Forecasting the Volatility of Australian Stock Returns
开发了多变量因子模型来预测澳大利亚股票的波动性，提出对跳跃稳健的估计方法，实证表明多变量模型优于单变量模型。
This article develops multivariate factor models for forecasting volatility in Australian stocks. We suggest estimation procedures for approximate factor models that are robust to jumps when the cross-sectional dimension is not very large, and also work with volatility measures that have been constructed so that they contain no jump components. Out-of-sample forecast analysis shows that multivariate factor models of volatility outperform univariate models, but there is little difference between simple and sophisticated factor models.