Efficiency of continuous double auctions under individual evolutionary learning with full or limited information

In this paper we explore how specific aspects of market transparency and agents’ behavior affect the efficiency of the market outcome. In particular, we are interested whether learning behavior with and without information about actions of other participants improves market efficiency. We consider a simple market for a homogeneous good populated by buyers and sellers. The valuations of the buyers and the costs of the sellers are given exogenously. Agents are involved in consecutive trading sessions, which are organized as a continuous double auction with order book. Using Individual Evolutionary Learning agents submit price bids and offers, trying to learn the most profitable strategy by looking at their realized and counterfactual or “foregone” payoffs. We find that learning outcomes heavily depend on information treatments. Under full information about actions of others, agents’ orders tend to be similar, while under limited information agents tend to submit their valuations/costs. This behavioral outcome results in higher price volatility for the latter treatment. We also find that learning improves allocative efficiency when compared to outcomes with Zero-Intelligent traders.     

Market behavior in bid,offer, and double auctions: A Reexmination

This study examines the robustness of the previously observed [Smith (1964)] differences in market behavior between double auctions (both buyers and sellers may enter price quotes), bid auctions (only buyers may enter price quotes), and offer auctions (only sellers may enter price quotes). Based on the results of twenty-one market experiments, we find little support for the a priori hypothesis that bid auction prices tend to be greater than double auction prices which tend to be greater than offer auction prices.     

Price flexibility and market performance in experimental markets

Summary Posted offer and double auction pricing institutions are examples of fixed and flexible pricing institutions, respectively. In laboratory settings, double auction markets dominate posted offer institutions in terms of the amount of potential surplus which traders are able to extract. These results are invariant with respect to the experience of traders or the perishability of the good traded. Performance differences across institutions may be due to information differences. The introduction of a second price-posting in the posted offer institution results in a pricing institution which is closer to the double auction's flexible pricing environment. Laboratory results suggest that under otherwise comparable conditions, this innovation is sufficient to remove the performance differences which have been demonstrated to exist between the posted offer and double auction institutions.We acknowledge the support of the Social Sciences and Humanities Research Council of Canada and the Arts Research Board of McMaster University. We thank Robert Moir and Susan Sabourin for computational work on this project and the fifteen undergraduates who worked as recorders and recruiters. We also thank Charles Plott and seminar participants at the Economic Science Association, Canadian Economics Association, and Waterloo University for the comments received on recent versions of this work.     

Optimality versus practicality in market design: A comparison of two double auctions

We consider a market for indivisible items with m buyers and m sellers. Traders privately know their values/costs, which are statistically dependent. Two mechanisms are considered. The buyer's bid double auction collects bids and asks from traders and determines the allocation by selecting a market-clearing price. It fails to achieve all possible gains from trade because of strategic bidding. The designed mechanism is a revelation mechanism in which honest reporting of values/costs is incentive compatible and all gains from trade are achieved. This optimality, however, comes at the expense of plausibility: (i) the monetary transfers among the traders are defined in terms of the traders' beliefs about each other's value/cost; (ii) a trader may suffer a loss ex post; (iii) the mechanism may run a surplus/deficit ex post. We compare the virtues of the simple yet mildly inefficient buyer's bid double auction to the flawed yet perfectly efficient designed mechanism.     

Convergence of outcomes and evolution of strategic behavior in double auctions

We study the emergence of strategic behavior in double auctions with an equal number of buyers and sellers, under the distinct assumptions that orders are cleared simultaneously or asynchronously. The evolution of strategic behavior is modeled as a learning process driven by a genetic algorithm. We find that, as the size of the market grows, allocative inefficiency tends to zero and performance converges to the competitive outcome, regardless of the order-clearing rule. The main result concerns the evolution of strategic behavior as the size of the market gets larger. Under simultaneous order-clearing, only marginal traders learn to be price takers and make offers equal to their valuations/costs. Under asynchronous order-clearing, all intramarginal traders learn to be price makers and make offers equal to the competitive equilibrium price. The nature of the order-clearing rule affects in a fundamental way what kind of strategic behavior we should expect to emerge.     

