. 34
( 39)


Section A. Appendix: A List of Some Recent Empirical Capital-Structure Related Publications.

Solutions and Exercises

1. Debt issuing and repurchasing activity, and the direct in¬‚uence of stock returns can each account for about
40% of the variation in debt-equity ratios. Long-Term debt net issuing can account for about 30%, and short-
term debt and equity net issuing can account for about 15%.
2. Non M&A related seasoned equity issuing activity is trivial in magnitude.
3. On the contrary. It may even mean that there is a lot of money left on the table by managers that have not
optimized their capital structures.

4. See Page 667.
5. It is pretty modest.
6. Firms with high tax obligations usually have low equity ratios, because they were highly pro¬table. These
¬rms have low debt ratios.
7. No”the theories tell us what should matter. There are a number of explanations why they may not translate
into observed corporate behavior.
8. The capital gains can be realized by investors who have little capital gains tax.
9. Executives and insiders may often not tender into a repurchase, but will enjoy the relatively higher share
price from a repurchase through executive compensation that is linked to the share price. Dividends tend to
be more regular than share repurchases. Some retail investors like dividends.

10. Taxes, credit ratings, ¬nancial ¬‚exibility, and earnings dilution.
11. No. Managers are con¬‚icted. They do not maximize shareholder wealth, but their own welfare.

(All answers should be treated as suspect. They have only been sketched, and not been checked.)
¬le=capempirical.tex: LP
682 Chapter 26. How Have Firms™ Capital Structures Evolved?.
Investment Banking

last ¬le change: Feb 23, 2006 (17:14h)

last major edit: Apr 2005

This chapter looks at the role of the investment bankers. The subject of investment banking
is important enough to deserve its own chapter”even if it is only a brief one. After all, in-
vestment banks are not only plum employers of ¬nance graduates, but they are also the most
important intermediaries through which corporations tap the capital markets. We will focus on
two important functions of investment banks: Facilitating the underwriting of securities and
advising ¬rms on mergers and acquisitions.

¬le=ibanking.tex: LP
684 Chapter 27. Investment Banking.

27·1. Investment Bankers

You can think of an investment banker as pretty much the same thing as an ordinary consumer
What an investment
bank is. banker. The primary di¬erence is that investment bankers advise corporations rather than in-
dividuals, and therefore operate on a larger scale. Like consumer bankers, investment bankers
provide the following services:

• lend capital to corporations;

• act as agents on behalf of other capital providers;

• orchestrate the legal and bureaucratic aspects of the capital raising process;

• o¬er investment advice”solicited and unsolicited.

The last function is especially important when a ¬rm seeks to undertake a large investment,
such as a merger or acquisition. We will therefore discuss this at length in the next section.

27·1.A. Underwriting Functions

If you go back to Table 25.1, you will see that with the exception of the direct role of stock
Underwriters help sell
corporate securities. returns and the payment of dividends, investment banks facilitate many of the ¬nancing activi-
ties that corporations undertake. In particular, the underwriting of new debt and equity issues
is one of the main businesses of many investment banks. Virtually all o¬erings of securities
for exchange-traded ¬rms are underwritten these days. The term underwriter originally came
from the guarantee of the issuing proceeds, equivalent to the fact that an insurance company
would underwrite a policy. These days, this guarantee is given only on the morning of the
o¬ering, when the underwriter practically knows for certain at what price investors are willing
to purchase shares. If it looks as if the issue cannot be sold, the underwriter will demand a
lower price and/or refuse to bring the issue to market altogether. So, the actual underwriting
guarantee itself is no longer as important as it was historically, when it could take weeks to
¬nd investors.
Instead, the main function of underwriters today is to provide the legal expertise in carrying
Underwriters help sell
corporate securities. through the process (issue origination); to make sure that the securities get placed, often with
speci¬c investors such as large institutional ones (issue placement); and to signal con¬dence
in the ¬rm by putting their own reputations on the line. Underwriters can help in many ways
throughout the process. For example, many investment banks have large brokerage arms, and
their analysts can spread “positive hype” through optimistic analyst reports for their invest-
ment banking corporate clients. This presumably increases the demand for investment in the
company, and is thus good for selling more shares and debt in the future.

Anecdote: The Analyst Recommends: Buy!
The number of analysts™ buy recommendations outnumbers the number of sell recommendations by a ratio
of about 5:1; when limited to strong buy and strong sell recommendations, this ratio changes to over 10:1.
The primary reason for this imbalance is a con¬‚ict of interest. Most brokerage ¬rms and by extension their
analysts are owned by investment banks. (They are even called “sell-side” analysts, even though their “advice”
goes to investors!) The investment banks are well aware that a sell recommendation is likely to induce the
targeted ¬rms not only to exclude the particular analyst from obtaining further information about the ¬rm, but
also to induce the targeted ¬rm to select a di¬erent underwriter. Therefore, the investment banks discourage
their analysts subtly and not so subtly from issuing sell recommendations. Although this analyst bias was always
widely recognized by professional investors, it had received scant attention in the press and little recognition
by small investors”until 2001, when it suddenly became a public scandal. It is still somewhat of a mystery why
then, but not before.
In April 2003, ten of the largest investment banks settled a lawsuit by setting aside funds for making indepen-
dent research available to brokerage clients and promising a separation of their brokerage analysis from their
investment banking functions. It is not yet clear how e¬ective these reforms have been.
¬le=ibanking.tex: RP
Section 27·1. Investment Bankers.

Nevertheless, it would be naïve for CFOs to consider underwriters as uncon¬‚icted agents on Underwriters are also
con¬‚icted agents.
behalf of their clients. Investment banks make their money from ¬nancing activity and thus
will push for the ¬rm to engage in activity”even if it is value-decreasing. They also sometimes
structure transactions in such a way that the valuation is not easy for the CFO to understand.
This can even occasionally allow the investment bank to obtain claims on the company at
below fair market pricing. (For example, many bonds include a strippable warrant kicker to
lower interest rates, and many CFOs may not understand as well as the bank does just how
much value they are giving away.) Of course, a good investment bank can work hard and create
value for its clients”but it is wise for the client to remain cognizant of the con¬‚icts involved.

27·1.B. The Major Underwriters

Table 27.1. Largest Underwriters in the United States in 2004

Underwriting M&A Advising IPOs
$ # $ # $ #
(in alphabetical order)
Bank of America $204 780 $73 109 $1.5 16
CSFB $362 1,359 $201 271 $3.6 23
Citigroup (Salomon-Smith-Barney) $534 1,892 $485 377 $3.6 19
Deutsche Bank $335 1,299 $247 218 - -
Goldman Sachs $286 855 $577 336 $7.1 29
JP Morgan $386 1,492 $511 396 $4.0 25
Lazard - - $230 207 - -
Lehman Bros $370 1,292 $308 175 $2.4 20
Merrill Lynch $374 1,564 $381 298 $4.5 31
Morgan Stanley $414 1,334 $381 298 $7.3 21
Rothschild - - $232 269 - -
UBS $300 1,175 $219 289 $2.3 16

Firms are in alphabetical order. All dollar ¬gures are in billions. A dash means the information is not available or
the ¬rm is not among the largest players. Underwriting amounts include all debt and equity o¬erings, corporate and
otherwise, for which the underwriter ran the book (was the lead underwriter). IPO underwriting is broken out in the
right-side columns. IPOs are small in absolute dollar amounts, but disproportionally rich in commissions and risk.
With the exception of Lazard and Rothschild, who are primarily M&A advisors, the remaining largest investment
bankers are also large underwriters.
Source: www.thomson.com (Thomson Financial).

Table 27.1 gives you an idea of who the big investment banks are. This market has an inter- How the large came to
be large.
esting history. Until November 1999, the Glass-Steagall Act of 1933 and a subsequent 1956
Act had prohibited interstate banking (keeping investment banks outside states with large
corporate presence, such as California, New York, Illinois, and Massachusetts, relatively small),
and the mixing of retail/commercial and investment banking. Glass-Steagall therefore made
it impossible for large consumer banks, such as Citibank or Chase Manhattan Bank, from
e¬ectively competing in the investment banking space. Many other countries never had this
distinction”they just had “banks” that performed both consumer/commercial and investment
banking. Therefore, the United States was unique in fostering a large number of relatively small
investment banks. Beginning in the 1980s, states allowed interstate banking, which was com-
pleted around 1994. The separation between ordinary and investment banking disappeared
with the repeal of Glass-Steagall in 1999, and the investment banking sector rapidly began
to consolidate. For example, in 1998, Citibank and Travelers Insurance group had merged to
become Citigroup. In the same year, Smith Barney purchased Salomon Brothers to become
Salomon Smith Barney. One year later, with Glass-Steagull repealed, Citigroup then purchased
Salomon Smith Barney, so the ¬ve formerly independent ¬nancial services providers are now
all just parts of one large ¬nancial conglomerate. Similarly, Chase Manhattan purchased J.P.
¬le=ibanking.tex: LP
686 Chapter 27. Investment Banking.

