Advisors who select individual securities use several sources of information when evaluating stocks. The forecasts of a company's earnings per share made by analysts and company management are the most commonly quoted predictions of earnings. However, an alternative measure has emerged over the last decade-whisper forecasts, which, while relatively new, appear to be more accurate. These forecasts, unpublished estimates from individual investors of quarterly earnings for publicly traded companies, can be used to discern future stock price movements, and thus offer new opportunities for financial advisors to add value to their client investment portfolios. It is also possible to earn above-market returns using a trading strategy based on the difference between whisper forecasts and analysts' estimates.
Over the last 10 years, these unpublished earnings estimates have accumulated on several Web sites, such as http://whispernumbers.com, where investors are able to access these forecasts and provide their own input. John Scheer, president of http://whispernumber.com, notes on the site, "Investors no longer need to be tied solely to the consensus estimate and can be confident using our data to help make better investing decisions."
We collected data on a random sample of 254 Standard & Poors 500 index companies and 251 non-S&P 500 firms that have whisper forecasts on Scheer's site, beginning in the second quarter of 2002 and ending in the second quarter of 2007. Figure 1 presents the frequency of whisper forecasts for both categories. On average, each S&P 500 firm had 11 whisper forecasts during this 20-quarter period. In contrast, each non-S&P 500 firm had only four.
The whisper forecasts were available for a wide variety of companies both inside and outside the S&P index, and we classified these by standard industry classification codes. Manufacturing had the largest industry representation in both categories. The finance industry was the second-largest group for the S&P 500 firms, while the service area was the second-largest group for the non-S&P 500 firms.
We assessed the whisper forecasts' accuracy by calculating the difference between the actual EPS and either the whisper forecast or the analyst prediction. Figure 2 shows the median forecast errors by industry for both S&P 500 and non-S&P 500 firms. Overall, whisper forecasts were more accurate than the analysts' forecasts for the full sample and for all industries except the non-S&P 500 mineral and construction sector. Interestingly, for non-S&P 500 companies, the median whisper forecast was perfectly accurate for three classifications: for manufacturing; for the finance, insurance and real estate area; and for public administration.
Given that whisper forecasts seem to be more accurate than analysts' forecasts, advisors could potentially detect the direction of future stock price movements using the difference between the two different forecasts. For the S&P companies, active management aficionados could examine a strategy we have tested, which involves buying firms with whisper forecasts greater than analysts' forecasts and selling short those firms whose whisper forecasts were less. This strategy, for fourth-quarter earnings, resulted in a small, positive risk-adjusted return for the six-day window surrounding the earnings announcement date, whereas the overall market risk-adjusted return was negative. We found that the trading strategy also earned returns above the market for the following six industries, regardless of the quarter of the earnings announcement: building construction; lumber and wood products; fabricated metals; retail; hotels; and business services.
Once advisors know which firms have whisper forecasts, these appear to be a useful alternative source of information that will allow them to examine earnings and future stock price movements.
Susan Machuga is an assistant professor of accounting at the University of Hartford. Karen Teitel is an assistant professor of accounting at the College of the Holy Cross. Ray Pfeiffer Jr. is a professor and the director of accounting and information systems at the University of Massachusetts. Brian Boyer is an MBA student at the University of Hartford.
