Asset Allocation and Stock Selection : Evidence from Static and Dynamic Strategies in Turkish Markets

We study stock-picking versus asset allocation decisions and static versus dynamic strategies for Turkish markets, from October 2007 to March 2012. We find portfolio performance results to be sensitive to extant conditions such as an asset class dominating others, and thresholds and constraints imposed to avoid over-concentration. Dynamic core-satellite strategies using the Sharpe Ratio optimizer minimized the downside risk and improved portfolio performance as compared to investing only in equities or in portfolios constructed by broad asset classes based solely on asset allocation, when the threshold parameters were determined dynamically. The second best portfolio was the naively diversified three-asset class portfolio. Confusion regarding the efficacy of strategies may arise due to the fact that optimizations or naïve allocations over restricted asset classes may actually worsen risk-adjusted performance; and we could either naïvely diversify over all asset classes, or we could dynamically optimize over all major asset classes and individual stocks