<<

. 84
( 208 .)



>>

390 Prospective Analysis: Forecasting




10-16
Prospective Analysis: Forecasting




indicates that recent increases in pro¬tability should usually not be extrapolated fully
into the future”for Nordstrom™s EPS, only 44 percent of such changes, on average, tend
to persist.


SUMMARY
Forecasting represents the ¬rst step of prospective analysis and serves to summarize the
forward-looking view that emanates from business strategy analysis, accounting analy-
sis, and ¬nancial analysis. Although not every ¬nancial statement analysis is accompa-
nied by such an explicit summarization of a view of the future, forecasting is still a key
tool for managers, consultants, security analysts, investment bankers, commercial bank-
ers and other credit analysts, and others.
The best approach to forecasting future performance is to do it comprehensively”
producing not only an earnings forecast but a forecast of cash ¬‚ows and the balance
sheet as well. Such a comprehensive approach provides a guard against internal incon-
sistencies and unrealistic implicit assumptions. The approach described here involved
line-by-line analysis, so as to recognize that different items on the income statement and
balance sheet are in¬‚uenced by different drivers. Nevertheless, it remains the case that a
few key projections”such as sales growth and pro¬t margin”usually drive most of the
projected numbers.
The forecasting process should be embedded in an understanding of how various ¬-
nancial statistics tend to behave on average, and what might cause a ¬rm to deviate from
that average. Absent detailed information to the contrary, one would expect sales and
earnings numbers to persist at their current levels, adjusted for overall trends of recent
years. However, rates of return on investment (ROEs) tend, over several years, to move
from abnormal to normal levels”close to the cost of equity capital”as the forces of
competition come to play. Pro¬t margins also tend to shift to normal levels, but for this
statistic, “normal” varies widely across ¬rms and industries, depending on the levels of
asset turnover and leverage. Some ¬rms are capable of creating barriers to entry that en-
able them to ¬ght these tendencies toward normal returns, even for many years, but such
¬rms are the unusual cases.
For some purposes, including short-term planning and security analysis, forecasts for
quarterly periods are desirable. One important feature of quarterly data is seasonality; at
least some seasonality exists in the sales and earnings data of nearly every industry. An
understanding of a ¬rm™s within-year peaks and valleys is a necessary ingredient of a
good forecast of performance on a quarterly basis.
There are a variety of contexts (including but not limited to security analysis) where
the forecast is usefully summarized in the form of an estimate of the ¬rm™s value”an
estimate that, after all, can be viewed as the best attempt to re¬‚ect in a single summary
statistic the manager™s or analyst™s view of the ¬rm™s prospects. That process of convert-
ing a forecast into a value estimate is labeled valuation. It is to that topic that we turn in
the following chapter.
391
Prospective Analysis: Forecasting




10-17 Part 2 Business Analysis and Valuation Tools




APPENDIX:
The Behavior of Components of ROE
In Figure 10-2, we show that the ROEs tend to be mean reverting. In this appendix, we
show the behavior of the key components of ROE”operating ROA, operating margin,
operating asset turnover, spread, and net ¬nancial leverage. These ratios are computed
using the same portfolio approach described in the chapter, based on the data for all U.S.
industrial ¬rms for the time period 1978“1998.

Figure A-1 Behavior of Operating ROA for U.S. Industrial Firms
for 1978“1998

30%
Top Fifth
20%
Operating ROA




Second Fifth
10%
Third Fifth
0%
“ 10% Fourth Fifth
“ 20% Bottom Fifth
“ 30%
1 2 3 4 5 6 7 8 9 10
Year


Figure A-2 Behavior of Operating Margin for U.S. Industrial Firms
for 1978“1998
Operating Profit Margin




20%
Top Fifth
10% Second Fifth
Third Fifth
0%
Fourth Fifth
“10%
Bottom Fifth
“20%
1 2 3 4 5 6 7 8 9 10
Year
392 Prospective Analysis: Forecasting




10-18
Prospective Analysis: Forecasting




Figure A-3 Behavior of Operating Asset Turnover for U.S. Industrial Firms
for 1978“1998
Operating Assets to Sales




