We use the earnings multiplier version of the dividend discount model to value the stock market because it is a theoretically correct model of value assuming a constant growth of dividends for an infinite time period, which is a reasonable assumption for the aggregate stock market.13Also, this valuation technique is consistently used in practice.
Recall that k and g are independent variables because k depends heavily on risk, whereas g is a function of the retention rate and the ROE. Therefore, this spread between k and g can and does change over time. The following equations imply an estimate of this spread at a point in time equal to the prevailing dividend yield:
Although the dividend yield gives an estimate of the size of the prevailing spread, it does not indicate the values for the two individual components (k and g) or what caused the change in the spread. More important, it says nothing about what the spread should be, which is the critical value that must be determined based upon estimating values for k and g.
P D
k g
P D k g
D P k g
j
j j
= −
= −
= −
1
1 1
1
/ /
/ Two-Part
Valuation Procedure
k D p g
= +
The ultimate objective of this microanalysis is to estimate the intrinsic market value for a major stock market series, such as the S&P Industrials Index. This estimation process has two equally important steps:
1. Estimating the future earnings per share for the stock market series
2. Estimating the appropriate earnings multiplier for the stock market series based on long- run estimates of k and g.14
Some analysts have concentrated on estimating the earnings for a market series with little con- sideration of changes in the earnings multiplier for the series. An investor who considers only the earnings for the series and ignores the earnings multiplier (i.e., the P/E ratio), assumes that the earnings multiplier will be relatively constant over time. If this were correct, stock prices would generally move in line with earnings. The fallacy of this assumption is obvious when one exam- ines data for the two components during the period from 1975 to 2000, as shown in Exhibit 13.8.
The year-end stock price is the closing value for the S&P Industrials Index on the last trading day of the year. The next column is the percentage change in price for the year. The earnings fig- ure is the earnings per share during the year for the S&P Industrials Index, and the next column shows the percentage change from the prior year. The fifth column is the historical earnings mul- tiplier at the end of the year, which is equal to the year-end value for the S&P Industrials Index divided by the historical earnings for that year. As an example, at the end of 1975, the S&P Industrials Index was equal to 100.88 and the earnings per share for the firms that made up the series were 8.58 for the 12 months ending 12/31/75. This implies an earnings multiplier of 11.76 (100.88/8.58). Although this may not be the ideal measure of the multiplier, it is consistent in its measurement and shows the changes in the relationship between stock prices and earnings over time. An alternative measure is the forward multiplier using next year’s earnings (i.e., stock price as of 12/31/75 versus earnings for the 12 months ending 12/31/76). This forward P/E series like- wise experiences substantial annual changes and is the multiple we will be estimating. Typically, it is also a smaller multiple because it considers future earnings that are generally higher.
There have been numerous striking examples where annual stock price movements for the S&P Industrials Index were opposite to earnings changes during the same year as follows:
➤ 1975 profit declined by 10 percent; stock prices increased by 32 percent.
➤ 1977 profits increased by 7 percent; stock prices declined by 12 percent.
➤ 1980 profits decreased by 1 percent; stock prices increased by over 27 percent.
➤ 1982 profits decreased by 21 percent; stock prices increased by 15 percent.
➤ 1984 profits increased by almost 23 percent; stock prices were basically unchanged.
➤ 1985 profits decreased by 15 percent; stock prices increased by about 26 percent.
➤ 1989 profits were almost unchanged; stock prices increased by over 25 percent.
➤ 1991 profits decreased by almost 32 percent; stock prices increased by over 27 percent.
➤ 1994 profits increased by almost 50 percent; stock prices were basically unchanged.
➤ 1997 profits increased about 2 percent; stock prices increased almost 29 percent.
➤ 1998 profits decreased by 9 percent; stock prices increased almost 32 percent.
During each of these years, the major influences on stock price movements came from changes in the earnings multiplier. The greater volatility of the multiplier series compared to the earnings per share series can be seen from the summary figures at the bottom of Exhibit 13.8 and from the graph of the earnings multiplier in Exhibit 13.9. The standard deviation of annual changes for the earnings multiplier series is much larger than the standard deviation of earnings changes Importance of
Both Components of Value
450 CHAPTER 13 STOCKMARKETANALYSIS
14Our emphasis will be on estimating future values for EPS, as well as k and g. We will show the relevant variables and provide a procedural framework, but the final estimate depends on the ability of the analyst.
