What Do Stock Markets Tell Us about Exchange Rates? (Digest Summary)

The authors find that exchange rate movements are unrelated to differentials in country-level equity total returns. Thus, a trading strategy that uses various predictors to produce long–short baskets of country equity indexes generates considerable alpha and solid Sharpe ratios. The returns are partly the result of extra compensation for global equity volatility risk. But after controlling for exposure to volatility, significant excess returns remain.

Correctly understanding the interplay between equity index and foreign exchange (FX) returns is crucial for investors, traders, and policymakers. The authors show that using factors to invest in the top expected return quintile portfolio of country equity indexes and to sell short the bottom expected return quintile basket generates annual excess returns of 7%–12% (in US dollars, without hedging). The authors forecast expected excess returns (versus the US equity market) using three distinct predictors: dividend yields, term spreads (the difference between 10-year and 3-month government bond yields), and momentum (cumulative 12-month returns).

The authors also show that returns are driven entirely by local currency equity market returns across countries. FX movements do not offset differentials in country-level equity returns. The evidence is similar to research on FX carry trade, which shows that exchange rate changes do not offset the profits from exploiting international interest rate differentials. Moreover, stock market returns tell us very little about exchange rates.

Importantly, the authors find that neither changing the source of the return data (they repeated the analysis using data on exchange-traded funds) nor the inclusion of transaction costs alters their main conclusions. Thus, their strategy is not merely theoretical but can also be implemented in the real world.

The authors use total return data from 42 MSCI country equity indexes from November 1983 to September 2011. The indexes and the dividend yield data are from Thomson Datastream. The authors obtain data on term spreads from Global Financial Data and exchange rate data from Barclays Bank International and Reuters.

The authors assign each country to one of five portfolios. The high-expected-return portfolio contains country indexes with strong momentum, high dividend yields, and wide term spreads. They build portfolios using each of these variables separately—that is, the variables are not combined and optimized to create the portfolios. All portfolios are held for one month, and returns are measured (in US dollars) in excess of the US equity market return. Returns are decomposed into two components: (1) the return on the international equity position in their local currency and (2) the FX component of the portfolio return.

The results stand up to robustness checks. For instance, the authors find that the results are not driven by a particular country or subset of countries. One concern is that the economic value of the factor-based country selection strategy has declined over time, especially during and after the 1990s. That said, the reduction in average returns is generally offset by a similar reduction in portfolio return volatility (with the exception of the term spread strategy).

Although there is ample research on the relationship between interest rate differentials and exchange rates, the market’s understanding of the link between international equity returns and FX is limited. The authors’ work fills this gap. Prior research has suggested that a positive relationship may exist between country equity index return differentials and FX because investors increase their holdings in equity markets that have recently outperformed. But the authors disprove this possibility, showing that country equity returns are not related to FX movements.