اثر بازگشت به میانگین سود دهی بانک بر عملکرد پس از ادغام در صنعت بانکداری

This research study examines the tendency for serial correlation in bank holding company profitability, finding significant evidence of reversion to the industry mean in profitability. The paper then considers the impact of mean reversion on the evaluation of post-merger performance of bank holding companies. The research concludes that when an adjustment is made for the mean reversion, post-merger results significantly exceed those of the industry in the first 5 years after the merger.

The banking industry has been undergoing an extensive period of restructuring as a result of technological innovations and regulatory changes. The number of banking mergers has accelerated in the last decade. A whole literature that analyzes the results of these mergers has developed; the general conclusion is that while mergers are good for the owners of the bank being acquired, the results for the acquirer are, at best mixed (Walter, 2004). Market reactions at the time of announcement tend to be either neutral or slightly negative. This result is counterintuitive. It seems unlikely that managers would continually make major mergers that are not clearly in the best interests of the shareholders. The shareholders would rebel and remove them from office, as has happened repeatedly in other industries when management fails to perform. Several studies have evaluated post-merger performance in the banking industry. These studies have generally indexed performance to some industry standard and then compared this indexed performance before and after the merger. The change in performance against the standard is ascribed as the effect of the merger. This approach does a good job of eliminating the effects of macroeconomic forces, regulatory changes, and other industry-wide phenomena. It assumes that absent a major event such as a merger, performance against the index is a random variable with a constant mean. Many, but not all academic research studies of banking mergers have found the average merger to be unsuccessful. This seems contrary to conventional wisdom as numerous mergers are continually being consummated. The implication is that something may have been omitted in the prior studies. The research paper of Fama and French (FF, 2000) suggests a solution to this conundrum. They find that profitability in most firms tends to revert to the mean in their industry. In most previous research on banking mergers no adjustment was made for mean reversion of post-merger bank performance as many of these studies were only event studies and the long term effects such as mean reversion were ignored. The omission of an adjustment for the mean reversion trend is a major part of the negative findings of prior post-merger studies. When we adjust for the mean reversion trend, we find significant improvements in merger related BHC ROE in 4 of the 5 post-merger years studied. The cash flow results are even stronger, with significant improvement versus the industry in all 5 years. One explanation of mean reversion is the basic economic argument that competitive forces, in the long-run, would bring about equality in the rates of return across the firms. The contribution of this paper lies in several areas. First, it extends the work of FF (2000) on mean reversion to cover BHCs that were expressly omitted because FF analyzed only non-regulated industrial firms. Second, it looks at post-merger performance results and compares them with the results of the acquirer in the year before the merger. Finally, and more importantly, this research demonstrates the importance of adjusting post-merger results for mean reversion trends when analyzing the effects of mergers.

This paper reaches three notable conclusions. First, the paper finds that similar to the results of FF (2000) there is mean reversion in corporate earnings over time for the BHCs. BHCs that outperform the industry in 1 year tend to move back toward the industry mean over time, as do those BHCs that under-perform the industry. The findings of significant mean reversion should not be a surprise, as the competitive nature of the industry makes it imperative for bankers to stay current with new developments. The second finding is that the trend toward mean reversion must be considered when examining changes over time to identify the impact of an unusual shock such as a merger. It is not sufficient to control for changes in industry performance over time and to say that everything else must be the result of the significant event. When there is a mean reversion trend in the data, ignoring the trend will misstate the impact of the merger. The post-merger impact appears stronger when measured against the acquirer’s results alone. In the banking industry, acquirers tend to be over-achievers, companies that have outperformed their peers on average. The mean reversion trend means that one would expect the acquirers’ industry-adjusted returns to decline in subsequent years and move toward the industry norms for a period of time. If this trend is ignored, all of its impact is ascribed to the result of the merger. In this study, correction of this error is the reason why significant improvements in IAROE related to the merger in 4 of the 5 years was found. It appears likely that this omission is a major part of the negative findings of prior post-merger studies. Third, and perhaps most important, this paper finds that when post-merger performance is measured correctly, BHC mergers add to profitability. The positive post-merger results found in this paper are consistent with the industry practice. There is extensive merger activity throughout the banking industry; merger activity has been taking place for a long time. It seems unlikely that banks would continue to acquire other banks if it were clearly not in the interests of the acquirer to do so. Furthermore, the tools for analysis of potential acquisitions are widely known as there are several textbooks on this topic. It would seem unlikely that all of these efforts would consistently produce bad decisions. Rather, it appears that the effects of mean reversion have masked the true positive impacts of the mergers.