Trading by Bank Insiders Before and During the Financial Crisis

In the paper, Trading by Bank Insiders Before and During the Financial Crisis, which was recently made publicly available on SSRN, I investigate whether managers of large U.S. banks foresaw the underperformance of their own bank prior to the recent financial crisis. To shed light on this question, I analyze the trades of bank managers in their own bank’s stock. Using banks’ performance during the crisis as an ex-post measure of risk exposure, the paper examines whether the bankers that took the most risk changed their insider trading before the onset of the crisis. The paper also links trading by bank insiders to the developments in the housing market, which played a crucial role in starting the crisis.

The role of bank managers in the crisis has been subject to considerable debate both in the academic literature and in the popular press. On the one hand, Fahlenbrach and Stulz (2011) do not find strong evidence to support the notion that incentive packages contributed to the crisis. Their results indicate that CEOs were holding sizeable equity stakes even as the crisis hit, and did not reduce their ownership in 2007 or during the peak of the crisis in 2008. They conclude that CEOs believed that the risks they took before the crisis would pay off, but that this turned out not to be the case. On the other hand, Bebchuk et al. (2010) criticize the incentive structures of bank managers. They point out that the top managers of Bear Stearns and Lehman Brothers cashed out a substantial amount of options in the period prior to the crisis. Bhagat and Bolton (2011) also dispute that managers had no awareness of the large risks they were facing. They analyze the compensation structure and CEO payoffs of the 14 largest US banks and argue that managerial incentives led to excessive risk-taking. This view is supported also by Cheng et al. (2009), who find a positive relation between excess executive compensation and risk taking. Their evidence suggests that overpaying bank managers who take high risks is positively associated with the level of institutional ownership of the bank.

The analysis of the paper covers the 100 largest U.S. banks. I match insider trading information from Thomson Reuters with stock price information from CRSP, accounting data from Compustat, and most importantly, information on executive compensation packages from ExecuComp. The main finding of the paper is that there are large differences in insider trading behavior between high- and low-exposure banks starting in 2006, when US housing prices indices first declined. During 2006, the number of insiders reducing their ownership increases by 12% in high-exposure banks, compared to low-exposure banks. Furthermore, insiders of high-exposure banks sell 5-7 million USD more of their bank’s stock, on average, than insiders of low-exposure banks. In relative terms, this represents an increase of 30-40% in the total yearly value of stock sales. This increase in insider sales precedes the drop in banks’ stock prices and the surge in banks’ CDS spreads by at least 12 months.

The paper is the first to document that developments in the housing market had a close temporal correlation with the trading decisions of bank insiders. My results show that sales by insiders in banks with high subprime exposures accelerated in the second quarter of 2006, when housing prices at the national level registered their first decline since 2000. The economic effects are sizeable, as bankers increase their sales by 30-45% during each of the first three quarters of housing price declines. The results of the paper are difficult to reconcile with the view that bank insiders continued to believe the investments they made prior to the crisis would pay off after housing prices started to fall in 2006. Rather, the evidence offered in this paper suggests that while bank insiders regarded investments in mortgage-backed securities profitable given the housing price growth, they altered their views on the profitability of these investments following the reversal in the housing market.

The full paper is available for download here.