Investors spend hours poring over SEC filings and financial statements seeking clues to the value and future financial performance of firms. The most sought-after items are those that assist in predicting future earnings and/or share prices, but that may have been traditionally overlooked by investors.
Professor Andrew Schmidt may have uncovered one of these oft-overlooked earnings signals. Schmidt observed that many firms described changes in their effective tax rates (ETRs) to explain changes in earnings per share (EPS), which suggested to him that managers think ETR information is value-relevant. For example, in the third quarter of 2000, Hewlett-Packard (HP) lowered its year-to-date ETR by one percentage point, which increased EPS by over three percent.
Many investors view ETR changes as attempts at aggressive accounting or outright earnings management — for example, if not for that one percentage-point ETR change, HP's EPS would have hit below analysts’ quarterly forecasts. But little attention has been directed at determining whether ETR changes can offer any clues to future earnings, and this is what Schmidt wanted to learn.
The ETR represents the amount of tax expense reported in the income statement divided by a firm's pretax income. The statutory rate on corporate income is 35 percent, but the ETR often varies due to state and local taxes, specialized tax breaks like the research and development tax credit, lower tax rates on foreign income, financial reporting items like the deferred tax asset valuation allowance, and tax-exempt income like municipal bond interest. These differences between the statutory rate and the ETR are disclosed in the ETR reconciliation of the income tax footnote in a firm's quarterly and annual reports.
To learn about the relationship between future earnings and income taxes, Schmidt separated earnings into two parts, the change in net income assuming the ETR remained the same as the prior year and the change in net income due to the change in the ETR. Schmidt called this second component the tax change component, or TCC. In order to capture the most useful portion of the TCC, Schmidt took advantage of quarterly financial reporting requirements to further decompose the annual TCC into an initial and revised portion based on the first quarter estimate of the annual ETR.
Accounting rules require firms to compute quarterly tax expense, in part, by estimating the ETR that is expected to be applicable for the full fiscal year. Interviews with tax and accounting professionals led Schmidt to believe that the items that would have the most persistent effect on management's estimate of the annual ETR would be reflected in the initial (first quarter) ETR estimate. Estimates in later quarters, he determined, would likely contain stale information and reveal potential attempts at aggressive accounting.
Schmidt used quarterly and annual data from Compustat for almost 13,000 firm-year observations from 1994 to 2001 to test his ideas about ETRs and earnings. Using this data, he determined that the annual TCC was useful in predicting both future earnings and future stock returns. This result essentially debunks the notion that earnings due to ETR changes are predominately related to aggressive accounting. His results then show that the initial TCC was more useful than the revised TCC in forecasting future earnings, which suggests that the first quarter ETR estimate is much more useful in earnings prediction than later ETR estimates. Schmidt’s data also suggests that ETR changes have contributed to the mispricing of shares, which closer consideration of both tax-change components can help correct.
“In the past people viewed these earnings changes caused by ETR changes as one-off earnings-management techniques that have no implication for future earnings,” Schmidt says. “ETR changes were thought only to influence short-term earnings.” His work corrects that misperception.
“Investors should take a close look at firms’ tax footnotes because the ETR and the ETR reconciliation — which are mandated by the SEC — are there,” Schmidt says. “Even though the data suggests that the first quarter ETR estimate captures useful information, investors should consider paying careful attention to the information in the ETR reconciliation given the evidence of its usefulness in forecasting future earnings.”
Andrew Schmidt is assistant professor of accounting at Columbia Business School.