We study the performance and behavior of Value at Risk measures used by a number of large U.S. banks before, during and after the financial crisis. Alternative benchmark VaR measures, including GARCH-based measures, are estimated directly from the banks’ trading revenues to explain the bank VaR performance results. While overly conservative in both the pre-crisis and post-crisis periods, bank VaR exceedances were excessive and clustered in the crisis period. This contrasted with mostly unbiased benchmark HS and GARCH VaRs in the pre-crisis and post-crisis periods, and vastly superior GARCH-based VaR performance in the crisis period with lower exceedance rates and no exceedance clustering. Our results document the bank VaRs very slow adjustment to changing market conditions and their systematic bias in all studied periods. Our results indicate that bank VaRs could be improved by the use of models with time-varying volatility, and built on banks’ knowledge of their current positions.

In this paper, we empirically estimate the costs of delay in the FDIC's closures of 433 commercial banks between 2007 and 2014 based upon a counterfactual closure regime. We find that the costs of delay could have been as high as 18.5 billion, or 37% of the FDIC's estimated costs of closure of 49.8 billion. We think that these findings call for a more aggressive stance by bank regulators with respect to the provisions for loan losses and write-downs of banks’ non-performing assets. More aggressive (and earlier) provisions and write-downs, or adoption of a capital ratio that penalizes nonperforming loans, would allow the concept of “prompt corrective action” (PCA) to play the role that it was meant to play in reducing FDIC losses from insolvent banks.

This paper analyzes a wide range of flexible drift and diffusion specifications of stochastic-volatility jump–diffusion models for daily S&P 500 index returns. We find that model performance is driven almost exclusively by the specification of the diffusion component whereas the drift specifications is of second-order importance. Further, the variance dynamics of non-affine models resemble popular non-parametric high-frequency estimates of variance, and their outperformance is mainly accumulated during turbulent market regimes. Finally, we show that jump diffusion models yield more reliable estimates for the expected return of variance swap contracts.

By examining the post-retirement outside board seats held by former CEOs of S&P 1500 firms, we find that CEOs’ post-retirement outside board memberships are influenced by the level and the tone of media coverage given to the CEOs’ firms while the CEOs were “on the job.” These results provide evidence of a direct economic link between media coverage of CEOs’ performance today and CEOs’ future opportunity sets. These results lend support to the proposition that the media can play a role in corporate governance by influencing the value of CEOs’ human capital.

We investigate the information content of stock correlation based network measures for systemic risk rankings, such as SIFIRank (based on Google’s PageRank). Using European banking data, we show that SIFIRank is empirically equivalent to a ranking based on average pairwise stock correlations as developed in this paper. The correlation based network measures complement currently available alternative systemic risk ranking methods based on book or market values. A further analytical investigation shows that the value-added appears to be mainly attributable to pairwise cross-sectional heterogeneity rather than to more subtle network relations and feedback loops.

Prior research argues that the process of intermediation is opaque and produces uncertainty about the riskiness of banks, which may adversely affect the efficiency of bank stock prices. Using the Hou and Moskowitz (2005) measure of price delay, which captures the inefficiency of stock prices, we test for, and find evidence supporting the idea that opacity is positively associated with price delay. Bank stocks have markedly higher delay than similar non-bank stocks. This higher level of delay is driven, in part, by market-based measures of informational opacity as well as the asset composition of the bank's balance sheet. Combined, our findings suggest that bank opacity reduces the efficiency of financial markets.

Simple analytical solutions for the prices of discretely monitored barrier options do not yet exist in the literature. This paper presents a semi-analytical and fully explicit solution for pricing discretely monitored barrier options when the underlying asset is driven by a general Lévy process. The explicit formula only involves elementary functions, and the Greeks are also explicitly available with little additional computation. By performing a Z-transform, we reduce the valuation problem to an integral equation. This equation is solved analytically with the solution expressed in terms of a Fourier cosine series. We then manage to analytically carry out the Z-transform inversion, and obtain a semi-analytical formula for pricing discrete barrier options. We establish the theoretical error bound and analyze the convergence order of our method. Numerical implementation demonstrates that our numerical results are accurate and efficient, and match up with the results from the benchmark methods in the literature.

We document a systematic seasonal component in the aggregate underperformance of active mutual funds. At the aggregate level, active funds underperform the market and other passive benchmarks only in the first month of a quarter. This intra-quarter performance seasonality holds across fund sizes and investment styles. The pattern is consistent with short-term stock return reversal effects along with aggregate window-dressing and, to a lesser extent, NAV-inflation practices around quarter-ends. We find marginal or no evidence of microstructure biases, fund investor flows, or cash distributions as sources of this seasonality. Our findings highlight new features of the active management underperformance puzzle.