ANNUAL CHANGES IN STOCK PRICES, CORPORATE EARNINGS, AND THE EARNINGS MULTIPLIER FOR S&P INDUSTRIALS INDEX: 1975–2000
YEAR-ENDSTOCK PERCENTAGE EARNINGS PERCENTAGE YEAR-END PERCENTAGE EARNINGS PERCENTAGE
YEAR PRICES CHANGE PERSHARE CHANGE EARNINGSMULTIPLE CHANGE MULTIPLEt+1 CHANGE
1975 100.88 31.9 8.58 –10.7 11.76 47.8 9.44 5.9
1976 119.46 18.4 10.69 24.6 11.17 –5.0 10.43 10.5
1977 104.71 –12.3 11.45 7.1 9.14 –18.2 8.03 –23.0
1978 107.21 2.4 13.04 13.9 8.22 –10.1 6.58 –18.1
1979 121.02 12.9 16.29 24.9 7.43 –9.6 7.51 14.1
1980 154.45 27.6 16.12 –1.0 9.58 28.9 9.23 22.9
1981 137.12 –11.2 16.74 3.8 8.19 –14.5 10.39 12.6
1982 157.62 15.0 13.20 –21.1 11.94 45.8 10.67 2.7
1983 186.17 18.1 14.77 11.9 12.60 5.5 10.28 –3.7
1984 186.36 0.1 18.11 22.6 10.29 –18.3 12.20 18.7
1985 234.56 25.9 15.28 –15.6 15.35 49.2 16.14 32.3
1986 269.93 15.1 14.53 –4.9 18.58 21.0 13.31 –17.5
1987 285.85 5.9 20.28 39.6 14.10 –24.1 10.75 –19.2
1988 321.26 12.4 26.59 31.1 12.08 –14.3 11.97 11.3
1989 403.49 25.6 26.83 0.9 15.04 24.5 16.29 36.1
1990 387.42 –4.0 24.77 –7.7 15.64 4.0 22.91 40.6
1991 492.72 27.2 16.91 –31.7 29.14 86.3 25.86 12.9
1992 507.46 3.0 19.05 12.7 26.64 –8.6 23.14 –10.5
1993 540.19 6.4 21.93 15.1 24.63 –7.5 16.45 –28.9
1994 547.51 1.4 32.83 49.7 16.68 –32.3 15.45 –6.1
1995 721.19 31.7 35.44 8.0 20.35 22.0 17.53 13.5
1996 869.97 20.6 41.15 16.1 21.14 3.9 20.65 17.8
1997 1,121.38 28.9 42.13 2.4 26.62 25.9 28.75 39.2
1998 1,479.16 31.9 38.37 –8.9 38.55 44.8 29.44 2.4
1999 1,841.92 24.5 50.25 31.0 36.66 –4.9 34.20 16.2
2000 1,527.86 –17.1 53.85 7.2 28.37 –22.6 NA NA
With Signs
Mean 13.2 8.5 17.7 8.4 15.9 7.3
Standard deviation 14.6 18.9 29.0 19.6
Coefficient of variation 1.1 2.2 3.4 2.7
Without Signs
Mean 16.6 16.3 23.1 17.5
Standard deviation 10.4 12.5 19.1 11.0
Coefficient of variation 0.6 0.8 0.8 0.6
EXHIBIT 13.8
NA—not available
Source: Financial Analysts Handbook (New York: Standard & Poor’s, 1998). Reprinted with permission.
452 CHAPTER 13 STOCKMARKETANALYSIS
40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00
40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00
Earnings Multiple
1977
1975 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year
EXHIBIT 13.9 YEAR-END EARNINGS MULTIPLE FOR THE S&P INDUSTRIALS INDEX BASED ON HISTORICAL EARNINGS: 1975–2000
(29.0 versus 18.9). The same is true for the relative volatility measures of the coefficient of vari- ability (3.4 versus 2.2). Also, the mean annual percentage change of the two multiplier series without signs has a larger mean annual percent change value (23.1 versus 16.3) and a larger stan- dard deviation of annual percentage change (19.1 versus 12.5). Therefore, these figures show that, of the two estimates required for market valuation, the earnings multiplier is the more volatile component. This is also true for the forward multiplier.
