New York – For a number of years Integrity Research has been analyzing the performance of research providers which are tracked by the Investars performance measurement system. The system is designed to enable investors to review the track records of various research providers that send their recommendations to Investars.
Typically, Integrity uses all of the research providers tracked regardless of the number of stocks they cover. This is to ensure that the data set is as large as possible so that summary statistics are reliable. Investars, on the other hand, uses three coverage size categories to track the performance of the different research providers: 1) those that cover less than 100 stocks; 2) those that cover from 100 to 499 stocks; and 3) those that cover more than 500 stocks. In this blog we assess the performance variation of the different coverage groupings.
Below is presented the 1-year BUY returns for the Investars companies presented in the aforementioned categories. It is clear that the firms with the lowest coverage do indeed have different return and volatility characteristics.
Source: Investars Performance Measurement system
Over the year studied the medium coverage companies had the best average return, followed by the high coverage firms and then the low coverage firms. This result is interesting because it tends to be obscured by both the Investars and the Integrity approaches. Notice, for example, that the low coverage firms had the highest total return at over 100%, meaning that they will always be at the top of the leader board if all firms are grouped together. Notice also that they will be at the bottom of the leader board since they have the lowest returns as well.
The above chart depicts the buy performance distributions of the three coverage classifications and demonstrates the differences in the data sets. But the story does not stop there. A real assessment as to whether the three groups are the same in terms of average return is a simple statistical test. This test indicates that, although there seems to be a consistent difference between the two sets, there is no statistical basis by which to infer that the three groups exhibit different returns (95% level of confidence). This analysis should be conducted at regular intervals to see that there is no change in this conlusion as new data comes to light.
Subsequent blogs will look into the types of firms that live in each of these coverage size camps and assess whether there are performance differences within the methodological approaches or not.