The economic research unit of the Bank of England recently published a paper where it investigates whether there is a correlation between equity returns and exchange rate returns. Hereafter a synopsis.
If a country’s equity market is expected to outperform that of other countries, should we expect its currency to appreciate or depreciate? The answer to this question is of great importance to international equity investors, policy makers and, of course, to academics. An investor holding foreign equities is naturally exposed to exchange rate fluctuations. Both portfolio performance and the decision regarding whether to hedge foreign exchange (FX) risk will depend on the covariance between equity and currency returns, as well as expected returns and return volatilities. The relation between equity and currency returns is also important for policy makers as valuation changes induced by FX and equity returns generate significant swings in international investment positions, and the recent crisis has been characterized by increased amplitude of these valuation swings. However, while a vast literature has investigated the link between interest rate differentials and exchange rates across countries, little is known about the relation between exchange rates and international equity returns.
This paper fills this gap by providing empirical evidence on whether expected returns on foreign equity portfolios are systematically associated with currency movements.
Upon a review of the existing economic literature, it appears that the sign of the correlation between equity and FX returns is not clear theoretically.
The first contribution of our paper is to establish the empirical correlation between these returns. There is not much evidence in the literature that exchange rate movements offset or substantially reduce expected differences in equity returns across countries. However, while existing studies have examined the correlation using statistical methods in a time-series setting, in this paper we take an economic value approach in a cross-sectional portfolio setting. In this paper the authors consider an investor who builds a portfolio designed to capture differences in the expected returns of international equity markets.
The second contribution of this paper is to assess whether the resulting portfolio returns can be explained as compensation for risk.
The results deliver several key messages. First, an investor can capture differences in expected equity returns across countries and make substantial returns in US dollars, in the range from 7% to 12% per annum. This finding clearly indicates that exchange rate changes do not offset expected equity return differentials, and the evidence is similar to that in the FX carry literature, which finds that exchange rate changes do not offset the profits available from exploiting international interest rate differentials. It is tempting to think of the strategy studied here as the FX carry trade using equities rather than bonds. However, this is not the case because the returns from the UEP strategy are virtually uncorrelated with the returns from the FX carry trade. (note: UEP stands for Uncovered Equity Parity, a condition which implies a correlation of minus unity between expected equity return differentials and currency returns, and a zero expected excess return to international equity investment).
These large returns may be due to a combination of risk premia arising in equity and FX markets. The asset pricing tests performed by the research unit, as part of the data validation process, suggest that there is some value to this argument. Global equity factors are useful in pricing the cross-section of 15 international equity portfolios.
Although all of these results point towards a risk explanation for the large returns of the UEP strategy, it is important to emphasize that risk premia only account for a fraction of the returns generated by the strategy over time. The time-series evidence tells us that, while risk exposures of our 15 portfolios are significant, positive and statistically significant excess returns of up to 10% per annum remain. This suggests that there may be additional drivers of our portfolio returns.
This paper investigates the relationship between international equity returns and FX returns using a portfolio approach that is designed to exploit differentials in expected equity returns across countries. In the empirical analysis we follow the recent literature on currency markets and carry trade strategies, and sort equity markets into portfolios according to their expected return differentials with the US equity market. Equity index returns are forecast using three different but well-known predictors: dividend yields, term spreads and 12-month momentum.
Using a sample of 42 countries, over a period spanning November 1983 to September 2011, we show that investing in the highest expected equity return quintile portfolio and shorting the lowest expected equity return quintile portfolio generates significant excess returns between 7% and 12% per annum across the three different predictors. The returns are entirely driven by differentials in local equity market returns across countries, with exchange rates not responding at all to relative stock market performance. These returns can be linked to exposures to some international risk factors, notably global equity market volatility risk, but even after accounting for these risk factors, sizeable average returns remain. In fact, the international equity strategy provides alphas of up to 10% per annum and larger Sharpe ratios than conventional currency and equity strategies.
Overall, this study provides evidence that exchange rate movements fail to offset differentials in country-level equity returns and, to return to the question in the title of this paper, stock market returns tell us very little about exchange rates.