<<

. 82
( 208 .)



>>

for 1979“1998
40%
Top Fifth
20%
Second Fifth
Third Fifth
ROE




0%
Fourth Fifth
“20%
Bottom Fifth
“40%
1 2 3 4 5 6 7 8 9 10
Year


THE BEHAVIOR OF COMPONENTS OF ROE. The behavior of rates of return on eq-
uity can be analyzed further by looking at the behavior of its key components. Recall
from Chapter 9 that ROEs and pro¬t margins are linked as follows:
Operating ROA + ( Operating ROA “ Net interest rate after tax )
ROE =
— Net financial leverage
ROE =
NOPAT margin — Operating asset turnover + Spread
ROE =
— Net financial leverage
ROE =
The time-series behavior of the components of ROE for U.S. industrial companies for
1979“1998 are shown in a series of ¬gures in the appendix to this chapter. The major
conclusions from these ¬gures are: Operating asset turnover tends to be rather stable, in
part because it is so much a function of the technology of the industry. Net ¬nancial le-
verage also tends to be stable, simply because management policies on capital structure
aren™t often changed. NOPAT margin and spread stand out as the most variable compo-
nent of ROE; if the forces of competition drive abnormal ROEs toward more normal lev-
els, the change is most likely to arrive in the form of changes in pro¬t margins and the
spread. The change in spread is itself driven by changes in NOPAT margin, since the cost
of borrowing is likely to remain stable if leverage remains stable.
To summarize, pro¬t margins, like ROEs, tend to be driven by competition to “nor-
mal” levels over time. However, what constitutes normal varies widely according to the
technology employed within an industry and the corporate strategy pursued by the
¬rm”both of which in¬‚uence turnover and leverage.5 In a fully competitive equilib-
rium, pro¬t margins should remain high for ¬rms that must operate with a low turnover,
and vice versa.
The implication of the above discussion of rates of return and margins is that a rea-
sonable point of departure for a forecast of such a statistic should consider more than
just the most recent observation. One should also consider whether that rate or margin
381
Prospective Analysis: Forecasting




10-7 Part 2 Business Analysis and Valuation Tools




is above or below a normal level. If so, then absent detailed information to the contrary,
one would expect some movement over time to that norm. Of course, this central ten-
dency might be overcome in some cases”for example, where the ¬rm has erected bar-
riers to competition that can protect margins, even for extended periods. The lesson from
the evidence, however, is that such cases are unusual.
In contrast to rates of return and margins, it is reasonable to assume that asset turn-
over, ¬nancial leverage, and net interest rate remain constant over time. Unless there is
an explicit change in technology or ¬nancial policy being contemplated for future peri-
ods, a reasonable point of departure for assumptions for these variables is the current pe-
riod level.
As we proceed below with the steps involved in producing a detailed forecast, the
reader will note that we draw on the above knowledge of the behavior of accounting
numbers to some extent. However, it is important to keep in mind that a knowledge of
average behavior will not ¬t all ¬rms well. The art of ¬nancial statements analysis re-
quires not only knowing what the “normal” patterns are but also expertise in identifying
those ¬rms that will not follow the norm.


ELEMENTS OF THE DETAILED FORECAST
Here we summarize steps that could be followed in producing a comprehensive forecast.
The discussion assumes that the ¬rm being analyzed is among the vast majority for
which the forecast would reasonably be anchored by a sales forecast.


The Sales Forecast
The ¬rst step in most forecasting exercises is the sales prediction. There is no gener-
ally accepted approach to forecasting sales; the approach should be tailored to the con-
text and should re¬‚ect the factors considered in the prior steps of the analysis. For
example, for a large retail ¬rm, a sales forecast would normally consider the prior year™s
sales, increases due purely to expansion of the number of retail outlets, and “comparable
store growth,” which captures growth in sales in already-existing stores. The forecast of
growth might consider such factors as customer acceptance of new product lines, mar-
keting plans, changes in pricing strategies, competitors™ behavior, and the expected state
of the economy. Another possible approach”and one that may represent the only feasi-
ble approach when little history exists”is to estimate the size of the target market,
project the degree of market penetration, and then consider how quickly that degree of
penetration can be achieved.
Table 10-1 presents a forecast of sales and earnings for Nordstrom for the ¬scal year
ending January 31, 2000 (¬scal 1999), produced by an analyst at Morgan Stanley in De-
cember 1998. At the time these forecasts were made, the analyst had information on
Nordstrom™s actual performance for the ¬rst three quarters of 1998 but not for the entire
1998 ¬scal year. As a result, some of the assumptions are driven by the actual perfor-
382 Prospective Analysis: Forecasting




