital through the costs of
capital quoted by the ¬nancial markets.

A summary of all capital structure e¬ects from Chapter 22 to here is in Table 23.7.
¬le=caprest.tex: RP
Section 23·9. Summary.

Table 23.7. Summary of Capital Structure E¬ects

E¬ect Favors

Personal Income Taxes Equity

Debt Expropriation Equity
Includes costs arising from the interaction of borrower credibility
and borrower ¬‚exibility. Includes complete contract speci¬cation

Financial Distress Costs Usually Equity
Includes ine¬cient operations, underinvestment problems, sup-
plier and customer incentives, failure to liquidate or sell at appro-
priate price, predatory policies by competitors, etc.

Corporate Income Taxes Debt

Too Much Cash Flow Debt
Sometimes called Moral Hazard. Includes overinvestment, free
cash ¬‚ow, excessive managerial perks, veri¬cation, etc.

Inside Information Debt
Sometimes called adverse selection. Sometimes called pecking

Behavioral Finance Situation-Dependent

Transaction Costs Situation-Dependent
¬le=caprest.tex: LP
618 Chapter 23. Other Capital Structure Considerations.

Solutions and Exercises

1. The ¬rm ¬rst pays taxes on money used for repurchase and dividends, but can use pre-tax money for interest
payments. Investors can easily shelter repurchases (as capital gains), but face the full brunt of Uncle Sam on
interest or dividend payments.
2. Stable and old: Pension Funds. Young and growing: Individuals.

3. Lower bankruptcy costs, both direct and indirect. Fewer incentive problems to put up extra “maintenance”
money. Fewer incentive problems to avoid liquidation and drag on, instead.
4. For example, legal fees and management attention.
5. Neglected maintenance reduces the value of assets relative to the ¬rst-best.
6. Shareholders do not want to sell the ¬rm if they are underwater, even if the o¬er is more than the ¬rm is
worth: all bene¬ts would go to the shareholders.
7. It can get the ¬rm to commit to undertake more risky projects. However, it can also make it more di¬cult
to respond when the company is already levered up.

8. Corporate Planes. Large headquarters. Large sta¬.

9. They help reduce the incentives of equity shareholders to expropriate bondholders.
10. It reduces the need for some bond covenants and thus gives the ¬rm more ¬‚exibility in case a great project
were to suddenly appear. Bondholders would be happy, because they, too, would bene¬t.
11. First, issuance of other securities that have an earlier or equal dip on the ¬rm™s cash ¬‚ows in distress. This
could be other bonds of equal or higher priority, or a straight out dividend payment. Second, the adoption of
risky projects. Numerical examples are in the text.
(a) The bond is convertible into 75% of the ¬rm™s equity.

Bad Luck Good Luck Future Ex- Today™s
1/2 1/2
Prob : pected Value Present Value

Project FM $100 $120 $110 $100

Convertible Bond with Face Value FV=$90
Bond(FV=$90) DT $90 $90
Converted to 75% EQ $75 $90
Best Choice DT $90 $90 $90 $81.82
Equity EQ $10 $30 $20 $18.18


Bad Luck Good Luck Future Ex- Today™s
1/4 1/4 1/4 1/4
Prob : pected Value Present Value

Project FM $100 $100 $120 $120 $110 $100.00
Project New $50 “$60 $50 “$60 “$5 “$4.54
Total Projects $150 $40 $170 $60 $105 $95.45

Straight Bond with Face Value FV=$90
Bond(FV=$90) DT $90 $40 $90 $60 $70 $63.64
Equity EQ $60 $0 $80 $0 $35 $31.82

(c) The bond is convertible into 75% of the ¬rm™s equity.
¬le=caprest.tex: RP
Section 23·9. Summary.

