CURRENCY CONUNDRUMS: JANUARY 2018 DATA UPDATE 4. Aswath Damodaran

Transcription

1 CURRENCY CONUNDRUMS: JANUARY 2018 DATA UPDATE 4 Aswath Damodaran

2 The Set Up There is perhaps no more mangled nor misunderstood part of financial analysis than dealing with currencies, and globalization has only made the problems worse. In financial analyses, we often make implicit (and bad) assumptions about currencies and these feed into our valuations. I look at three of the most common misconceptions related to currencies and use them to update currency related numbers at the start of

3 Misconception 1: The Government Bond Rate is always the Risk Free Rate I was taught, in my first finance class, to use the US treasury bond rate as the risk free rate, when estimating expected return, reflecting the dollar-centric world of my MBA studies. The notion, though, that the government bond rate, denominated in the local currency, is the risk free rate in that currency persists, albeit expanded to include other currencies. Thus, we are told to use the Brazilian $R bond as the $R risk free rate and the Indian rupee government bond as the Indian rupee risk free rate. Its proponents argue that governments control the printing of money and hence never have to default. 3

4 But the facts contradict them.. In the last three decades, a significant proportion of all sovereign defaults have been local currency defaults; of the 58 sovereign defaults between 1996 and 2012, 31 were in the local currency. As to why countries would choose to default in the local currency, when they can print enough money to pay off debt, the answer is straight forward. Printing more money debases your currency and countries, faced with a choice between defaulting and debasing their currencies. often conclude that default is a better option. 4

5 So what? If a sovereign entity (or government) can default even on local currency debt, it stands to reason that a bond issued by it is no longer a risk free rate. In fact, using that government bond rate, as if it were a risk free rate, can lead to the double counting of risk, especially if analysts use a higher equity risk premium to capture additional country risk. For instance, consider the Nigerian Naira 10-year government bond, trading to yield 14.12% on January 1, If that rate is used as the risk free rate in Naira for a Nigerian company, in conjunction with a high equity risk premium for Nigeria, you are counting risk twice in your computation, once in your risky risk free rate and again in your equity risk premium. To avoid double counting, you have to cleanse the government bond of default risk and to do so, you have to estimate how much of the interest rate on the bond can be attributed to default risk. 5

6 Cleansing the Government Bond Rate There are three ways to get the default spread: Use a US dollar or Euro government bond issued by the sovereign in question and to net out the US Treasury Bond rate or the German Euro bond rate as the risk free rates. The sovereign Credit Default Swap on the sovereign in question, which is a measure of the default spread of the sovereign. Use the sovereign rating estimated by S&P, Moody s or Fitch for a country and to estimate the typical spread at which other bonds with the same rating trade at. The problem with all these spreads is that they are dollar-based, not local currency-based, and netting these spreads out from the local currency government bond can create an inconsistency. 6

7 Risk free Rates in Currencies 7 Aswath Damodaran 7

8 Misconception 2: There is a Global Risk Free Rate It is perhaps understandable that an analyst who looks at the differences in rates across currencies, before or after adjusting for default risk, will conclude that since the risk free rate is the lowest of the rates at which an entity can borrow money, the lowest of the rates across currencies, perhaps the Swiss Franc or the Japanese Yen, is the global risk free rate and that all other currencies are risky. That is a dangerous delusion, since there is a simple and powerful reason why risk free rates vary across currencies. The risk free rate in a currency is a reflection of expected inflation in that currency, and risk free rates will be higher in high-inflation currencies than in low-inflation ones, and can become negative in deflationary currencies. 8

9 Implication 1: No blending allowed! When valuing a company that has operations in many countries and derives its revenues in multiple currencies, you cannot create blended averages of risk free rates or growth rates in different currencies. Doing so would be akin to averaging the temperature in New York (measured in Fahrenheit) with the temperature in Frankfurt (measured in Celsius) to arrive at an average temperature for the two cities. 9

10 Implication 2: Differential Inflation can yield risk free rates The risk free rate in any currency can be estimated by using the differential inflation rate between that currency and one where a risk free rate is observable (like the US dollar or the Euro).\ There is a short cut Local Currency Riskfree rate = US $ Risk free Rate + (Expected Inflation rate in local currency Expected Inflation rate in US dollars) 10

11 Misconception 3: Currency choices will drive value When valuing a company, analysts often default to valuing it in the local currency, but currency is a choice. I can value Severstal, a Russian steel company, in rubles, US dollars or Euros. At first sight, the fact that risk free rates are lower in US dollars and Euros, relative to the ruble, may seem to suggest that you could inflate Severstal s value, by switching the valuation currency from rubles to dollars, but you will not. The answer lies in the last section, where we noted that risk free rates vary across currencies, because of differential inflation. 11

12 The Currency Consistent Valuation When estimating growth rates, remember that these growth rates include the inflation rate and be consistent with your currency choice. If you are using expected exchange rates to convert cash flows from one currency to another, those expected exchange rates have to reflect differential inflation, i.e., use purchasing power parity. 12

13 The Macro Pushback The other push back that you will get is that your views on a currency can cause the value estimated in that currency can diverge. For instance, if you expect the Brazilian Reai to appreciate against the US dollar for the next few years, the value of Embraer will be higher, estimated in $R than in US$. While that is true, the reason for the divergence has nothing to do with the use of the currency and everything to do with bringing your currency views into your valuation. Thus, if the value of Embraer is 20% higher, when you value it in $R than in US$, that 20% difference is entirely due to your view that the $R will strengthen, which if true, you can profit from in simpler and more direct ways than buying Embraer. 13

14 In conclusion If there are lessons to be learned from looking at the dangers of mixing currencies, it is that it is time that we stopped being casual about the currencies that we measure and use in our analysis. At the risk of sounding pedantic, we should never estimate a growth rate, without being explicit about what currency that growth rate is estimated in, or talk about cash flows, without thinking about the inflation that we have embedded in them. 14