
Based on his research, Mr. Zacks set up a company, Zacks Investment Research. The company gathers information on about 4,400 companies from 3,000 analysts at 150 brokerage firms. From the analysts reports Zacks develops a consensus Earnings Per Share (EPS) estimate. As earnings dates approach Zacks tracks which companies have rising EPS estimates, which are unchanged and which are falling. Zacks looks at four factors in the estimates to arrive at a ranking for each company. They look at agreement among the analysts, the magnitude of the change forecast, the upside or difference between the new forecast and the consensus and if the earnings surprise on the upside or downside.
With all this data and the above analysis Zacks ranks 5% of the companies as "Strong Buy", 15% as a "Buy", 60% as a "Hold", 15% as a "Sell" and 5% as a "Strong Sell." Buying a portfolio of Zacks #1 Rank stocks over the last 17 years since 1988 has generated an average annual return of 33% versus the S&P 500 return of 12% over the same period. Here are returns for the Zacks #1 Rank and the S&P 500 for the last five years
Zacks #1 Rank S&P 500
2000 14.31% -9.10%
2001 24.27% -11.88%
2002 1.22% -22.10%
2003 74.74% 28.69%
2004 28.79% 10.87%
I don't have the 2005 results yet, buy I am confident it will have outperformed again.
You can find the Zacks Rank at Zacks.com. Subscribers to their premium web site, ZacksAdvisor.com have access to daily updates through both the web and personalized daily email. They offer a free two week trial at http://at.zacks.com/?id=2089.







The only problem with using changes in earnings estimates is that the population, quality and other critical dimensions of the analyst community changes over time. If you'll notice, there are A LOT more analysts in a bull market than in a downturn. Additionally, with recent reforms, the traditional economic structure of equity research is changing drastically. Therefore, any research based on research in the 1990's should viewed skeptically as lot has changed and what used to be true may not be true anymore.
Posted by: TechTrader | January 26, 2006 8:44 PM | Permalink to Comment