Stock returns, quantile autocorrelation, and volatility forecasting

Abstract We examine stock return autocorrelation at various quantiles of the returns' distribution and use it to forecast volatility. Our empirical results show that strength autoregression varies across in terms both magnitude persistence. Specifically, order coefficients is lower left tail comparison with right tail. Additionally, we quantile autoregressive (QAR) framework proposed this study improves out-of-sample volatility forecasting performance compared generalised conditional heteroscedasticity (GARCH)-type models other quantile-based models. also observe greater outperformance QAR estimates during periods financial turmoil. Moreover, method explains stylized ‘leverage effect’ associated asset returns presence asymmetry.