Managing Mood
December 8, 2017
By Ruth Paarmann

Auditors, regulators, and investors, take note: A CFO’s state of mind can influence financial statements. Research published this year provides evidence that the loan loss provision—an important estimate of future loan charge-offs that banks make each quarter—can be affected by a CFO’s state of optimism or lack thereof.

In February, Assistant Professor of Accounting Sam Melessa and his co-authors, including colleagues Professor Paul Hribar and Assistant Professor Jaron Wilde, published research in the Journal of Accounting & Economics demonstrating that a banking executive’s optimistic state can cause them to underestimate the loan loss provision, the most significant and economically important accrual in the banking industry.

“Sentiment is kind of a fuzzy construct, but with how we measure it, it’s similar to a market-wide optimism factor that is specific to managers,” Melessa says. “My research interests include the quality of information in financial statements as well as the macro factors that affect financial reporting. Managerial sentiment is one of those factors.”

The loan loss estimates affect net income and profit and receive scrutiny from auditors and regulators. According to Melessa, the potential errors that can result from optimism or pessimism are worthy of attention from investors and creditors alike because the errors can cause net income to be a less reliable measure of firm performance. This is important given that net income is widely used in compensation contracts, valuation models, debt covenants, etc.

Gauging Managerial Sentiment

To dive deeper, the research team turned to the Duke University/CFO Magazine Business Outlook Survey to gauge managerial sentiment—a measure of optimism or pessimism—in relation to loan loss provisions provided by banks. The survey asks hundreds of senior financial executives to rate their level of optimism about their firms’ business prospects.

The research found that when sentiment is high and managers are overly optimistic, they underestimate the loan loss provision (i.e., future loan charge-offs), and when managers feel overly pessimistic, they overestimate the provision. In addition to identifying a correlation between managerial sentiment and these estimates, the research team also discovered that this effect is stronger for firms that have loan charge-offs that are more difficult to estimate.

“The managers of banks with charge-offs that are easy to predict aren’t as affected by sentiment. For banks that have difficulty predicting future charge-offs, we see that sentiment plays a bigger role and has a bigger effect,” says Melessa, who worked in finance at Merrill Lynch and Morgan Stanley before entering academia in 2012. “And managers are unaware they are being affected.”

“The types of effects we document could be manifest in financial statements such as the income statement or balance sheet,” he explains. “And sentiment can affect the estimates made by non-banking firms like Microsoft or Walmart.”

The study suggests that during times of high sentiment, when managers are overly optimistic, they are more likely to make accounting estimates that artificially inflate reported net income. For example, managers might unknowingly underestimate bad debts, refrain from making appropriate impairments and overstate the fair values of Level 3 assets. In contrast, during times of low sentiment—when managers are overly pessimistic—they are more likely to overprovision for bad debts, impair asset values, and understate fair values, potentially resulting in a deflated net income number.

“Although our study focuses on the loan loss provision estimate, we expect that other estimates, and therefore reported profits, will also be affected by managerial sentiment,” notes Melessa.

Leveraging This Knowledge

According to Melessa, the research results are important for students as well as regulators, auditors, and others in the industry.

“Academics are generally interested in understanding the factors that affect the quality of numbers in financial statements,” he notes. “Before learning of the research in this area, I was unaware of the effects of these psychological biases, but I’m now more aware of these issues.

“More importantly, there is a debate about how banks should account for the loan loss provision, and this study speaks directly to that issue,” he says. “The accounting rules of loan loss provision have changed over time. Our study demonstrates that factors like managerial sentiment should be taken into account when standard setters and regulators consider the accounting rules for the loan loss provision.”

Melessa also teaches this information to the juniors and seniors in the Financial Accounting and Reporting course to inform and educate them about managerial biases.

“This is important for students, because it demonstrates that you can’t just take numbers on the financial statements at face value. There are different factors and incentives that can influence the numbers, because many of the numbers are estimates. Managerial sentiment is one of the factors that students should be aware of.”

He adds, “We hope this research is informative to users of financial statements so they, too, are aware that this specific factor can affect the statements.”

This article first appeared in the 2017 edition of The Iowa Ledger, a magazine for alumni and friends of the UI Department of Accounting.