<<

. 109
( 208 .)



>>

Implied ROE in 2003 55% 51% 68% 63% 83% 77%
Implied price, based on
cost of capital of:
8% $127 $125 $157 $155 $193 $192
9% $101 $99 $125 $123 $153 $151
10% $83 $82 $103 $101 $126 $124
512 Equity Security Analysis




13-10
Equity Security Analysis




Table 13-1 shows several combinations of assumptions that are consistent with an ob-
served market price of $126. One is 15 percent earnings growth for the next ¬ve years
and an equity cost of capital of 8 percent. Under these assumptions, earnings per share
will effectively double in the next ¬ve years, from $3.29 in 1998 to $6.62 in 2003. In
addition, IBM™s return on equity soars from 33 percent in 1998 to between 51 and 55 per-
cent, depending on the rate of growth in book value. It is interesting to note that changes
in the book value growth rate do not have a strong impact on the valuation. Other com-
binations of assumptions that generate a value consistent with the market price include
(1) a ¬ve-year earnings growth of 20 percent and a 9 percent cost of capital, and (2) a
¬ve-year growth rate of 25 percent and a 10 percent cost of capital. Of course, these as-
sumptions imply that IBM will earn extremely high returns on equity by 2003, between
63 and 83 percent.
Security analysis need not involve such a detailed attempt to infer market expecta-
tions. However, whether the analysis is made explicit or not, a good analyst understands
what economic scenarios could plausibly be re¬‚ected in the observed price


Key Analysis Questions
By using the discounted abnormal earnings/ROE valuation model, analysts can in-
fer the market™s expectations for a ¬rm™s future performance. This permits ana-
lysts to ask whether the market is over- or undervaluing a company. Typical
questions that analysts might ask from this analysis include the following:
• What are the market™s assumptions about long-term ROE and growth? For ex-
ample, is the market forecasting that the company can grow its earnings with-
out a corresponding level of expansion in its asset base (and hence equity)?
If so, how long can this persist?
• How do changes in the cost of capital affect the market™s assessment of the
firm™s future performance? If the market™s expectations seem to be unexpect-
edly high or low, has the market reassessed the company™s risk? If so, is this
change plausible?



Developing the Analyst™s Expectations
Ultimately, a security analyst must compare his or her own view of a stock with the view
embedded in the market price. The analyst™s own view is generated using the same tools
discussed in Chapters 2 through 12: business strategy analysis, accounting analysis, ¬-
nancial analysis, and prospective analysis. The ¬nal product of this work is, of course, a
forecast of the ¬rm™s future earnings and cash ¬‚ows and an estimate of the ¬rm™s value.
However, that ¬nal product is less important than the understanding of the business and
its industry that the analysis provides. It is such understanding that enables the analyst
to interpret new information as it arrives and to infer its implications.
513
Equity Security Analysis




13-11 Part 3 Business Analysis and Valuation Applications




Key Analysis Questions
In developing expectations about the ¬rm™s future performance using the ¬nancial
analysis tools discussed throughout this book, the analyst is likely to ask the fol-
lowing types of questions:
• How profitable is the firm? In light of industry conditions, the firm™s corpo-
rate strategy, and its barriers to competition, how sustainable is that rate of
profitability?
• What are the opportunities for growth for this firm?
• How risky is this firm? How vulnerable are operations to general economic
downturns? How highly levered is the firm? What does the riskiness of the
firm imply about its cost of capital?
• How do answers to the above questions compare to the expectations embed-
ded in the observed stock price?



The Final Product of Security Analysis
For ¬nancial analysts, the ¬nal product of security analysis is a recommendation to buy,
sell, or hold the stock (or some more re¬ned ranking). The recommendation is supported
by a set of forecasts and a report summarizing the foundation for the recommendation.
Analysts™ reports often delve into signi¬cant detail and include an assessment of a ¬rm™s
business as well as a line-by-line income statement, balance sheet, and cash ¬‚ow fore-
casts for one or more years.


FINANCIAL STATEMENT DATA AND SECURITY PRICES
While security analysis clearly involves much information beyond the ¬nancial state-
ments, those statements play an important role. Much research over the past three de-
cades has helped describe the role of ¬nancial statement data in the setting of security
prices. An understanding of that role provides an appreciation for the importance of that
data in security analysis, as well as market agents™ ability to digest such data.
A thorough review of research on ¬nancial statement data and security prices lies
well outside the scope of this chapter. However, we can summarize a few of the key ¬nd-
ings from the literature.


