Security Analysis (Text reference: Chapter 14) discounted cash flow techniques price-earnings ratios other multiples example #1: U.S. retail stores more on price to book value multiples more on price to sales multiples investment strategies based on multiples example #2: Bank of Montreal macroeconomic factors and industry level analysis 1 Discounted cash flow techniques dividend discount models the intrinsic value of a stock may be estimated as the PV of all of the future cash flows to which the owner is entitled: P D 1 0 1 k D 1 P 1 1 k D 2 1 k 2 D 3 1 k 3 the constant growth model assumes that dividend payments will grow forever at some rate g. In this case: P D 1 0 k g 2
more realistically, there are differential growth models. For example: D 0 $2, g 1 20% per year for three years, g 2 5% per year thereafter, β 0 8, E r M r f 6%, r f 3% k 03 8 06 078 2 1 2 P 0 1 078 2 1 2 2 1 078 2 2 1 3 2 1 078 3 2 1 3 2 1 05 078 05 1 078 3 $110 92 note that it is important to do scenario analysis g 1 25% P 0 $125 06 g 2 4% P 0 $82 97 T 4 years P 0 $125 70 E r M 10% P 0 $86 05 3 net investment total investment economic depreciation 0 only if some earnings are retained (note: we are assuming here that the firm will not issue new shares or borrow more funds to finance new investments) if retention ratio = 0, g 0 g retention ratio return on retained earnings P 0 D 1 k EPS 1 k typically firms do have growth opportunities: P 0 EPS 1 k PVGO example: EPS 1 10, retention ratio = 50%, ROE = 20%, k 15% D 1 P 0 k g is growth necessarily good? 5 15 10 $100 4
earnings-based valuation most analysts focus on earnings rather than dividends since the latter are more discretionary since D t E t RE t E t I t, P 0 t 1 when ROE = k, P 0 E 1 k D t 1 t k t 1 E t 1 t k t 1 I t 1 t k the plowback ratio (a.k.a. earnings retention ratio) is the fraction of earnings reinvested in the firm (denoted by b) the dividend payout ratio is the percentage of earnings paid out as dividends and is equal to 1 b note that g ROE b. Example: E t $100, ROE = 12%, b 60. The firm reinvests $60 and earns a return of 12%, so E t 1 $100 $60 12 $107 20. The growth rate g 7 2% 12 60. 5 price-earnings ratios since P 0 E 1 k also since PVGO, P 0 1 E 1 k 1 k 1 PVGO E 1 PVGO E 1 k P D 1 E 1 1 b 0 k g k g P 0 1 b E 1 k ROE b 6
P/E ratios depend on: k PVGO ROE b choice of accounting methods P/E ratios are also affected by inflation and business cycle effects note that earnings here should be economic earnings, but accounting earnings are used in practice also note that future earnings are what really matters, but historical earnings are often used ( trailing vs. leading P/E ratios) a recent variation (commonly used for high-tech firms) is the P/E ratio divided by the growth rate (called the PEG ratio). A common rule of thumb is that the PEG ratio of a fairly priced firm equals one. 7 free cash flow allows for financing alternatives other than retained earnings often used in takeover/restructuring analysis basic idea is to estimate the value of the firm as a whole (PV of cash flows assuming all-equity financing plus NPV of financing alternatives (e.g. debt tax shields)) equity value is then total firm value less the market value of all non-equity claims recall that leverage will affect returns, risk, and β. Since we are initially calculating the value under all-equity financing, we have to use formulas like: β EQUITY β ASSETS 1 1 T C D E 8
More on multiples in addition to P/E, some commonly used multiples include price to book value (P/BV) and price to sales (P/S) many others are used in specific industries: e.g. passengers per mile flown, sales per franchise, etc. general procedures: 1. Select comparable firms tradeoff between similarity and sample size exclude abnormal firms (merging, restructuring, changing strategic direction, etc.) commonly used criteria include: industry classification technology clientele size leverage 9 2. Select appropriate bases for multiples which to use? earnings? sales? book values? general observation: the higher up the income statement, the less it is subject to choice of accounting methods, but the less it reflects differences in operating efficiency across firms 3. Average multiples across comparable firms allows the calculation of a fair price for a dollar of earnings, or book value of equity, or sales, etc. important implicit assumption: ability of firms to convert the bases into cash is the same 4. Forecast future bases examples: in order to use earnings to estimate value, we need to project next year s earnings and multiply by the average P/E ratio (but what about future years?) if we want to use sales/franchise, we have to forecast the number of franchises 10
