The strategy uses machine learning to rank bonds by boosted momentum, focusing on equity returns, market value, and liquidity, selecting top quintile bonds, rebalanced monthly for optimized returns.
ASSET CLASS: bonds | REGION: United States | FREQUENCY: Monthly | MARKET: bonds | KEYWORD: Corporate Bonds
I. STRATEGY IN A NUTSHELL
Use machine learning (Boosted Regression Trees) to rank corporate bonds by equity momentum, market value, and liquidity. Go long on the top quintile, equally weighted, and rebalance monthly for enhanced return prediction.
II. ECONOMIC RATIONALE
Equity momentum predicts corporate bond returns due to slower bond market adjustments. Liquidity and market value affect signal translation speed, while machine learning identifies profitable patterns and key economic drivers.
III. SOURCE PAPER
Boosting Momentum [Click to Open PDF]
Hendrik Kaufmann, Deka Investment GmbH; Philip Messow, Robeco Institutional Asset Management; Jonas Vogt, Quoniam Asset Management GmbH
<Abstract>
Machine learning techniques have gained popularity in recent years, but only to a limited extent in fixed income research. We do some new work in the application of Boosted Regression Trees to the equity momentum factor in the corporate bond market. We report significant performance gains to investors using machine learning-driven forecasts, roughly doubling the alpha and information ratio to better known equity momentum strategies. In addition to past equity returns, we include size and liquidity of stocks and bonds into our model framework.
IV. BACKTEST PERFORMANCE
| Annualised Return | 4.21% |
| Volatility | 2.46% |
| Beta | N/A |
| Sharpe Ratio | 0.92 |
| Sortino Ratio | N/A |
| Maximum Drawdown | N/A |
| Win Rate | N/A |
