At this point, we have determined that industry analysis helps an investor select profitable invest- ment opportunities and we have completed a thorough macroanalysis of the industry. Our next question is, How do we go about valuing an industry and estimating the rate of return that an investment in it will provide? Again, we consider the two equity valuation approaches introduced in Chapter 12—the present value of cash flows and the relative valuation ratios. Beginning with the present value of cash flow models, we demonstrate the DDM with the two-stage growth assumption and then assume constant growth for the retail drugstore industry. Following this, we consider the present value of free cash flow (FCF) model. Subsequently, we will analyze the alternative relative valuation techniques with the price/earnings ratio and analysis of the P/BV, P/CF, and P/S ratios compared to the relative valuation ratios for the market presented in Chapter 13.
Although our investment decision is always the same, the form of the comparison depends on which valuation approach is being used. In the case of the present value of cash flow tech- niques, we derive a present value for the industry using our required rate of return for the industry—that is, we compare the present value of the specified cash flow versus the prevail- ing value of the index. If our estimated present value exceeds the prevailing index value, we should overweight the industry. Alternatively, if the PV of cash flows is less than the industry index, it implies that the industry is overvalued (i.e., the industry will not provide our required rate of return if acquired at the prevailing market price) and we should underweight this indus- try in our portfolio.
In contrast, if we use the two-step P/E ratio approach, we compute a current intrinsic value for the industry and compute an expected rate of return for the holding period based on this intrinsic value and the expected dividend return during the period compared to the current mar- ket price. If this expected rate of return exceeds the required rate of return (k), you should over- weight the industry; if the return is below k, you should underweight the industry.
To demonstrate industry analysis, we use Standard and Poor’s Retail Store–Drug index to rep- resent industrywide data for this industry. This retail store–drug index (hereinafter referred to as the retail drugstore [RDS] industry) contains three companies: (1) Longs Drug Stores, (2) Rite- Aid, and (3) Walgreen Company. The industry should be reasonably familiar to most observers, and it is consistent with the subsequent company analysis of Walgreens.
Recall that the reduced form DDM is
➤14.1
where:
Pi=the price of industry i at time t
D1=expected dividend for industry i in period 1 equal to D0(1 +g) k=the required rate of return on the equity for industry i
g=the expected long-run growth rate of earnings and dividend for industry i
As always, the two major estimates for any valuation model are k and g. We will discuss each of these at this point in the chapter with the understanding that we will also use these estimates subsequently when applying the two-step, price/earnings ratio technique for valuation.
Estimating the Required Rate of Return (k) Because the required rate of return (k) on all investments is influenced by the risk-free rate and the expected inflation rate, the differentiating factor in this case is the risk premium for the RDS industry versus the market. In turn, we discussed the risk premium in terms of fundamental factors, including business risk (BR), financial risk (FR), liquidity risk (LR), exchange rate risk (ERR), and country (political) risk (CR). Alternatively, you can estimate the risk premium based on the CAPM, which implies that the risk premium is a func- tion of the systematic risk (beta) of the asset. Therefore, to derive an estimate of the industry’s risk premium, you should examine the BR, FR, LR, ERR, and CR for the industry and compare these industry risk factors to those of the aggregate market. Alternatively, you can compute the system- atic risk (beta) for the industry and compare this to the market beta of 1.0. Prior to calculating a beta for the industry, we briefly discuss the fundamental risk factors for the industry.
Business risk is a function of relative sales volatility and operating leverage. As we will see when we examine the sales and earnings for the industry, the annual percentage changes in retail drugstore sales were less volatile than aggregate sales as represented by PCE. Also, the OPM (operating profit margin) for retail drugstores was less volatile than the S&P Industrials Index OPM. Therefore, because both sales and the OPM for the RDS industry have been less volatile than the market, operating profits are substantially less volatile. This implies that the business risk for the RDS industry is below average.