Morgan in 2000, and merged with Bank One Corporation (a large credit card issuer) in 2004.
CSFB is the combination of Credit Suisse, a very large Swiss bank, and First Boston, an old
investment bank. Lehman Brothers and Goldman Sachs still remain independent”and there
are even some players that remain independent in just one sub-¬eld of investment banking
(speci¬cally, Lazard and Rothschild are just M&A advisors).
Much competition is no longer just in the investment banking ¬eld or even just across a whole
Like the rest of the
world, we now have range of ¬nancial services, but even transcends national borders. The same cast of bankers
competes not only in the United States, but also in the bigger global market. Even though
the United States remains the biggest underwriting market (debt plus equity), it was followed
closely by Europe (with the Middle East and Africa), Asia (without Australia and Japan), Japan,
Australia, and Latin America. In billions of dollars, the 2004 market was

United Europe Asia Japan Australia Latin
Middle East and (excl. Japan)
States America

$173 166 $69 $61 $17 $4

Moreover, Europe is now the biggest equity o¬ering market, with the United States being the
biggest debt o¬ering market. But, again, these are arti¬cial distinctions: investment banking
has truly gone global, with many foreign companies issuing securities in the United States
and many domestic companies issuing securities in Europe or Japan. About 70% of securities
issues in the United States are sold to the public, and 30% are sold to private parties (without
SEC involvement).
Let us now enlarge our perspective and consider commercial banking in general. Investment
UK and US banks are the
top competitors. banking is just one of the many businesses commercial banks engage in. The general trend
has been for consolidation of many activities”few investment banks remain standalone op-
erations. Table 27.2 shows the rankings of global banks by assets under management as of
2004. The table shows some remarkable di¬erences between the more market-oriented banks
in the United Kingdom and United States on the one hand, and the more institutionally ori-
ented banks in Europe and Japan on the other hand. For example, Bank of America is at the
bottom of this list with “only” $736 billion under management. Mizuho of Japan is at the top
of this list with $1.285 trillion under management. Yet Bank of America is worth more than
three times as much as Mizuho! There is much evidence that foreign banks have been largely
unable to translate their ¬nancial reserves (a natural scale advantage) into pro¬tability equal
to their American and British counterparts. Indeed, many of them are struggling against the
more nimble competition.
Solve Now!
Q 27.1 How important is the guarantee of securities placement success that underwriters provide
their clients?

Q 27.2 What are the main functions of underwriters today?

Q 27.3 How are the interests of investment banks di¬erent from those of their clients (investors
and ¬rms)?

Q 27.4 How good and unbiased are brokerage buy recommendations?

Q 27.5 Name some of the largest underwriters in the United States today.

Q 27.6 In relative terms, how important is the American market compared to the European

Q 27.7 Name some of the largest global banks today. Does it matter whether the criterion is
market value or book value of assets under management?
¬le=ibanking.tex: RP
Section 27·2. The Underwriting Process.

Table 27.2. The Largest 15 Global Banks in 2003, by Assets under Management

Rank Bank Country Assets Capital Value
1 Mizuho Japan $1,285 $38 $50
2 Citigroup USA $1,264 $69 $243
3 UBS Switzerland $1,120 $24 $87
4 Credit Agricole France $1,105 $55 $36
5 HSBC Holdings UK $1,105 $55 $164
6 Deutsche Bank Germany $1,014 $27 $43
7 BNP Paribas France $989 $32 $55
8 Mitsubishi Japan $975 $47 $57
9 Sumitomo Japan $950 $34 $41
10 Royal Bank, Scotland UK $806 $35 $87
11 Barclays Bank UK $791 $27 $57
12 Credit Suisse Group (CSFB) Switzerland $778 n/a $42
13 JP Morgan Chase USA $771 $43 $78
14 UFJ Holdings Japan $754 $21 n/a
15 Bank of America USA $736 $44 $171

All numbers are in billions of U.S. dollars. Assets are assets under management. Capital is “tier one capital” (also
called core equity), which is common stock, disclosed reserves, and retained earnings. Although based on book
value and therefore unreliable, it is the most common regulatory de¬nition for bank capitalization. Value is market
value as of June 17, 2004. The table shows that American and British ¬rms had relatively high capitalizations.

Source: The Banker magazine, July 2, 2004.

27·2. The Underwriting Process

We now turn to the question: How much does it cost to issue securities?

27·2.A. Direct Issuing Costs

Table 27.3 describes cost data from 1990 to 1994. (Although no data are available for issuing Typical real-world fees.
costs after 1994, it is unlikely that much has changed in relative terms since then.) The table
shows that selling equity shares is more expensive than selling bonds. For example, placing
a $100 million bond o¬ering may cost the ¬rm direct expenses of $2.3 million in total, while
placing a $100 million equity o¬ering may cost $4.2 million. From the perspective of the ¬rm,
there seems to be a price for issuing “capital at risk.” The more risk for sale, the higher the
underwriting costs. There is usually more value at risk in a $1 million equity o¬ering than in a
$10 million bond o¬ering. In this context, you can often think of a typical $10 million corporate
bond o¬ering as the combination of, say, a $9.5 million risk-free Treasury bond (with very low
commissions, below even what corporate bonds charge) and $0.5 million in highly levered
equity (with high commission). Therefore, the cost of issuing corporate debt lies between the
cost of issuing Treasury-like securities and the cost of issuing risky equity.
In turn, underwriters have their reasons for charging more for placing riskier securities. First, Why underwriters
charge for risk.
investors can be found a lot more easily if the securities are safer. In the extreme, safe bond
issues are almost substitutes for Treasury bonds, and so investors are not very concerned about
risk analysis and so are easy to engage. Second, the underwriter carries some of the risk of
the securities he places with his own reputation capital. For example, when an underwriter
takes a ¬rm public in an IPO, he partly carries the risk that the ¬rm will go bankrupt later on,
which will not play well with investors that the bank solicited. After a couple of such repeats, the
underwriter would probably no longer be able to ¬nd IPO investors. Therefore, when companies
¬le=ibanking.tex: LP
688 Chapter 27. Investment Banking.

Table 27.3. Typical U.S. Fees, 1990-1994, in Percent.

Initial Convertible Plain
Public Bond Bond
Proceeds O¬erings O¬erings O¬erings O¬erings
($ millions) UWC TC UWC TC UWC TC UWC TC
2“10 337 9.1 17.0 167 7.7 13.3 4 6.1 8.8 32 2.1 4.4
10“20 389 7.2 11.6 310 6.2 8.7 14 5.5 8.7 78 1.4 2.8
20“40 533 7.0 9.7 425 5.6 6.9 18 4.2 6.1 89 1.5 2.4
40“60 215 7.0 8.7 261 5.1 5.9 28 3.3 4.3 90 0.7 1.3
60“80 79 6.7 8.2 143 4.6 5.2 47 2.6 3.2 92 1.8 2.3
80“100 51 6.5 7.9 71 4.3 4.7 13 2.4 3.0 112 1.6 2.2
100“200 106 6.0 7.1 152 3.9 4.2 57 2.3 2.8 409 1.8 2.3
200“500 47 5.7 6.5 55 3.3 3.5 27 2.0 2.2 170 1.8 2.2
500“ 10 5.2 5.7 9 3.0 3.2 3 2.0 2.1 20 1.4 1.6
11.0 7.1 3.8 2.2
All 1,767 7.3 1,593 5.4 211 2.9 1,092 1.6

Source: Lee-Lochhead-Ritter, 1996. N is the number of observations. UWC is underwriter compensation (in percent
of proceeds). TC is total costs, which includes registration fees, printing fees, and legal and auditing costs (in percent
of proceeds).

¬rst sell shares in an initial public o¬ering (IPO), which is the most risky investment banking
business around, the costs are usually highest”a fact that Table 27.3 shows quite nicely. As
a sidenote, IPO commissions have become considerably more uniform after 1994. Almost all
underwriters are now charging exactly 7.0% in commission, a fact that has made some observers
wonder about how competitive the underwriting market for IPOs truly is. Although there are
several dozen underwriters, it could be that the market is segmented enough along the size and
industry dimension, so that each IPO really has only one to three natural choices from whom
an underwriter can reasonably be selected.
The costs listed in Table 27.3 are, of course, not even complete as far as direct costs are con-
Management Fees!
cerned. Importantly, they do not account for the time and focus that management spends on
the issuing process, which could otherwise have been spent more productively (an opportunity
cost). The e¬ort is relatively more modest in safer o¬erings”for IPOs, it is a very lengthy and
time-consuming task. Similarly, we have no evidence as to how expensive any time delay in
funding would be on the values of projects. These two costs are conceivably just as important,
but we cannot assess them because we have no data on the costs of management time and pro-
ject delay. Finally, there are the indirect costs and bene¬ts that the revised capital structure
itself creates”the subject of our earlier chapters and subsection 27·2.C.
There is one additional direct cost to issuing debt that is worthwhile breaking out. We have
Bond Rating Agencies.
already mentioned bond rating agencies in Section 6·2.C on Page 119. Issuers can pay Moody™s,
Standard&Poors, or Fitch to rate their bonds. This typically costs $5,000 to $25,000 per bond
issue. Having a public bond rating helps potential investors gauge the risk. Indeed, many
institutions are prohibited from buying any unrated bonds, making ratings a necessity for
many large bond o¬erings. Only the largest and most stable ¬rms can issue investment grade
bonds. All other ¬rms can only issue high-yield bonds, that is, bonds rated BB or worse (see
also Table 6.1 on Page 121). To get a better impression of issuing activity, please browse the
issuing calendar in the Wall Street Journal, as well as a Moody™s or S&P Bond Manual in your
local library. (The Moody™s descriptions are now published by Mergent, a sister company of
¬le=ibanking.tex: RP
Section 27·2. The Underwriting Process.

