Why Scenario DCF Is More Accurate than Risk Premiums  693 and came to a similar valuation—an EBITDA multiple of around 4.5—despite using a very high country risk premium of 11 percent on top of the WACC. The result was similar because the second adviser made performance assump- tions that were far too aggressive: real sales growth of almost 10 percent per year and a ROIC increasing to 46 percent in the long term. Such long-term performance assumptions are unrealistic for a commodity-based, competitive industry such as chemicals. In another, broader set of analyst forecasts from 2015 to 2018, 30 percent of industries were expected to achieve growth rates more than 20 percent, while in the United States, only 5 percent were expected to achieve similar results. It’s hard to imagine 30 percent of industries growing more than 20 percent per year. These are among the reasons we favor a scenario DCF approach to valu- ing emerging-markets companies. It allows you to focus on company-specific risks, not generic risks. Our empirical research also shows that there isn’t much of a country risk premium built into the valuation of stocks in some emerging markets. If there were a substantial country risk premium, we’d expect price-to-earnings ratios (P/Es) to be much smaller than they are. Consider Brazil. Over the past decade, many valuations we’ve seen have incorporated country risk premiums of 3 to 5 percent, plus an inflation dif- ferential versus U.S. companies of about 2 to 3 percent. That leads to a cost of equity of 15 to 18 percent. From 2015 to 2018, the P/E for the major Brazilian market index has been in the range of 10 to 17 times. Going back to the value driver formula derived in Chapter 3, we can solve for the expected growth in earnings, given estimates for the other values: P E g k g e = −     − ( ) 1 ROE / where g is the growth rate of earnings, ROE is return on equity, and ke is the cost of equity. If we assume a P/E of 12 times, a cost of equity of 15 percent, and a mar- ginal return on equity of 20 percent (above historical averages), the implied growth rate of earnings in perpetuity would have to be about 11.5 percent nominal, or about 7.5 percent in real terms (assuming 4 percent inflation, based on 2 percent in the United States and two percentage points higher inflation in Brazil). But 7.5 percent real growth in perpetuity is clearly unrealistic. Looked at another way, if we assume 3.5 percent real growth in earnings in perpetuity (an optimistic view), the implied P/E at a 15 percent cost of equity is 8.3 times, which is about 30 percent lower than current P/Es. It’s impossible to come up with a consistent set of assumptions that ties together a P/E of 12 and 15 percent cost of equity. 694 EmErging markEts If we eliminate the country risk premium, our results work mathematically and economically. We’ll use 2016 as an example and solve for the implied cost of equity. The P/E was about 13 times. Assuming 3.5 percent long-term real growth plus 4 percent infl ation and a 14 percent ROE, we calculate a nominal cost of equity of about 11 percent. Subtracting infl ation at 4 percent gives us a 7 percent real cost of equity, not very different from the real cost of equity of the United States (see Chapter 15). Of course, these results are highly sensitive to small changes in some of the assumptions. The key point is that it is very diffi cult to reconcile current P/Es with a country risk premium of 3 percent or more. Exhibit 35.1 shows a time series of implied costs of equity for the Brazilian index over the ten years from 2009 to 2018. The nominal cost of equity stays within a range of 10 to 12 percent, or about 6 to 8 percent real. An implied country risk premium for Brazil is more likely to be closer to 1 percent than 3 to 5 percent. One reason country risk premiums should be lower than levels used by many is that many country risks, including expropriation, devaluation, and war, are largely diversifi able. Consider the international consumer-goods player illustrated in Exhibit 35.2 . Its returns on invested capital were highly volatile for individual emerging markets. Taken together, however, these mar- kets were hardly more volatile than developed markets; the corporate port- folio diversifi ed away most of the risks. Finance theory clearly indicates that the cost of capital should not refl ect risk that can be diversifi ed. This does not mean that diversifi able risk is irrelevant for a valuation: the possibility of ad- verse future events will affect the level of expected cash fl ows. But once this has been incorporated into the forecast for cash fl ows, there is no need for an additional markup of the cost of capital if the risk is diversifi able. EXHIBIT  35.1 Low Variability in Implied Cost of Equity Brazilian implied cost of equity, nominal1 % 2009 11.6 2010 12.3 2011 12.4 2012 10.9 2013 10.6 2014 10.7 2015 10.9 2016 11.1 2017 10.0 2018 11.5 1 Assuming average prior 10-year ROE, 4.0% long-term inflation, 3.5% long-term real profit growth. Source: Capital IQ.