Bad Luck Good Luck Future Ex- Today™s
1/4 1/4 1/4 1/4
Prob : pected Value Present Value
Total Projects $150 $40 $170 $60 $105 $95.45

Bond(FV=$90) DT $90 $40 $90 $60
Converted to 75% EQ DT $112.50 $30 $127.50 $45
Best Choice DT $112.50 $40 $127.50 $60 $85 $77.27
Equity EQ $37.50 $0 $42.50 $0 $20 $18.18

Note that the shareholders are no longer better o¬ if the project is taken, because they receive $18.18
either way. If we made the debt convertible into 75.1% of the ¬rm™s equity, then the shareholders would
be outright worse o¬.

13. You should not bid anything (except perhaps $500). If you bid $750, then you will get the house only if it is
worth $500, and you would therefore lose 33%.

14. Managers may behave irrationally, and pay dividends even though this is expensive from a personal income
tax perspective. Investors may indeed like dividends irrationally, even if it is not in their self-interest.

15. No! Even this investor is better o¬ with a share repurchase, but his increase in wealth now comes from in
increase in unrealized (and therefore still untaxed) capital gains.
16. None. If absolutely necessary, it should be ¬rms with many tax-exempt investors.

(All answers should be treated as suspect. They have only been sketched, and not been checked.)
¬le=caprest.tex: LP
620 Chapter 23. Other Capital Structure Considerations.
Clinical Observations About Capital Structure

Debt and Equity in Subject Firms
last ¬le change: Mar 19, 2006 (15:38h)

last major edit: Apr 2005

You now know the considerations that should help management determine the ¬rm™s optimal
debt-equity ratio. Before we dive into the process, let us ¬rst look at the ¬nancial structure
of one particular company, IBM, to get a better feel of the capital structure in a real company.
Thereafter, we shall look at some capital structure summary statistics for other publicly-traded

¬le=capclinical.tex: LP
622 Chapter 24. Clinical Observations About Capital Structure .

24·1. Tracking IBM™s Capital Structure From 2001 to 2003

Perhaps the best way to understand capital structure is to follow one company. Table 24.1
We will follow IBM from
2001 to 2003. shows how IBM™s capital structure evolved from 2001 to 2003 . The “change” lines in the table
refer to changes in the rows above, and make it easy to see where big changes were happening.
Your ultimate goal is to understand the debt ratios at the bottom of the table.
The top part of Table 24.1 shows how IBM™s liabilities evolved. In the context of debt ratios, it
Debt”Total or just
Financial Debt? is common to consider debt as either all liabilities or just the ¬nancial debt. The latter is the
sum of long-term ¬nancial debt plus any ¬nancial debt within short-term liabilities. IBM has
debt of

2001 2002 2003
Financial Debt, in billions $27.151 $26.017 $23.632
Total Liabilities, in billions $66.855 $73.702 $76.593

