The investment universe consists of S&P 500 index and 3-months treasury bills. Firstly, the daily open-to-open realized volatility is decomposed into open-to-close and close-to-open volatility. For open-to-close volatility, apply the pre-averaging RV estimator of Jacod, Li, Mykland, and Podolskij (2009). The close-to-open volatility is estimated as the squared overnight return. Therefore, the final estimate for daily volatility is open to close pre-averaged RV plus the squared overnight return.
ASSET CLASS: bonds, ETFs, futures | REGION: United States | FREQUENCY:
Weekly | MARKET: bonds, equities | KEYWORD: Machine Learning, Equity Indices
I. STRATEGY IN A NUTSHELL
Invest in S&P 500 and T-bills using weekly volatility predictions from Lasso regression. Adjust allocations to target risk, increasing exposure in low-volatility periods and reducing it in high-volatility periods.
II. ECONOMIC RATIONALE
Volatility targeting protects investors during high-risk periods and exploits low-risk periods. Machine learning improves prediction accuracy over historical volatility, enabling robust forward-looking allocation decisions using Lasso regression.
III. SOURCE PAPER
Forecasting Stock Market Volatility and Application to Volatility Timing Portfolios [Click to Open PDF]
Dohyun Chun, Yonsei University; Hoon Cho, Korea Advanced Institute of Science and Technology (KAIST); Doojin Ryu, Sungkyunkwan University
<Abstract>
This study predicts stock market volatility and applies them to the standard problem in finance, namely, asset allocation. Based on machine learning and model averaging approaches, we integrate the drivers’ predictive information to forecast market volatilities. Using various evaluation methods, we verify that those high-dimensional models have better predictive performance relative to the standard volatility models. Furthermore, we construct volatility timing portfolios and discover that portfolios based on high-dimensional models mostly yield higher Sharpe ratios compared with the market. Among others, the least absolute shrinkage and selection operator (LASSO) generates the most accurate forecasts, leading to outstanding investment performance, regardless of the forecasting horizon.
IV. BACKTEST PERFORMANCE
| Annualised Return | 14.49% |
| Volatility | 23.04% |
| Beta | N/A |
| Sharpe Ratio | 0.63 |
| Sortino Ratio | N/A |
| Maximum Drawdown | N/A |
| Win Rate | N/A |