The point of this discussion is not to reduce the importance of the earnings estimate but to note that the estimation of future market value requires two separate estimates and both are important and necessary. Therefore, we will begin by considering a procedure for estimating aggregate earnings. Later, we discuss the procedure for estimating the aggregate market earnings multiplier.
ESTIMATING EXPECTED EARNINGS PER SHARE
The estimate of expected earnings per share for the market series will consider the outlook for the aggregate economy and for the corporate sector. This requires the following steps:
1. Estimate sales per share for a stock market series, such as the S&P Industrials Index. This estimate of sales involves a prior estimate of gross domestic product (GDP) because of the relationship between the sales of major industrial firms and this measure of aggregate eco- nomic activity. Therefore, prior to estimating sales per share, we will consider sources for an estimate of GDP.
2. Estimate the operating profit margin for the series, which equals operating profit divided by sales. Given the data available from Standard and Poor’s, we will define operating profit as earnings before interest, taxes, and depreciation (EBITDA).
3. Estimate depreciation per share for the next year.
4. Estimate interest expense per share for the next year.
5. Estimate the corporate tax rate for the next year.
These steps will lead to an estimate of net earnings per share that will be combined with an esti- mate of the forward earnings multiplier to arrive at an estimate of the current intrinsic value for the stock market series.
GDP is a measure of aggregate economic output or activity. Therefore, one would expect aggre- gate corporate sales to be related to GDP. We begin our estimate of sales for a stock market series with a prediction of nominal GDP from one of several banks or financial service firms that reg- ularly publish such estimates.15Using this estimate of nominal GDP, we can estimate corporate sales based on the historical relationship between S&P Industrials Index sales per share and aggregate economic activity (GDP).16
As noted, we will use a sales figure for an existing stock market series—the S&P Industrials Index.17The plot in Exhibit 13.10 shows the relationship between the annual percentage changes in GDP and S&P Industrials Index sales per share contained in Exhibit 13.11. Generally, there Estimating Sales
per Share for a Market Series Estimating Gross Domestic Product
15This would include projections by Standard & Poor’s appearing late in the year in The Outlook; and projections by sev- eral of the large investment firms, such as Goldman, Sachs, & Company (“The Pocket Chartroom”) or Merrill Lynch, as well as by banks. The Wall Street Journal publishes a survey of over 50 economists every 6 months that includes esti- mates of various interest rates, GDP, inflation, and the value of the dollar versus the Japanese yen. For a sample survey, see Fred R. Bleakley, “Economy’s Strength Is Seen Cooling in Second Half,” The Wall Street Journal, 1 July 1996, A2.
16Because GDP includes imports and exports, we also considered a pure domestic series entitled “Final Sales of Domestic Product.” Because an analysis of both series indicated that the GDP series provided superior regression results, it is used.
17Sales per share figures are available from 1945 in Standard & Poor’s Analysts Handbook (New York: Standard & Poor’s Corporation). Because the composite series include numerous companies of different sizes, all data are on a per-share basis. The book is updated annually, and some series are updated quarterly in a monthly supplement.