2.00
Top Fifth
1.50 Second Fifth
Third Fifth
1.00
Fourth Fifth
0.50
Bottom Fifth
0.00
1 2 3 4 5 6 7 8 9 10
Year

Figure A-4 Behavior of Operating Asset Turnover for U.S. Industrial Firms
for 1978“1998

0.30
Top Fifth
0.20
Second Fifth
0.10
Spread




0.00 Third Fifth
“ 0.10
Fourth Fifth
“ 0.20
Bottom Fifth
“ 0.30
1 2 3 4 5 6 7 8 9 10
Year

Figure A-5 Behavior of Operating Asset Turnover for U.S. Industrial Firms
for 1978“1998

6.0
Net Financial Leverage




Top Fifth
4.0
Second Fifth
2.0
0.0 Third Fifth
“ 2.0 Fourth Fifth
“ 4.0
Bottom Fifth
“ 6.0
1 2 3 4 5 6 7 8 9 10
Year
393
Prospective Analysis: Forecasting




10-19 Part 2 Business Analysis and Valuation Tools




DISCUSSION QUESTIONS
1. Merck is one of the largest pharmaceutical firms in the world. In the period 1985 to
1995 Merck consistently earned higher ROEs than the pharmaceutical industry as a
whole. As a pharmaceutical analyst, what factors would you consider to be important
in making projections of future ROEs for Merck? In particular, what factors would
lead you to expect Merck to continue to be a superior performer in its industry, and
what factors would lead you to expect Merck™s future performance to revert to that
of the industry as a whole?
2. John Right, an analyst with Stock Pickers Inc., claims: “It is not worth my time to
develop detailed forecasts of sales growth, profit margins, etcetera, to make earnings
projections. I can be almost as accurate, at virtually no cost, using the random walk
model to forecast earnings.” What is the random walk model? Do you agree or dis-
agree with John Right™s forecast strategy? Why or why not?
3. Which of the following types of businesses do you expect to show a high degree of
seasonality in quarterly earnings? Explain why.
• a supermarket
• a pharmaceutical company
• a software company
• an auto manufacturer
• a clothing retailer
4. What factors are likely to drive a firm™s outlays for new capital (such as plant, prop-
erty, and equipment) and for working capital (such as receivables and inventory)?
What ratios would you use to help generate forecasts of these outlays?
5. How would the following events (reported this year) affect your forecasts of a firm™s
future net income?
• an asset write-down
• a merger or acquisition
• the sale of a major division
• the initiation of dividend payments
6. Consider the following two earnings forecasting models:
Et ( EPS t + 1 )
Model 1: = EPS t
5
‘ EPS t
1
Et ( EPS t + 1 ) --
-
Model 2: =
5
t=1
E t ( EPS ) is the expected forecast of earnings per share for year t+1, given informa-
tion available at t. Model 1 is usually called a random walk model for earnings,
whereas Model 2 is called a mean-reverting model. The earnings per share for Ford
Motor Company for the period 1990 to 1994 are as follows:
394 Prospective Analysis: Forecasting




10-20
Prospective Analysis: Forecasting




Year 1990 1991 1992 1993 1994
EPS $0.93 $(2.40) $(0.73) $2.27 $4.97

a. What would be the 1995 forecast for earnings per share for each model?
b. Actual earnings per share for Ford in 1995 were $3.58. Given this information,
what would be the 1996 forecast for earnings per share for each model? Why do
the two models generate quite different forecasts? Which do you think would bet-
ter describe earnings per share patterns? Why?
7. Joe Fatcat, an investment banker, states: “It is not worth my while to worry about de-
tailed long-term forecasts. Instead, I use the following approach when forecasting
cash flows beyond three years. I assume that sales grow at the rate of inflation, capital
expenditures are equal to depreciation, and that net profit margins and working cap-
ital to sales ratios stay constant.” What pattern of return on equity is implied by these
assumptions? Is this reasonable?


NOTES
1. See Patricia O™Brien, “Analysts™ Forecasts as Earnings Expectations,” Journal of Account-
ing and Economics (January 1988): 53“83.
2. See George Foster, “Quarterly Accounting Data: Time Series Properties and Predictive

<<

. 84
( 208 .)



>>