Abstract

Market microstructure based proxies for information asymmetry have been widely used for over two decades. However, their empirical validation is surprisingly scarce. We attempt to address this gap by empirically testing two of the more popular proxies, namely, Glosten and Harris (1988) adverse selection cost component of the bid-ask spread (lambda), and Easley, Hvidkjaer, and O’Hara (2002) probability of informed trading (PIN). We estimate these proxies across three portfolios: broad-based ETFs, sector ETFs, and common stocks. Arguably, information asymmetry about broad-based ETFs should be primarily related to market wide information asymmetry, while information asymmetry about stocks should be mostly related to firm characteristics. We find that while PIN is highest for the broad-based portfolio and lowest for the stock portfolio, lambda is highest for the stock portfolio and lowest for the broad-based portfolio. To the extent that information asymmetry about stocks should be greater than information asymmetry about systematic factors, we provide evidence in support of lambda as a measure of information asymmetry. PIN is more closely related to liquidity.