27·2.B. Underwriter Selection

Much of our interest in underwriting is from the perspective of ¬rms wanting to raise capital. So, Firms are often selected
by history. Industry
how do ¬rms select investment bankers? Most of the time, by simple inertia! Firms tend to go
expertise is the next
with the investment banker that they have always done business with. It is often only when they most important factor.
consider switching”which itself usually occurs after a ¬rm has “outgrown” in capitalization
the reputation and size of its original underwriter”that they undertake a comparative analysis.
If they do such an analysis, an important factor in the selection process is “industry expertise.”
Such expertise can help the underwriter navigate the process more smoothly, communicate
and understand better the concerns of top management, connect the ¬rm to the right potential
investors, and o¬er the services of specialized analysts who can help hype the o¬ering more
to potential retail investors.
However, there is a puzzle. There is empirical evidence suggesting that it is cheaper for a Agency issues between
managers and the ¬rm
corporation to ask several investment banks to compete for the underwriting of an issue”but
may also play a role.
most ¬rms just hire one investment bank (and typically their old investment bank) and stick
with it. (The exception are utilities, which are obliged by regulation to bid out their capital
raising activities.) Why this inertia? One view is that ¬rms are willing to pay more because
hired investment banks provide better service along other dimensions than just underwriting
at lowest cost. For example, it may take less management time if the existing underwriter
is already well informed about the company through previous interactions. Another view is
that ¬rms select their underwriters based on such considerations as personal friendships and
convenience. The most cynical view (see Chapter 28 on corporate governance) is that executives
like underwriters who “wine and dine” them, give them better and cheaper personal banking
services (such as allocations to shares in initial public o¬erings), and provide a placement
network for executives if they want or have to move. After all, an investment banker is not
likely to recommend to bigger and better companies those CFO executives who they barely
know and who have minimized the investment bank™s pro¬ts.

27·2.C. Sum-Total Issuing Costs ” The Financial Market Reaction

We have already brie¬‚y discussed the ¬nancial market reactions to issuing activity in the previ- Capital market reactions
are a better measure of
ous chapter (with more depth in our web chapter), but it is important to bring these reactions
the overall net cost/gain
into our discussion of the costs of issuing here. The average company dropped by about 1.5% of an issue.
when it announced a new equity issue, which was about 15% in dilution. In a sense, this may be
a better measure of the cost of issuing than the direct fees we measured in Table 27.3, because
it includes the direct fee. For example, start with a perfect market in which a $100 million ¬rm
raises $50 million and pays the underwriter $30 million in commissions. It is the old share-
holders who pay the $30 million. The new shareholders participate only if they can buy at
the appropriate price. Because the ¬rm will be worth $120 million, new shareholders demand
41.7% of the ¬rm. Old shareholders experience an announcement price drop from 100% · $100
million to 58.3% · $120 million = $70 million. In an e¬cient market, this value drop must

Anecdote: Legal Monopolies: Bond Ratings
Prior to 2003, federal securities laws had recognized just three “nationally recognized statistical rating agencies”
(NRSRO): Moody™s, Standard&Poor™s, and Fitch. (In 2003, the SEC added Dominion; in 2005, it added A.M. Best.)
In the second half of the twentieth century, the SEC began to rely on ratings to determine what sort of securities
certain regulated ¬nancial institutions could own. The raters had not always enjoyed such privileged status. At
the beginning of the twentieth century, they were simply investment service agencies that provided investors
with research for a fee. In the 1970s, the revenue model changed, and Moody™s and S&P (by far the larger and
more important) began to charge issuers instead of investors.
In 1994, the Je¬erson County School District No R-1 of Colorado decided not to obtain a Moody™s ranking. To
their surprise, Moody™s decided to publish an unsolicited and unusually detailed “Special Comment,” anyway.
It was a negative rating that downgraded the school district, and interestingly it occurred on the day of the
pricing of the bond. Although Je¬erson County sued, a judge later ruled that Moody™s was protected by the
First Amendment™s freedom of speech clause.
This legal protection also helped the three major credit rating agencies in Enron™s case. They received substantial
fees from Enron. Interestingly, even when Enron was already trading at $3 per share and the market was aware
of its trouble, all three agencies still gave investment grade ratings to Enron™s debt.
¬le=ibanking.tex: LP
690 Chapter 27. Investment Banking.

occur in the instant of the announcement, not at the actual issuing. CFOs sometimes make the
mistake of adding up the direct cost and the dilution cost to arrive at a total cost of issuing, but
this example illustrates that the dilution (announcement drop) is not an additional cost, but a
total cost.
If we now were to observe that the value of outstanding equity had dropped from $100 million
Adding up costs and
bene¬ts. to $60 million instead of to $70 million, then the ¬rm must have lost another $10 million in
value through the issuing of the equity elsewhere. In contrast to the direct fees, we usually do
not immediately know the causes for the remaining dilution. It could be that existing owners
believe that the ¬rm gave away too much in features, or that it chose the wrong securities
features, or that the ¬rm or shareholders will now pay more in taxes, or that shareholders
learned the bad news that management was doing so poorly that the ¬rm needed to raise more
money. Actually, it is not one or the other, but the sum of all the possible value e¬ects. In
the end, the point is that the extra loss of $10 million is a cost just like the direct cost of $30
million paid to the underwriter. Of course, if the value of outstanding equity had dropped
from $100 million to $80 million, the issue must have cost the $30 million in commissions,
but created $10 million of value elsewhere. In the extreme, if the ¬rm value increased upon
the announcement from $100 million to $110 million (and we know that some ¬rms increase
in value upon the announcement of a new issue), we know that the issue cost $30 million in
underwriting fees, but created $40 million in value.


• A ¬rm that seeks to maximize shareholder value should minimize all costs of
issuing”whether underwriter related or deadweight costs (such as taxes)”
and maximize all value created by issuing.

• In an e¬cient and perfect market, the instant dilution is a measure of all the
costs and bene¬ts of an issue.

In real life, it is not so easy to just look at the announcement reaction. First, when managers
Real Life Dif¬culties
consider whether to issue, they have not announced it yet, and so they do not know the market™s
value reaction. Could they perhaps merely announce their intent, and then see what the value
consequence is? Unfortunately, if the market can anticipate that managers are just ¬‚oating a
trial balloon, it may not react at all. If the market response is a function of what it believes
managers will do, and if what managers will do is a function of what they believe the market
will do, then the blind may be leading the blind. The outcome could be anything. If the market
believes the managers will carry through an equity o¬ering, responds negatively, and managers
then cancel the issue, the net e¬ect is not as bad as carrying through with the bad issue”but
¬rms still end up worse o¬ than if they had never announced an issue to begin with.
But can we learn at least something from what other corporations have experienced in the
Learning from the
market response same situation? Yes, we can”and we have. In the previous chapter, we described the following
outcomes for publicly traded corporations:

• On average, when ¬rms raise more external capital, it is bad news and the stock price
drops. Conversely, when a ¬rm pays out capital, it is good news and the stock price rises.

• On average, when ¬rms replace equity with debt, keeping ¬rm size steady, the stock
price generally increases. Conversely, when these ¬rms replace the (remaining) debt with
equity, the stock price generally decreases.

We interpreted this evidence that investors seem to infer that management will waste the extra
money, or learn that the ¬rm can no longer produce as much money as they thought it could.
Our analysis was based on the following “anchor” estimates for the announcement price reac-
tions for three important and frequent capital markets events for publicly traded ¬rms in the
United States from 1980 to 2000:
¬le=ibanking.tex: RP
Section 27·2. The Underwriting Process.

• The two-day announcement price change for an equity issue (increasing ¬rm size and
decreasing debt-equity ratio) was a drop of about 1.5-2.0%, with a standard deviation of
about 6%.

• The equivalent announcement price change for the typical debt issue was about 0.2%, with
a standard deviation of about 3%.

• The equivalent announcement price change for a dividend announcement was a price
gain of about 0.25%, with a standard deviation of about 4%. The larger the dividend
announcement, the more positive the ¬nancial market reaction.

Again, these estimates include all costs and bene¬ts of issuing (or paying dividends), including
but not limited to underwriting and expert fees.
Unfortunately, the historical event studies also have considerable practical limitations when it Where our event study
knowledge is limited.
comes to our assessing the costs and bene¬ts of investment banking activity.

