With quarterly earnings calls a routine corporate ritual, it’s imperative not to let the cat out of the bag. That may be easier said than done. How can top management guard against unintentionally revealing potentially damaging inside information in the course of responding to analyst and investor queries?

An obvious way is through scripting — anticipating questions that may be troublesome and preparing answers that will yield no hint of problems ahead. But new research raises red flags regarding this practice for both managers and investors.

A study in the new issue of the American Accounting Association journal The Accounting Review finds “a negative current market reaction to Q&A scripting, suggesting that investors discern the lack of spontaneity and view it as a negative signal, thus calling into question the usefulness of this strategy as a means of delaying the disclosure of bad news for investors.”

Prof. Joshua Lee

Based on a sample of earnings calls of 2,861 companies over 10 years, the paper’s author, Florida State University professor Joshua Lee, found that the stock of companies that relied most heavily on scripting (the top quintile in that regard) sustained a two-day, market-adjusted return significantly below the sample mean on the day of and day following conference calls.

While the mean two-day, market-adjusted return for the sample was a gain of 0.1%, heavily scripted firms lost 0.2%, putting them 0.3 percentage points below the average. In marked contrast, shares of firms that employed scripting minimally or not at all rose 0.3 percentage points above the mean.

And these market responses were not misguided, the study reveals. Probing unexpected earnings (the amount that earnings diverge from analyst predictions), Lee found that “investors’ reaction to managers’ lack of spontaneity is consistent with negative future unexpected accounting performance in the next two quarters.” In other words, it’s consistent with [weak] future fundamentals for the company.

Over those two following quarters, mean unexpected earnings of the most-scripted quintile of companies were 12.5% lower than mean unexpected earnings of the full sample.

Script Detector

The researchers used linguistic software to compare the difference in managerial speaking style as evidenced by call transcripts between the two portions of earnings calls: the management discussion (MD), in which the CEO, CFO, or other top executive reads from a prepared text, and the Q&A with analysts and investors.

“Research suggests that the style of written language differs fundamentally from spontaneous speech,” Lee wrote. “If the manager adheres to a script when responding to analysts’ questions, the manager’s speaking style during the Q&A is more likely to resemble the style of the MD portion. If, on the other hand, the manager speaks spontaneously, speaking style is likely to significantly differ between the two portions of the call.”

Employing an intricate methodology that combines trigonometry and vector math, the professor gauged the degree of difference between the MD and Q&A portions of 40,820 earnings calls and the relationship of that difference to various company-specific stock market and earnings measures.

In testing whether his findings were statistically significant, Lee controlled for an array of factors that can affect investors’ response to calls, including company size and growth; share-price volatility; surprises in current earnings; surprises in future-earnings guidance issued during calls; the tone of calls (balance between positive and negative words); and whether a management spokesperson refused to respond to a query.

In addition to discovering that scripting was negatively related to two-day market returns and to longer-term unexpected earnings, Lee found the following:

• The greater the amount of scripting, the further downward analysts revised their earnings forecasts for companies during the 30 days following a call. This is evidence that, as the study puts it, that they “view managers’ lack of spontaneity as a negative signal.”
• Greater scripting occasioned greater bid-ask spreads on company stock in the three days following a call as compared with spreads in the three days preceding the call.
• Companies that relied heavily on scripting (the top decile in reliance) were about 20% less likely than those that were least scripted (bottom decile) to issue future earnings guidance in the course of the call. This finding “suggest[s] that firms provide less, not more, information when their managers adhere to a Q&A script.”

While the study’s findings amount to a red flag for investors who detect scripting in a CEO’s or CFO’s response to queries, Lee cautions against jumping too quickly to conclusions.

“A CEO or CFO who is not especially fluent,” he says, “may simply prefer to have answers laid out in advance, so one can’t be sure that some kind of concealment is going on without knowing [that executive’s] normal modus operandi. A better bet is to listen for shifts between spontaneous answers and what sound like scripted ones.” Such shifts suggest a sensitive issue, Lee notes.

He also concedes that in some circumstances scripting may be the lesser of two evils for managers, its negative effects notwithstanding. “Managers may find themselves between a rock and a hard place, where the disadvantages of inadvertently disclosing something unfavorable are greater than sounding scripted. It’s a dilemma that more than a few top managers are likely to face at some point.”

The new study, entitled “Can Investors Detect Managers’ Lack of Spontaneity? Adherence to Predetermined Scripts during Earnings Conference Calls,” is in the January/February issue of The Accounting Review.