Investors will turn to less credible sources of information about stocks they hold when more credible sources provide unwelcome information, according to a new study.
If a highly credible source bears unwelcome tidings — for example, that an investor's stock is not likely to meet an earnings target — the credibility of that source does not count for much, according to new research by Jane Thayer, an assistant professor at the J. M. Tull School of Accounting at the University of Georgia. Her research appears in the current issue of the American Accounting Association’s Accounting Review.
"Investors receiving unfavorable information regarding a recent investment decision will seek additional information that substantiates their decision, even if it requires them to forfeit a certain amount of credibility in that information," wrote Thayer.
An investor in this situation, the study continues, "will not gather a balanced set of credible information, but will seek information that bolsters his/her chosen position," even though the outcome "is a set of information that can result in biased estimates and foregone profits."
The study breaks significant new ground in assessing a phenomenon introduced in the psychology literature more than a half century ago and called "cognitive dissonance," a mental inconsistency that causes individuals discomfort which they seek to alleviate as naturally as they would act to reduce hunger or thirst. While cognitive dissonance has been studied in many contexts, the new study focuses the most specifically on how it affects investor information-seeking, and the distinctive nature of the paper's findings helped earn it the Competitive Manuscript Award of the American Accounting Association.
“It's not as if credibility gets thrown to the winds completely, just that it loses some of the weight it generally has,” said Thayer. “Confronted by bad news about their investments, people seem impelled to seek out good news even if it doesn't come from a particularly credible source. The finding, I believe, should foster increased appreciation of how difficult, as well as important, it is to maintain discipline in investing."
The paper's findings derive from a single-session, Web-based experiment involving 92 second-year students from a highly rated MBA program, the vast majority of whom had at least some experience with stock investing. Participants were randomly assigned to either a long position or short position, after which they viewed basic information about two unnamed pharmaceutical companies modeled after actual firms (including revenue growth, profit margins, R&D expenses, and major recent developments in governance and operations) and were asked to choose between them as investments. The subjects were then randomly assigned an earnings-per-share benchmark to which actual EPS would be compared to determine whether they gained or lost from their investments — a benchmark of either $3.84 or $3.76.
The MBA candidates then proceeded to the heart of the experiment, where they learned that the analyst consensus for their stock was $3.80, which was either favorable news or unfavorable news depending on what benchmark participants were assigned and whether they were long or short. Then four analyst forecasts were displayed, two for $3.88 and two for $3.72, one forecast for each figure emanating from a well-known firm, and one from an unknown firm. The forecasts were linked to a corresponding analyst report, and the key experimental measures for Thayer were what reports participants clicked on and how much time they devoted to each report within an overall time limit of three minutes.
The professor found markedly different results, depending on whether the $3.80 consensus forecast (the initial forecast to which subjects had been exposed) constituted favorable or unfavorable news.
Participants for whom the consensus forecast was unfavorable devoted about one-third more time to viewing the analyst reports than did the favorable group, and they divided their time in strikingly different ways. Whereas the favorable group actually spent significantly more time on high-credibility reports that ran counter to their preference than to those that coincided with their preference, the unfavorable group devoted over two and a half times more viewing time to the latter than to the former.
Perhaps even more striking, the unfavorable group actually spent slightly more time on low-credibility reports that coincided with their preference than high-credibility reports that ran counter to it.
"Participants who initially received favorable information chose to view a more balanced set of preference-consistent and preference-inconsistent information compared to participants receiving unfavorable information,” said the study. “Participants receiving initial unfavorable information spent a majority of their viewing time on preference-consistent information...[and] preferred to view at a similar rate less-credible information that was preference-consistent and more-credible information that was preference-inconsistent."
As the last step in the experiment, after they had the chance to peruse analyst reports, participants were asked to make an earnings-per-share estimate for their company. Here too there were sharp differences, as "participants in the unfavorable condition provided final EPS forecasts that were on average less accurate than those in the favorable condition. These findings suggest that, in spite of the additional time participants in the unfavorable condition spent gathering information, their choice to view additional preference-consistent information...[led to] final EPS forecasts [that] were biased in a manner consistent with their preferred outcome."
The paper addressed the difference between its principal findings and those of a large body of psychological research on cognitive dissonance, noting, "These psychology studies provide experimental participants with information that varies at extreme levels of credibility. The capital markets information environment, on the other hand, offers investors a vast amount of information from sources representing a wide range of credibility. My findings suggest that investors looking to support a previous investment decision are willing to forego a certain level of source credibility to find supportive information. However, at the point at which the variance in two sources' credibility becomes too extreme, investors will likely choose to acquire information from the more credible source."
The study, entitled "Determinants of Investors' Information Acquisition: Credibility and Confirmation," is in the January/February issue of The Accounting Review, published six times a year by the American Accounting Association, a worldwide organization devoted to excellence in accounting education, research and practice. Other journals published by the AAA or its specialty sections include Accounting Horizons, Issues in Accounting Education, AUDITING: A Journal of Practice and Theory, Behavioral Research in Accounting, The Journal of the American Taxation Association, and The Journal of Management Accounting Research.
