Quantile Regression Methods and Applications in Finance for Style Analysis and Market Timing

The performance of actively managed mutual funds is often described in terms of style exposures and market timing. Classical approaches address both dimensions through mean regressions, which summarize behaviour by average factor loadings and mean excess returns. This perspective, however, does not fully reflect the heteroskedastic, asymmetric and heavy-tailed nature of equity returns and provides limited information about behaviour in the tails of the distribution. This thesis studies quantile regression methods and their application to style analysis and market timing. On the timing side, the analysis replaces a single market beta and a single timing coefficient with quantile paths that track how sensitivity and curvature vary across the conditional return distribution. A dual-benchmark rule based on tail confidence intervals is introduced to judge whether apparent timing signals are confirmed once both a broad market and a style-aligned benchmark are considered and once regressions are repeated on volatility-standardized returns. On the style side, the thesis refines returns-based style analysis through rolling constrained decompositions and a ridge-regularized version that manages concentration and turnover through a transparent selection rule. The research analyzed six U.S. equity mutual funds including large-cap growth, large-cap value, small-cap equities and real estate. Quantile paths occasionally indicated convex or concave behaviour in specific parts of the distribution, but once timing was evaluated under the dual-benchmark tail rule and checked for robustness to volatility standardization, no fund displayed confirmed timing at conventional confidence levels. By contrast, returns-based style analysis reveals stable average style with low to moderate turnover. The ridge-regularized style program reduced concentration without a material loss of fit and with bounded turnover, which supports the interpretation of these funds as primarily style-focused mandates with limited distribution-aware timing.