20.0
15.0
10.0
5.0
0.0
–5.0
20.0
15.0
10.0
5.0
0.0
–5.0
Percent Change in S&P Industrials Index Sales
2.0
0.0 4.0 6.0 8.0 10.0 12.0 14.0
Percent Change in GDP
EXHIBIT 13.10 SCATTER PLOT OF ANNUAL PERCENTAGE CHANGES IN S&P INDUSTRIALS INDEX SALES AND GDP
is a strong relationship between the two series whereby a large proportion of the percentage changes in S&P Industrials Index sales per share can be explained by percentage changes in nominal GDP. The relationship is not stronger because (1) the S&P Industrials Index sales series is more volatile than the GDP series and (2) the GDP series never experienced a decline. The equation for the least-squares regression line relating annual percentage changes (% ∆) in the two series for the period 1975–2000 is
$ ∆S&P Industrials Index Salest=–2.40 +1.16 (% ∆in Nominal GDPt) (–0.85)(4.06)
Adj. R2=0.36 454 CHAPTER 13 STOCKMARKETANALYSIS
NOMINAL GDP; FINAL SALES OF DOMESTIC PRODUCT, AND STANDARD AND POOR’S INDUSTRIALS INDEX SALES PER SHARE: 1975–2001
FINALSALES TO S&P INDUSTRIALSINDEX
NOMINALGDP PERCENTAGE DOMESTICPURCHASERS PERCENTAGE (DOLLARVALUE OF PERCENTAGE
YEAR (BILLIONS OFDOLLARS) CHANGE (BILLIONS OFDOLLARS) CHANGE SALES PERSHARE) CHANGE
1975 1,635.2 8.9 1,627.9 9.2 185.2 1.7
1976 1,823.9 11.5 1,809.1 11.1 202.7 9.4
1977 2,031.4 11.4 2,032.7 12.4 224.2 10.6
1978 2,295.9 13.0 2,296.2 13.0 251.3 12.1
1979 2,566.4 11.8 2,572.4 12.0 292.4 16.3
1980 2,795.6 8.9 2,816.8 9.5 327.4 12.0
1981 3,131.3 12.0 3,116.5 10.6 344.3 5.2
1982 3,259.2 4.1 3,294.7 5.7 333.9 –3.0
1983 3,534.9 8.5 3,592.3 9.0 334.1 0.1
1984 3,932.7 11.3 3,969.3 10.5 379.7 13.7
1985 4,213.0 7.1 4,305.4 8.5 398.4 4.9
1986 4,452.9 5.7 4,578.2 6.3 387.8 –2.7
1987 4,742.5 6.5 4,857.6 6.1 430.4 11.0
1988 5,108.3 7.7 5,196.1 7.0 486.9 13.1
1989 5,489.1 7.5 5,542.1 6.7 541.4 11.2
1990 5,803.2 5.7 5,860.1 5.7 594.6 9.8
1991 5,986.2 3.2 6,007.1 2.5 586.9 –1.3
1992 6,318.9 5.6 6,331.7 5.4 601.4 2.5
1993 6,642.3 5.1 6,681.7 5.5 603.6 0.4
1994 7,054.3 6.2 7,078.9 5.9 626.3 3.8
1995 7,400.5 4.9 7,451.7 5.3 676.6 8.0
1996 7,813.2 5.6 7,872.1 5.6 701.9 3.7
1997 8,318.4 6.5 8,344.8 6.0 750.7 7.0
1998 8,781.5 5.6 8,860.1 6.2 755.5 0.6
1999 9,268.6 5.5 9,460.9 6.8 812.0 7.5
2000 9,872.9 6.5 10,187.5 7.7 853.9 5.2
2001 10,205.6 3.4 10,595.5 4.0 NA NA
Average 7.4 7.6 6.3
EXHIBIT 13.11
Source: Economic Report of the President, 2002 (Washington, DC: U.S. Government Printing Office, 2002); and Financial Analysts Handbook (New York: Standard & Poor’s, 2001). Reprinted with permission.
These results indicate that about 36 percent of the variance in percentage changes in S&P Indus- trials Index sales can be explained by percentage changes in the nominal GDP. Thus, given an estimate of the expected percentage change in nominal GDP for next year, we can estimate the percentage change in sales for the S&P Industrials Index series and therefore the amount of sales per share. For example, assume the consensus estimate by economists is that nominal GDP next year will increase by approximately 6 percent (a 3 percent increase in real GDP plus 3 percent inflation). This estimate, combined with the regression results, would imply the following esti- mated increase in S&P Industrials Index sales:
% ∆S&P Industrials Index Sales =–0.024 +1.16 (0.06)
=0.046
=4.6%
Notably, this is referred to as a point estimate of sales because it is based on a point estimate of GDP. Although we know there is actually a distribution of estimates for GDP, we have used the mean value, or expected value, as our point estimate. In actual practice, you would probably con- sider several estimates and assign probabilities to each of them.
