New York, NY – Stock Sorting Ability – Each quarter, the team at Integrity Research Associates tabulates data on performance for the companies in the Investars equity research performance evaluation database (www.investars.com). The data is as of the end of December 31, 2006 and covers a full year of performance.
One of the more compelling metrics that has been looked at is the ability of a research firm to allocate stocks into the correct categories – BUY, SELL or HOLD.
The rationale for this metric is that if a research company is worth its salt its BUY recommendations should consistently outperform its HOLD recommendations, its HOLD recommendations should outperform its SELL recommendations and its BUY recommendations had better outperform its SELL recommendations.
Having said that, there are a few ways in which one could go about measuring this. At Integrity Research we take both a simplistic approach and a more complex approach. The simple approach is as follows:
BUYS > SELLS = 50 points
BUYS > HOLDS = 25
HOLDS > SELLS = 25
ELSE = 0
The index value is the sum of the individual index values.
While this metric is appealing in its simplicity when examining the attributes of a particular research company, it is not very interesting when comparing a number research companies against each other. The points system is valid but, like the batting average, has no way of accounting for the magnitude of the differences between the various portfolios.
For example, there were 30 companies that had perfect 100 scores. See below table:
|
Sorting Ability |
Number of Companies |
|
|
Score |
Q4 2006 |
|
|
100 |
30 |
|
|
75 |
37 |
|
|
50 |
0 |
|
|
25 |
27 |
|
|
0 |
26 |
|
In the interests of ranking companies in a more granular manner we have added a wrinkle to the sorting ability metric. Rather than a simple points system, we have created a weighted average sorting ability metric.
The weights to the above metric are the actual differences between, for example, the BUYS and the HOLDS of a particular research companies returns. Just to be clear this metric has nothing to do with return, it is simply a raw score. The higher the score the better the job the research company did in sorting stocks into the appropriate buckets.
Below we tabulate the results of our analysis:
10 of the Top IRPs and IBs for the year ended December 31, 2006
Stock Sorting Score
Over 500 Stocks Covered
| Rank | Research Company | IRP / IB |
| 1 | When2Trade | IRP – Quant |
| 2 | Jefferies | IB – Fund |
| 3 | Market Profile Theorems (MPT) | IRP – Quant |
| 4 | Raymond James | IB – Fund |
| 5 | Standpoint Research | IRP – Quant |
| 6 | RBC Capital Markets | IB – Fund |
| 7 | Columbine Capital | IRP – Quant |
| 8 | Sabrient Systems | IRP – Quant |
| 9 | Zacks Investment Research | IRP – Quant |
| 10 | Merrill Lynch | IB – Fund |
10 of the Top IRPs and IBs for the year ended June 30, 2006
Stock Sorting Score
Under 500 Stocks Covered
| Rank | Research Company | IRP / IB |
| 1 | Punk & Ziegel | IB – Fund |
| 2 | Greenwich Investment Research | IRP – Fund |
| 3 | Miller Johnson Steichen Kinnard | IB – Fund |
| 4 | American Microcap Inst | IRP – Fund |
| 5 | Taglich Brothers | IB – Fund |
| 6 | Beonning & Scattergood | IB – Fund |
| 7 | Morgan Joseph | IB – Fund |
| 8 | Ladenburg Thalmann & Co. Inc. | IB – Fund |
| 9 | Emerging Growth Equities | IRP – Fund |
| 10 | Dougherty & Co | IB – Fund |
The over 500 coverage firms are dominated by the quantitative IRPs, With When2 Trade taking top honors. Independent Quant shops took 6 of the top ten spots in the over 500 stocks covered category.
The under 500 stocks covered were dominated by sell-side research firms using a fundamental approach to analyzing shares. Fully 7 of the top ten were sell side shops.
Note: The data used in the above article, comes from the Investars web site. The Analysis and conclusion of this article are those of Integrity Research Associates. Investars has re-engineered its product offering, which now includes a full suite of services to the buy- and sell-side to track research performance, manage research and to integrate this into the broker vote system.