Earnings and book value are good indicators of stock prices.
Accounting earnings and book values ignore important aspects of the ¬rm™s economic
landscape, are subject to distortion by managers, and are not adjusted for in¬‚ation in the
U.S. and most other countries. One could (and the ¬nancial press frequently does) rea-
514 Equity Security Analysis




13-12
Equity Security Analysis




sonably question whether accounting numbers are good indicators of the expected cash
¬‚ows that should drive stock prices.
It turns out that, in spite of the widely discussed shortcomings of accounting systems,
earnings and book value offer a good re¬‚ection of much of the information in security
prices. In the U.S., the combination of book value (per-share) and earnings explains, in
a typical year, nearly two-thirds of the cross-sectional variation in stock prices.6 Such a
¬nding indicates that book value and earnings provide good starting points for predict-
ing the cash ¬‚ows that should drive prices.
That book value and earnings do not summarize the information in prices more com-
pletely should not be surprising. There are a number of factors that in¬‚uence prices that
accounting systems are not designed to capture well, including, for example, the value
of brand assets, growth opportunities, and research and development.
Explaining variation in the level of a ¬rm™s stock prices is one thing; explaining stock
returns, which depend on changes in those levels, is quite another. The latter is clearly
the more challenging task. It is necessary to not only identify factors that explain value,
but also to determine to what extent information about the factors became known to mar-
ket agents within the interval over which the price changes are measured. Researchers
have in fact had dif¬culty explaining more than a small fraction of the variance in stock
returns over years or shorter intervals. Earnings data are the most powerful of the factors
that have been studied, but even so, the explanatory power is relatively low. A combina-
tion of earnings and earnings changes (both expressed relative to price at the beginning
of the year) explains only about 5“15 percent of the variation in annual stock returns.7
To summarize, the picture that emerges is that earnings data provide somewhat noisy
indicators of value”good enough to approximate whether the stock price should be
closer to, say, $10 than $5, but not suf¬ciently precise to provide clear indications of
whether that price level might have changed by, for example, 10 percent rather than 5
percent over the past year. Thus, while the earnings number is a good starting point for
analysis, more information is certainly required to track stock prices.


Market agents can anticipate much of the information in earnings.
To say that ¬nancial statement data re¬‚ect much of the information in prices does not
necessarily mean that when those data are reported, they convey new information. In-
deed, market agents have access to a variety of information sources more timely than ¬-
nancial statements, and they use these sources to anticipate the data ultimately revealed
in ¬nancial statements.
Figure 13-1 describes the extent to which the key ¬nancial statement datum”earn-
ings”is anticipated by market agents.8 In the ¬gure, ¬rms are divided into 10 groups,
based on the extent to which quarterly earnings have changed from prior quarters. (The
earnings change is labeled SUE, for standardized unexpected earnings.) The importance
of the earnings information is evident in how much the stock price performance differs
across the groups. The top performers experience a three-month stock price increase
515
Equity Security Analysis




13-13 Part 3 Business Analysis and Valuation Applications




Figure 13-1 Stock Price Movements Before and After Quarterly Earnings
Announcements
Stock return relative to control group (%)

6
Pre-announcement period Post-announcement period
SUE
deciles
10
4
9

SUE
8
deciles
2 10
7 9
8
6
7
6
0
5
5
4
4 3
“2 2
1
3



2
“4




“6
1




“8
“40 “20 0 0 20 40
“60 60

Event time in trading days relative to earnings announcement day
Explanation: Firms are grouped into ten portfolios based on “standardized unexpected earnings,” or SUE: actual earn-
ings less a statistical forecast, and scaled by the standard deviation of past unexpected earnings. Stock returns (less
those for a size control group) are then cumulated over the 60 days before and after the earnings announcement for
each of the ten groups.



4.2 percent greater than a control group, while those at the bottom underperform by
6.1 percent. However, note that most (about 60 percent) of this movement occurs before
516 Equity Security Analysis




13-14
Equity Security Analysis




the week of the earnings announcement. This underscores that there are sources more
timely than earnings that re¬‚ect the same information that will ultimately be re¬‚ected in
earnings.
How does the market anticipate the earnings announcement? In some cases, manage-
ment itself reveals information. For example, management makes statements to the press
and in ¬nancial analysts™ meetings about the ¬rm™s progress. That information should
improve market agents™ ability to forecast earnings. Sometimes management will make
explicit statements about the range in which earnings are likely to be. Even in the ab-
sence of such direct information channels, however, it should be possible to anticipate
to some extent how well a ¬rm is performing. One could learn through discussions with
retail outlets, suppliers, competitors, and industry news sources. Even general informa-
tion about the state of the economy and the industry should permit more educated
guesses about how well a ¬rm is performing.
The ¬ndings summarized in Figure 13-1 offer an important lesson for security ana-
lysts. Speci¬cally, it™s not good enough to be aware of earnings as soon as they are an-
nounced. A good analyst also tracks more timely information sources.
A ¬nal comment on Figure 13-1 pertains to stock price movements after the earnings
are announced. Note that for those ¬rms with earnings increases, the stock prices con-
tinue to rise, and for those with earnings decreases, the prices continue to fall (relative
to the control group). This is the phenomenon that was mentioned brie¬‚y in the section
on market ef¬ciency. The ¬gure suggests that even though most of the response to earn-
ings occurs on a timely basis, some portion appears to be delayed.


Financial statement details matter.
Throughout our discussion of business strategy analysis, accounting analysis, ¬nancial
analysis, and prospective analysis, we drew on ¬nancial statement information beyond

<<

. 109
( 208 .)



>>