Risk-Budget Optimization for Factor-Based Index Enhancement
Summary
This report summary examines how to turn a supplied composite stock factor into a long-only index-enhancement portfolio. It contrasts common linear optimization, which is simple to compute but ignores relationships among stocks and can produce concentrated holdings, with quadratic optimization, which can incorporate covariance information. The reported tests found little improvement from covariance estimated using historical returns; estimates based on style factors modestly raised the information ratio but also increased volatility and drawdown, while results were sensitive to parameter choices.
The proposed alternative adapts risk-budgeting ideas from asset allocation and uses cyclic coordinate descent and the alternating direction method of multipliers to solve constrained portfolio problems. An example using a quality factor compares enhanced portfolios against two Chinese equity indices over a stated historical test period. The summary reports improved information ratios, lower drawdowns and turnover, and slightly higher diversification ratios versus linear optimization. These findings are specific to the example and period; the source summary does not provide enough implementation detail to assess robustness or generalize the results.
Key ideas
- Linear factor portfolio optimization is computationally simple but can ignore cross-stock relationships and concentrate holdings.
- Quadratic optimization can use covariance information, but its gains depend on covariance estimation and parameter choices.
- The proposed method frames portfolio construction as a constrained risk-budgeting problem.
- Cyclic coordinate descent and the alternating direction method of multipliers are used to solve the optimization.
- The reported index-enhancement example shows better risk and turnover measures than linear optimization over its test period.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.