Country v sector effects in equity returns and the roles of geographical and firm-size coverage

Since Roll (The Journal of Finance 47(1):3–41, 1992) and Heston and Rouwenhorst (Journal of Financial Economics 36:3–27, 1994), there has been a debate whether country factors in international stock returns are typically more variable than sector factors. The addition of emerging markets (EMs) does boost the ratio of country-factor variance relative to industry-factor variance: these markets have a higher variability, but are also less related to global factors. Investigating to what extent this phenomenon can be tracked down to the impact of adding more small firms, we find the following. (1) Small firms do have higher volatility, but only after controlling for country and sector affiliation. (2) Small firms do have weaker sector affinity, as expected. (3) Small firms unexpectedly have weaker local-market sensitivities than large firms. Facts (2) and (3) mean that adding more small firms to the data base has a diversifying effect on both the sector- and country-factor variance; while the impact on sector variance is larger, the net effect turns out to be tiny. (4) Adding emerging markets has a very marked impact on the variance ratio. In fact, the addition of small stocks to the sample hardly dents the effect of adding EMs. Thus, the role of EMs cannot be reduced to just a small-firm phenomenon.