The financial risk for this industry is difficult to judge because of widespread use of building leases in the industry. Still, on the basis of the reported data on debt to total capital or interest coverage ratios, the FR for this industry is substantially below the market. Assuming substantial use of long-term lease contracts, when these are capitalized, this industry probably has financial risk about equal to the market.
To evaluate the market liquidity risk for an industry, it is necessary to estimate the liquidity risk for all the firms in the industry and derive a composite view. The fact is, there is substantial variation in market liquidity among the firms in this industry. Walgreens is very liquid, whereas Longs Drug Stores and Rite-Aid are relatively illiquid. A conservative view is that the RDS industry probably has above-average liquidity risk.
Exchange rate risk (ERR) is the uncertainty of earnings due to changes in exchange rates faced by firms in this industry that sell outside the United States. The amount of ERR is deter- mined by what proportion of sales is non-U.S., how these sales are distributed among countries, and the exchange rate volatility for these countries. This risk could range from an industry with very limited international sales (e.g., a service industry that is not involved overseas) to an indus- try that is clearly worldwide (e.g., the chemical or pharmaceutical industry). For a truly global industry, you need to examine the distribution of sales among specific countries because we know that the exchange rate risk varies among countries based on the volatility of exchange rates
P D
k g
i=
−
1
Valuation Using the Reduced Form DDM
500 CHAPTER 14 INDUSTRYANALYSIS
with the U.S. dollar. The ERR for the RDS industry would be quite low because sales and earn- ings for these drugstore firms are almost wholly attributable to activity within the United States.
The existence of country risk (CR) is likewise a function of the proportion of foreign sales, the specific foreign countries involved, and the stability of the political/economic system in these countries. As noted, there is very little CR in the United Kingdom and Japan, but there can be substantial CR in China, Russia, or South Africa. Again, for the RDS industry, country risk would be very low because of limited foreign sales.
In summary, for the RDS industry, business risk is definitely below average, financial risk is at best equal to the market, liquidity risk is above average, and exchange rate risk and country risk are almost nonexistent. The consensus is that the overall risk for the RDS industry should be lower than for the aggregate market on the basis of fundamental characteristics.
The systematic risk for the retail drugstore industry is computed using the market model as follows:
➤14.2 %∆RDSt= αi+ βi(%∆S&P 500t) where:
% DRDSt=the percentage price change in the retail drugstore (RDS) index during month t
`i=the regression intercept for the RDS industry
ai=the systematic risk measure for the RDS industry equal to Covi,m/r2m
To derive an estimate for the RDS industry, the model specified was run with monthly data for the five-year period 1997 to 2001. The results for this regression are as follows:
The systematic risk (β =0.82) for the RDS industry is clearly below unity, indicating a low- risk industry (i.e., risk less than the market). These results are quite consistent with the prior analysis of fundamental risk factors (BR, FR, LR, ERR, CR).
Translating this systematic risk into a required rate of return estimate (k) calls for using the security market line model as follows:
➤14.3 ki=RFR+ βi(Rm– RFR)
Recall that in Chapter 13 we derived three estimates for the required market rate of return based upon alternative risk premiums (0.040–0.092–0.116). For our purposes here, it seems like the midpoint is reasonable—that is, a nominal RFR of 0.052 and an Rmof 0.092. This, combined with a beta for the industry at 0.82, indicates the following:
k=0.052 +0.82 (0.092 – 0.052)
=0.052 +0.82 (0.04)
=0.052 +0.0328
=0.088 =8.48%
For ease of computation we will use a k of 8.5% A microestimate of fundamental risk below average and a risk estimate using the CAPM likewise below average implies an industry earn- ings multiple above the market multiple, all other factors being equal.
