Death, Destruction, and Manipulation

Abstract

We propose during periods of natural disasters, there is greater scope for the frequency and severity of market manipulation due to distraction and investor misestimations. For the population of U.S. stocks from 2007-2018, we merge intra-day manipulation data from SMARTS / NASD with National Oceanic and Atmospheric Administration's National Weather Service data. Disasters are measured by the dollar value of property damage and crop damage, and the number of injuries and deaths. Damages to property and crops are further scaled by county-level GDP. The data indicate that market manipulation is more common, and the trading value surrounding the manipulation is more pronounced, among firms that are headquartered in disaster zones. We further examine industry-specific effects, which shows more pronounced effects in intuitive ways; for example, crop and property damages enhance manipulation in agricultural sectors, while injuries enhance manipulation in the manufacturing sectors. These findings are robust to controls for firm idiosyncratic and systematic risk, as well as various methods that include but are not limited to using an alternative proxy of manipulation and difference in difference (DiD) analysis.