Published online by Cambridge University Press: 06 April 2009
This paper assesses the importance of the role ofprices as aggregators of private information in theS&P 500 future market. We estimate primitiveparameters of the Hellwig (1980) noisy rationalexpectations model, when both prices and terminalvalues are observable. The variance-covarianceparameters governing futuers prices and terminalvalues can be inverted to obtain estimate of theprimitive parmeters, including the precision ofprivate infromation and the variance ofliquidity-motivated trades. We also estimatecoefficients in the linear price conjecture, weightsthat agents place on different sources ofinformation, and the informativeness of prices. Wefind that the variance of the error term in agents'private signals is several orders of magnitudelarger than the variance of liquidity-motivatedtrades. But, in a large market, prices are still soinformative that the market as a whole appears toweight them more than prior beliefs.
Dupree College of Management, Georgia Institue ofTechnology, Atlanta, GA 30332 and RutgersUniversity, 96 New England Avenue, #18, Summit, NJ07901, respectively. This paper has benefited formearlier work by Krishnan with Uday Chandra. Theauthors are grateful for the comments,suggestions, criticism, and encouragement of CelalAksu, Bruno Biais, Jordi Caballé, Rick Green, MarkGrinblatt, Guy Laroque, Fallaw Sowell, andparticipants at the 1996 European summer Symposiumin Financial Market at Gerzenzee, Swizerland,where a previous version of this paper waspresented. This study began when Krishnan was atthe University of Minnesota, and the computer workhas benefited in part from facilites provided bythe Minnesota supercomputing Institue. The authorsalso thank Peter Bossaerts (a referee) and GuofuZhou (associate editor and a referee).