An in-depth examination of trading volume and return volatility dynamics

Purpose This study investigates the relationship between return volatility and the trading volume for DJIA and the S&P 500 indices using intraday, overnight and daily volatility measures. Design/methodology/approach This paper uses several volatility measures, including overnight (close to open), intraday (Parkinson, Garman–Klass and Roger–Satchell), overnight and intraday (Yang-Zhang), historical volatility, VIX and VIX 9-day and conditional volatility found using the GARCH model and investigates the relation of volume and volatility using the vector autoregression (VAR) and various GARCH methods. Findings Daily returns show persistent, positive lagged volume effects, indicating momentum trading. Overnight volatility leads to higher volumes followed by reversals, while intraday volatility initially increases volume but reverses as informed traders adjust speculative trades. Combining intraday and overnight volatility highlights informed trading during speculative activity, whereas historical volatility shows no significant effect on volume. Volatility Granger-causes volume in most cases except intraday measures. Higher volatility initially raises returns but reverses, with overnight and intraday volatility reducing returns short-term before recovery. Originality/value This is a comprehensive study where nine different measures of volatility are used, and the relationship is examined using the VAR as well as various GARCH models. This helps us understand how volume plays a role in the arrival of information, both symmetric and asymmetric, which could give rise to overreaction/underreaction to good versus bad news.