Conscious
Investor: Wizard
Investing made simple with a 3-step expert Wizard
USA, Canada and Australia: Over 10,000 stocks examined every day.
| Put together a portfolio of companies whose aggregate
earnings march upwards over the years, and so will the portfolio’s
market value. Earnings, earnings, earnings ... ultimately the earnings
will decide the fate of a stock. Earnings are probably the most closely followed aspect
of a firm's income statement as an indicator of profitability. The unique measurement of stability as developed by
Dr. John Price allows investors to draw considered conclusions about
the management of a company to sustain consistent financial performance
without having to resort to estimates or guesswork. |
For the technically minded
David Dreman manages over $10 billion through Dreman Value Management. After extensive research, he and Michael A. Berry concluded that the errors in analysts' earnings forecasts are too large to be useful.[1] Many other researchers have reached the same conclusion.[2]
The punch line is that by filtering companies using staegr extensive back tests show that we cut the average error size by a factor of 5 or more. One study looking at all stocks listed on the NYSE, the NASDAQ and the AMEX showed that errors were as low as 15.36% compared to errors of 90% and higher shown in various studies.
It is true that analysts have a difficult job with widely varying results. They have to come up with forecasts for the earnings that even the senior management of the companies would have a hard job making. The following is a brief explanation of how the accuracy of analyst forecasts can be measured and how the staegr approach can squash these errors.
Many websites display consensus forecasts. These are the average of the forecasts of the analysts following the company. As investors we want to know if we can rely on them as a reasonable forecast of earnings.
If the actual earnings per share (eps) exceed the consensus forecast, it is called a positive earnings surprise. In the opposite direction, if the actual eps is below the consensus forecast, it is called a negative earnings surprise.
The absolute value of one of these surprises divided by the forecasts of the earnings is a common way of measuring accuracy. It is referred to as SURPF. For example, if the actual eps is $3.00 and the forecast of the eps is $4.00, then SURPF = |(4-3)/4| = 25.00%.
If SURPF is calculated for many companies, the mean of the results is referred to as the mean relative absolute error. Lawrence Brown calculated that mean relative absolute error over the years from 4Q 1983 (4Q) to 1996 (2Q) was 91.60%.
At
Conscious Investor we looked at all stocks traded on the NYSE, the NASDAQ
and AMEX. Then we discarded those that did not have at least 10 years
of data. Next we ranked them from the highest staegr to the lowest using
the last 6 years of data. These stocks were then divided into 10 deciles
or groups of equal size from the highest staegr to the lowest staegr.
Earnings forecasts were made by using the historical growth rate calculated using hgrowth. The mean relative absolute error was calculated for each of the deciles.[3] The error for the decile with the highest staegr was only 15.36%.
Conclusion By screening on stability using staegr, this study and many others show that consensus forecasts errors can be reduced by a factor of 5 or more. By concentrating on these companies, investors can make earnings forecasts with heightened confidence.
Footnotes
- David N, Dreman and Michael A. Berry, "Analyst Forecasting Errors and Their Implications for Security Analysis," Financial Analysts Journal, vol. 51, no. 3 (May/June 1995):30- 41.
- Lawrence D. Brown, "Analyst Forecasting Errors: Additional Evidence", Financial Analysts Journal, November/December 1997: 81-88.
- The forecasts were made using "eforecast", a proprietary tool for making forecasts based on historical growth.
Trademarks The following trademarks are licensed for use on this site. Conscious Investor® is a registered trademark in the US and Australia. Staegr™® and hgrowth™® are trademarks in the US and registered trademarks in Australia. Eforecast™ is a trademark. You may only use these marks with proper attribution.You may only use these marks with proper attribution.
Staegr, hgrowth and eforecast were developed and coded by Dr. John Price.