5. Value the firm seems straightforward, but we may need to do further analysis to resolve different estimates from different multiples valuation: entire firm or shareholders equity? to mitigate leverage-related problems, analysts sometimes use multiples to value the entire firm the same basic procedures apply, except that the multiples must be redefined appropriately, e.g. rather than P/E, we might use Total Firm Value/Operating Income to obtain the value of equity, subtract the value of all non-equity claims (e.g. debt, preferred stock) from estimated total firm value 11 earnings multiples trailing earnings multiple AVG P 0 E 0 E 1 leading earnings multiple AVG P 0 the leading multiple is more appropriate for valuation, but to use it we need to forecast E next year for all of the comparable firms using the leading multiple: using the trailing multiple: estimated value AVG P 0 E 1 estimated value AVG P 0 E 0 E firm 1 E firm 1 12
this is more properly seen as an estimate of next year s price under the assumption that the multiple is stable over time to estimate price today, we must discount: P firm 0 P AVG 0 E 0 E1 firm 1 k note that it is sometimes claimed that valuation using multiples is easy because you don t have to estimate future cash flows and discount rates. But for leading multiples, you have to forecast earnings for every comparable firm, whereas for trailing multiples you do need discount rates. 13 Example #1: U.S. retail stores some data for firms in the industry is provided below on slide 15 suppose we want to value Dayton Hudson using an earnings multiple. We take as given Value Line s projection that EPS will increase by $1.13 to $4.85 in 1992. which are the comparable firms? DH operates 770 upscale department stores across the U.S. A reasonable subset of the firms which might be appropriate is Dillard, May, Nordstrom, Jacobson, J.C. Penney, L. Luria, and Sears. the table at the top of slide 16 provides a comparison of these firms in terms of size and leverage. If there are returns to scale, larger firms should be more efficient. Sales also proxies for the geographical distribution of stores in the chain. On this basis we eliminate L. Luria and Jacobson because they are too small/regionally concentrated and Sears because it is too large. 14
Retail Store Industry Value Line Data for May 29, 1992 Retail Store Share Price Trailing P/E β M/B P/CF per share Dillard Dept. 41 22.28 1.20 2.89 3.55 Dayton Hudson 65 17.47 1.45 2.41 20.31 Dollar General 20 24.39 1.00 3.38 14.29 Family Dollar 17 23.29 1.25 4.17 14.17 Jacobson Stores 16 21.92 1.30 1.07 3.60 J.C. Penney 67 16.75 1.20 2.23 23.10 Kmart Corp. 23 11.39 1.20 1.57 2.71 L. Luria & Son 6.6 30.00 1.10 0.46 1.83 May Dept. Stores 55 13.68 1.35 2.83 44.00 Fred Meyer 24 13.26 1.10 1.05 3.20 Nordstrom, Inc. 33 19.88 1.25 2.87 10.31 Pic N Save 16 14.29 1.15 2.51 8.21 Sears, Roebuck 45 12.13 1.10 1.12 5.96 Service Merchandise 11 14.47 1.70 10.38 6.29 Wal-Mart Stores 53 37.86 1.25 8.72 23.56 Woolworth Corp. 28 18.54 1.25 1.80 6.36 15 Debt Pref. Stock Equity Total Sales per Retail Store (book value) (book value) (mkt value) Value Leverage Stores Sales store Dayton Hudson 4,227 377 4,630 9,234 49.86% 770 16,115 21 Dillard Dept. 1,038 0 4,572 5,611 18.51% 198 4,036 20 Jacobson Stores 98 0 92 190 51.33% 24 396 16 J.C. Penney 3,354 684 7,816 11,854 34.07% 1,813 16,201 9 L. Luria & Son 2 0 35 38 6.38% 53 208 4 May Dept. Stores 3,918 394 6,790 11,102 38.84% 3,613 10,615 3 Nordstrom, Inc. 482 0 2,701 3,183 15.15% 63 3,180 50 Sears, Roebuck 19,200 325 15,486 35,011 55.77% 1,800 57,242 32 Total Value Value Retail Store Value Leverage EBIT to EBIT Sales to Sales Dillard Dept. 5,611 18.51% 412 13.619 4,036 1.390 J.C. Penney 11,854 34.07% 1050 11.290 16,201 0.732 May Dept. Stores 11,102 38.84% 1061 10.464 10,615 1.046 Nordstrom, Inc. 3,183 15.15% 263 12.103 3,180 1.000 Average 11.869 1.042 16
a comparison of the firms left in the sample shows that they vary a lot in terms of leverage and DH is one of the most heavily levered firms DH should have a lower P/E than comparable firms since its equity is riskier as an alternative, we value entire firms using firm-level multiples (see the second table on slide 16) the firm-level multiples show no leverage-related patterns, so we can use their average to calculate a value for DH. We use the industry average Value/EBIT ratio of 11.869. we use the Value Line projection of 1992 Sales of $18 billion. Assuming the EBIT/Sales ratio remains at its 1991 level of 4.78%, the increase in sales implies 1992 EBIT of $860 million. This implies a 1992 value of $860 11 869 or $10.21 billion for DH. 17 since this is based on a trailing multiple, we need to discount to get an estimate of current (1991) value. what is the appropriate discount rate? We need a rate for the entire firm, not just the equity. Recall β ASSETS β EQUITY E V β D DEBT 1 T c V and since for small leverage, β DEBT 0, β ASSETS β EQUITY E V we use this to calculate an average β ASSETS for the comparable firms of 0.91. Assuming a corporate tax rate of 37%, a risk free rate of 6%, and a market risk premium of 8%, the risk-adjusted discount rate for DH is estimated as: k r f 1 T c β ASSETS market risk premium 06 1 37 91 08 11 06% 18