Obviously, the di¬erences are large, but either measure serves a purpose, and this is worth
some contemplation.
Total liabilities include other debt, such as pension or other current liabilities (e.g., accounts
Claimants become so by
providing assets. payable). How are these di¬erent from claims that creditors and shareholders receive? Credi-
tors and shareholders provide assets (value through cash in¬‚ows) and receive their claims in
exchange. It is this contribution that makes them part owners of the ¬rm. But the same is
true for most other obligations. For example, the pension fund is a creditor. Its claim on the
assets came about because IBM did not fully pay its employees in cash, but e¬ectively deferred
some employee compensation. IBM has, in e¬ect, made its employees creditors. And just
like creditors and shareholders, who provided funds directly, the pension fund has provided
value (assets) in terms of making IBM™s employees produce goods. In other words, the pension
obligation arose together with the positive impact of employees on the ¬rm™s assets. Just like
¬nancial debt, the employees™ pension fund should be considered both in terms of its impact
on the scale of the ¬rm (increasing both assets and liabilities) and in terms of its impact on the
mix of debt and equity”it is not just an obligation that pops out of a vacuum.
It may seem arti¬cial to distinguish liabilities such as pensions from ¬nancial debt. Both are
Other Liabilities are very
similar to ¬nancial debt. subject to the same theoretical issues”for example, both are paid from pre-tax earnings, and
both require servicing (or else, bankruptcy looms). Many liabilities also require interest pay-
ments and many require repayment on equally as rigid a timetable as ¬nancial debt. In this
respect, de¬ning debt as just ¬nancial debt can seem arti¬cial and perhaps even misleading.
However, ¬nancial debt and non-¬nancial obligations do di¬er, at least a little. First, ¬nancial
varies”¬nancial debt claims are mediated through the ¬nancial sector (banks and markets). Second, sharp changes in
can be easily rearranged;
¬nancial claims are relatively easy and common. For example, a ¬rm can pay o¬ its short-term
pension bene¬ts cannot.
debt (e.g., by issuing long-term debt). In contrast, although not always and necessarily the case,
obligations like pensions or payables tend to march on pretty steadily. You can consider this
di¬erence in claims to be linked to the speed of control that ¬nancial managers have over these
di¬erent components of the ¬rm™s scale and debt-equity ratio. CFOs probably have relatively
little short-term control over the ¬rm™s pension obligations, more short-term control over the
¬rm™s equity market value and stock price (e.g., by retaining equity), and a lot of short-term
control over the ¬rm™s debt and equity issuing and repurchasing, and over its dividend policy.
This is one reason why we are especially interested in ¬nancial debt ratios.
By now, I hope it has become clear to you that characterizing capital structure cannot be accu-
Consider both debt
ratios” and book or rately accomplished by just one number. Instead, capital structure must be seen from di¬erent
market value?
angles. And the distinction between ¬nancial debt and total liabilities is merely the ¬rst set of
angles. The next perspective you should consider is whether you should rely on book values
or market values. Unfortunately, market values for liabilities are not readily available, so we
must rely on the accounting book values. Fortunately, there is often little di¬erence between
the two as far as liabilities are concerned.
¬le=capclinical.tex: RP
Section 24·1. Tracking IBM™s Capital Structure From 2001 to 2003.

In contrast, for equity, you do have a real choice: do you want to measure equity in terms of Equity”Book or Market
book value or market value? In IBM™s case,

2001 2002 2003
Equity, Book Value, in billions $23.488 $22.782 $27.864
Equity, Market Value, in billions $208.437 $133.484 $157.047

Again, both book-value and market-value based equity ratios are in common use. Yet, as you
can see, when it comes to equity, your choice can make quite a di¬erence. Does IBM carry three
times as much debt as equity (per its book value of equity relative to total liabilities), or just
one-third to one-half as much debt as equity (per its market value)? Which is “correct”? This
book favors the market value. The book value is not a true value. It is the “plug-in” number
that equalizes both sides of the balance sheet. In fact, despite equity™s limited liability, it can
even be negative. If you look deeper, you will discover that the book value of equity is heavily
linked to past retained earnings and depreciation, and not to the value of future opportunities.
Therefore, our book focuses more on the market value of equity”but be aware that this is a
judgment call that not everyone shares.
Finally, you can quote debt ratios either relative to the equity value, as in a debt-equity ratio, or Debt-Equity or
Debt-Asset Ratios? And
relative to the entire ¬rm value, as in a debt-asset ratio. Assets are the sum of total liabilities
what assets?
and total shareholders™ equity. The assets on the balance sheet rely on the accounting book
value of equity. Thus, it is also common to compute a market value of the ¬rm™s assets by
adding the book value of total liabilities to the market value of equity.
The bottom of Table 24.1 shows the e¬ects of some of these choices for IBM. Clearly, in con- Some examples.
sidering whether a ¬rm carries a lot or just a little debt, it matters which de¬nition you choose.
If you compute a market-based liabilities-to-equity ratio, IBM ¬rst saw its debt-equity ratio
increase from
$66, 855 (24.1)
D/E 2001 = ≈ 32%
$208, 437
$73, 702 (24.2)
D/E 2002 = ≈ 55% ,
$133, 484
and then decrease back to
$76, 593 (24.3)
D/E 2003 = ≈ 49% .
$157, 047
In 2003, you could have also quoted as low a debt ratio as 15%, if you worked with the ¬nancial
debt/equity ratio. And, if you instead computed a book based measure, you could have instead
$76, 593 (24.4)
D/BE 2003 = ≈ 275% .
$27, 864

