Stock Sorting – Yes Virginia, Holds Matter


New York – In our final quarterly performance offering, we tackle the issue of stock sorting ability. This performance metric measures the research firms’ ability to get the stocks it covers into the right buckets–Buy, Sell or Hold. The data for this study comes from the Investars database, while the conclusions and additional calculations are those of Integrity Research Associates.

In previous incarnations, we utilized the Buy, Sell and Hold return data to determine a points system index. This index gave more points to a firm’s ability to distinguish buys from sells and less points if the firm was able to distinguish a buy from a hold or a hold from a sell. Our new approach is simply and index based on relative ranking of the firm in each of these categories. For example,  a firm that had the largest spread between its buys and sells would be awarded a rank of 1. This process is repeated until each firm has a rank in the buy-sell, buy-hold and hold-sell categories. These ranks are simply added together and then the resultant points distribution is turned into an index of sorting ability.

The Table below details the top ten firms in sorting ability from the Investars universe of research providers.

Integrity Research Associates Sorting Ability Index
Sourced from Investars Data

Rank Company Sorting Ability Index
1 Taglich Brothers 100.0
2 Hilliard Lyons 97.2
3 Fundamental Research Corp 96.6
4 Rodman & Renshaw 96.4
5 Wedbush Morgan 95.8
6 American Microcap Inst. 88.8
7 Kintisheff Research 88.5
8 Cohen Brothers & Co 87.7
9 Cantor Fitzgerald 85.4
10 Merrill Lynch 84.3

In sorting ability, it would seem to be easier to sort stocks from a smaller list of stocks than from a larger list, since the conviction associated with each stock could be far less for a larger coverage list. Indeed, 6 of the 10 ten firms have coverage of less than 100 stocks, three have coverage of between 100 and 500 stocks and only one has coverage of greater than 500 stocks.

Given the preponderance of low coverage firms in the top ten, if one wants specific stock recommendations, one may be better off to seek the advice of a low coverage firm that covers this stock. However, the question is whether the firms with small coverage lists are simply more volatile (less reliable) over time, or whether certain firms are consistently at the top of the leader board. We already know from cross section data (i.e. data across all firms for one time period) that the small coverage firms tend to populate the top and the bottom of the performance scale. But this does not reflect any consistency, or lack thereof, in performance.

As a test of inter-temporal consistency, we assess the performance of the 6 low coverage RPs over time. To do this we are using one year returns of the BUY – SELL recommendations for the past 4 quarters.

Performance of the BUY minus the SELL Recommendations Over Time
Basis points per Day

Company Q306 Q406 Q107 Q207 Avg
1 Taglich 22.84 28.12 58.12 24.27 33.84
2 Fundamental
Research Corp
2.91 5.11 51.32 14.62 18.49
3 American Micro 0 30.3 5.05 24.23 14.85
4 Rodman & Renshaw 3.32 11.28 2.5 16.97 8.51
5 Cantor NA NA -1.06 11.05 5.0
6 Kintisheff 2.34 4.38 -051 12.41 4.65

The surprising answer as to whether the best low coverage research firms exhibit any consistency over time, is that they do. All of the top ten firms in sorting ability in the most recent study had solid performances along with reasonable volatility of returns over the 4 quarters reviewed. In the top spot was Taglich Brothers with an average return (BUY – SELL) of 33.84 basis points per day over the for time periods studied.

To be fair, the time periods, defined as the year ended at that particular quarter end, have significant overlap in the time series data considered, which may tend to overstate consistency. Still the results demonstrate that the good firms ted to consistently perform well over time, rather than oscillate from the top of the performance table to the bottom as previously posited.

Note: The data or this study was from the Investars Web Site. The analysis, additional calculations and interpretation is that of Integrity Research Associates alone.


About Author

Leave A Reply