Periodic Reporting for period 4 - INFORMATIVEPRICES (Market Selection, Frictions, and the Information Content of Prices)
Période du rapport: 2020-11-01 au 2021-04-30
An important reason to trust markets is the belief that market prices accurately summarize a vast array of information. Whether prices efficiently aggregate information is a central economic question. Informative prices require an efficient price formation mechanism as well as informed market participants. In this project, we explore which market configurations result in prices that aggregate information dispersed among market participants, and we analyze the methods by which economic agents acquire information. In particular, we study (1) The information content of prices when imperfectly informed bidders choose between multiple common-value auctions; (2) The pattern of market selection by informed and uninformed bidders across multiple auctions; (3) Information aggregation in auctions where bidders acquire information before bidding; (4) The effectiveness of acquiring information through communication.
Our work on common-value auctions demonstrates how frictions in one auction can hinder information aggregation in another frictionless auction, even in circumstances where information is aggregated in both markets under complete information. In particular, we show that frictionless markets attract uninformed bidders while frictional markets attract informed bidders, and this pattern of market selection precludes information aggregation in both markets.
Our work on costly information acquisition in auctions establishes that the information content of the auction price is completely characterized by the cost-benefit ratio of information as the amount of information acquired by an agent converges to zero. Moreover, we show that information is aggregated in the auction if this cost-benefit ratio converges to zero.
In addition to investing in information, an economic decision-maker can acquire information by consulting informed experts. However, conflicts of interest between experts and the decision-maker frequently hinder effective communication. Our work on communication shows that a decision-maker can elicit truthful information from an expert by gradually increasing the stakes in the relationship. In other words, starting small and progressively increasing the stakes facilitates truthful communication between an expert and a decision-maker. Furthermore, if the decision-maker has multiple experts to consult, our results show that she can again obtain truthful information from the experts. This is because the decision-maker can check the veracity of her advice by comparing one expert’s recommendation with another’s.
Our work on common-value auctions demonstrates how frictions in one auction can hinder information aggregation in another frictionless auction, even in circumstances where information is aggregated in both markets under complete information. In particular, we show that frictionless markets attract uninformed bidders while frictional markets attract informed bidders, and this pattern of market selection precludes information aggregation in both markets.
Our work on costly information acquisition in auctions establishes that the information content of the auction price is completely characterized by the cost-benefit ratio of information as the amount of information acquired by an agent converges to zero. Moreover, we show that information is aggregated in the auction if this cost-benefit ratio converges to zero.
In addition to investing in information, an economic decision-maker can acquire information by consulting informed experts. However, conflicts of interest between experts and the decision-maker frequently hinder effective communication. Our work on communication shows that a decision-maker can elicit truthful information from an expert by gradually increasing the stakes in the relationship. In other words, starting small and progressively increasing the stakes facilitates truthful communication between an expert and a decision-maker. Furthermore, if the decision-maker has multiple experts to consult, our results show that she can again obtain truthful information from the experts. This is because the decision-maker can check the veracity of her advice by comparing one expert’s recommendation with another’s.
This project developed a game-theoretic model to study information aggregation in an economy where bidders choose between a frictionless common-value auction and an outside option based on their private information. The outside option is also an auction market subject to frictions resulting from a positive reserve price. (“Market Selection and the Information Content of Prices,” joint with Mehmet Ekmekci, Econometrica, forthcoming.) This framework allowed us to highlight the interplay between self-selection into an auction, bidding behavior, and the information content of prices. Our main finding shows no equilibrium aggregates information in either market if the object-to-bidder ratio in the frictionless market exceeds a certain cutoff. If, on the other hand, the object-to-bidder ratio is less than the cutoff, then we show that information is aggregated in every equilibrium. Furthermore, if the market is comprised only of perfectly informed and uninformed bidders, then only very few informed bidders select the frictional market. The pattern of market selection into the two markets ensures that the expected prices are equalized across markets and precludes information aggregation.
Market participants often need to invest in information before deciding whether to trade. To investigate the incentive to invest in information, we studied a common-value auction in which a large number of identical, indivisible objects are sold to a large number of ex-ante identical bidders with unit demand. Before the auction, potential bidders can choose to buy information from various sources at a cost. (“Price Discovery in a Large Auction with Costly Information” with Mehmet Ekmekci.) This framework allowed us to derive the limit price distribution as a closed-form function of the costs and benefits of information. This price distribution implies that information is aggregated if the cost of information converges to zero faster than the benefit of information as the amount of information acquired converges to zero.