10-8
Prospective Analysis: Forecasting




Table 10-1 Analyst™s Forecast of 1999 Income Statement for Nordstrom

1999 Forecast 1998 Actual
.................................. ..................................
$ Millions % of Sales $ Millions % of Sales
.........................................................................................................................
Total sales 5627 100.0 5028 100.0
Cost of sales 3760 66.8 3345 66.5
SG&A expense 1537 27.3 1405 28.0
Other income 107 2.1 107 2.1
Earnings before interest and taxes 437 7.8 385 7.7
Net operating pro¬t after taxes
(NOPAT) 275 4.9 238 4.7
Net interest expense after taxes 37 0.7 31 0.6
Net Income 238 4.2 207 4.1
.........................................................................................................................
Tax expense forecasted by the analyst has been allocated to operations and interest expense.
Source: “Nordstrom: Shareholders should be as satisfied as customers,” by B. Missett et al., Morgan Stanley Dean Witter,
December 2, 1998.


mance in 1997 rather than the performance in 1998. The actual results for 1998 are also
shown in the table for comparison.
The 1999 sales growth forecast is signi¬cantly larger than the 6 percent growth rate
in 1998, and similar to the growth rate in 1997. In commenting on the forecast, the ana-
lyst recognized at least two factors that might support a more optimistic outlook on sales.
He viewed the comparable store sales growth in 1998 to be unusually low (in fact, neg-
ative) as resulting from Nordstrom™s focus in that year on better inventory management
and reducing markdowns. The analyst expected the comparable store sales in 1999 to
bounce back to a higher level, 3 percent, but below the level of 4 percent in 1997. The
rest of the sales growth in 1999 is forecasted to come from opening new stores.
The Morgan Stanley forecast appears to be based largely on analysis that views the
¬rm as a whole. An alternative approach”not feasible for all ¬rms”is to build a sales
forecast on a product line-by-product line basis, or by major business segments of a ¬rm


The Forecast of Expenses and Earnings
Expenses should be forecast item by item, since different expenses may be driven by dif-
ferent factors. However, most major expenses are clearly related to sales and are there-
fore naturally framed as fractions of sales. These include cost of sales and SG&A
expenses. R&D need not track current sales closely; however, R&D generally tracks sales
at least roughly over the long run. Other expenses are more closely related to drivers
other than sales. Interest expense is driven by debt levels and interest rates. Depreciation
expense should be forecast in a way consistent with the ¬rm™s depreciation policy; under
straight-line depreciation, the expense would tend to be a fairly stable fraction of begin-
383
Prospective Analysis: Forecasting




10-9 Part 2 Business Analysis and Valuation Tools




ning depreciable plant. Tax provisions are driven by pretax income and factors (such as
tax rates applicable to certain foreign subsidiaries) that have a permanent impact on tax
payments. Equity in the income of af¬liates is determined by whatever drives the af¬li-
ate™s earnings.
In the case of Nordstrom, the two largest expenses”cost of sales and SG&A expense
”were forecast by the analyst as fractions of sales (see Table10-1). Cost of sales was
expected to marginally decrease as a fraction of sales, causing the gross margin percent-
age to increase to 33.5 percent from 33.2 percent. This projected increase is a continua-
tion of the margin improvement in 1998, a re¬‚ection of the view that management™s new
strategy will continue to cut purchase costs. SG&A is also expected to decrease from 28
percent to 27.3 percent. Here, the analyst assumed that the SG&A costs increased tem-
porarily in 1998 and that they will return in 1999 to the levels experienced in 1997 and
1996.
The analyst appears to assume that the net interest expense and other income will re-
main approximately unchanged as a percent of sales. Tax expense is projected as 39.4
percent of pretax income”35 percent for federal and 4.4 percent for state taxes.
The forecasts of sales and expenses produce an expected net margin of 4.2 percent, a
small improvement over the 1998 net margin of 4.1 percent. The analyst is betting that
Nordstrom™s emphasis on cost cutting and value-based management will continue to
produce improvements in the company™s bottom line.