Now look again at the ingredients that go into the liability-based debt ratios. Table 24.1 shows Ingredients.
that debt-equity ratio changes themselves are also multidimensional”all sorts of debt and
equity changes participated. To understand the evolution of debt ratios, you need to realize the
role of these ingredients. Our next goal is to look over each of the participating components
one by one to get a feel for how changes in ¬rms™ debt ratios come about. This additional
information is usually found in the footnotes accompanying ¬nancial statements, so we have
to rely on the footnotes from IBM™s 2002 and 2003 annual reports. If you wish to read any IBM
historical ¬nancials, you can ¬nd them at www.ibm.com/annualreport/. It is not important for
you to understand every little detail”IBM is just one company, and every company looks a little
di¬erent. Your goal is to follow the basics, and be able to look up what else you might want to
¬le=capclinical.tex: LP
624 Chapter 24. Clinical Observations About Capital Structure .

Table 24.1. Major Components of Debt and Equity for IBM, 2001“2003

2001 2003
Long-Term Debt $15,963 $19,986 $16,986 see Table 24.2

+ $4, 023 ’$3, 000
Short-Term Liabilities $35,119 $37,900 see Table 24.3
’$569 +$3, 350
includes Short-Term Financial Debt $11,188 $6,031 $6,646
’$5, 157 +$615
Pension Liabilities $10,308 $13,215 $14,251
+$2, 907 +$1, 036
Other Liabilities $5,465 $7,456 see Table 24.4
+$330 +$1, 175
Minority Interest ” none
Negative Goodwill ” none
Total Debt $66,855 $73,702 $76,593
+$9, 003 +$2, 891
Financial Debt $27,151 $26,017 $23,632
+$1, 134 +$2, 385

2001 2003
Stockholder™s Equity (BV) $23,448 $22,782 $27,864
’$666 +$5, 082
Total Issued Shares 1920.96 1937.39

“ Treasury Shares = “ 198.59 “ 242.88

Number of Shares 1723.19 1722.38 1694.51
’0.81 ’27 .87
Price/Share $120.96 $77.50 $92.68
’$43.46 +$15.18
’ Market Value $208,437 $133,484 $157,047
’$74, 953 +$23, 563

Total Firm Value
’ Book Value of Assets $88,147 $96,484 $104,457
+$8, 337 +$7 , 973
’ Market Value of Assets $273,136 $207,186 $233,640
’$65, 950 +$26, 454

Market-Value Based Debt Ratios
2001 2002 2003
Total Liabilities-Equity Ratio (MV) 0.32 0.55 0.49
+0.24 ’0.06
Total Liabilities-Total Assets Ratio (MV) 0.24 0.35 0.33
+0.11 ’0.02

Financial Debt-Equity Ratio (MV) 0.13 0.19 0.15
+0.06 ’0.04
¬le=capclinical.tex: RP
Section 24·2. IBM™s Debt.

24·2. IBM™s Debt

Liabilities are usually more colorful than equity, and IBM™s debt is no exception. A glance at Let™s look at the four
non-zero components of
Table 24.1 tells you that there are four sources contributing to IBM™s debt: long-term debt,
IBM™s liabilities.
short-term (or current) debt, pension liabilities, and other liabilities. There are two smaller
components: minority interest of our business owned by third parties (which is therefore al-
most like equity) and negative goodwill (related to an accounting discount at which IBM might
have purchased other companies). These two rarely play a large role, and they did not play an
important role for IBM, either.