1. Disentangling Value E¬ects from Noise In our example, it was easy to compute that our
¬nancial market reaction was the sum of $30 million in underwriting fees, plus other
bene¬ts and costs. In the real world, stock prices are very noisy. A $100 billion company
may drop by as much as a $1 billion in value on a day on which it announces nothing.
If the ¬rm had happened to announce an issue, we would have erroneously concluded
that our issue created total costs of $1 billion. (Some of this noise can be reduced if we
knew the exact minute of the day when the issue announcement hit the ¬nancial market,
but this is usually not the case”if only because many announcements occur after trading
These problems are borne out by what we see in the data: tremendous heterogeneity in
¬nancial market responses. One ¬rm may experience a +5% response, while another ¬rm
may experience a ’5% response to the same action. It is not clear whether we can disen-
tangle the part of this heterogeneity caused by noise in stock returns from the valuation
e¬ect due to the issue.

2. Comparability A big assumption in learning from other companies™ responses is that our
company is just like these other companies. If you recall our Chapter 10 discussion
of comparables, how con¬dent are you in this assessment? Probably not very”every
company is di¬erent, and every day is di¬erent. Are you more like the corporation that
increased by 0.4% in value upon the announcement of a new equity issue two years ago,
or more like the corporation that decreased by 0.8% in value ¬ve years ago?
As indicated, there are important time and industry e¬ects. For example, at certain times
and in certain industries, investors may very much like or dislike issues, i.e., react posi-
tively or negatively. A good example is high-tech ¬rms around the turn of the millennium.
In 1999, many equity issues by almost any Internet ¬rm were greeted very warmly; in
2001, not even the best and most solid Internet businesses could raise funds.

So, would the average 0.2% drop that we see in the data for debt issues apply to you if you is-
sued debt? Given the plus or minus 3% standard heterogeneity, the context of the debt o¬ering
is probably more important than the average reaction. Put di¬erently, with this large a stan-
dard deviation and this low a mean, about 48% of the ¬rms would have experienced a positive
response and 52% a negative response, and it is not clear where your company would fall. For
debt issues, in particular, this means that, as ¬nancial manager, you should probably not care
too much that the average debt o¬ering dropped by 0.2%. Similarly, how would your stock
react if you announced a dividend? The average response is positive, but plenty of ¬rms expe-
rience negative stock market reactions. (The exception are very, very large dividend initiations,
which are met by a clearly positive shareholder response.) On the other hand, the dividend
announcements produced large value increases relative to the money being paid out. The one
event where the ¬nancial markets have been teaching us that the net costs of issuing are very
high are equity o¬erings. Yes, even though this dilution applies to many ¬rms, it does not
apply to every ¬rm, and perhaps not to your ¬rm. With empirical estimates of dilution ranging
¬le=ibanking.tex: LP
692 Chapter 27. Investment Banking.

between 10% and 20%, issuing costs can be quite substantial. (If we subtract the 7% total direct
costs documented in Table 27.3, we still ¬nd an additional 3-13% of value destruction.) Net in
net, the bene¬t of equity issuing for the average ¬rm does not seem to outweigh its costs.
We noted that you can reduce the return noise by shortening the event window. If you do,
Pinpointing the event
reduces noise and event you are leaning more on the e¬cient markets assumption”that stock markets immediately
contamination, but leans
react”and on your own ability to pinpoint the time or times when the market fully learned of
more heavily on
this information. This brings up yet another important drawback:
immediate market

Event Anticipation Recall that the markets respond only to the unanticipated part of an event”
and even though it is still useful to learn whether the e¬ect was net-positive or net-
negative, it would be better to learn exactly what the market assesses the value conse-
quences to be.

Most studies look at the 2-3 day announcement return, because they rarely know exactly when
the information of the new issue is released.
Given these shortcomings, you can draw two conclusions:
Use this information,
but be careful!

1. The measured market reaction provides very useful information as far as the costs of
issuing are concerned.

2. The measured market reaction is so noisy that you cannot get around computing your
own independent value estimates of any contemplated investment banking transactions.

Solve Now!
Q 27.8 Roughly, how expensive is selling $100 million worth of equity? $100 million worth of
bonds? $100 million in an IPO?

Q 27.9 Why is it more expensive to place equity than debt?

Q 27.10 What factors seem important when ¬rms select underwriters?

Q 27.11 How would you measure the whole cost of issuing, including deadweight costs that
capital structure changes create, direct fees, and everything else? Do you see any implementation

Q 27.12 What do you expect the price reaction to be on the day that the new seasoned equity
o¬ering shares are sold into the market? (This is not the announcement day.)
¬le=ibanking.tex: RP
Section 27·3. Mergers and Acquisitions.

27·3. Mergers and Acquisitions

Though not exclusively a capital raising function, facilitating and advising on mergers and acqui- M&A ¬ts and does not
sitions (M&A) constitutes one of the major functions of the investment banker. It also overlaps
with the world of underwriting, because we know that much issuing”and almost all seasoned
equity issuing for Fortune-100 corporations”occurs only in the M&A context. Although our
focus is the role of the investment banker, we shall start with a brief introduction to M&A.
A merger occurs when two corporations agree to marry on an equal basis. An acquisition Various types of M&A.
occurs when one company purchases another. Most mergers and acquisitions are friendly,
i.e., the target management agrees to be acquired (often in exchange for extra consideration
for managers personally, which is a perfectly legal form of a bribe). Acquisitions are further
classi¬ed by whether the acquirer pays with cash (a “cash o¬er”) or with the corporation™s
shares as currency (a “stock o¬er”). The typical method of execution is the tender o¬er, which
simply invites shareholders to present their shares in exchange for cash or stock. Its execution
can be contingent on enough shares being tendered. In a leveraged buyout (LBO), the acquirer
is ¬nancing the buyout mostly with debt (often high-yield or junk bonds), and the LBO ¬rm
usually delists to become a privately owned company. In a management buyout (MBO), the
management is the LBO buyer.
Takeovers naturally increase the scale of the ¬rm (which managers tend to like), but whether Why Acquire or Merge?
Value Gains.
they create value (or just more revenue) depends on the situation. Value can come from several
sources, the most prominent of which are the following:

Synergies The merging of systems, skills, structures, departments, and sta¬ can improve ef-
¬ciency. Economies of scale can come from the elimination of duplicate departments or
more e¬cient production and distribution. For example, the merging of ATM networks
can attract more bank customers; the elimination of double bureaucracies can reduce
overhead; and the reduction in competition can make raising prices easier.

Shutdown E¬ciencies Sometimes, it is better to shrink or liquidate a ¬rm, and the current
management is unwilling or unable to execute such a drastic measure. A takeover by
individuals with less of an institutional history often makes this easier.

Expropriation A transfer of management can allow breaking implicit promises that ¬rms have
made but not put into writing. After all, it is di¬cult to contract out every promise that
employers make to employees and vice-versa. Consequently, all companies rely on at
least some employee loyalty and all employees rely on at least some company loyalty.
An example of how breaking implicit promises allows a ¬rm to become more pro¬table
is a company that has attracted employees by paying less, but by implicitly promising
long-term employment stability and generous pension and health bene¬ts, as long as the
¬rm does well. But as the company and its workers age, these liabilities can become
quite signi¬cant, and a takeover could allow new management to save money by ¬ring
employees, now older and more expensive, or by replacing an overfunded pension fund
with a less costly alternative.

Two more value gains come about through the higher leverage often assumed in acquisitions,
especially in leveraged or management buyouts:

Tax Bene¬ts Higher debt ratios reduce the amount of pro¬ts collected by the IRS.

Better Governance The need to service debt usually makes it easier to convince both managers
and employees that they have to work harder and spend less on pet projects”or the ¬rm
will go bankrupt. Ironically, management buyouts are often contemplated by the most
wasteful managers, who themselves have the incentives to make their own corporations
look bad, so that they can buy them on the cheap and improve them.

