Does Accounting Measurement Influence Market Efficiency: A Laboratory Market Perspective

Abstract

Using laboratory markets where accounting regimes can be directly compared with equivalent economic parameters, we test whether and how two different accounting measurement bases – historical cost and mark-to-market – influence trader perceptions and asset mispricing. Across three experiments, our results show that traders perceive otherwise equivalent assets differently by regime. In the mark-to-market regime traders perceive stronger links between performance and market price changes, and weaker links between performance and asset fundamentals. We also observe that traders in the mark-to-market regime prefer information about future market prices but traders in the historical cost regime prefer information about future dividends. These perceptions correspond with greater market-level mispricing/bubbles in the mark-to-market regime. Our results suggest that accounting regimes can, on their own, contribute to price bubbles and their subsequent collapse.