24·2.A. Long-Term Debt

IBM is a large Fortune-100 company. Many such large companies have a myriad of publicly Long-Term debt
increased and then
traded long-term bonds outstanding. (Small ¬rms tend to rely more on bank debt.) Table 24.2
decreased, mostly driven
shows how IBM™s long-term debt ¬rst increased by $4.023 billion and then decreased by $3.0 by IBM™s notes.

Straight Bonds The top part of Table 24.2 are IBM™s straight long-term bonds (debentures).
(Note how one of IBM™s bonds has 90 years remaining to maturity!) These bonds seem not
to have either an active call feature (or IBM would surely have retired its 8.375% bond due
in 2019), or a sinking-fund provision (because in most of these, the outstanding principal
remained constant from 2001 to 2003). The only bond on which IBM retired any principal
was its 6.5% bond, due in 2028. As to new debt, on October 1, 2003, IBM issued a 5.875%
bond for $600 million dollars at 97.65 (i.e., below par”given the 5.875% coupon, this
bond was a discount bond).
Net in net, IBM did not change its straight bond borrowing from 2001 to 2002, and in-
creased it by only $219 million from 2002 to 2003.

Notes There was more ¬nancing action in IBM™s notes. Notes are very similar to bonds. The
di¬erence is that they are usually not issued in one big, underwritten chunk, but instead
are sold into the market as the ¬rm wants to raise more money”in other words, “o¬ the
shelf.” Notes are also often callable. Together, these two features make it easy to expand
or contract long-term debt, as needed.
IBM increased its medium-term notes by $3.5 billion from 2001 to 2002, and then de-
creased it by $2.4 billion from 2002 to 2003. (Relatively lower interest rates may help
explain some of the shift from longer-term notes into medium-term notes in 2002, but
not in 2003. In any case, the two do not exactly o¬set one another.)
Net in net, $3.5 billion of IBM™s $4 billion increase in long-term borrowing in 2002 and
$2.4 billion of IBM™s $3 billion decrease came from its medium-term notes. Other notes
were used to o¬set some of this, but, nevertheless, IBM seems to have mostly used its
notes program to expand or contract its long-term borrowing needs.
Hybrid Borrowing: Note also that IBM had one hybrid debt-equity instrument”a con-
vertible 3.43% note. It was issued by IBM to the partners of Price-Waterhouse-Coopers
Consulting (PwCC), a ¬rm that IBM acquired in late 2002.

Foreign Borrowing Over this time period, IBM repurchased a good deal of Euro debt. The
Euro appreciated from about 1.1e/$ in 2001 to about 0.9e/$ by 2002, but the decline in
the value of IBM™s Euro debt obligations is even steeper. IBM also reduced its Canadian
debt, and eliminated its Swiss Franc debt. In contrast, IBM continues to rely heavily on
¬nancing in Yen. Nevertheless, you cannot interpret these changes as speculation on
exchange rates, because IBM described elsewhere in its ¬nancials how it hedges some of
its currency risk. Moreover, not only IBM™s obligations, but also many of its assets were
overseas, so the net exposure of IBM to foreign currency is not easy to determine.

Fair Value Adjustment Usually, long-term debt is carried at historical value, not market value.
However, some of IBM™s debt was “hedged””that is, IBM had ¬nancial contracts that would
¬le=capclinical.tex: LP
626 Chapter 24. Clinical Observations About Capital Structure .

change opposite in value to those of some or all of its bonds. From 2001 to 2003, short-
term interest rates fell, while long-term interest rates remained around 5%. The fair value
adjustment re¬‚ects the change in value of the hedged bonds. (Somewhere else on IBM™s
balance sheet will be an opposite item”an asset measuring the value change experienced
by the hedge instruments.)

Current Maturities Some of IBM™s long-term debt became current (had less than one year left
before coming due), and therefore was reclassi¬ed. This could account for about $1.1
billion less in long-term borrowing in 2002, and $543 million in 2003.