Once sales per share for the market series have been estimated, the difficult estimate is the profit margin. Three alternative procedures are possible depending on the desired level of aggregation.
The first is a direct estimate of the net profit margin based on recent trends. As shown in Exhibit 13.12, the net profit margin series is quite volatile because of changes in depreciation, interest, and the tax rate over time. As such, it is the most difficult series to estimate.
The second procedure would attempt to estimate the net before tax (NBT) profit margin. Once the NBT margin is derived, a separate estimate of the tax rate is obtained based on recent tax rates and current government tax pronouncements.
The third method estimates an operating profit margin, defined as earnings before interest, taxes, and depreciation (EBITDA), as a percentage of sales. Because this measure of operat- ing earnings as a percentage of sales is not influenced by changes in depreciation allowances, interest expense, or tax rates, it should be a more stable series compared to either the net profit margin or net before tax margin series. Our analysis begins with estimating this operating profit margin series.
After we estimate this operating profit margin, we will multiply it by the sales estimate to derive a dollar estimate of operating earnings (EBITDA). Subsequently, we will derive separate estimates of depreciation and interest expenses, which are subtracted from the EBITDA to arrive at earnings before taxes (EBT.) Finally, we estimate the expected tax rate (T) and multiply EBT times (1 – T) to get our estimate of net income. The following sections discuss the details of esti- mating earnings per share beginning with the operating profit margin.
Finkel and Tuttle hypothesized that the following four variables affected the aggregate profit margin:18
1. Capacity utilization rate 2. Unit labor costs 3. Rate of inflation 4. Foreign competition Estimating
Aggregate Operating Profit Margin Alternative Estimates of Corporate Net Profits
18Sidney R. Finkel and Donald L. Tuttle, “Determinants of the Aggregate Profit Margin,” Journal of Finance 26, no. 5 (December 1971): 1067–1075.
Capacity Utilization Rate One would expect a positive relationship between the capacity utilization rate and the profit margin because if production increases as a proportion of total capacity, there is a decrease in per-unit fixed production costs and fixed financial costs. The rela- tionship may not be completely linear at very high rates of capacity utilization because operat- ing diseconomies are introduced as firms are forced to use marginal labor and/or older plant and equipment to reach the higher capacity. The figures in Exhibit 13.13 indicate that capacity uti- lization ranged from a peak of over 87 percent in 1978 to a trough of about 64 percent in 1982 and less than 70 percent during the recent recession of 2000–2002.
Unit Labor Cost The change in unit labor cost is a compound effect of two individual fac- tors: (1) changes in wages per hour and (2) changes in worker productivity. Wage costs per hour typically increase every year by varying amounts depending on the economic environment. As shown in Exhibit 13.13, the annual percentage increase in compensation per hour varied from 456 CHAPTER 13 STOCKMARKETANALYSIS
S&P INDUSTRIALS INDEX SALES PER SHARE AND COMPONENTS OF OPERATING PROFIT MARGIN: 1977–2000
EBITDAa DEPRECIATION INTEREST INCOME TAX NET INCOME
SALES PER PERCENT PER PERCENT PER PERCENT PER TAX PER PERCENT
YEAR PERSHARE SHARE OFSALES SHARE OFSALES SHARE OFSALES SHARE RATE SHARE OFSALES
1977 224.24 34.34 15.31 8.53 3.80 3.22 1.44 11.14 49.31 11.45 5.11
1978 251.32 38.63 15.37 9.64 3.84 3.81 1.52 12.14 48.21 13.04 5.19