αt = 0.003 R2 = 0.62 βt = 0.82 DW = 1.83 t-value= 7.40 F = 68.37
Estimating the Expected Growth Rate (g) Recall that earnings and dividend growth are determined by the retention rate and the return on equity.
g=f (Retention Rate and Return on Equity)
We have consistently broken down return on equity into the following three components:
Therefore, we need to examine each of these variables in Exhibit 14.5 to determine if they imply a difference in the expected growth rate for RDS as compared to the aggregate market (S&P Industrials Index).
Earnings Retention Rate The retention rate data in Exhibit 14.5 indicate that the RDS industry has a higher retention rate (69 percent versus 55 percent). This means that the RDS industry would have a potentially higher growth rate, all else being the same (i.e., equal ROE).
Return on Equity Because the return on equity is a function of the net profit margin, total asset turnover, and a measure of financial leverage, these three variables are examined individually.
Historically, the net profit margin for the S&P Industrials Index series has been consistently higher than the margin for the RDS industry. This is not surprising because retail firms typically have lower profit margins but higher turnover.
As noted, one would normally expect the total asset turnover (TAT) for a retail firm to be higher than the average industrial company. This expectation has typically been confirmed because the average TAT for the S&P Industrials Index was 1.06 versus 2.70 for the RDS indus- try. Beyond the overall difference, the spread between the two series changed over the period.
This change occurred because the TAT for the S&P Industrials Index series declined steadily over the period while the TAT for the RDS industry experienced a secular decline from 1977 to 1999 but a major increase in 2000, as shown in Exhibit 14.6. Multiplying these two ratios indicates the industry’s return on total assets (ROTA).8
When we do this for the two series, the results in Exhibit 14.5 indicate that the return on total assets (ROTA) for the S&P Industrials Index series went from 6.54 percent in 1977 to 5.29 per- cent in 2000 and averaged 5.06 percent, whereas the ROTA for the RDS industry went from 11.56 percent to 8.89 percent and averaged 8.35 percent. Clearly, the industry ROTA results were superior on average.
The final component is the financial leverage multiplier (total assets/equity). As shown in Exhibit 14.5 and Exhibit 14.7, the leverage multiplier for the S&P Industrials Index increased
Net Income Sales
Sales Total Assets
Net Income Total Assets
× =
Net Profit Equity
Net Income Sales
Sales Total Assets
Total Assets Equity Profit
Margin
Total Asset Turnover
Financial Leverage
= × ×
= × ×
502 CHAPTER 14 INDUSTRYANALYSIS
8The reader is encouraged to read Appendix 14C to this chapter, which contains a discussion of an article by Selling and Stickney wherein they analyze the components of ROA and relate this to an industry’s economics and its strategy:
Thomas Selling and Clyde Stickney, “The Effects of Business Environment and Strategy on a Firm’s Rate of Return on Assets,” Financial Analysts Journal 39, no. 1 (January–February 1983).
EARNINGS MULTIPLIER FOR THE S&P INDUSTRIALS INDEX AND THE RDS INDUSTRY, AND INFLUENTIAL VARIABLES: 1977–2000 EARNINGS RETENTION NET PROFIT TOTAL ASSET RETURN ON TOTAL ASSETS/RETURN ON MULTIPLIER (t+1)RATEMARGINTURNOVERTOTAL ASSETSEQUITYEQUITY YEARS&P INDRDSS&P INDRDSS&P INDRDSS&P INDRDSS&P INDRDSS&P INDRDSS&P INDRDS 19778.399.4256.8079.905.154.071.272.846.5411.562.081.5313.6017.68 19786.588.7158.8076.105.194.031.272.816.5911.322.151.5214.1717.21 19797.187.6163.7072.605.573.471.303.007.2410.412.201.6515.9317.18 19808.137.4759.7070.804.923.471.313.046.4510.552.231.6614.3717.51 198110.729.2158.1069.004.863.451.283.036.2210.452.251.6514.0017.25 19829.279.2446.0068.903.953.441.172.974.6210.222.311.6610.6816.96 19839.6613.8350.4070.704.423.791.152.865.0810.842.281.8811.5920.38 198411.7613.8258.5062.404.773.041.222.655.828.062.391.9313.9115.55 198514.3813.9348.5059.203.842.961.152.684.427.932.541.8811.2214.91 198612.5215.1544.0067.103.753.111.072.714.018.432.581.9410.3516.35 198712.2015.7057.0068.404.772.881.082.805.158.062.622.0413.5016.45 198811.2813.9163.1069.405.512.900.982.825.408.183.032.0716.3616.93 198914.7214.0155.9767.715.012.790.972.824.867.873.162.0315.3615.97 199023.1913.8749.8068.324.262.890.972.824.138.133.272.0013.5016.26 199122.5115.5928.1967.192.972.910.942.852.808.313.241.839.0915.25 199222.5022.5133.8666.413.272.890.962.833.138.193.501.8210.9514.89 199315.8515.4341.9653.173.732.150.922.793.426.003.761.8812.8611.27 199415.1413.8361.0666.225.242.890.952.634.897.603.512.0017.1415.17 199515.5315.4660.6166.335.242.840.972.665.087.553.322.0316.8715.33 199618.8726.2869.1471.195.863.210.932.025.456.483.262.1717.7714.06 199726.2226.3660.3159.055.611.960.942.235.274.373.332.1817.559.53 199825.5934.8254.9675.265.142.580.852.044.395.273.702.4916.2613.12 199930.6743.0065.3775.096.272.700.842.095.265.653.322.4917.4614.08 200069.1980.626.333.040.842.925.298.893.092.0016.3617.74 Mean15.3416.4954.7968.794.823.061.062.705.068.352.881.9314.2015.71
EXHIBIT 14.5 Source:Financial AnalystsHandbook(New York:Standard & Poor’s,2001). Reprinted with permission.
504 CHAPTER 14 INDUSTRYANALYSIS
3.5
3.0
2.5
2.0
1.5
1.0
0.5
3.5
3.0
2.5
2.0
1.5
1.0
0.5
Total Asset Turnover
Year
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 S&P Industrials Index RDS
EXHIBIT 14.6 TIME-SERIES PLOT OF TOTAL ASSET TURNOVER FOR THE S&P INDUSTRIALS INDEX AND THE RDS INDUSTRY: 1977–2000
4.0
3.5
3.0
2.5
2.0
1.5
1.0
4.0
3.5
3.0
2.5
2.0
1.5
1.0
Total Assets/Equity
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year
S&P Industrials Index RDS
EXHIBIT 14.7 TIME-SERIES PLOT OF FINANCIAL LEVERAGE FOR THE S&P INDUSTRIALS INDEX AND THE RDS INDUSTRY: 1977–2000
overall from 2.08 to 3.09, whereas the leverage multiplier for the RDS industry went from 1.53 to 2.00. Although these higher financial leverage multipliers imply greater financial risk for both the S&P Industrials Index series and the RDS industry, they also contribute to a higher ROE, all else being the same.