this implies a total current DH value of $10.21/1.1106 = $9.193 billion. The estimated equity value is this amount less the value of debt ($4.227 billion) and the value of preferred stock ($377 million), or $4.589 billion. since the current market value of DH s equity is $4.63 billion, there is no reason to assume it is not correctly valued suppose we used the whole-firm Value/Sales ratio rather than Value/EBIT. The average of comparable firms is 1.019. Applying this to the 1992 projected sales of $16.115 billion gives a 1992 value of $19.962 billion, just about double our estimate using EBIT. historically, DH s Value/Sales ratio has been very different from the rest of the sample: the firms are not comparable using this ratio. One reason is that sales numbers don t tell very much about operating efficiency. For example, DH may be trying a lower margin/higher volume strategy than its competitors. 19 More on price to book value multiples for a stable firm, P D 1 0 k g EPS 1 1 b k g defining ROE as EPS 1 BV 0 where BV 0 is the current book value of equity, we have P BV 0 0 ROE 1 b P 0 ROE 1 b k g BV 0 k g moreover, since g ROE b, P 0 ROE g BV 0 k g 20
P/BV ratios are determined by risk, growth prospects, and ROE firms with low P/BV and high ROE are potentially undervalued; firms with high P/BV and low ROE are possibly overvalued advantages: BV provides a relatively stable and simple benchmark given consistent accounting standards, comparable across firms can be used for cases where E 0 disadvantages: affected by accounting decisions on inventories, depreciation, etc. not comparable across jurisdictions with different accounting standards not very useful in cases where there are not a lot of fixed assets BV can become negative after a series of losses 21 More on price to sales multiples for a stable firm, P D 1 0 k g EPS 0 1 b 1 g k g letting sales per share be S 0 and defining profit margin as EPS 0 S 0, we have P S 0 0 profit margin 1 b 1 g k g P 0 profit margin 1 b 1 g S 0 k g P/S increases with profit margin, the payout ratio, and the growth rate, and it decreases with risk 22
firms with high P/S and low profit margins might be overvalued, firms with low P/S and high profit margins could be undervalued P/S multiples are widely used to value privately held firms and to compare value across publicly traded firms advantages: can always calculate it (unlike P/E and P/BV which might be negative) revenue is less distorted by accounting decisions more stable than P/E multiples since S doesn t fluctuate as much as E over the business cycle disadvantages: can be very misleading if the firm has cost control problems may not be comparable across firms with different strategies 23 Investment strategies based on multiples growth vs. value investing: these terms may be defined in various ways, but a common distinction is in terms of P/BV (firms with relatively low P/BV are considered to be value firms) some people argue that growth is hard to forecast accurately, and you should only pay for proven performance ; others contend that investors are systematically too optimistic regarding growth either way, you should avoid high P/E firms and invest in low P/E firms there is some evidence in support of this (Basu, Journal of Finance, 1977) also recall the Fama and French results factor in size and book-to-market equity P/E doesn t matter if you already there is relatively little evidence regarding P/S based strategies: portfolios based on low P/S ratios do well, but not as well as low P/E ones do low P/S firms are often small, so this could be the size effect again 24
