Wavelet decomposition approach for understanding time-varying relationship of financial sector variables: a study of the indian stock market

In this paper, we study the effect of overall stock market sentiment in India on sectoral indices and on individual stock prices in terms of co-movement, dependence and volatility transmission along with the magnitude and persistence of the effects. The study uses wavelet decomposition framework for breaking down different financial time series into time-varying components. Quantile Regression, Wavelet Multiple Correlation and Cross-Correlation analysis, and Diebold-Yilmaz spillover analysis are then applied to investigate the nature of dependence, association, and spillover dynamics. For further focus, we have considered different time periods separately to identify the effect of market phases. Interesting results are obtained with respect to persistence of shocks, both across and within time periods. These have implications with respect to understanding market behavior and also perception of sectors and stocks. Keywords: Wavelet Decomposition, SENSEX, Quantile Regression, Wavelet Multiple Correlation, Wavelet Multiple Cross Correlation, Diebold-Yilmaz Spillover.