some studies have pointed out that regression models used to compare different accounting standards (e.g., before and after IAS/IFRS adoption) may be mis-specified because the relationship between prices and accounting measures is not linear. Ashbaugh and Olsson (2002) provide consistent evidence by showing that the violation of clean surplus accounting makes regressions based on the Ohlson model (1995) mis-specified. Clarkson et al. (2011) also document increased nonlinearity in the relationship between share prices and accounting data subsequent to IFRS adoption, which alters statistical inference based on a traditional linear pricing model.