An International Meeting of the American Accounting Association American Accounting Association 2006 Annual Meeting August 6–9, 2006 Washington, D.C.

Thomas D. Dowdell
North Dakota State University

Abstract: Investors use analyst forecasts in investing decisions and researchers use them in earnings-response-coefficient studies. Yet, prior studies suggest that analyst forecasts might be systematically biased, reducing their usefulness to both investors and researchers. Researchers have examined the relation between expected growth and forecast bias and found mixed evidence, in that forecast bias differs significantly between high-growth and low-growth firms when the P/B ratio is used to measure expected growth, but not when the P/E ratio is used in this capacity. In this paper, I use P/B and P/E ratios jointly to measure expected growth (following Fairfield, 1994; Penman, 1996; and Danielson and Dowdell, 2001) to more precisely investigate its relation with forecast errors. I observe that the relation between P/E and forecast errors depends on the P/B level, explaining the low overall relation between P/E and forecast errors found in previous studies.

Back to Session Listing

AAA Home Page