QUESTIONS

1. When constructing stock portfolios, what negative side effects can result from ignoring country membership?
2. What are the two types of country bets that may arise in stock portfolios and what are the drawbacks of each?
3. Why do “country noise” bets arise and why is it impossible to avoid them without directly considering country membership?
4. How do these issues differ between developed and emerging markets?
5. What approach do we recommend for avoiding unintended country bets in stock portfolios?

^* We thank Michael Katz, John Liew, Lasse Pedersen, and Prasad Ramanan for helpful comments. The views and opinions expressed herein are those of the authors and do not necessarily reflect the views of AQR Capital Management, LLC, its affiliates, or its employees.

¹ See Jianguo Chen, Ting Zheng, and Andrea Bennett, “Sector Effects in Developed vs. Emerging Markets,” Financial Analysts Journal 62, no. 6 (2006): 40–51; Steven L. Heston and K. Geert Rouwenhorst, “Industry and Country Effects in International Stock Returns,” Journal of Portfolio Management 21, no. 3 (1995): 53–58; Kate Phylaktis and Lichuan Xia, “Sources of Firms' Industry and Country Effects in Emerging Markets,” Journal of International Money and Finance 25, no. 3 (2006): 459–475; Ana Paula Serra, “Country and Industry Factors in Returns: Evidence from Emerging Markets' Stocks,” Emerging Markets Review 1, no. 1 (2000): 127–151; Frank Nielsen, “Emerging Markets: A 2009 Update,” MSCI Barra webinar, August 13, 2009; and MSCI Barra, “Country and Industry Effects in Global Equities,” MSCI Barra Research Bulletin, October 2008, http://www.mscibarra.com/products/analytics/models/RB_Country_Global_Effects_Global_Equities.pdf.

² As an interesting comparison, sector membership explained on average 13% of returns in developed markets but only 5% in emerging markets.

³ All statistics cited in this paragraph, as well as the results shown in Exhibit 16.1, are based on AQR internal analysis, using Barra, MSCI, and XpressFeed data for stock returns and country and sector classifications.

⁴ We perform our analysis using emerging and developed markets large-cap universes and monthly data from 1995 to 2009. Financial statement data are from Worldscope, and market capitalization and returns are from Barra, MSCI, and XpressFeed. We calculate all quantities (book value, market capitalization, and 1-year returns) in units of USD, using currency exchange rate data provided by Factset. As is typical in the academic literature, when measuring one-year performance for the momentum strategy, we exclude the most recent month. See, for example, Cliff S. Asness, Tobias J. Moskowitz, and Lasse H. Pedersen, “Value and Momentum Everywhere,” working paper, AQR Capital Management, 2008. This exclusion allows us to avoid the effect of short-term price reversions that often occur after large price moves. To control for outliers, we Winsorize the stock data, at each month in the sample, to the 1st and 99th percentiles.

⁵ We use GIC sectors, based on data from MSCI and XpressFeed.

⁶ We use BARRA's GEM risk model.

⁷ To avoid multicollinearity, we omit one dummy variable corresponding to the sector, or country, having the smallest number of stocks.

⁸ We view each manager in this interpretation as targeting the same level of expected risk.

⁹ Importantly, this conclusion holds regardless of how the country selection portfolios are expressed through stocks (i.e., regardless of the weighting scheme used to distribute country bets to the underlying stocks).