All of these are important M&A drivers, though not equally in each and every takeover. In some
takeovers, it may be primarily synergies; in another, it may be primarily better governance. It
¬le=ibanking.tex: LP
694 Chapter 27. Investment Banking.

is also often di¬cult to distinguish synergies from expropriation or better governance. If a
long-employed but now unproductive worker (or entire department) is eliminated, is this gain
expropriative or e¬ciency-value-enhancing?
Some takeovers succeed; others fail. The principal negative when a large company takes over
Long-Term Success and
Failure. a smaller company is often worse focus and poorer management. There is good evidence that
takeover activity in the 1960s and 1970s was driven by the desires of managers to increase ¬rm
size and form conglomerates, many of which were then run more poorly after the acquisition
than before. In the 1980s, the situation reversed: many of these conglomerates were taken over
and again dismembered. In addition, many small companies began buying bigger companies.
The principal negatives when a small company takes over a larger company, as in an LBO, are
the loss of the bene¬ts of access to capital and the lack of diversi¬cation. That is, the buyer
often has so much of his capital at stake in the ¬rm that he may have to ration capital even for
positive NPV projects, and forego the insurance of diversi¬cation.
Of course, in both cases, the acquirer can make the mistake of overpaying for the target. It turns
The cost of acquisition.
out that the poster child example for the end of the LBO wave of the 1980s was Campeau™s
1988 purchase of Federated Department Stores (e.g., Macy™s and Bloomingdales). Campeau
went bankrupt in 1992. It is true that Campeau had lost a lot of money in buying Federated,
but it seems that even this LBO made money”just not for Campeau. Federated had traded for
$4.25 billion in 1988. When it emerged after bankruptcy in 1992, it became clear that Campeau
had managed to raise the value to $5.85 billion (adjusting for market movements over the same
period)”a $1.5 billion value increase. Unfortunately for Campeau, he had paid $7.67 billion.
This raises the important question of who bene¬ts most from the value gains in M&A™s”the
Who gets the gains?
acquirer or the target? If the acquirer purchases the target at the prevailing market price before
his appearance, all bene¬ts would accrue to the acquirer. If the acquirer purchases the target
at a much higher price, many bene¬ts would accrue to the target. Indeed, if the price is high
enough, the acquirer may lose money and the target shareholders may gain money.
The empirical evidence suggests that targets make out like bandits. A study by Ernst and
It is the target! The
acquirer often loses. Young suggests that the typical announcement price gain is about 25%. From 1996 to 2000, this
premium even shot up to between 40% and 50%! It is no surprise, then, that most of the takeover
value gains have not accrued to the acquirer. A recent study by Moeller, Schlingemann, and
Stulz (2005) looked at publicly trading acquirers. From 1980 to 1998, they lost about 1.6 cents
in value for every acquisition dollar. From 1998 to 2001, this shot up to 12 cents per acquisition
dollar. (As usual, there was a lot of heterogeneity across M&A, but most of the 12 cent ¬gure
was driven by some really bad outliers!) For public acquisitions, the total acquisition value
gain”the dollar bene¬t to target shareholders plus the dollar cost to acquiring shareholders
(the acquirer is usually larger!)”seems to be just about zero.
Still, there is much heterogeneity in value changes for acquirers, just as there was much hetero-
OK, not all acquirers
lose. But why do so geneity in value changes when ¬rms issued securities. Some acquisitions are very good not only
many acquirers do the
for the target, but also for the acquirer. You have to judge acquisitions on a one-by-one basis.
wrong thing for their
Nevertheless, you might ask: Why are bad acquisitions not unusual? The reason is that they
are often not only in the interest of the investment banks, who push them because they reap
good fees from M&A ¬nancing and advice, but also in the interest of the acquiring managers.
Running a bigger company usually means more prestige and compensation for target managers.
The target management naturally often resists, even though they should be thrilled for their
shareholders. The best example of this con¬‚ict of interest may have been the merger between
Chase and Bank One. The negotiation took place between the Bank One CEO, Dimon, and Chase
CEO Harrison, both of whom wanted to become CEO immediately. The original plan was for
Dimon to succeed Harrison after two years. Dimon o¬ered to sell Bank One at a zero premium
if he just were to become the merged company™s CEO immediately. Harrison rejected this o¬er,
and instead paid a $7 billion premium from Chase shareholders to Bank One shareholders in
order to retain his post for these two extra years. Chapter 28 will look at M&A activity from a
corporate governance perspective.
¬le=ibanking.tex: RP
Section 27·3. Mergers and Acquisitions.

27·3.A. M&A Participants, Deal Characteristics, and Advisory Fees

In Table 27.4, there are hard statistics for almost all domestic acquisitions that involved a
publicly traded corporations from 1980 to 2003, and classi¬ed by the quality of advisor (within
the industry in which the takeover occurred). Still, the data are not complete. There were many
mergers and acquisitions among ¬rms that were not public, and even for the roughly 15,000
acquisitions having involved a public corporation, we have good data on fees for only 6,000.
The typical acquirer was about three to four times as large as the typical target. Also, the mean Participating Firms
¬rm size was much larger than the median ¬rm size, suggesting some disproportionally large
¬rms were in the sample. About one-half to two-thirds of M&A™s occurred between ¬rms in the
same industry (classi¬ed by the “two-digit SIC [standard industry classi¬cation] code”). About
half to two-thirds of M&A™s involved public acquirers or targets.
The average deal size was about $800 million, but the top-tier investment banks advised on Deal Characteristics
disproportionally larger deals. About one in ¬ve takeovers occurred through a tender o¬er
(the alternative being a negotiated merger with the target, not involving an o¬er to sharehold-
ers). Only a small fraction of all deals were classi¬ed as hostile, where the target management
resisted. About one third of all deals were paid for in “all cash,” and about one-third were
paid for with “all stock” (in which the acquirer paid target shareholders with its own shares).
Somewhere between about 10% and 15% of acquisitions were abandoned. If successful, it took
the typical deal about four months to complete. Note that when the deal was hostile, a much
larger fraction of targets seem to have engaged top-tier underwriters.
The median advising fees were just about half a percent of the amount of the transaction Fees
(usually the target size), on average. The mean fee was much larger, suggesting that there
were a few large fee outliers. Remarkably, top-tier investment bankers charged about the same
proportional fees as their lower-tier brethren”the reason why they earned more fees is simply
that their deals were larger.
Solve Now!
Q 27.13 Why do ¬rms like to acquire other ¬rms?

Q 27.14 Do acquirer or target shareholders gain more in a takeover?

Q 27.15 What are possible sources of value gains in takeovers?

Q 27.16 How large is the typical acquirer relative to target?

Q 27.17 What is the typical commission for M&A advice that investment bankers earn? How
does it di¬er across the tier of investment bank retained, and across acquirer and target?
¬le=ibanking.tex: LP
696 Chapter 27. Investment Banking.

Table 27.4. Descriptive Statistics of U.S. M&A Transactions from 1980 to 2003

Acquirer Adviser Tier Target Adviser Tier
Top Mid Bot Total Top Mid Bot Total
Firm Value ($ million) $7,642 $5,084 $1,020 $4,916 $2,106 $1,237 $265 $1,395
Median ($ million) $1,765 $711 $213 $736 $440 $251 $65 $241
Acq and Tgt in same industry 63.6% 62.7% 65.9% 64.0% 49.0% 45.5% 60.5% 52.2%
Prop of public acquirers 64.5% 62.0% 72.0% 66.6%
Prop of public targets 58.5% 50.4% 43.3% 51.3%
Deal (Tgt) value ($ million) $1,357 $659 $127 $761 $1,821 $663 $126 $840
Median ($ million) $275 $132 $37 $120 $403 $138 $48 $127
Prop of tender o¬ers 19.7% 17.7% 9.7% 16.1% 24.9% 23.1% 15.3% 20.8%
Prop of hostile deals 3.6% 3.9% 0.8% 2.9% 10.4% 5.3% 2.0% 5.7%
Num of acquirer advisers 1.20 1.11 1.03 1.12 0.84 0.67 0.49 0.66
Num of target advisers 0.90 0.77 0.59 0.76 1.34 1.16 1.06 1.18
Prob of Completion 88.9% 89.2% 91.8% 90.0% 73.6% 79.5% 85.6% 79.8%
Days to Completion 116 100 102 106 141 132 148 141
Prop of all-cash deals 37.6% 38.3% 32.8% 36.3% 42.8% 48.6% 42.8% 44.5%
Prop of all-stock deals 28.8% 27.8% 39.1% 31.6% 23.4% 22.1% 38.9% 28.9%
Pct of cash 47.3% 48.7% 42.2% 46.2% 53.0% 58.2% 48.8% 53.0%
Pct of other 14.5% 14.3% 10.1% 13.1% 16.1% 14.0% 6.4% 11.8%
Pct of stock 38.1% 36.9% 47.7% 40.7% 30.9% 27.8% 44.8% 35.2%

Fees Paid to advisers, in millions of dollars
Mean $4.83 $2.65 $0.77 $2.89 $6.47 $2.79 $0.97 $3.06
Median $2.38 $1.00 $0.25 $1.00 $3.70 $1.40 $0.44 $1.13
Deal Value, in millions of dollars
Mean $2,494 $1,092 $208 $1,345 $2,177 $749 $150 $899
Median $416 $195 $55 $177 $525 $181 $58 $144
Fees paid as percentage of deal value
Mean 0.91% 0.90% 0.93% 0.91% 0.87% 1.13% 1.15% 1.06%
Median 0.47% 0.58% 0.52% 0.52% 0.67% 0.80% 0.82% 0.76%
N 733 672 591 1,996 1,124 1,113 1,695 3,932

Rows report means (except where noted otherwise), and can have di¬erent numbers of observations. In the top
rows, there are typically about 15,000 acquisitions; in the middle panel there are typically about 15,000 acquisitions.
These are roughly equally split across categories. There is fee information for only about 6,000 acquisitions, and
the distribution is somewhat biased, which is why N is reported in the last column, and why the deal values here do
not match deal values above.
Source: Walter, Yawson, Young, undated (June 2005).
¬le=ibanking.tex: RP
Section 27·4. Summary.

