Rising Conservatism: Implications for Financial Analysis

Abstract

We provide evidence that is consistent with an increase in reporting conservatism by U.S. companies in the past few decades. Using a constant sample of almost 900 companies, we examined several measures of accounting conservatism, including the level and rate of accumulation over time of negative nonoperating accruals, the differential timeliness of incorporating good news versus bad news in reported earnings, the skewness and variability of the earnings distribution relative to the cash flows distribution, and changes in the market-to-book ratio. The increased conservatism has contributed to a persistent and prevalent decline in reported profitability, an increase in the incidence of losses, and an increase in the dispersion of earnings. Increased conservatism affects financial ratios and P/E multiples. Thus, incorporating information on the level of a company's reporting conservatism improves valuations and the yield to investment strategies that are based on these ratios.

Anecdotal evidence suggests that financial reporting by U.S. companies has become more conservative in recent decades. For example, most of the new accounting pronouncements have had the effect of accelerating expense recognition and further deferring the recognition of revenues. Furthermore, U.S. capital markets have become more litigious, which induces company managers to be less aggressive in their financial reporting and auditors to be more cautious and prudent in their audits. The observation of an increase in the frequency of losses in recent years, although undoubtedly caused by a number of factors, is consistent with an increase in reporting conservatism.

The fact that generally accepted accounting principles have a built-in conservative bias is widely recognized, but the recent trend toward an even greater conservatism has not been systematically documented. We examine the change in the degree of conservatism in financial reporting over the 1950–98 period and discuss its implications for financial statement analysis. Using a constant sample of almost 900 companies, we identify several measures of conservatism, including the level and rate of accumulation over time of negative nonoperating accruals (defined as the difference between net income and cash flows from operations, excluding depreciation and changes in the balance of noncash working capital accounts), the differential speed of incorporating good and bad news in reported earnings, the skewness and variability of the earnings distribution relative to the cash flows distribution, and changes in the market-to-book ratio (M/B).

The results of a series of tests are consistent with an increase in reporting conservatism, particularly since about 1980. For example, reported profitability gauged by such ratios as return on assets shows a persistent decline over time without a parallel drop in operational cash flows. In fact, the gap between reported earnings and operational cash flows has widened in recent years, which reflects a systematic and material accumulation of negative nonoperating accounting accruals. This trend is consistent with a transition to a more conservative reporting regime. Other measures of conservatism, among them the speedier incorporation in the financial statements of bad news relative to good news, indicate a similar trend.

The finding of increased conservatism suggests that the high M/B values and P/E multiples that peaked in the late 1990s arose, in part, because of changes in the financial reporting regime. Thus, they may indicate more than overpricing. When adjusted for conservatism, an adjustment that takes the form of removing from the income numbers the accumulation of negative nonoperating accruals, the sharp rise in M/Bs and P/Es in the 1980s and 1990s becomes much more modest.

We provide at least one measure of change in reporting conservatism that is readily available to analysts as a way to improve fundamental analysis—the current accumulation of nonoperating accruals. We demonstrate that this measure can be used to characterize the reporting regime of individual companies and, therefore, to adjust the earnings and equity multiples computed for a company.