This brings us to the final value of ROE, which is the product of the three ratios (profit mar- gin, total asset turnover, and financial leverage), or the product of return on total asset and the financial leverage multiplier. The data in Exhibit 14.5 and the plot in Exhibit 14.8 indicate that the ROE for the RDS industry was generally higher than the market except for the period 1993–1999. The average annual ROE was 15.71 percent for the RDS industry versus 14.20 per- cent for the S&P Industrials Index series. These average percentages are quite consistent with what would be derived from multiplying the averages of the components from Exhibit 14.5 as follows:
Although examining the historical trends and the averages for each of the components is impor- tant, you should not forget that expectations of future performance will determine the ROE value for the industry. In the current case, this analysis of expectations is very important because of the change in relative ROE during the period 1993–1999. As an analyst, it is necessary to determine whether the change during this period is a permanent change in the relative performance of this
ROEESTIMATE BASED ON TOTAL PERIOD AVERAGES (1977–2000)
PROFIT TOTALASSET TOTALASSETS/
MARGIN TURNOVER EQUITY ROE
S&P Industrials Index 4.82 × 1.06 × 2.88 = 14.71
RDS Industry 3.06 × 2.70 × 1.93 = 15.95
24.00 22.00 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
24.00 22.00 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
Net Profit Margin
Year
1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 S&P Industrials Index RDS
EXHIBIT 14.8 TIME-SERIES PLOT OF THE RETURN ON EQUITY FOR THE S&P INDUSTRIALS INDEX AND THE RDS INDUSTRY: 1977–2000
industry versus the market. In this case, you should be very concerned because of the inferior performance of the RDS industry during 1993–1999. Specifically, if you use the results for the recent five-year period (1996–2000), the ROE results are:
Notably, using the recent results, the ROE results are reversed. Combining these recent ROE results with alternative retention rates provides interesting growth estimates:
The point is, using full-period retention results indicates almost equal expected g. Alternatively, using the retention rates for the recent five-year period indicates a higher g for the market. Given the decline in g for the industry when we consider the recent results, it is probably appropriate to use a growth estimate for the RDS industry that is below the recent conservative estimate—
that is, 9.5 percent. Obviously, the best estimate of g would be based upon an estimate of the three components of ROE for the future five years.
Combining the Estimates At this point, we have the following estimates:
k=0.085 g=0.095
D0=$4.80 (Recent 12-Month Dividends)
D1=$4.80 ×1.095 =$5.26 (Estimated Dividend for 2002)
Because of the inequality between k and g (this g is above k and above the long-run market norm of about 7 percent and probably cannot be sustained), we need to evaluate this industry using the temporary growth company model discussed in Chapter 12. We will assume the following growth pattern:
2003–2005 0.095
2006–2007 0.090
2008–2009 0.080
2010–onward 0.070
GROWTH ESTIMATES BASED ON RECENT ROEWITH HISTORICAL AND RECENT RETENTION RATES RECENT* HISTORICAL** ESTIMATED RECENT* RECENT* ESTIMATED
ROE RR g ROE RR g
S&P Industrials Index 17.16 .55 9.44 17.16 .64 10.98
RDS Industry 13.85 .69 9.56 13.85 .72 9.97
*Recent five-year average.
**Total period average.
ROEESTIMATE BASED ON RECENT FIVE-YEAR AVERAGES (1996–2000)
PROFIT TOTALASSET TOTALASSETS/
MARGIN TURNOVER EQUITY ROE
S&P Industrials Index 5.84 × 0.88 × 3.34 = 17.16
RDS Industry 2.70 × 2.26 × 2.27 = 13.85
506 CHAPTER 14 INDUSTRYANALYSIS
Using these estimates of k and this growth pattern, the computation of value for the industry using the DDM is contained in Exhibit 14.9.
These computations imply a value of $423.45 compared to a price for the industry index of about $640.00 in mid-2002. Therefore, according to this valuation model and these k and g esti- mates, this industry is about 50 percent overvalued at this time. As will be shown, a fairly small change in the k – g spread can have a large effect on the estimated value.
If we assumed a constant growth rate of 7 percent from the beginning and a D1of $5.76, the value would be even lower as follows:
This is clearly a low estimate of value since it assumes the base growth rate of 7 percent from the beginning. To demonstrate the effect of a change in the k – g spread, if we assume a decline in this spread from 0.015 to 0.01 because of either a lower k or a higher g or some combination of the two, the value of the industry would be:
This is still below the current price of $640, but if we reduce the spread to 0.009, the value becomes $640 (5.76/0.009), which exceeds the prevailing price. The question becomes: how do you justify such a k – g spread in terms of the two components? The point is, using this constant growth model, if you cannot justify a lower k than 8.50 percent or a higher long-run growth rate (g) than 7.00 percent, you would conclude that this industry is overvalued and should be underweighted in your portfolio.
