Knowledge that Transforms

With grouping, a sample is sorted by an observable variable and the mean values of the dependent variable in the extreme-ranked groups are compared. We show that test power is maximized when the two extreme groups each contain 27% of the sample, a much larger percentage than that typically used in the literature. This result is not sensitive to the distribution of the dependent variable. We also show that regression is unambiguously more powerful than grouping, even when the independent variable is measured with error.

Most models of market volatility use either past returns or ex post volatility to forecast volatility. In this paper, the dynamic behavior of market volatility is assessed by forecasting the volatility implied in the transaction prices of Standard & Poor's 100 index options. We test and reject the hypothesis that volatility changes are unpredictable. However, while our statistical model delivers precise forecasts, abnormal returns are not possible in a trading strategy (based on daily out-of-sample volatility projections) which takes transaction costs into account, suggesting that predictable time-varying volatility is consistent with market efficiency.

We examine employee stock ownership plan (ESOP) announcements to study the effects of an increase in managerial voting rights without a proportional increase in the ownership of cash flow claims. Our finding that when managers initially control few votes firm value increases with the fraction of shares contributed to the ESOP supports the view that managerial vote control serves shareholder interests. Conversely, the decrease in firm value with larger contributions to the ESOP when managers initially control many votes reflects a divergence of incentives that increases the agency problems between managers and outside shareholders.

We assess potential information transfers by examining the association between the earnings announcements of early and late announcers in an industry. Our earnings prediction models are statistically significant much more frequently than would be expected by chance. The models suggest potential positive information transfers on average, but there is substantial cross-industry variation in the strength of this relation. We find that the greatest price reactions by nonannouncers to same-industry earnings announcements occur in industries with the greatest earnings comovement

This paper re-examines the Ou and Penman (1989) conclusion that fundamental analysis identifies equity values not currently reflected in stock prices, and thus systematically predicts abnormal returns. Their fundamental summary measure Pr, the estimated probability of an earnings increase, also proxies for firm size and CAPM risk. After controlling cross-sectional differences in CAPM beta and firm size, no significant incremental predictive ability is attributable to Pr. The Pr measure is interpreted as a proxy for expected return differences rather than as new evidence of a systematic market underreaction to the future earnings signal inherent in current financial statements.

We examine the profitability of a trading strategy which is based on a logit model designed to predict the sign of subsequent twelve-month excess returns from accounting ratios. Over the 1978–1988 period, the average annual excess return produced by the trading strategy ranges between 4.3% and 9.5%, depending on the specific measure of excess return and weighting scheme involved. However, our implementation of the Ou and Penman (1989) trading strategy in the 1978–1988 period, which is based on a logit model that predicts subsequent unexpected earnings- per-share from accounting ratios, does not earn excess returns.

This paper examines the response of managers of property-casualty insures to the differential costs and benefits of understanding the liability for outstanding claim losses. The primary hypothesis is that the incentive to underestimate the liability is a decreasing function of the insurer's actual financial position. Empirical tests suggest that managers of financially weak insurers bias downward their estimates of claim loss reserves relative to other insurers after controlling for exogenous economic factors. Evidence also reveals that managers of insurers ‘close’ to receiving regulatory attention understate reserve estimates to an even larger degree.

This study examines whether annual financial statements filed with the Securities and Exchange Commission are timely sources of information for investors. We examine a summary measure, the probability of bankruptcy, through which the release of financial statements might communicate information to investors. The results indicate that a significant association exists between revisions in the probability of bankruptcy due to nonearnings data and security returns over the fiscal year, but that investors have largely revised their estimates of the probability of bankruptcy prior to the release of the full financial statements.

This study seeks to identify economic and financial characteristics that distinguish three groups of companies classified by response to the U.K.'s mandatory CCA standard (SSAP 16) as loyal compliers, early defectors and hard-line noncompliers. Three major explanatory variables emerge – leverage, firm size, and the fixed assets to total assets ratio. Overall results suggest that a major motivation for compliance was lower income reporting, especially as an argument for continuing recognition of CCA for tax and regulatory purposes. The belief that noncompliance was largely motivated by preparation costs is discounted.