In sum, there are many long-term ¬nancing instruments that can play a role. In IBM™s case, the
most important factor in¬‚uencing changes in borrowing was its expansion and contraction of
its medium-term notes program.

24·2.B. Current Liabilities

Table 24.3 breaks out current (i.e., short-term) liabilities, which are due to be paid within one
Note the many different
short-term obligations! year. The CFO has most in¬‚uence over short-term debt”at least commercial paper and short-
term loan borrowing. We also see the long-term debt that fell into short-term debt. The remain-
ing liabilities are mostly incurred in the course of the ¬rm™s operations.
We can see that IBM actively reduced its short-term borrowing from 2001 to 2002, and then
expanded it from 2002 to 2003.

24·2.C. Other Liabilities

Table 24.4 shows a set of other obligations that can have an impact on the amount of corporate
debt. For IBM, only changes in restructuring actions really mattered in 2002. In 2003, however,
both changes in IBM™s deferred taxes and deferred income played important roles.

24·2.D. Other Observations and Discussion

Table 24.1 also shows that just under 20% of IBM™s obligations in 2003 are pension obligations to
Pension obligations are
very important for ¬rms its more than 300,000 current employees (and also former employees). This is a very important
with many
part of IBM™s liabilities, but it would be di¬cult to discuss in less than a chapter in itself”and
employees”almost as
it would lead us far away from ¬nance.
important as long-term
debt for IBM!
Interestingly, Table 24.1 can also tell us how IBM™s shifted its obligations from short-term debt
The time dimension of
IBM™s obligations, and into medium- and long-term debt in 2002, and then reversed (or no longer continued) this
the prevailing yield
trend in 2003. This can be seen both in IBM™s arrangement of long-term vs. current liabilities
(Table 24.1), and within its long-term liabilities, in its arrangement between long-term notes
and medium-term notes (Table 24.2). However, the passing of time itself makes outstanding
obligations shorter-term, so we might like to know how its ¬nancial obligations for each year
developed. If we dig deeper into the ¬nancial statement footnotes, we can ¬nd these, too:

2001 2002 2003 2004 2005 2006 2007 2008 2009
← ←
As of 2001 $11,188 $5,186 $3,106 $1,501 $1,904 $2,261 $6,471

As of 2002 $6,031 $3,949 $3,613 $1,670 $2,705 $846+$9,940
As of 2003 $6,646 $4,072 $3,113 $2,760 $1,289+$225+$7,942

This shows that IBM changed its capital structure dynamically (e.g., it always ¬nanced itself
with some short-term debt, so the ¬rst-year term is large), but for any given year further out, a
static shift (i.e., in a given year, like 2006) is not as obvious. When thinking about obligations,
IBM™s CFO did not operate in a vacuum, but was probably very concerned with the yield-curve.
Relative to 2000, short-term and medium-term interest rates had dropped signi¬cantly, but
¬le=capclinical.tex: RP
Section 24·2. IBM™s Debt.

Table 24.2. IBM™s Long-Term Liabilities

Maturities 2001 2002 2003
At Dec 31

U.S. Dollars:
5.875% 2032 “ “ $600
6.22% 2027 $500 $500 $500
6.5% 2028 $700 $700
7.0% 2025 $600 $600 $600
7.0% 2045 $150 $150 $150
7.125% 2096 $850 $850 $850
7.5% 2013 $550 $550 $550
8.375% 2019 $750 $750 $750
$4,100 $4,100 $4,319
±$0 +$219

3.43% conv.notes— 2007 “ $328 $309
Notes, 6%, 5.9%‡ 2003-32 $2,772 $2,130 $3,034
Med Term Notes, 4%, 3.7%‡ 2003-18 $3,620 $7,113 $4,690
+$3, 493 ’$2, 423

Other: 4.9%, 4.0%‡ 2003-09 $828 $610 $508
$11,320 $14,281 $12,860
+$2, 961 ’$1, 421

