Research Area
Equity risk-model diagnostics, factor-structure analysis and covariance forecasting tools for Australian stock portfolios.
Principal-component analysis of Australian equity returns, including diagnostics on the number of statistical factors and the stability of latent factor structure across rolling five-year windows.
Evaluation of economic, style and statistical factors using factor-mimicking portfolios, return diagnostics and benchmark comparisons against randomly formed portfolios.
Correlation diagnostics showing where factor-mimicking portfolio returns overlap, where exposures are distinct and where apparent factor families may be capturing similar underlying risks.
Market-purged diagnostics comparing industry returns with factor returns to identify where style effects overlap with sector structure and where factor signals add information beyond industry risk.
Nested cross-sectional regressions estimating the incremental explanatory power of industries, observable attributes and statistical factors for Australian stock returns.
Tests of alternative covariance models, including factor-based covariance forecasts, residual covariance structure, shrinkage methods and realised-risk comparison over forward horizons.
Covariance-forecasting results are currently being prepared for publication.