We present a pricing model of bank bailout insurance guarantees against tail risk and empirical evidence that provides a rational explanation why big bank equities “underperform” relative to small banks during normal times while they “overperform” during crises. A new measure accounting for left-tail risk protection against losses conditional on a crisis explains the “underperformance” of large banks during normal periods. Over the long-term spanning several economic cycles, bank assets are fairly priced regardless of size. Our empirical evidence supports our model’s predicted pattern of excess bank return reversals across economic cycles following Too-Big-To-Fail (TBTF) bailout policy in 1984.

We show how firm-level real options lead to idiosyncratic skewness in stock returns. We then document empirically that growth option variables are positive and significant determinants of idiosyncratic skewness. The real option impact on skewness is more significant in firms with lottery-type features, small size, high volatility, distressed, low return on assets, and low book-to-market ratio. We also find that expectation on idiosyncratic skewness is associated with lower Sharpe ratios. This suggests investors are willing to sacrifice mean-variance portfolio efficiency for greater skewness deriving from real options. Furthermore, financial flexibility has a positive incremental effect, enhancing the beneficial role of asset flexibility on idiosyncratic skewness.

Considerable debate surrounds how the US government's TARP bailout intervention has affected the risk-taking and moral hazard behavior of U.S. banks around the global financial crisis. We examine this issue with a focus on lottery behavior introducing MAX/MIN as a new measure of lotteryness in banking to capture the loss protection from bank bailout guarantees. We find that the TARP bailout increased the likelihood of bank lotteryness and risk shifting. Lottery-like bank equities are riskier after TARP and exhibit fatter right to left tails. A consistent pattern of risk taking and lottery behavior extends both before and after the 2008–2009 crisis, engulfing the largest systemic banks (SIFIs). While confirming that lottery-like bank equities have lower short-term return, we find they exhibit better cumulative long-term return performance. Our findings have important policy implications regarding government intervention in banking crises.