Realized or Still on the Paper? Fair Value Accounting and the Realization Effect

• Applicant: Jan Müller-Dethard
• Subject Area: Accounting and Finance
• Term: from 2021 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 458150472

In this research proposal, I aim to contribute to the ongoing discussion in economics on what influences the dynamics of risk-taking. There is broad consensus that gains and losses which an agent has experienced over the past affect subsequent risk-taking. In this vein, a large body of literature has analyzed how prior outcomes affect subsequent risk-taking. However, the existing evidence on the direction of this influence, i.e. whether prior gains/losses lead to more or less risk-taking, is seemingly less clear and somehow inconsistent even after decades of research. Some studies find that individuals engage in more risk-taking following a loss, while others find less risk-taking after a prior loss. Similar inconclusive findings are found for gains.Recently, a very exciting new development in this field has received substantial attention, namely the distinction between realized and unrealized outcomes (Imas, 2016). Using the framework of realization, Imas (2016) shows that the puzzling findings of differential risk-taking after one and the same loss can be reconciled. Merkle, Müller-Dethard and Weber (2020) extend this framework.A field which is particularly concerned about the distinction between realized and unrealized outcomes is accounting (e.g. the realization principle). With this research proposal, I aim to apply recent findings in behavioral economics to the field of accounting. In particular, I test whether the realization effect from economics (larger risk-taking after a paper loss than after a similar realized loss) applies to the context of different accounting standards. I argue that the way changes in the value of assets are measured and disclosed under fair value versus historical cost may have a direct impact on how managers frame investment episodes, how they evaluate risky prospects, and consequently how much risk they take.Using an experiment, I want to analyze how the dynamics of managerial risk-taking change conditional on how a loss is measured and disclosed in the financial statements. While fair value accounting measures the asset at the price that the entity would realize if it sold the asset at the measurement date (similar to a realized loss in case of a decrease in value), historical cost accounting measures the asset at the initial purchase price (consistent with the idea that outcomes are in fact unrealized). Analogous to the theoretical predictions of the realization effect, I expect to find lower risk-taking after an unrealized loss under fair value accounting than after the same unrealized loss under historical cost accounting.My work contributes to two strings of literature. First, I expect to extend recent work on the realization effect in the sense that framing can lead to similar patterns in risk-taking as compared to “real” realization. Second, I believe that this work contributes to the literature on how accounting measures can prevent managers from loss-chasing behavior.

• DFG Programme: WBP Fellowship
• International Connection: USA
• Participating Institution: The University of Chicago ; Booth School of Business
• Host: Professor Dr. Alex Imas