Kaplan and Roll

The “Kaplan and Roll” study is a study in 1972, it was used to test the Mechanistic Hypothesis which to test the relationship between the stock market’s reaction and accounting changes (which changed the reported earnings but had no effect on cash flow) with respect to investment credit and depreciation. It was a 60-week study, both 30 weeks before and after the announcement. Use the changing method from accelerated to straight line method as an example, the result shows that the abnormal return was insignificant from 0 immediately after the announcement and a a decline in market price sharply, having a negative abnormal return from week 36 onwards.

From the results, we can see that the market does not react to the announcement as there should be a positive abnormal return if the market does react to the announcement but not a negative one as what we see in the results. This proves the Efficient Market Hypothesis since it believes that we cannot gain abnormal return as the market has already reacted to it. In addition, the negative abnormal return proves that the market reacts to all new, established pieces of information to the market but not to the financial statements solely.

However, the negative abnormal return might be only a result of the firms’ declining earnings. The firms may have been having trouble and choose to change the method to hide their decline earnings. As we can see, there is a slightly positive abnormal return between the announcement week and week 36. This can prove that the market reacts to the announcement positively (having positive abnormal return when increase in reported income) and highly even the firms are experiencing decline earnings. It may even have a higher abnormal return if the firms are not having difficulties.

As a conclusion, a statistically significant relationship between the accounting changes and market price changes were not found and there is still discrimination between the Efficient Market Hypothesis and the Mechanistic Hypothesis. It also cannot develop and test theories in just any one study.