Rethinking Volume
Professor Philippe van der Beck
Assistant Professor in Finance Unit
Harvard Business School
Harvard University
Total trading volumes in financial markets are enormous and far exceed return volatility. In contrast, “net volume” – trading from long-term portfolio reallocations – is not inflated by transitory round-trip trades and is substantially lower. If return volatility is high, while net volume is low, then market participants either agree with each other (they are “homogeneous”), or they are not sensitive to price changes (they are “inelastic”), implying large price impacts for demand shocks. We formalize this tradeoff and show that the ratio of return volatility to net volume is a lower bound on the price impact, conditional on a given level of investor heterogeneity. Using a host of different measures of investor heterogeneity from survey and portfolio data, we document that heterogeneity is substantial, implying meaningful lower bounds on the price impact. The bounds align closely with reduced-form estimates from a variety of quasi-experiments, such as price impacts from index reconstitutions. Traditional liquidity measures based on total trading volume – such as the Amihud ratio – perform poorly when demand shocks are persistent and liquidity must ultimately come from persistent portfolio reallocations, that is, from net volume. Our bounds are particularly useful in settings where event study evidence is difficult to obtain: We illustrate how they vary over time, across individual assets, and at various levels of aggregation, such as the aggregate stock market, and discuss their implications for asset pricing models and the macro structure of liquidity.