27·4. Summary

The chapter covered the following major points:

• The market for securities underwriting has become part of a global market for investment
banking services. The largest U.S. investment banks underwrite between $300 and $500
billion in securities in a good year.

• The direct costs of underwriting are a function of o¬ering size. An IPO costs around 10%
in total costs, a seasoned equity o¬ering about 7%, a convertible bond o¬ering about 4%,
and a straight bond o¬ering about 2%.

• There is large heterogeneity in the reaction of the ¬nancial markets to the announcement
of a new debt or equity issue (or dividend payment).

• The typical ¬rm drops about 2% when it announces a new equity issue. This corresponds
to a 10-20% dilution cost for existing shareholders.

• Investment bankers also serve an important advisory function in M&A activity. In an
active year, for many investment banks, the M&A they are advising on is often similar in
magnitude to the amount they underwrite”$200 to $500 billion. Acquirers often overpay
for targets, which means that most of any value gains accrue to target shareholders, not
acquirer shareholders.

• Based on information from M&A deals among publicly traded corporations between 1980
and 2003, it appears that:

“ Average advisory fees are about 1% of the target (transaction) size.
“ Median advising fees are about 0.5“0.7% of the transaction size.
“ The 80-90% of proposed deals that ultimately carry through take about 4 months to
“ Fewer than about 5% of all acquisitions are hostile (and most of these occurred in the
“ The typical acquirer is about three or four times larger than the target.
“ Between one-half and two-thirds of acquisitions are within the same industry.
“ About one-third to one-half of acquisitions are paid for with all cash, and about 30%
are paid for with all stock.

Anecdote: An Investment Banking Job?
In “The Making of an Investment Banker,” Paul Oyer followed Stanford MBA graduates from the classes of 1960
through 1997. Investment bankers enjoy between $2 million and $6 million in discounted lifetime income (in
real 1996 dollars). This is much higher pay than what they can earn if they enter other professions. The average
I-banker earned 60% more than a management consultant at graudation, and 300% more than the average
Stanford MBA 15 years after graduation.
More interestingly, Oyer ¬nds that stock market conditions at graduation time play a big role not only in
obtaining a ¬rst job in I-banking, but also in the probability that an individual will ever end up on Wall Street.
(Many of the individuals graduating in bear years end up as entrepreneurs!)
His ¬ndings allow Oyer to conclude that random factors beyond capability are very important in determining
individuals™ life-time path and compensation; and that there is a very deep pool of potential I-bankers in any
given Stanford MBA class.
¬le=ibanking.tex: LP
698 Chapter 27. Investment Banking.

Solutions and Exercises

1. The guarantee of securities placement sales success is usually fairly unimportant, because it is only given on
the day of the o¬ering.
2. Formal process management, selling, certi¬cation.
3. Investment bankers like transaction volume and fees, not value creation for their clients. The latter matters
primarily to the extent that it helps the former. An investment banker who continually costs its clients
money will eventually lose many. For investors, investment bankers are often interested merely in selling
securities to, whether they are good or bad investments. For ¬rms, investment bankers are often interested
in restructuring, whether it makes sense or not. Furthermore, investment banks sometimes structure issues
in a way that makes it hard for ¬rms to value what they are giving away.
4. Most recommendations were not very good, as evidenced by the fact that most recommendations are “buy.”
This helps the investment banker attract corporate clients.
5. See Table 27.1.
6. They are about equal sized.
7. See Table 27.2. By market value, the U.S. and British banks are far more important.

8. About $4 million (SEO), $1.5 million (bond), and $7 million (IPO) in terms of underwriting costs. Add about
$1 million for other costs in an IPO, and $300,000 to $500,000 for bond and seasoned equity o¬erings.
9. There is more capital at risk, which in turn means that the underwriter has to put more of its reputation on
the line and work harder to place the securities.
10. First, inertia. Second, their “outgrowing” their previous underwriter. Third, industry expertise. Fourth,
personal relations.
11. Through the degree of dilution at the announcement price reaction! Unfortunately, it is not known by man-
agers before hand, and so depends on comparability assumptions, and stock returns are noisy.
12. It should be about zero, because the share sale is an event that was announced earlier and thus should be
almost perfectly anticipated. If the market did not use this information e¬ciently, and the share price were
to go down on the day of the o¬ering, you could short the equity shares the day before the o¬ering, and
repurchase them the day after the o¬ering for a pro¬t.

13. Because it can be in the interest of the ¬rm (creating value), or because it can be in the interest of managers
(and advising bankers).
14. Target shareholders.
15. Synergies, shutdown e¬ciencies, and expropriation. In addition, if ¬nanced by debt, it could be tax bene¬ts
and better governance.
16. About three to four times.
17. The mean is about 1% (0.9% for acquirer, 1.1% for target). The median is about 0.5% (for the acquirer) and
0.8% (for the target). The di¬erences across tier and between target and acquirer seem fairly small.

(All answers should be treated as suspect. They have only been sketched, and not been checked.)
Corporate Governance

More Agency Con¬‚ict
last ¬le change: Feb 23, 2006 (15:31h)

last major edit: May 2005

For the most part, we have assumed that managers act on behalf of owners and maximize ¬rm
value. We have mentioned agency con¬‚ict before, as in Chapter 7. In this chapter, we describe
the con¬‚ict between corporate owners and top management in more detail. In theory, we know
that debt should be paid ¬rst; equity should receive the residual; and managers should be
compensated according to their marginal value to the company. But we did not ask the simplest
of all questions: Why do managers return any money to investors? After all, a speci¬c investor
contributes little to the corporation after the corporation has her money.
This question is the domain of corporate governance, which concerns itself with the con¬‚ict
of interest between the managers of the corporation and its capital providers.
It is also important to understand what corporate governance is not”it is not good manage-
ment. Instead, governance is the mechanism to control management if it is bad. Governance
mechanisms may never have to spring into action if management is good.

¬le=governance.tex: LP
700 Chapter 28. Corporate Governance.

28·1. Less Fact, More Fiction: In Theory

Most companies start out with few con¬‚ict of interest problems”if only because the en-
Firms typically start out
tightly controlled but trepreneur owns the entire ¬rm, provides most capital, and makes all decisions. As the com-
eventually become
pany grows, the entrepreneur usually needs to raise more outside capital, either to expand the
diffusely held, which
¬rm or just to start enjoying the new riches. To be able to get investors to part with their cash,
brings many problems.
the entrepreneur must create a corporate charter and install safeguards that satisfy potential
investors, legal requirements, and common practice. Eventually, the founder™s personal role
begins to fade, more and more capital is raised from the outside, and management becomes
“professional” in the sense that it no longer acts solely based on the whim of the entrepreneur.
These professional managers also bring with them unique quali¬cations and specialization ben-
e¬ts. Legally, these managers now become “agents” acting on behalf of the investors (primarily
shareholders), the “principals.” Practically, as the distinction between capital providers and
decision makers grows over time, so do con¬‚ict of interest (“agency”) problems”and not just
between them, but also among them. Although even two or three co-owners can squabble,
when there are thousands of shareholders, as in a publicly traded company, the coordination
problems take on an entirely new dimension. Shareholders usually agree that they prefer more
money to less money, but often on little else. In any case, they are not capable of constantly
voting and communicating their desires to their agent managers, much less checking over what
their managers are doing day-to-day”and managers know this, too. Thus, investors in large,
publicly traded corporations are typically represented not only by managers, but also by a set
of institutions (discussed in this chapter), so that they themselves can fade into a more passive
Solemn promises of both corporate value maximization and eventual pro¬t participation are
To get money back,
investors need control not enough for shareholders and creditors. Investors must be able to coerce their managers to
rights. To get money in
honor their promises. This is primarily (but not exclusively) achieved by control rights, which
the ¬rst place, ¬rms
give investors power over managers, and especially when managers do not act appropriately.
need to give control
Giving up such control rights is in the interest of owners, because the terms under which they
can obtain capital in the ¬rst place are better when the control rights are better. This argument
is really the same that we used in Chapters 22-23 to justify why owners want to set up an
optimal debt/equity ratio.
In real life, control rights are not perfect”it would be impractical or impossible to protect
The equilibrium
outcome is second-best, capital providers perfectly. The cost of preventing all managerial opportunism would be pro-
permitting unavoidable
hibitive. It would not maximize ¬rm value if the ¬rm spent $10 in audits to prevent $1 in fraud.
So corporations and capital providers must live with a second-best solution, in which there is
a constant tension between investor protection and managerial self-enrichment.
Debt and equity are distinct. You already understand the di¬erences in cash ¬‚ow rights be-
Control rights differ for
debt and equity. tween them”debt has ¬rst dibs on the promised payments, and the residual cash ¬‚ows go to
equity. But their control rights are distinctly di¬erent, too:

Equity Shareholders are legally in charge of decisions, but their primary power is their ability
to vote and appoint representatives, usually once a year during the annual meetings. Their
elected corporate board has the power to hire and ¬re managers.

