Our experiments investigate the extent to which traders learn from the price, differentiating between situations where orders are submitted before versus after the price has realized. In simultaneous markets with bids that are conditional on the price, traders neglect the information conveyed by the hypothetical value of the price. In sequential markets where the price is known prior to the bid submission, traders react to price to an extent that is roughly consistent with the benchmark theory. The difference’s robustness to a number of variations provides sights about the drivers of this effect
We provide both an *axiomatic* and a *neuropsychological* characterization of the dependence of choice probabilities on deadlines in the softmax form, with time-independent utility function and time-dependent accuracy parameter.
The softmax model (also known as Multinomial Logit Model or Power Luce Model) is the most widely used model of preference discovery in all fields of decision making, from Quantal Response Equilibria to Discrete Choice Analysis, from Psychophysics and Neuroscience to Combinatorial Optimization. Our axiomatic characterization of softmax permits to empirically test its descriptive validity and to better understand its conceptual underpinnings as a theory of agents rationality. Our neuropsychological foundation provides a computational model that may explain softmax emergence in human multialternative choice behavior and that naturally extends the dominant Diffusion Model paradigm of binary choice.
Daniel Martin is an Assistant Professor in the Managerial Economics and Decision Sciences (MEDS) department at Northwestern University’s Kellogg School of Management. He is a behavioral and experimental economist who studies the processing and disclosure of information. For example, he investigates why firms do not voluntarily and clearly disclose information about product quality and why consumers do not pay full attention to information about prices or product quality.
Framing effects are often attributed to misperceptions. In this study, however, we document a large and robust framing effect that is not reflective of misperceptions. Our framing effect persists when agents gain experience, pay attention, and are provided with information that prevents miscalculations. We propose and provide evidence as to why our framing effect persists: the majority is driven by self-serving motives. Our results suggest that framing effects, as well as other behavioral biases driven by self-serving motives, may be notably robust to de-biasing conditions.