The Forecast of Balance Sheet Accounts
Since various balance sheet accounts may be driven by different factors, they are usually
best forecast individually. However, several asset accounts, including operating working
capital accounts and operating long-term assets, are driven over the long run by sales ac-
tivity. Thus, these accounts can be forecast as fractions of sales, allowing for any expect-
ed changes in the ef¬ciency of asset utilization. If management plans for capital
expenditures are known, they would clearly be useful in forecasting plant assets. Liabil-
ity and equity accounts will depend on a variety of factors, including policies on capital
structure, dividends, and stock repurchases.
While it is useful to project balance sheet accounts in detail for some purposes, it may
be adequate sometimes to project a summary balance sheet that contains major catego-
ries of assets and liabilities along the lines discussed in the ¬nancial analysis chapter”
operating working capital, net operating long-term assets, net debt, and shareholders™
equity. Such projections are useful for valuing a company. One simple approach to pro-
jecting a summary balance sheet is as follows:
First, one can project operating working capital and operating long-term assets by
making assumptions about these two asset categories as a fraction of sales. The sum of
these two items is net operating assets. Next, by making an assumption about net ¬nan-
cial leverage (ratio of net debt to equity), one can project the amount of debt and equity
needed to support these net operating assets. Therefore, to project summary balance
384 Prospective Analysis: Forecasting




10-10
Prospective Analysis: Forecasting




sheets, one needs to make only three critical assumptions: ratio of operating working
capital to sales, ratio of operating long-term assets to sales, and the ratio of net debt to
equity.
Table 10-2 presents Morgan Stanley Dean Witter analysts™ forecast (as of December
1998) of the 1999 balance sheet for Nordstrom. The balance sheet accounts on the asset
side are primarily driven by the analyst™s assumptions on Nordstrom™s turnover ratios as-
sets. The analyst assumed, for 1999 relative to the levels achieved in 1998, higher levels
of accounts receivable and inventory and lower levels of accounts payable . Since these
forecasts were made prior to the release of the fourth quarter results for 1998, they do
not re¬‚ect the unexpected signi¬cant reduction in working capital achieved by Nord-
strom in that quarter. The forecast on net property, plant, and equipment is based on an
assumption that capital expenditures will be slightly lower in 1999 relative to 1998, and
that the depreciation expense will remain the same as a proportion of gross PP&E. Recall
that the analyst made an assumption regarding the number of new stores that will be
opened in 1999 in making the sales forecast, and it is presumably the basis for the capital

Table 10-2 Analyst™s Forecast of Nordstrom™s 1999 Balance Sheet

1999 Forecast 1998 Actual
.................................. ..................................

Net Operating Assets $ Millions % of Sales $ Millions % of Sales
.........................................................................................................................
Accounts receivable 771 13.7 587 11.7
Inventory 902 16.0 750 14.9
Other operating current assets 96 1.7 102 2.0
Accounts payable (373) (6.6) (340) (6.8)
Other operating current liabilities (338) (6.0) (287) (5.7)
Operating working capital 1058 18.8 812 16.1
PP&E, net 1479 26.3 1362 27.1
Other long-term assets 18 0.3 73 1.5
Other operating long-term liabilities (179) (3.2) (225) (4.5)
Net operating long-term assets 1318 23.4 1210 24.1
Total net operating assets 2376 42.2 2022 40.2

% of Net % of Net
Net Capital $ Millions Capital $ Millions Capital
.........................................................................................................................
Total short-term and long-term debt 959 40.4 946 46.8
Cash and short-term investments (15.4) (0.7) (241) (11.9)
Net debt 945 39.7 705 34.9
Total shareholders™ equity 1431 60.3 1317 65.1
Total net capital 2376 100.0 2022 100.0
.........................................................................................................................
Source: “Nordstrom: Shareholders should be as satis¬ed as customers,” by B. Missett et al., Morgan Stanley Dean
Witter, December 2, 1998.
385

<<

. 82
( 208 .)



>>