Practical Considerations in Quantitative Short Extension Strategies

This article analyzes short extension strategies under realistic scenarios in portfolio construction, where the portfolio strategy includes constraints such as tracking error, industry bets, asset level bounds, and turnover. Further, it considers a systematic tool, called Constraint Attribution, to analyze the impact of these constraints on the realized active returns of portfolios in a historical backtest study. For example, the article compares a 130/30 strategy to its long-only counterpart in backtest studies using two different expected return data as well as varying tracking error limits. It further studies the effect of using a fixed leverage ratio versus a dynamic strategy where the leverage can change over time. Finally, the article analyzes the effect of transaction costs, namely in the form of leverage costs (which may be a combination of borrow costs, short rebate costs, and short-selling costs). It concludes that the appropriate leverage ratio varies through time and depends largely on the underlying risk of the market, the tracking error limit, the forecasted returns, and the distribution of the information contained in them.