B. Dong
Finance, School of Business and Economics In my previous blog ( Altruists going on an ego trip of pursuing wealth and fame ), we find that investors may become to be irrational when they take into account stocks’ ESG information. Toward high ESG stocks, they expect a higher return, perceive lower risk, and become less ambiguity-averse, which means they tend to believe that high ESG stocks have a higher sharp ratio than non-ESG stocks and positively invest in ESG. However, does it mean that the ESG label can induce less disagreement of stock performance between investors and therefore improve market efficiency?
Disagreement toward ESG
The efficient market hypothesis assumes financial securities are always priced correctly by investors, and hence there is no disagreement toward a security performance among investors in an efficient market. But, in reality, the financial market has never been an efficient one because of, for instance, irrationality and limits to arbitrage. When encountering ESG information, on the one hand, investors may interpret the relationship between ESG and financial performance differently, meaning they price the same security even far away from each other. Therefore, ESG maybe increase disagreements between investors and cause the market to be more inefficient; On the other hand, given ESG information, investors may also build a consensus on the relationship between ESG and financial performance because they hold a similar understanding of ESG. Therefore, ESG maybe also decrease disagreements between investors and improve market efficiency.It’s important to know the degree of disagreement toward ESG between investors. In the case of the ESG disagreement is less compared with non-ESG disagreement, it could be a signal of better market efficiency because investors have more consensus toward ESG. But this could also be a signal of a "Green Bubble" because investors may overall overestimate ESG performance. Oppositely, if the ESG disagreement is bigger than the non-ESG disagreement, there will be more opportunities for arbitraging between ESG and non-ESG securities.
Measurement of ESG disagreement
We easily see the differences in the average level of investors’ expectations toward ESG and non-ESG financial performance (investors expect a higher return toward ESG than non-ESG), but how large is the ESG disagreement we do not see? To answer this question, we need a "scoring rule" to measure the ESG disagreement, a way to weigh and combine individual expected ESG return into a single measure of the average expected ESG return. We use the method of least squares, proposed by Carl Friedrich Gauss, to score the contribution of individual ESG disagreement. The overall ESG disagreement, called mean squared disagreement (MSD), is the average of the squares of the individual ESG disagreement.With MSD, we can compare the degree of disagreement between ESG and non-ESG and tell whether ESG can improve market efficiency. To know investors’ return expectations toward ESG and non-ESG, We randomly assigned 320 subjects equally to two groups, the ESG group, which gives the ESG information about a high ESG fund, and the non-ESG group, which doesn’t provide any ESG information about the same fund. We showed both groups the past returns of the fund and asked them to predict the fund’s next year’s return. By this method, we obtain subjects’ return expectations toward ESG and non-ESG. We showed the same fund to the subjects, so the fundamental information of the fund is the same for subjects except for the ESG information. This setting allows us to observe exactly the ESG’s impact on market efficiency.
Experiment results
If the ESG group’s MSD is lower (higher) than the non-ESG group, there is less (more) disagreement between the subjects in the ESG group toward ESG return than the non-ESG group. As a result, we could not find a significant difference in disagreement regarding the fund return between the two groups. The ESG group’s MSD is 124.56, and the non-ESG group’s MSD is 118.36. Figure 1 shows the distribution of the individual subject’s disagreement. Therefore, based on our experiment, we find that ESG information doesn’t improve market efficiency. Meanwhile, even though, on average, subjects expected a higher return on the fund given the high ESG information, there is a significant disagreement among the subjects. Therefore, subjects may understand the ESG concept differently and hold heterogeneous beliefs toward ESG performance.We could not show the composition of ESG disagreement in the lab experiment. But, theoretically, we can assume the individual ESG disagreement comes from subjects’ biases toward ESG information and the noisy error. Subjects are irrational when given ESG information, and they may overestimate or underestimate ESG performance. The average of noisy errors is zero.
To summarize, we launched a lab experiment to elicit subjects’ expected returns toward a fund associated with/without ESG information. By comparing subjects’ mean squared disagreement of their return expectations of the fund, we find that ESG information doesn’t improve market efficiency.
