The results of the empirical analysis were quite surprising, considering the findings of previous researches. In most part, the results were relatively modest and not statistically significant. Therefore, it is determined that none of the four hypotheses received sufficient support; Thus, it is stated that the zero hypothesis is failed to reject in all of the four hypotheses. This result is somewhat inconsistent with the previous research, which has documented either significant negative or positive response from the stock market to CSR performance-related information. The hypotheses were constructed to be extensions from the findings of previous researches and to cover similar areas yet focus on areas that had received little attention from the prior studies. However, the method and the samples used did not provide significant support for the hypotheses.
The first hypothesis, which was tested in the univariate analysis with the market model, suggested that the companies appearing on the Global 100 -list would receive positive abnormal returns for their stocks. However, though there were some disturbances in the event day AARs, especially in the USA-model event day AARs which showed weak statistical significance, but which was not supported by the generalized rank test, no material coherent reaction was received from the stock market. Therefore, the results were partly in line with previous research indicating that different markets in different geographies might react differently to information related to CSR performance (e.g. Auer
& Schuhmacher, 2016). There are two potential main avenues which could explain the results of this study in the context of the first hypothesis:
The first avenue would be that the markets do react to CSR performance information by corporate Knights, but the event windows of this study did not capture the reaction.
However, it was considered to be quite unlikely to happen, since it would mean that there has been significant information leakage prior to the event date or that the stock market would have reacted only after two days from the event day. The latter would be highly
controversial to efficient market hypothesis and to the notion of McWilliams and Siegel (1997, 636), who discussed that there is evidence of the market reacting to new information in 15 and 90 minutes.
The second avenue would be that markets will not react to CSR performance information by Corporate Knights. This would be because markets are not interested in CSR performance information in general, or they are not specifically interested in Corporate Knights’ information. The first claim would be contrarian to previous research since lots of previous research have found that CSR performance information has had value implications, either negative or positive (e.g. Halme & Niskanen, 2001; Kaspereit &
Lopatta, 2016; Klassen & McLaughlin, 1996; Krüger 2015), and since Amato and Amato (2012) and Yadav, et al. (2016) illustrated the market reacting to environmental performance information in highly similar settings. Additionally, since Orlitzky, et al.
(2003, 415) illustrated that environmental performance has a smaller relationship with financial performance than the other dimensions of CSR, markets should be increasingly interested in CSR performance information compared to just information considering environmental performance. However, the findings of Auer and Schuhmacher (2016) could back the argument that markets are not interested in CSR performance information in general since they found that CSR-criteria based investment generated no extra returns.
Nevertheless, the suggestion, that the markets are not specifically interested in Corporate Knights’ information could be the most potential one. It can be because markets see that specifically, the CSR view of Corporate Knights has no value implications or that the CSR view of Corporate Knights does not generate any extra information for the markets.
It could be controversial that markets would consider the Corporate Knights’ notion of CSR performance as irrelevant for investor value since CSR defined in a similar context affected positively to investor value (e.g. Ameer & Othman, 2012; Eccles, et al., 2014;
McWilliams & Siegel, 2011). Therefore, markets would not share the views of the prior studies by not reacting to the information. However, the disregard could be also because the markets see that the Global 100 -list provides little additional information considering the CSR performance levels of companies. This concept will be discussed more thoroughly with the third hypothesis.
The second hypothesis in which was assumed that the company’s relative position in the list would affect the volume of abnormal returns. This hypothesis was tested in the
cross-sectional analysis. On the contrary to what Amato and Amato (2012) and Yadav, et al.
(2016) showed, the relative position did not affect to abnormal returns. Additionally, this is somewhat contrary to the notion of Aouadi and Marsat (2018) who showed that CSR performance information has only value implications for highly visible companies since the top companies could be argued to have more visibility. Additionally, this notion was controlled by including a variable measuring company size in the regression, which was not statistically significant.
The third hypothesis was considering if the Global 100 can provide additional information to markets. The hypotheses were continuing the notion of prior studies (e.g. Klassen &
Mclaughlin, 1996; Yadav, et al., 2016) which linked higher value implications to events where the markets received information for the first time about a company’s level of CSR performance. Therefore, it was assumed that the companies, which had not any CSR performance information in Thomson Reuters Eikon -database and were not previously ranked by Corporate Knights would receive higher abnormal returns. However, this hypothesis was not supported by the cross-sectional analysis, which showed inconsistent and statistically insignificant results. The assumption that a single database would describe the CSR performance information available for the markets is rather naïve, and therefore it is possible that it could not provide a true illustration of what can be considered as new information for the markets. Conversely, if the information was new it could then suggest that the markets did not consider the information provided by Corporate Knights as important.
However, according to the efficient market hypothesis, even according to semi-strong terms, if the information is publicly available it is already valued in the stock prices. And since most of the information Corporate Knights use is publicly available, the part of the information which is public should be already valued to stock prices. The efficient market hypothesis is often counterargued by suggesting that with some information the costs of acquiring it will surpass the benefits of trading based on it (Grossman & Stiglitz, 1980).
It can be argued that such a phenomenon did not happen with the CSR performance information made easily accessible by Corporate Knights. Therefore, the fourth hypothesis, which was an extension of the third hypothesis, was showing similar evidence, which supports the notion of efficient markets, since in the cross-sectional test the results were also inconsistent with the “newcomer” coefficient. Thus, the null
hypothesis was failed to reject with the fourth hypothesis, which was similarly the case for all the hypotheses.
Since all the null hypotheses were not rejected it suggests that the CSR performance disclosure by Corporate Knights does not have value implications. However, Corporate Knights can still have additional effects for the markets and to the society by publishing the Global 100 -list. First, CSR reporting activity has increased significantly throughout the years, and though superior CSR reporting performance does not have any value implications either, parties such as Corporate Knights can be able to steer the discourse of CSR performance towards more objective direction, since the CSR reporting purposes have largely stayed the same (Cho, Michelon, Patten & Roberts, 2015). Second, Corporate Knights can provide robustness checks to the information other CSR performance raters provide, which can be beneficial for the construct validity of all the measures, as Semenova and Hassel (2015, 250) argue. Finally, however, as Chelli and Gendron (2013) point out that idealizing the quantification of everything can have its risks. What certainly is not optimal for the entire ecosystem is that companies seek to better in the metrics of the rankings and not in CSR performance itself.