P= 5 76 = 0 010. 576
. $
P=
− = =
5 76 0 085 0 070
5 76
0 015 384 00 .
( . . )
.
. .
DIVIDEND DISCOUNT CALCULATIONS
DISCOUNT
ESTIMATED FACTOR PRESENTVALUE
YEAR DIVIDEND @ 8.5% OFDIVIDEND
2002 5.26 — —
2003 5.76 0.9217 5.31
2004 6.31 0.8495 5.36
2005 6.91 0.7829 5.41
2006 7.53 0.7216 5.43
2007 8.21 0.6650 5.46
2008 8.86 0.6129 5.43
2009 9.57 0.5649 5.41
Cont. Valuea 682.67 0.5649 385.64
Total Value $423.45 EXHIBIT 14.9
aConstant Growth Rate =7%
Continuing Value: $ . ( . )
. .
$ .
. $ .
D k g
1 9 57 1 07 0 085 0 070
10 24
0 015 682 67
− =
− = =
Similar to the presentation in Chapter 13, we initially define the FCFE series and present the series for the recent 15-year period, including an estimate for 2001 in Exhibit 14.10. Given these data, we will consider the historical growth rates for the components and for the final FCFE series as inputs to estimating future growth for the valuation models. You will recall that FCFE is defined (measured) as follows:
Net Income
+Depreciation expense – Capital expenditures – ∆in working capital – Principal debt repayments +New debt issues
As noted, the FCFE data inputs and final annual value of FCFE for the RDS industry for the period 1987–2001 is contained in Exhibit 14.10 along with 5-year and 10-year growth rates of the components. Using this data, we derive an estimate using the FCFE model under two sce- narios: (1) a constant growth rate from the present, and (2) a two-stage growth rate assumption.
The Constant Growth Rate FCFE Model We know that the constant growth rate model requires that the growth rate (g) be lower than the required rate of return (k), which we have spec- ified as 8.50 percent. In the current case, this is difficult because the 10-year growth rate exceeds Industry Valuation
Using the Free Cash Flow to Equity (FCFE) Model
508 CHAPTER 14 INDUSTRYANALYSIS
COMPONENTS OF FREE CASH FLOW TO EQUITY FOR THE RETAIL DRUG STORE INDUSTRY
NET DEPRECIATION CAPITAL CHANGE IN PRINCIPAL NEWDEBT
YEAR INCOME EXPENSE EXPENDITURE WORK. CAPITAL REPAYMENT ISSUES TOTALFCFE
1987 5.53 2.63 5.80 2.37 — 2.00 1.99
1988 6.30 3.09 6.72 2.50 — 0.65 0.82
1989 6.69 3.39 6.02 9.36 — 7.00 1.70
1990 7.67 4.15 7.62 2.37 — 0.83 2.66
1991 8.26 4.03 7.73 0.16 1.20 — 3.20
1992 8.96 4.39 7.19 1.96 1.03 — 3.17
1993 7.09 4.75 9.24 0.14 — 2.71 5.17
1994 10.51 5.32 11.40 2.20 — 3.93 6.62
1995 11.76 6.02 13.97 3.30 — 4.16 4.67
1996 13.92 6.07 15.00 14.33 — 20.14 10.80
1997 10.77 8.45 17.40 –5.04 3.74 0.00 3.12
1998 15.81 9.66 24.95 0.65 — 20.58 20.45
1999 19.07 10.25 26.06 –4.93 — 3.39 11.58
2000 20.94 10.80 24.24 –19.06 30.50 — –3.94
2001E 21.10 11.15 25.00 –5.00 — — 12.25
5-Year Growth Rate 8.60 13.10 10.80 NM NM NM 2.60
10-Year Growth Rate 9.80 10.50 12.40 NM NM NM 14.35
EXHIBIT 14.10
E =estimate.
NM =not meaningful.