Many investors are strongly affected by a company's corporate social responsibility whether they know it or not, according to a new study by a group of accounting professors, and should take into account a company’s CSR record when picking stocks.

The study, which appears in the current issue of the American Accounting Association journal The Accounting Review, found “an unintended causal relationship” between CSR performance and investors’ estimates of fundamental value.

“Absent an explicit assessment of CSR performance, investors’ estimates of fundamental value significantly increase with positive CSR performance,” said the study. “However, consistent with this effect being unintentional [these investors] have relatively low awareness about the significant influence of CSR performance on their estimates.”

“Market participants who ponder a company’s CSR record as well as its financial performance turn out to make less extreme estimates of the company's fundamental value than those who only assess financial results," said Mark E. Peecher or the University of Illinois at Urbana-Champaign, who conducted the research with his colleagues W. Brooke Elliott and Kevin E. Jackson of the University of Illinois and Brian J. White of the University of Texas at Austin. “While our findings do not determine which group's estimate  is right or wrong, they do indicate that the latter group is less aware of the CSR's impact on  their thinking, even though the impact is greater, a combination that does not inspire a lot of confidence in their evaluations.”

When companies have a strong CSR record, investors who focus almost exclusively on financials estimate a company's fundamental value to be about 25 percent higher than those who divide their thinking more equally between financials and CSR. When a firm has a poor CSR record, the former group's estimate is about 9 percent lower.

The study’s findings emerge from an experiment involving 87 graduate business students with prior experience using financial statements to evaluate companies’ performance. The students were randomly divided into five groups that were all asked to estimate the fundamental value per share of a fictional retail company.

All the groups were given a press release providing data on the company’s income statements for the past three years; its balance sheet for the past two years; and its most recent annual revenue growth, same-store sales growth, earnings growth, and return on assets as compared to actual retail companies (such as Kohl’s and Sears) and industry averages. The participants were asked to respond to a series of statements assessing the company’s financial performance and then proceed to a matrix in which they entered forecasts for the next four years that were processed by software to arrive at an estimated fundamental value per share.

Where the groups diverged is in what came before this: in the first stage of the experiment they received widely different profiles of the fictional company’s CSR, its record with regard to labor, product quality and safety, and the environment. Two groups received page-long profiles describing the firm’s CSR as well below industry averages; two received profiles in which it was well above industry averages; and one pictured the company as a middling CSR performer.

A further key difference was that the groups receiving above-average and below-average profiles were each divided in two, with half being asked to respond to a series of statements assessing the firm’s CSR performance (the so-called explicit-assessment condition)  and the remainder (the no-explicit assessment condition) simply being directed to proceed to the financial press release.

The heart of the experiment was comparison of these two groups: those in the explicit-assessment condition, who were prompted to ponder the company's CSR record, and those who knew what that record was but were not prompted to assess it.

The researchers found that if the CSR profile was positive, there was a dramatic difference in the estimates of fundamental value between the participants in the explicit-assessment condition and those in the no-explicit-assessment condition, with the latter group estimating the fundamental value of the company to be, on average, $25.92 per share, compared to a $20.82 average estimate for those in the explicit-assessment condition.

There was a similar pattern of results when the professors told the participants to imagine they received a $10,000 inheritance and asked how much they would be willing to invest in the company. The researchers found that those in the no-explicit-assessment condition were likely to be willing to spend significantly more per share.

If the CSR profile was poor, the negative rankings had more impact on the no-explicit-assessment group, which made an average estimate of $19.14 a share compared to $21.12 average estimate for the explicit-assessment group.

When the participants in the experiment were asked to what extent CSR affected their fundamental-value estimates, the average awareness score was significantly lower for the no-explicit-assessment group, even though CSR clearly had a greater influence on their estimates.

In short, the relationship between CSR and their evaluations was largely unintended.

In accounting for these results, the researchers found that the study applies to the field of financial investment insights that have been developed over decades as part of what is called affect-as-information theory. This theory holds that when people are prompted to attribute a feeling to its source, it diminishes the feeling's effect on their judgments. Otherwise, it can subconsciously influence those judgments. In a classic experiment in this genre, telephone calls were placed on either a rainy or sunny spring day to random individuals, who were asked to answer questions on life satisfaction as part of a survey.

Participants who were called on rainy days reported themselves less satisfied with their lives than those called on sunny days, unless the caller indicated an interest in how the weather affects people's moods or even asked offhandedly how the weather was at the subject's locale. In either case, the mention of the weather diminished its emotional effect by prompting the participant to realize that the weather rather than true dissatisfaction with life was the likely source of their negative feelings.

Peecher and his fellow researchers believe a similar psychological mechanism is at work among the participants in their investment experiment. The feelings engendered by low- or high-performance CSR are attenuated by prompting the participants to think about that record rather than merely react to it emotionally.

“Evaluating CSR doesn’t change its emotional impact on investors, but it does enforce a measure of perspective and detachment on them,” Peecher said. “Whatever the CSR record, it is certainly not the whole story when it comes to fundamental value, and one wants to avoid being overly swayed by it one way or the other.”