This paper considers inference in a broad class of nonregular models. The models considered are nonregular in the sense that standard test statistics have asymptotic distributions that are discontinuous in some parameters. It is shown in Andrews and Guggenberger (2009a) that standard fixed critical value, subsampling, and m out of n bootstrap methods often have incorrect asymptotic size in such models. This paper introduces general methods of constructing tests and confidence intervals that have correct asymptotic size. In particular, we consider a hybrid subsampling/fixed-critical-value method and size-correction methods. The paper discusses two examples in detail. They are (i) confidence intervals in an autoregressive model with a root that may be close to unity and conditional heteroskedasticity of unknown form and (ii) tests and confidence intervals based on a post-conservative model selection estimator. Copyright 2009 The Econometric Society.

This paper provides evidence on the risk factors that are priced in bank equities. Alternative empirical models with precedent in the nonfinancial asset pricing literature are tested, including the single-factor CAPM, three-factor Fama–French model, and ICAPM. Our empirical results indicate that an unconditional two-factor ICAPM model that includes the stock market excess return and shocks to the slope of the yield curve is useful in explaining the cross-section of bank stock returns. However, we find no evidence that firm specific factors such as size and book-to-market ratios are priced in bank stock returns. These results have a number of important implications for the estimation of the banks’ cost of capital as well as regulatory initiatives to utilize market discipline to evaluate bank risk under Basel II.

This paper examines the potential impacts of artificial smoothing (abnormal accruals) and real smoothing (derivatives) on firm value. We find that the value of the firm decreases with the magnitude of abnormal accruals and increases with the level of derivative use. Moreover, the accrual discount is more pronounced in firms with weak investor protection and the hedging premium is greater for poorly governed firms. These results suggest that although managers can engage in real smoothing to improve the informativeness of firms’ earnings and thus reduce agency costs, they might use artificial techniques to cosmetically improve the income stream in order to expropriate minority shareholders. In further support of agency theories, we report that poor corporate governance motivates the use of abnormal accruals and discourages derivative use.

ABSTRACT: Optimal agency contracts pay the lowest wage necessary to induce profit-maximizing effort. Employees could view such contracts as violating reciprocity because, relative to more reciprocal contracts, they offer a lower wage in exchange for higher effort. Consequently, the profit-maximizing effectiveness of optimal contracts could be impaired if employees reject them or reduce their effort. We use experimental labor markets to examine (1) how employees respond to an optimal versus a suboptimal reciprocity-based contract when each contract is the only contract available, (2) how employees respond to these contracts when firms choose which one to offer, (3) whether the firms' contract offers depend on employees' reactions to those offers, and (4) how employees and firms react to a hybrid contract that incorporates features of both contracts. We find that the optimal contract is less effective than agency analysis predicts, the reciprocity-based contract can be equally effective, and the hybrid contract dominates a market in which all three contracts are available. Implications of these results are discussed.