Example #2: Bank of Montreal almost all of the information used is based on text Figure 14.2 (p. 510), which is a Value Line report on BMO dated August 10, 2001 trailing P/E as noted above, research has found that low P/E stocks tend to do relatively well a common rule is to buy a stock if its P/E is more than one standard deviation below the industry average, sell if it is more than one standard deviation above the industry average, and do nothing otherwise (i.e. hold: if owned, don t sell unless a better opportunity is available; don t buy if not owned) BMO s trailing P/E is reported by Value Line as 13.8 over the past 15 years, the banking industry average P/E is 11, with a standard deviation of 3 (taken from Datastream) hold 25 median P/E some analysts believe that trailing P/E should be compared not to the industry but rather to the firm s own past history when E is small for a given year, P/E will be inflated, distorting the average P/E for the firm use median P/E since it is less sensitive to outliers a common rule is sell (buy) when P/E is significantly higher (lower) than median P/E the median P/E for BMO is 9 since BMO s trailing P/E is 13.8, this suggests either selling or holding 26
relative P/E this involves comparing how the firm is doing compared to other firms relative to how it has done in the past for Value Line, relative P/E is based on all stocks covered (not just the industry) for historical relative P/E, usually use the median of the past 5-15 years according to Value Line, BMO currently has relative P/E ratio of 0.71 and we can calculate that the median relative P/E for 1990-2000 was 0.56 since relative P/E is above historical relative P/E, this also indicates either selling or holding 27 forward P/E note that Value Line reports a P/E of 12.8 for BMO; this is a mixture of trailing and forward because it is based on the last 6 months earnings and forecasted earnings for the next 6 months, i.e. 40 44 40 44 12 8 76 80 84 75 3 15 BMO s forward P/E (based on the next 4 quarters) is 40 44 40 44 12 1 84 75 85 90 3 34 since this is close to the historical industry average of 11, we again have a hold signal 28
P/B ratio this is viewed as a buy signal if P/B is less than one the rationale for this is that B is an accounting estimate of the per share liquidation value of the firm: if the firm can be purchased for less than its estimated liquidation value, this indicates a buying opportunity for BMO, P/B is 40.44/20.85 = 1.94, which is obviously not a buy signal Graham s criteria very conservative; a firm must meet 4 criteria to have a buy recommendation: 1. The firm must have paid a dividend in each of the last 12 years. 2. Positive earnings must have been achieved in each of the last 10 years. 3. P/B 1.5. 4. P/E 15, where E is average EPS over the past 3 years. although BMO has a consistent record of paying dividends, has achieved positive earnings over the past decade, it does not get a buy recommendation because P/B exceeds 1.5 29 earnings multiplier idea is to compare expected return and required return if expected ( ) required, buy (sell) to calculate expected return, we need to forecast next year s share price and dividends we calculate Pˆ 1 P E E 1, where P/E is the firm s historical median value in this case, ˆ P 1 9 3 34 $30 06 predicted dividends per share for 2002 are $1.16, so expected return 30 06 1 16 40 44 40 44 0 23 required return can be based on personal assessment, or from a pricing model such as CAPM BMO has β 0 95; in August 2001, r f 4 25%; using 8% for the historical market risk premium gives a required return of 0425 0 95 08 11 85% sell 30
dividend discount model (text pp. 507-509) forecasted dividends for 2002-2005 are $1.16, $1.21, $1.26, and $1.32 assume dividend growth is constant after 2005; Value Line forecasts a retention ratio of 71% and ROE of 15.5%, so g 155 71 11% the estimated stock price for 2005 is P 2005 1 32 1 11 1185 11 $172 38 therefore today s intrinsic value is P 0 1 16 1 1185 1 21 1 1185 2 1 26 1 1185 3 1 32 172 38 1 1185 4 $113 89 since intrinsic value exceeds price, this is a buy signal however, note that Value Line is predicting a price range of $40-$56 for 2004-2006 using historical average ROE of 13.9% instead of 15.5% changes today s value estimate to $50.55 31 Macroeconomic factors and industry level analysis analysts must consider the economic environment when assessing the firm s current situation and future prospects P/E ratios tend to fall when inflation rises (because reported earnings overstate true economic earnings) export-oriented and import-oriented firms and industries will be dramatically affected by changes in exchange rates cyclical firms and industries (e.g. cars) will tend to outperform in expansions, underperform in recessions defensive firms and industries (e.g. insurance) will tend to underperform in expansions, outperform in recessions highly leveraged firms will tend to have performances that move strongly with interest rates (in the opposite direction) also should consider industry life cycle and competitive structure 32