Debt The bond contract not only speci¬es how much the ¬rm obligates itself to repay in the
future, but also the immediate legal remedy if the lender fails to do so or fails to meet
any number of pre-speci¬ed covenants. Usually, this means that the lender receives pos-
session of the ¬rm (collateral)”no ifs, ands, or buts.

A ¬rm that has no corporate board may not ¬nd shareholders willing to purchase equity shares;
When corporate
governance usually a ¬rm that does not give the right to force bankruptcy upon default may not ¬nd creditors
works and when it does
willing to lend money. But control rights are not all black-and-white. If the ¬rm does not o¬er
not work.
perfect protection to its capital providers, it may still be able to obtain capital, although at
worse terms that require the surrender of a higher percentage of the ¬rm or the payment of a
higher interest rate. This is one important reason why entrepreneurs and ¬rms are interested
in good corporate governance in that it enables them to raise capital at better terms. But it
¬le=governance.tex: RP
Section 28·2. Managerial Temptations.

also explains when corporate governance is likely to break down”when the entrepreneur (or
shareholder-owners) are no longer solidly in charge, and self-interested managers have already
taken control. After all, if managers do not care for shareholders, they may not care if they have
to give away a larger fraction of the ¬rm to get control of more money from new shareholders.
And the problem gets even worse when the ¬rm generates a lot of internal cash and no longer
needs to raise much new external capital.

28·2. Managerial Temptations

Although the legal ¬ction is that managers act solely on behalf of the ¬rm and that shareholders The human mind for
scheming is in¬nite.
own the ¬rm after creditors are paid o¬, the fact is that all parties act primarily in their own
interests. But why is this a problem? How might a manager do harm? Unfortunately, there
is a whole battery of tactics managers can employ to enrich themselves at the expense of
shareholders. This section outlines a slew of possible managerial self-enrichment schemes. In
the next section, we will discuss mechanisms that seek to restrain such managerial behavior.

28·2.A. Illegal Temptations

The most obvious method is simple theft. For example, in March 2003, the Mercury News Simple Theft is rare, but
it does occur.
reported that 58-year-old C. Gregory Earls, head of an investment company called USV Partners,
simply funneled investor money into a trust fund for his children. What prevents corporate
managers from taking corporate diamonds out of the corporate safe? For the most part, it is
the law, which criminalizes simple theft, which is therefore fairly rare. (Mr. Earls competes for
the prize of “dumbest criminal””it is hard to leave a clearer paper trail than he did.)

The next step up is fraud, because it is more complex, and therefore more di¬cult to detect and Fraud is fairly common...
prove. For example, in 2003, Hop-on Wireless claimed to sell disposable cell phones. It turns
out that the prototypes were Nokia phones with plastic cases around them. The CEO raised
funding, promising not to take a salary”but promptly used the funds to pay o¬ his credit
card debts (see theft above) and gave a company he owned a $500,000 contract (see transfer
payments below).
Usually, fraud involves manipulation of ¬nancials. Unlike Hop-on™s extreme case, many ac- ...but earnings
management is legal.
counting choices are not so black-and-white”the line between illegal accounting manipulation
and legal earnings management is both wide and gray. There are many judgment calls that
corporate executives have to make. There is empirical evidence that legal corporate earnings
management is particularly aggressive just before the corporation issues more equity, for obvi-
ous reasons, and that the most aggressive earnings managers later perform worse. So, appro-
priate conservatism may not be in the interest of owners, but it could be wise. Nevertheless, too
much conservatism is also not in the interest of shareholders. Painting too bleak a picture may
make the business collapse. What prevents rosy picture painting? GAAP and SEC scrutiny limit
the discretion of managers to legally manipulate the ¬nancials. And again, there are criminal
penalties against fraud.
¬le=governance.tex: LP
702 Chapter 28. Corporate Governance.

Insider Trading
One more step up”and a surprisingly common form of agency con¬‚ict”is insider trading. For
Insider trading is very
common. example, the most well publicized recent insider trading scandal involved Sam Waksal, CEO of
ImClone (IMCL). Waksal received advance bad news about clinical tests of an ImClone cancer
drug, and proceeded to tip o¬ his family and friends (including Martha Stewart) that they should
immediately sell their shares. His next seven years are all booked up now.
Like earnings management, insider trading can be either legal or illegal”and again there is a
Some insider trading is
legal and should be wide gray line. Managers almost always have more information than shareholders. They love
allowed”but which?
to trade on it before the public learns it, and naturally, this does not make other shareholders
better o¬. It would be unwise to prohibit all insider trading, because insiders do need to be
able to sell and buy shares just like the rest of us, if only to diversify some of their wealth.
But it is illegal for them to trade on information that is not yet public, which is easy to prove
if this entails an impending news release. (Because this is so easy to detect and prove, it is
curious how someone as smart as Waksal could have made such a big mistake.) More often, the
information that executives have is “soft,” and the empirical evidence shows that they indeed
do well in their private, legal insider trading. They generally tend to buy before the ¬rm gets
better and sell before the ¬rm get worse.

The next step up is yet more di¬cult to detect and prove”the pilfering of corporate resources
Complex theft through
transfers is more through transfers. The CEO of a public company can own private companies that do business
with the public company on favorable terms. On occasion, the terms become so favorable
that they warrant criminal indictment. For example, on May 1, 2003, the U.S. Department
of Justice alleged that “in 1997, Fastow [CFO of Enron] conspired with others, including his
wife, to create an [entity owned by the Fastows] in order to reap for themselves the pro¬ts
generated by certain Enron wind farms, while simultaneously enabling Enron to fraudulently
receive government ¬nancial bene¬ts to which it was not entitled.” Naturally, the smarter the
manager, the more complex will be the arrangements, so that the true costs and true bene¬ts
to the public company are more di¬cult to assess. Criminal prosecution of such schemes is
fairly rare, especially if the corporate executive has followed legal procedures to the letter (e.g.,
by asking for board approval).

Side Note: Transfers can also occur to friends of the management or to large shareholders, who then owe
more loyalty to the CEO. The ambiguous role of large shareholders in corporate governance is described below.

Yet another way for executives to get rich at the expense of shareholders, and again one step
Third parties bribe
managers all the time. more di¬cult to detect is bribes. Managers of publicly traded companies need not even create
their own temptations: they practically come to them. For example, during the 1998-2000 tech-
nology bubble, receiving IPO share allocations was practically like getting free money. (Normal
¬rst-day rates of return were around 50%. Ordinary brokerage clients would rarely receive any
allocations.) Citigroup was eager to do investment-banking business with WorldCom, a publicly
traded company. It therefore allocated $17 million in 21 o¬erings into CEO Bernie Ebbers™ per-
sonal account. In one IPO (Rhythms Net Connections) alone, Ebbers was allegedly handed $16
million. De facto, Ebbers was “courted” to direct the business of the publicly traded company,
WorldCom, to Citigroup.
Preferential allocations to and treatment of executives™ personal accounts were and continue
to be common practice. Ebbers was an extreme case, but not the only one. Lesser methods of
bribing executives are so commonplace in business that they are considered almost ordinary.
For example, there is evidence that competitive bids for high-level professional services (such
as the hiring of a search ¬rm or the placement of a bond or equity issue) usually result in better
contract terms than negotiated contracts for the ¬rm”and yet most companies negotiate rather
than bid out contracts. Although negotiation can be better for other reasons, more commonly
the reason lies elsewhere, and here is why. Executives of smaller ¬rms naturally want to be on
¬le=governance.tex: RP
Section 28·2. Managerial Temptations.

the candidacy list to become executives of bigger companies. It is therefore in their interests
to form good relationships with investment banks and executive search ¬rms. An executive
who uses competitive bids, which minimize the pro¬ts of the professional service ¬rms, and
who constantly switches from one low bidder to the next, is unlikely to build much loyalty and
subsequent quid-pro-quo support.

28·2.B. Legal Temptations

If you now have the impression that fraud, theft, insider trading, and bribes are the most Most managers are not
criminals, though.
important agency con¬‚icts between shareholders and managers, you would be wrong. The
most important con¬‚icts arise in the day-to-day execution of business, and are more judgment
calls than outright unethical behavior”few CEOs actively seek out behavior that is obviously

Empire Building
Most managers see it as their task to grow the business. There are a whole slew of reasons why Most managers want to
grow the ¬rm at all cost.
they want to do so. Most managers are less loyal to an abstract, ever-changing shareholder, as
they are to their very tangible companies. Running bigger companies is also in the self-interest
of managers: Executives of bigger companies are more prominent and receive more compen-
sation. Some decades ago, this was even explicit: managerial compensation schemes were
directly tied to sales, not earnings! These days, it is more implicit, and comes about through
the choice of “comparable managers” when executive pay schemes are set. Unfortunately, cor-
porate growth is not necessarily shareholder value maximizing. For example, in the 1980s, oil
companies were ¬‚ush with both cash and assets (oil reserves). The industry had overcapacity.
Most oil companies did not return excess cash to shareholders, but chose to spend $20/barrel
exploring for more oil reserves, even though oil could be purchased in the marketplace for
$6/barrel. Managers naturally are paid for operating a company”di¬cult tasks such as oil ex-
ploration, growing the ¬rm, acquiring other companies. It is not in the interests of managers to
return cash to shareholders, especially when it entails drastic shrinking (asset sales) or when
it means being taken over by another company. The reward for being a good manager who
maximizes shareholder wealth would often be unemployment!
Many academics believe that the highest agency costs in American companies today (in terms This growth is
enormously costly to
of expected costs to shareholders) are not illegal actions, but failure to direct corporate assets
towards the activities that maximize shareholder wealth. The agency cost is particularly high
for ¬rms that have lots of cash and cash ¬‚ow (e.g., from prior pro¬table activities) and few
good new opportunities.