1979 292.38 45.71 15.63 10.82 3.70 4.55 1.57 14.02 46.26 16.29 5.57
1980 327.36 48.11 14.70 12.37 3.78 5.95 1.82 13.67 45.89 16.12 4.92
1981 344.31 51.00 14.81 13.82 4.01 7.49 2.18 12.95 43.62 16.74 4.86
1982 333.86 47.68 14.28 15.30 4.58 5.23 2.47 10.95 45.34 13.20 3.95
1983 334.07 50.18 15.02 15.67 4.69 7.62 2.25 12.12 45.07 14.77 4.42
1984 379.70 57.11 15.04 16.31 4.30 8.54 2.25 14.15 43.86 18.11 4.77
1985 398.42 56.39 14.15 18.19 4.57 9.24 2.32 13.68 47.24 15.28 3.84
1986 387.76 54.70 14.11 19.41 5.01 9.75 2.51 11.01 43.11 14.53 3.75
1987 430.35 64.59 15.01 20.21 4.70 10.14 2.36 13.96 40.77 20.23 4.71
1988 486.92 80.02 16.43 23.59 4.84 14.84 3.05 15.00 36.07 26.59 5.46
1989 541.38 85.56 15.80 24.21 4.47 18.79 3.47 15.73 36.96 26.83 4.96
1990 594.55 87.52 14.72 26.31 4.43 20.17 3.39 16.27 39.64 24.77 4.17
1991 586.86 75.35 12.84 27.50 4.59 18.74 3.19 12.20 41.91 16.91 2.88
1992 601.39 76.74 12.76 29.48 4.90 16.20 2.69 12.01 38.61 19.05 3.17
1993 603.62 78.67 13.03 28.72 4.76 14.66 2.43 13.36 37.56 21.93 3.63
1994 626.26 94.06 15.02 29.58 4.72 12.77 2.04 18.88 36.51 32.83 5.24
1995 676.62 103.50 15.30 33.06 4.89 14.21 2.10 20.79 36.97 35.44 5.24
1996 701.91 115.45 16.45 36.11 5.14 14.32 2.04 23.87 36.71 41.15 5.86
1997 750.71 123.76 16.49 39.77 5.30 14.44 1.98 27.02 39.07 42.13 5.61
1998 755.48 106.68 14.12 32.89 4.35 14.56 1.93 20.86 35.22 38.37 5.08
1999 812.00 134.86 16.61 41.70 5.14 15.50 1.91 27.41 35.29 50.25 6.19
2000 853.86 146.03 17.10 43.50 5.09 16.39 1.92 32.29 37.49 53.85 6.31
EXHIBIT 13.12
aThis is used as an estimate of operating earnings.
Source: Financial Analysts Handbook (New York: Standard & Poor’s, 2001). Reprinted with permission.
2.1 percent to 10.8 percent. If workers did not become more productive, this increase in per-hour wage costs would be the increase in per-unit labor cost. Fortunately, because of advances in tech- nology and greater mechanization, the worker units of output per hour (the measure of labor pro- ductivity) have increased over time—our labor force has become more productive. If wages per hour increase by 5 percent and labor productivity increases by 5 percent, there would be no increase in unit labor costs because the workers would offset wage increases by producing more.
Therefore, the increase in per-unit labor cost is a function of the percentage change in hourly wages minus the increase in productivity during the period. The actual relationship typically is not this exact due to measurement problems, but it is quite close as indicated by the data in VARIABLES THAT AFFECT THE AGGREGATE PROFIT MARGIN: CAPACITY UTILIZATION RATE, PERCENTAGE CHANGE IN COMPENSATION, PRODUCTIVITY, UNIT LABOR COST, AND CONSUMER PRICE INDEX: 1975–2001(P)
COMPENSATION/ OUTPUT/ UNIT LABOR
WORK HOURS WORK HOURS COSTS
UTILIZATION PERCENTAGE PERCENTAGE PERCENTAGE RATE OF
YEAR RATE(MFG) CHANGE CHANGE CHANGE INFLATION
1975 72.2 10.1 2.7 7.2 6.9
1976 78.0 8.6 3.7 4.7 4.9
1977 82.1 8.0 1.5 6.4 6.7
1978 87.3 8.9 1.3 7.6 9.0
1979 83.0 9.5 –0.4 10.0 13.3
1980 78.6 10.8 –0.3 11.1 12.5
1981 71.9 9.7 1.2 8.3 8.9
1982 63.9 7.5 –0.6 8.1 3.8
1983 74.5 4.3 4.5 –0.2 3.8
1984 77.1 4.3 2.2 2.1 3.9
1985 76.0 4.7 1.3 3.3 3.8
1986 76.3 5.2 3.0 2.1 1.1
1987 80.3 3.8 0.4 3.4 4.4
1988 83.7 4.5 1.3 3.2 4.4
1989 80.3 2.7 0.8 1.9 4.6
1990 75.6 5.5 1.1 4.3 6.1
1991 74.9 4.9 1.2 3.6 3.1
1992 77.8 5.3 3.7 1.6 2.9
1993 80.5 2.2 0.5 1.7 2.7
1994 83.6 2.1 1.3 0.8 2.7
1995 81.0 2.1 0.9 1.2 2.5
1996 81.4 3.1 2.5 0.5 3.3
1997 82.5 3.0 2.0 0.9 1.7
1998 80.7 5.4 2.6 2.7 1.6
1999 81.2 4.4 2.3 2.0 2.7
2000 77.7 6.5 3.3 3.1 3.4
2001 69.7 4.5 1.1 3.3 1.6
EXHIBIT 13.13
Source: Economic Report of the President, 2002 (Washington, D.C.: U.S. Government Printing Office, 2002).