An economic agent can also acquire information by communicating with an expert. As part of this project, we studied communication between an informed expert and an uninformed decision-maker, whose incentives are not aligned, by constructing a repeated-cheap talk game. ( “Starting Small to Communicate,” with Levent Kockesen and Elif Kubilay, Games and Economic Behavior, 2020). Either the decision-maker or the expert controls the stakes involved in their relationship. In both cases, if the potential conflict of interest is large, then the stakes increase over time in a truthful equilibrium. Moreover, as the potential conflict of interest increases, the extent to which the stakes are back-loaded increases, i.e. stakes are initially smaller but grow faster.
This project also studied communication between a decision-maker and multiple agents by constructing a cheap-talk game where two experts first choose what information to acquire and then offer advice to a decision-maker whose actions affect the welfare of all. ("Cross-verification and Persuasive Cheap Talk" with Mehmet Ekmekci and Ludovic Renou.) In our framework, the experts cannot commit to reporting strategies. Yet, we show that the decision-maker’s ability to cross-verify the experts’ advice acts as a commitment device for the experts. In particular, we construct an equilibrium, where an expert’s equilibrium payoff is equal to what he would obtain if he could commit to truthfully revealing his information.
Market participants often need to invest in information before deciding whether to trade. To investigate the incentive to invest in information, we studied a common-value auction in which a large number of identical, indivisible objects are sold to a large number of ex-ante identical bidders with unit demand. Before the auction, potential bidders can choose to buy information from various sources at a cost. (“Price Discovery in a Large Auction with Costly Information” with Mehmet Ekmekci.) This framework allowed us to derive the limit price distribution as a closed-form function of the costs and benefits of information. This price distribution implies that information is aggregated if the cost of information converges to zero faster than the benefit of information as the amount of information acquired converges to zero.
An economic agent can also acquire information by communicating with an expert. As part of this project, we studied communication between an informed expert and an uninformed decision-maker, whose incentives are not aligned, by constructing a repeated-cheap talk game. ( “Starting Small to Communicate,” with Levent Kockesen and Elif Kubilay, Games and Economic Behavior, 2020). Either the decision-maker or the expert controls the stakes involved in their relationship. In both cases, if the potential conflict of interest is large, then the stakes increase over time in a truthful equilibrium. Moreover, as the potential conflict of interest increases, the extent to which the stakes are back-loaded increases, i.e. stakes are initially smaller but grow faster.
This project also studied communication between a decision-maker and multiple agents by constructing a cheap-talk game where two experts first choose what information to acquire and then offer advice to a decision-maker whose actions affect the welfare of all. ("Cross-verification and Persuasive Cheap Talk" with Mehmet Ekmekci and Ludovic Renou.) In our framework, the experts cannot commit to reporting strategies. Yet, we show that the decision-maker’s ability to cross-verify the experts’ advice acts as a commitment device for the experts. In particular, we construct an equilibrium, where an expert’s equilibrium payoff is equal to what he would obtain if he could commit to truthfully revealing his information.
This project makes several contributions to the information aggregation literature: (1) We are the first to study bidding behavior in a multi-object common-value auction where bidders have outside options, and the distribution of types is endogenously determined. (2) In this context, we highlight a new mechanism based on self-selection that can lead to the failure of information aggregation. (3) We are the first to analyze information acquisition in a multi-object common-value auction. (4) Our results show that large auctions can aggregate information even if information is costly.
This project also makes several contributions to the communication and cheap-talk literature: (1) Previous work showed that truthful communication is not possible between a decision-maker and an expert in a one-shot cheap-talk game. In contrast, we show that truthful communication is possible if the cheap-talk game is repeated and the stakes in the relationship gradually increase. (2) Prior work on persuasion characterized the most preferred information structure for an expert who faces a decision-maker that chooses the course of action. However, past work also found that the expert cannot communicate this information structure to the decision-maker in a cheap-talk game. In contrast, we show that if the decision-maker faces two experts instead of one, then truthful communication is possible.
This project also makes several contributions to the communication and cheap-talk literature: (1) Previous work showed that truthful communication is not possible between a decision-maker and an expert in a one-shot cheap-talk game. In contrast, we show that truthful communication is possible if the cheap-talk game is repeated and the stakes in the relationship gradually increase. (2) Prior work on persuasion characterized the most preferred information structure for an expert who faces a decision-maker that chooses the course of action. However, past work also found that the expert cannot communicate this information structure to the decision-maker in a cheap-talk game. In contrast, we show that if the decision-maker faces two experts instead of one, then truthful communication is possible.