Other currencies†
Euros (5.4%, 5.3%)‡ 2003-09 $3,042 $2,111 $1,174
Yen (1.0%, 1.1%)‡ 2003-15 $4,749 $4,976 $4,363
Canadian (5.8%, 5.8%)‡ 2003-11 $441 $445 $201
Swiss (4.0%, 4.0%)‡ 2003 $151 $180 “

Other (6.6%, 6.0%) 2003-14 $726 $730 $770
$20,429 $22,723 $19,368
+$2, 294 ’$3, 355

Unamort. (Prem)/Disc $47 “$1 $15
SFAS #133 Fair Value Adj.† $396 $978 $806
$20,778 $23,702 $20,159
+$2, 924 ’$3, 543

Less current maturities $4,815 $3,716 $3,173
Total $15,963 $19,986 $16,986
+$4, 023 ’$3, 000

: The appendix contains the Standard&Poor™s Bond Report on this particular issue. — : These convertibles notes

were issued in the 2002 acquisition of PwCC to PwCC partners, and some began converting into equity in 2003.

The ¬rst interest rate is the average from 2001 to 2002, the second from 2002 to 2003. : This item “marks to
market” the value of the debt instruments when interest rates change. The IBM footnotes footnote this further as
In accordance with the requirements of SFAS No. 133, the portion of the company™s ¬xed rate debt obligations that
is hedged is re¬‚ected in the Consolidated Statement of Financial Position as an amount equal to the sum of the debt™s
carrying value plus a SFAS No. 133 fair value adjustment representing changes recorded in the fair value of the
hedged debt obligations attributable to movements in market interest rates and applicable foreign currency exchange
¬le=capclinical.tex: LP
628 Chapter 24. Clinical Observations About Capital Structure .

Table 24.3. IBM™s Current (Short-Term) Liabilities

2001 2003

Short-term Debt $11,188 $6,031 $6,646
’$5, 157 +$615

Commercial Paper $4,809 $1,302 $2,349

’$3, 507 +$1047

+ Short-Term Loans $1,564 $1,013 $1,124

’$551 +$111

+ Long-Term Debt, Current $4,815 $3,716 $3,173

’$1, 099 ’$543

Taxes $4,644 $5,476 $5,475
+$832 ’$1

Accounts Payable $7,047 $7,630 $8,460
+$583 +$830

Comp and Bene¬ts $3,796 $3,724 $3,671
’$72 ’$53

Deferred Income $4,223 $6,492

+$1, 053 +$1, 546

Other Accrued Liabilities $4,221 $6,413 $7,156
+$2, 192 +$743

Total Current $35,119 $37,900

’$569 +$3, 680

This revision shifted $330 from deferred income into other liabilities, which can be seen in Table 24.4.

long-term rates were somewhat sluggish. Here is how the economy-wide rates changed over
this period.

Maturity 2000 2001 2002 2003
Treasury, Short-Term 1 month 2.47% 1.63% 1.02%
Treasury, Medium-Term 3 year 6.22% 4.09% 3.10% 2.10%
Treasury, Long-Term 20 year 6.23% 5.63% 5.43% 4.96%
Corporate, Short-Term 1 Month 6.3% 3.8% 1.7% 1.1%
Aaa Bonds Medium Term 7.6% 7.1% 6.5% 5.7%

Finally, the ¬nancials also tell a little bit about IBM™s interest payments and unused credit lines.
Some more interesting

2001 2002 2003
Interest Paid and Accrued $1,235 $815 $663
Unused Credit Lines $16,121 $16,934 $15,883
¬le=capclinical.tex: RP
Section 24·2. IBM™s Debt.

Table 24.4. IBM™s Other Liabilities

2001 2003

Deferred Taxes $1,485 $1,450 $1,834
’$35 +$384

Deferred Income $1,145 $1,842


. 31
( 39)