International characteristic-based asset pricing

In this paper, we develop characteristic-based asset-pricing models for international stocks. We price stocks using passive portfolios created based on observable characteristics: market capitalization, book-to-market, prior-year return, growth of total assets, and operating profitability, each separately created for a given geographical region of the world. As such, our approach allows for segmentation in characteristic-based asset pricing among regions. Using a resampling micro-portfolio approach recently introduced by Barras (2018), we find that market capitalization is the most powerful characteristic in pricing international stocks, and that a threecharacteristic model based on market capitalization, book-to-market, and prior-year return has the lowest pricing errors. We also show that characteristic-based benchmarks exhibit much lower pricing errors, relative to global factor-based models. We further apply our characteristic models to the equity holdings of U.S. funds that invest in international stocks. International index funds exhibit zero abnormal returns, while active funds that charge higher fees and that mainly invest in emerging markets and small or mid-capitalization stocks exhibit positive and significant abnormal returns. These results indicate that U.S.-domiciled active managers are able to generate abnormal returns in less-efficient sectors of non-U.S. stock markets, when expected returns are measured using characteristic-based pricing.