Exhibit 13.13. For example, during 1983, productivity increased by slightly more than the hourly compensation did so there was basically no change in unit labor cost. In contrast, during 1980, wage rates increased by 10.8 percent, productivity declined by 0.3 percent because of the recession, and, therefore, unit labor costs increased by 11.1 percent. Because unit labor is the major variable cost of a firm, one would expect a negative relationship between the operating profit margin and percentage changes in unit labor cost—that is, a small (below-average) change in unit labor cost, similar to what we experienced during the mid-1990s (1994–97), should cor- respond to an above-average operating profit margin.
Rate of Inflation The precise effect of inflation on the aggregate profit margin is unresolved.
Finkel and Tuttle hypothesized a positive relationship between inflation and the profit margin for several reasons. First, it was contended that a higher level of inflation increases the ability of firms to pass higher costs on to the consumer and thereby raise their profit margin. Second, assuming the classic demand-pull inflation, the increase in prices would indicate an increase in general eco- nomic activity, which typically is accompanied by higher margins. Finally, an increase in the rate of inflation might stimulate consumption as individuals attempt to shift their holdings from finan- cial assets to real assets, which would contribute to an expansion.
In contrast, many observers doubt that most businesses can consistently increase prices in line with rising costs. Assume a 5 percent rate of inflation that impacts labor and material costs. The question is whether all firms can completely pass these cost increases along to their customers.
If a firm increases prices at the same rate as cost increases, the result will be a constant profit margin, not an increase. Only if a firm can raise prices by more than cost increases can it increase its margin. Many firms are not able to raise prices in line with increased costs because of the elasticity of demand for their products.19Such an environment will cause the profit margin to decline. Given the alternative scenarios, it is contended that most firms will not be able to increase their profit margins or even hold them constant. Because many firms will experience lower profit margins during periods of inflation, it is expected that the aggregate profit margin will probably decline when there is an increase in the rate of inflation.
Given the contrasting expectations, one would need to consider the empirical evidence to determine the relationship between inflation and the operating profit margin.
Foreign Competition Finkel and Tuttle contend that export markets are more competitive than domestic markets so export sales are made at a lower margin. This implies that lower exports by U.S. firms would increase profit margins. In contrast, Gray believed that only exports between independent firms should be considered and they should be examined relative to total output exported.20Further, he felt that imports could have an important negative impact on the operating profit margin because they influence the selling price of all competing domestic prod- ucts. Therefore, there is a divergence of expectations regarding the ultimate effect of foreign trade on the operating profit margin, so it is likewise an empirical question.
Analysis of the annual data for the period 1977 to 1997 by the authors confirmed that the rela- tionship between the operating profit margin and the capacity utilization rate was always signif- icant and positive, whereas the relationship between the unit labor cost and the operating profit margin was always negative and significant. Alternatively, the rate of inflation and foreign trade variables were never significant in the multiple regression. Finally, the simple correlation between the profit margin and inflation was consistently negative.
458 CHAPTER 13 STOCKMARKETANALYSIS
19An extreme example of this inability is regulated industries that may not be able to raise prices at all until after lengthy hearings before regulatory agencies. Even then, the increase in rates may not match the cost increase.
20H. Peter Gray, “Determinants of the Aggregate Profit Margin: A Comment,” Journal of Finance 31, no. 1 (March 1976):
163–165.