Accrual accounting—recording revenues and expenses when they are earned and incurred, rather than when cash is actually transferred—is integral to the accuracy of a company’s financial reports. The primary role of accruals is to help evaluate a company’s economic performance more accurately, but it is difficult to tell when accruals do this job in a trustworthy way.
How well do accruals perform their primary role? Neither practitioners nor academics have developed a good understanding of how accurately accruals measure performance, says Chicago Booth’s Valeri Nikolaev. However, he suggests an approach that could help.
Say a manufacturer delivers a truckload of shirts to a retailer, which won’t submit a payment until the next fiscal quarter. In accrual accounting, the manufacturer would go ahead and recognize the expenses and revenue immediately, before any cash arrives. Accordingly, accrual accounting requires accountants to estimate cash flows associated with revenue and expense transactions, which creates a possibility that accruals are untrustworthy.
Existing accrual-quality measures are of limited use, says Nikolaev, because there’s no obvious way to separate the company’s underlying performance from accounting errors, which could occur due to several factors.
“First, performance measurement requires making assumptions, estimates, and judgments, which give rise to estimation error,” writes Nikolaev. “Second, error may occur because GAAP imposes constraints on what accountants may report. Even if a firm’s management observed true performance, they must follow GAAP measurement rules, which are aimed at the minimization of aggressive accounting, rather than just telling investors what the performance is. Finally, error can arise due to intentional earnings manipulations.”
Nikolaev’s model suggests a way to identify accruals and earnings quality and, in doing so, isolates performance and accounting error.
To identify accounting errors and separate them from performance, Nikolaev introduces a model of accruals that explicitly captures their performance measurement role but allows for the presence of accounting errors. He also introduces an econometric framework that uses a flexible set of assumptions to identify accounting quality and its components.
His model considers institutional properties of earnings, cash flows, and accruals, as needed, as a basis for identifying accounting-quality parameters. For example, if accountants overstate receivables in a given quarter, future earnings should be lower by the same amount. This is a self-correcting effect commonly known as a reversal. At the same time, true performance—untainted by accounting errors in accruals—will generally not exhibit such reversals. Taking advantage of this and other properties of accounting information, Nikolaev’s model can statistically distinguish between performance and accounting error.
Chicago Booth’s Ray Ball and London Business School’s Lakshmanan Shivakumar, among others, have noted that discretionary accruals may tend to accelerate a company’s recognition of unrealized economic losses, but not of gains that arise from changes in expected future cash flows. While this counteracts incentives to over-state earnings, it could potentially make a company’s results appear weaker than they actually are.
Nikolaev says his approach levels out this kind of disparity. His model can account for different treatments of economic shocks—one-time events and other unusual occurrences—to performance. It can also be modified to identify and exclude “income smoothing” and other “managed” or manipulated accrual components.
Nikolaev’s model suggests a way to identify accruals and earnings quality and, in doing so, isolates performance and accounting error. But the researcher says the approach is meant not as the final word in analyzing accounting quality, but rather as a guide for future research.