If country risk is not diversifiable, either because the marginal investor is not globally diversified or because the risk is correlated across markets, we are left with the task of measuring country risk and estimating country risk premiums. In this section, we will consider two approaches that can be used to estimate country risk premiums.

One approach builds on historical risk premiums and can be viewed as the historical risk premium plus approach. In the other approach, we estimate the equity risk premium by looking at how the market prices stocks and expected cash flows – this is the implied premium approach.

Historical Premium

Most practitioners, when estimating risk premiums in the United States. look at the past. In this approach, we look at what we would have earned as investors by investing in equities as opposed to investing in riskless investments. We will consider why this approach cannot be used in emerging market and possible modifications. Historical Premiums in Emerging Markets The argument for using historical risk premiums is a simple one.

If investors have, on average, earned 5% more by investing in stocks than government bonds in the past, this is a reasonable estimate of what they will continue to make in the future. Notwithstanding the logic in this statement, the problem with historical risk premiums remains their imprecision. After all, the historical premiums are extracted from returns on stocks and bonds, which are volatile over time. As a consequence, the historical premium of 5% may come with a standard error that is so large as to make it useless.

How can we estimate the standard error? Roughly speaking, the standard error in the risk premium is a function of the annual standard deviation in stock returns and the number of years of data that we have:

To illustrate, the standard error in the historical risk premium for the United States, computed using 75 years of data is calculated below:

With an annualized standard deviation in stock returns of 20%, the standard error, even with 75 years of data, is roughly 2.31%. With emerging markets, we will almost never have access to as much historical data as we do in the United States.

In markets like Brazil, Russia and Indonesia, we can consider ourselves lucky if we can find 10 to 20 years of reliable historical data. If we combine this with the high volatility in stock returns in these markets, the conclusion we have to draw is that historical risk premiums can be computed for these markets but they will be useless because of the large standard errors in the estimates¹.

Modified Historical Risk Premiums

While historical risk premiums for markets outside the United States cannot be estimated with much precision, we still need to estimate a risk premium for use in these markets. To approach this estimation question, let us start with the basic proposition that the risk premium in any equity market can be written as: Equity Risk Premium = Base Premium for Mature Equity Market + Country Equity Risk Premium The country premium could reflect the extra risk in a specific market, reflecting the fact that it is not a mature equity market.

To estimate the base premium for a mature equity market, we can look at the US market. Between 1928 and 2002, stocks in the United States delivered a premium of 4.53% over government bonds. A more expansive look at all equity markets in the 20th century suggests an equity risk premium of about 4%². To estimate the country equity risk premium, however, we need to measure country risk and convert the country risk measure into a country risk premium.

---

Note:

1 Consider the Brazilian market, where we have about 10 years of reliable historical data on stock returns since the Real Plan in 1994. With an annualized standard deviation of about 30% in returns, the standard error in the risk premium, using 10 years of data, is about 10%.

2 See “Triumph of the Optimistis” by Dimson, Marxh and Staunton.

Prof. Aswath Damodaran

Next: Measuring Country Risk Premiums

Summary: Index