Corporate Perks
A closely related, though smaller, problem is that managers disproportionately enjoy spending Perks are goodies that
managers have the ¬rm
money on perks. A public company may buy a corporate airplane that costs $100 million
buy for themselves.
and increases productivity by the equivalent of $10 million”just because it gives managers $1
million worth of pleasure. Plush corporate headquarters and ¬‚eets of corporate aircraft are
usually sure signs of publicly traded companies, especially in slow-growth industries.
¬le=governance.tex: LP
704 Chapter 28. Corporate Governance.

Executive Pay
Naturally, executives are most con¬‚icted when it comes to higher executive pay. Executive
Executive Pay is often
excessive. compensation comes in many forms: salary, bonus, stock grants, option grants, retirement
bene¬ts, perks, and severance packages. For example, in 2002, Business Week reported that
Archie Dunham of Conoco (COP) earned $30 million in salary and bonus alone; Alfred Lerner
of MBNA (the credit card lender, KRB) earned $200 million in salary, bonus, and long-term
compensation grants; Je¬ Barbakow of Tenet Healthcare (THC) earned $300 million (Tenet
is currently under indictment); and Larry Ellison of Oracle (ORCL) earned $800 million from
2000-2002, mostly from stock and options. Since then, executive compensation has grown
every year. In December 2005, the Wall Street Journal further reported that the income taxes
on corporate perks (e.g., cars, jets, loan forgiveness) that many CEOs receive are also often paid
by the corporations and reported only as relatively obscure “tax gross-ups.” (52% of companies
report some gross-ups.) Other recent evidence suggests that pension packages that are usually
not reported are larger than the reported executive compensation.
Why do shareholders in the United States pay so much money? How much worse a manager
Questions about
executive pay. would the next-best executive have been, and would he have done the job for half the com-
pensation? Why do second-in-commands usually earn only a very small fraction of the CEO™s
salary, even just one year before they succeed the CEO? And why do European CEOs of even
the most successful corporations earn only about 10% as much as their U.S. comparables?

• It could be that U.S. managerial talent is scarcer than European managerial talent. Tougher
competition for good management in the United States requires paying higher compensa-

• It could be that a large salary is necessary to attract and retain top CEOs in the United
States, and to motivate lower-ranked executives to strive for this prize”but less so in

• It could be that American CEOs are operating in a governance structure that has allowed
them to receive higher salaries than their European counterparts. (This could be because
they could capture the corporate board better, or because social norms are di¬erent.)

• It could be simple error that is not corrected by the market place. It could be that the
Europeans have it wrong and are simply paying too little. Or it could be that Americans
have it wrong and are simply paying too much.

As to the ¬rst argument, it is hard to believe that executive talent is generally scarcer in the
United States than it is in Europe. It is also hard to believe that the marginal contribution (the
di¬erence between the best executive and the next-best executive) is typically much larger in
the United States than in Europe. As to the second argument, it is possible that European social
norms are di¬erent from American norms, but there is little evidence one way or another. The
latter two arguments seem more plausible. The cultural, ethical, and legal constraints in Europe
are di¬erent from those in the United States. In most European countries, the chief executive
is not the chairman of the board, and social norms prevent too high a managerial salary. (Of
course, the same social norms and legal regulations make it more di¬cult for managers to
take drastic actions on behalf of shareholders, e.g., when it comes to downsizing and employee
¬le=governance.tex: RP
Section 28·2. Managerial Temptations.

Managers, like all employees, like to be indispensable. If they decide to take projects for which Take projects where
your expertise will be
they are presumably indispensable, their own personal value to the ¬rm and therefore their
compensation will likely go up. If they decide to build redundancy”that is, hire someone who
can step in for them, thereby making themselves dispensable”their own value to the ¬rm will
likely go down. In fact, they might even be replaced by the board. The ability to hold up
the company, once managers have become indispensable (or at least, very di¬cult to replace),
probably plays an important role in the awarding of high executive compensation contracts. In
this case, no managerial board capture is necessary for high executive compensation to occur.
The board will not have a choice and will award high compensation “voluntarily.”
Bureaucracy often helps promote entrenchment. It can discourage shareholder wealth maxi-
mization, but help managers to become indispensable (knowledgeable of the internal process),
and even undertake bizarre projects, internally justi¬ed by “proper procedure.” In contrast,
¬ghting bureaucratization on behalf of shareholders is a painful and prolonged process, with
few rewards for the involved executives unless the ¬rm is in dire straits.

Friendship and Loyalty
Most managers prefer to have loyal friends working around them, instead of gad¬‚ies and po- Nepotism abounds. Few
boards have non-friends
tential replacements. This natural characteristic naturally promotes nepotism (in the broad
on them.

Perverse Incentives
Though rare, managers can even have the incentive to drive down ¬rm value: they can then ne- Sometimes, managers
even prefer low values.
gotiate better incentive compensation contracts or even acquire the ¬rm in a leveraged buyout,
either of which is often followed by seemingly miraculous turnarounds. The most prominent
example is that of Ross Johnson, CEO of RJR Nabisco, whose actions are chronicled in the
bestselling book Barbarians at the Gate.

Ethical Con¬‚icts
A manager may feel special obligations toward many factors that are not in the interest of
shareholders: the town in which the factories are located, the workers employed, charitable
causes, and so on. Under these circumstances, managers may explicitly or implicitly donate
the shareholders™ money toward causes that they deem to be more worthwhile than paying

I am descriptive here and am not stating what is necessarily the most appropriate. The
web chapter on ethics deals with this question in more detail.

Side Note: Some economists™ models assume that executives prefer working less (called “shirking”). However,
lack of work ethics among executives is rarely a problem in the real world”instead, it is self-enrichment that
is the problem.
¬le=governance.tex: LP
706 Chapter 28. Corporate Governance.

28·2.C. The Incentive of the Entrepreneur to Control Temptations

Are all these problems real? Yes, they are. So, in order to get you, as a private outside investor,
We now show that
entrepreneurs (i.e., to give your money to a ¬rm, it must create mechanisms that will curtail its misbehavior. If
coordinated owners)
these mechanisms are not in place, the ¬rm will be unable to raise more capital at reasonable
bene¬t from good
costs. (Yet, as we shall see later, if manager-agents are already ¬rmly in charge of shareholder-
corporate governance.
principals, they may not care about “reasonable” costs of capital, either.)

An Example of the Entrepreneur™s Incentives
The argument that it is in the interest of the owners to get the governance right is really the
same as the argument that it is in the interest of the owners to get the debt-equity ratio right.
Here is a simple illustration of how good corporate governance can bene¬t the original owner:

• Start with a penniless entrepreneur who has an invention that requires $25 million invest-
ment but that will be worth $100 million in today™s dollars. If the entrepreneur cannot
raise the money, the project will be captured by the competition. Without capital, the
entrepreneur™s wealth would be $0. There are large gains to ¬nding investors. This is
why companies go public to begin with: gains from diversi¬cation and capital outweigh
the costs of agency con¬‚icts.

• If the owner does not need to raise any funds and could avoid all professional manage-
ment, his net worth is $75 million. Similarly, if the owner can somehow commit to avoid
all agency problems, investors can give him $25 million for 25% of the company, leaving
him again with $75 million.

• Investors may rationally believe that agency problems will appear as soon as money is
raised. The particular agency problem in our example is that the entrepreneur cannot
prevent the grabbing of $30 million by management-in-charge after the capital has been
raised”including by himself. That is, he cannot commit the corporation to control the
agency problem. Furthermore, to hide the managerial scheming, this “theft” will require
another $10 million of waste.
If the entrepreneur must now raise external capital, will investors be satis¬ed with 25% of
the company for a $25 million investment? Probably not: any future manager”including
the entrepreneur himself ”will want to steal the money. Investors expect this, and there-
fore value the company only at $100 ’ $30 ’ $10 = $60 million. To raise $25 million re-
quires the entrepreneur to part with $25/$60 = 42% of the company. The entrepreneur™s
net worth, if he can remain in charge, will be the 58%·$60 = $35 million that he will
own, plus the $30 million that he can steal. The $65 million is $10 million less than
what he could have gotten if he could have committed to zero future managerial agency
problems”still better than the $0 if he were not to raise any external funding.


. 34
( 39)