The Effect of Margin Requirement on Expected Return: A Natural Experiment

Abstract

This paper tests directly the theoretical result of Gârleanu and Pedersen (2011) and Ashcraft, Gârleanu and Pedersen (2010) which state that the required return of a security depends on its market beta and its margin requirement. I find strong empirical evidence in support of the theoretical result that the security’s required return increases with its margin requirement, by conducting a natural experiment with event study setting that exploits unique margin requirement system in National Stock Exchange (NSE) India, where specific threshold rules determine a stock's margin trading eligibility monthly, which cause exogenous variation in the stock-level margin requirements over time in the dataset. The asset pricing effect of margin requirement change shows as high positive abnormal return during an event period when a stock becomes margin trading eligible. Conversely, stocks that become margin trading ineligible face highly negative abnormal returns during the margin requirement change event. I also find that some investors with levered margin positions are able to forecast margin requirement changes in the context of NSE’s margin requirement system, which shows as unwinding of the margin positions prior the margin ineligibility event occurs. In addition to this, I show both theoretically and empirically how it is possible in practice to forecast these margin requirement change events one month in advance in the context of NSE’s margin requirement system, by constructing a forecasting model that utilizes machine learning logistic regression algorithm that uses publicly available data.