Knowledge that Transforms

We examine the accounting choice decision in the context of the timing of adoption of SFAS 87. Unlike most prior studies, we consider interactions between firm characteristics and the magnitudes of the financial statement effects of an accounting decision. We expect that including interactions will both enhance the ability to explain accounting choice, and facilitate distinction between omitted variables (Ball and Foster, 1982) and hypothesized relations. Results are consistent with these expectations.

This paper studies accounting choice in 76 NYSE firms with persistent losses and dividend reductions (40% forced by binding covenants). We find that managers' accounting choices primarily reflect their firms' financial difficulties, rather than attempts to inflate income. Firms with and without binding covenants exhibit minor accrual differences in the ten years before the dividend reduction. In the dividend reduction and following three years, the full sample exhibits large negative accruals that likely reflect the fact that 87% of sample firms engage in contractual renegotiations -with lenders, unions, government, and/or management-that provide incentives to reduce earnings.

Differences between SIC codes assigned to companies by COMPUSTAT and CRSP are examined. Large differences are observed at two-, three-, and four-digit levels. Correlations of intra-industry monthly stock returns are larger, and variances of intra-industry financial ratios are smaller for industries based on COMPUSTAT codes. Replication of a portion of Freeman and Tse (1992) produces significant results using COMPUSTAT codes, consistent with the original research, but insignificant results for CRSP codes.

We present evidence that 20 percent of the 10-Ks in our sample are filed with the SEC after the 90-day statutory due date. Firms that delay filing their 10-K are not a random sample of firms; up to 25 (10) percent of the firms experiencing unfavorable (favorable) economic events delay their 10-K. Firms that delay their 10-K are, on average, small, have negative accounting rates of return, negative earnings changes, low liquidity, and high financial leverage; they also experience negative market- adjusted stock returns.

Controlled experiments provided evidence that (1) employees are more likely to accept incentive contracts described in bonus terms than contracts that appear identical except for being described in penalty terms, and (2) when employees' judgment of their past performance is dependent on memory, the preference for bonus over penalty contracts increases with experience. These phenomena are explained in terms of the human information processing costs of communicating and evaluating the contract terms, and further implications are drawn for the empirical study of contracting.

This paper tests Jensen's (1986a) free cash flow theory that managers overinvest retained earnings by undertaking negative-NPV projects. The test is based on the theory's implication that earnings response coefficients depend positively on payout ratios. It is supported by the data for a sample of 112 firms over the 1951–1986 period. Another test finds that the returns on retained earnings are lower than proxies for the market's required rate of return, as implied by the free cash flow theory. However, the sample firms' Jensen alphas are reliably positive, suggesting that caution is warranted in interpreting the results.

This research investigates whether volume reactions to a public announcement are related to changes in the risk of securities (i.e., investors undertake portfolio rebalancing when the risk of their portfolio becomes misaligned with their respective risk preferences). Our results document that an average (40 percent) change in beta is associated with a 0.10 percent increase in the number of shares traded in a ten-day period around the earnings announcement. Although risk clientele effects are less important than information effects, they are empirically significant.

This study documents one effect of the theory of implicit taxes, providing evidence that a change in the tax rate results in a change in pre-tax returns. Yield spreads of pairs of Treasury bills maturing in the last week of December and the first week of January are examined. Year-ends not affected by rate changes show a significant positive yield spread between these pairs of bills, reflecting an upward-sloping yield curve. However, for year-ends coinciding with the tax rate reductions of 1981 and 1986 there is a significant negative yield spread between these pairs of bills.

As in other countries, Australian accounting for unincorporated joint ventures varies systematically between firms in different phases in the extractive industries (explorer vs. producer). We argue that these differences in accounting method can be explained by differences in the type of assets and the manner in which they are owned and financed. We hypothesize that when unincorporated joint ventures are financed on a with-recourse basis they will be proportionately consolidated; when they are financed on a non-recourse basis one-line reporting is expected. Empirical results are consistent with this hypothesis and appear to generalize to the real estate industry.

This study reexamines competition, cost structure, and growth opportunities effects on earnings response coefficients and extends this literature in several ways. First, it presents a more refined theoretical motivation for investigating competition and cost structure effects. Second, it introduces new economic factor proxies that confirm prior findings with respect to competition but differ from prior findings with respect to cost structure and growth opportunities. Overall the evidence suggests that accounting earnings reflect information about future economic rents generated by firms' assets-in-place. The evidence also suggests, contrary to prior studies, that accounting earnings are not very informative about firms' growth opportunities.