Industry Concentration, Analysts' Earnings Forecasts, and Bid-Ask Spread

Abstract

We examine the economic consequences of corporate disclosure using industry concentration as a proxy for disclosure. The use of this proxy allows us to circumvent problems inherent in such examinations carried out by prior studies using analysts' ratings as a proxy for disclosure. Specifically, analysts' ratings are based on analysts' perceptions and are therefore susceptible to a halo effect (Kerlinger [1986]), which may lead to a spurious association between analysts' ratings and the economic consequence variables. Furthermore, firm performance can simultaneously affect disclosure and analyst behavior/market liquidity, making it difficult to isolate the impact of disclosure on analyst behavior/market liquidity (Healy and Palepu [2001]). We document that firms in more concentrated industries have a smaller analyst following, higher dispersion in analysts' earnings forecasts, higher forecast errors, higher volatility in forecast revisions, and higher bid-ask spreads. These results help alleviate Healy and Palepu's [2001] concern that because of the limitations of the analysts' ratings measure, conclusions of prior studies on the economic consequences of disclosure may not be valid. Our results also contribute to the literature on industry concentration by suggesting that disclosure quality is not responsible for Hou and Robinson's [2003] finding that concentrated industries have low cost of equity capital. Finally, based on ours and Hou and Robinson's results we provide an explanation for the weak association between disclosure and the cost of equity capital that has been reported in recent accounting studies.