This paper uses an accounting-based approach to identify growth characteristics that can help explain equity returns at the country level for a sample of 30 countries over the past two decades.
Most previous country level empirical research has focused on dividends, perhaps due to the availability of long-run data on dividends or to prior theories being more focused on dividends rather than accounting fundamentals. Dividends are problematic because: (i) dividend policies are endogenously chosen by firms, (ii) dividends tend to vary only gradually due to signaling incentives and clientele effects, (iii) there is a declining trend in dividends and an increasing substitution by share repurchases, and (iv) dividends create potentially distorting tax effects.
Furthermore, a focus on dividends only provides partial coverage of firms in the cross-section as the dividend ratio (dividend/price) would be irrelevant for firms that do not pay dividends. At the country level, this firm level heterogeneity in dividends introduces noise. As a result, dividends are a poor measure of fundamental value.
Instead, we examine whether fundamental accounting characteristics associated with future earnings growth, in particular earnings and book values, explain cross-sectional variation in country level returns.
Using a panel of 6,600 country-month return observations over the period March 1993 through to June 2011 covering 30 countries, we find strong evidence that earnings-to-price ratios (E/P) and book-to-price ratios (B/P) jointly explain cross-sectional variation in country level returns.
A key implication of our results is that combining measures of value, such as E/P and B/P, offers a theoretically superior way to measure expected returns.