Nonlinear Relationship Between Investor Sentiment and Conditional Volatility in Emerging Equity Markets

The present study aims to identify the non-linear relationship of bullish and bearish investor sentiment with conditional volatility. It is conducted in emerging equity markets of Brazil, India, Pakistan, Russia, Indonesia, South Africa, and China. The data regarding share prices, shares outstanding, and trading volume is collected from the representative indices for a period from 2001 to 2020. Investor Sentiment Index is constructed using Principal Component Analysis and decomposed into bullish and bearish investor sentiment. The GARCH model is applied to generate conditional volatility and the Non-linear Auto Regressive Moving Average model is applied to analyze the asymmetric relationship between conditional volatility and investor sentiment at the country level. The Panel GARCH model is applied to generate conditional volatility for panel data, and the Non-linear Dynamic Auto Regressive Moving Average model is applied to investigate the nonlinear relation of investor sentiment with volatility. Bullish and bearish investor sentiments show a significant effect in generating conditional volatility in the markets in both linear as well as nonlinear settings.