Conservative accounting and linear information valuation models

Purpose - To improve on previous linear information valuation models (LIMs) by adding conservative accounting methods to reduce bias. Design/methodology/approach - Assumes expected future residual income (RI) as a linear function of the RI in the previous period, and of other information (OI) not yet revealed in the accounts. Explains the downward bias of estimated intrinsic value relative to market value. Modifies the LIM to include a conservatism-correction term that includes the effects of positive RI and OI. Tests against a model with no conservatism-correction and one with no correction for OI . Takes a sample of 130,359 cases between 1951 and 1995. Findings - Finds that correcting for conservatism reduces the bias in estimates of value significantly, but only in the presence of OI. Without OI, shows that bias is even worse. Research limitations/implications - Notes no effect on the inaccuracy of estimates, and recommends including a variable based on analyst forecasts about future RI. Originality/value - Advances the development of an accurate and true valuation model that predicts RI.