Evolution of markets: a simulation with centralized,decentralized and posted offer formats

We apply standard evolutionary dynamics to study trader participation in three competing market formats —call market (CM), posted offer (PO) and decentralized market (DM). In our framework, heterogeneous buyers and sellers seek to transact a homogeneous good, which can be done by allocating their time among three different market formats. Our results show that (i) the final participation of traders in the CM is much higher, (ii) the PO can coexist with the CM, (iii) the DM unravels when competing against CM and (iv) the DM can coexist with the PO, depending on the initial participation conditions. Thus, we demonstrate that agent-based micro-simulations coupled with replication dynamics can reproduce most of the previously known theoretical results and have the potential to investigate other commonly used hybrid auction protocols.     

Co-learning patterns as emergent market phenomena: An electricity market illustration

The definition of emergence remains problematic, particularly for systems with purposeful human interactions. This study explores the practical import of this concept within a specific market context: namely, a double-auction market for wholesale electric power that operates over a transmission grid with spatially located buyers and sellers. Each profit-seeking seller is a learning agent that attempts to adjust its daily supply offers to its best advantage. The sellers are co-learners in the sense that their supply offer adjustments are in response to past market outcomes that reflect the past supply offer choices of all sellers. Attention is focused on the emergence of co-learning patterns, that is, global market patterns that arise and persist over time as a result of seller co-learning. Examples of co-learning patterns include correlated seller supply offer behaviors and correlated seller net earnings outcomes. Heat maps are used to display and interpret co-learning pattern findings. One key finding is that co-learning strongly matters in this auction market environment. Sellers that behave as Gode-Sunder budget-constrained zero-intelligence agents, randomly selecting their supply offers subject only to a break-even constraint, tend to realize substantially lower net earnings than sellers that tacitly co-learn to correlate their supply offers for market power advantages.     

Zero-intelligence versus human agents: An experimental analysis of the efficiency of Double Auctions and Over-the-Counter markets of varying sizes

We study two well-known electronic markets: an over-the-counter (OTC) market, in which each trader looks for the best counterpart through bilateral negotiations, and a double auction (DA) market, in which traders post their quotes publicly. We focus on the DA–OTC efficiency gap and show how it varies with different market sizes (10, 20, 40, and 80 traders). We compare experimental results from a sample of 6400 undergraduate students in Economics and Management with zero-intelligence agent-based simulations. Simulation results show that the traded quantity increases with market size under both DA and OTC. Experimental results confirm the same tendency under DA, while the share of periods in which the traded quantity is lower than the efficient one increases with market size under OTC, ultimately leading to a DA–OTC efficiency gap increasing with the market size. We rationalize these results by putting forward a simple model of OTC market as a repeated bargaining procedure under incomplete information on buyers' valuations and sellers' costs. We show that efficiency decreases slightly with size due to two counteracting effects: acceptance rates in earlier interactions decrease with size, and earlier offers increase, but not always enough to compensate for the decrease in acceptance rates.     

Harassment, corruption and tax policy: reply

Saha [Eur. J. Polit. Econ. (2003)] has raised some points about our paper [Eur. J. Polit. Econ. 16 (2000) 75]. We herewith reply to these points.     

Run theorems for low returns and large banks

In this paper, we revisit the issue of bank fragility in the Diamond and Dybvig (J Polit Econ 91:401–419, 1983) model with sequential service and finite traders. We provide a precise condition under which banks are susceptible to a run when the return on investment is low, and we show that sufficiently large banks are always susceptible to a run. One interpretation of the condition is that exposure to runs occurs when desire for consumption smoothing or predictability of preference profiles are relatively high.     

Riskiness for sets of gambles

Aumann–Serrano (J Polit Econ 116:810–836, 2008) and Foster–Hart (J Polit Econ 117:785–814, 2009) suggest two new riskiness measures, each of which enables one to elicit a complete and objective ranking of gambles according to their riskiness. These riskiness measures were created with a risky world in mind, but not an uncertain one. We apply similar arguments to models of decision under uncertainty and develop complete and objective rankings of sets of gambles, which arise naturally in many such models. Clearly, these results extend the previous riskiness measures, and they have a natural interpretation in terms of those measures even when uncertainty does play a significant role.     

Bank incentives, contract design and bank runs

We study the Diamond-Dybvig [Bank runs, deposit insurance, and liquidity, J. Polit. Econ. 91 (1983) 401-419] model as developed in Green and Lin [Implementing efficient allocations in a model of financial intermediation, J. Econ. Theory 109 (2003) 1-23] and Peck and Shell [Equilibrium bank runs, J. Polit. Econ. 111 (2003) 103-123]. We dispense with the notion of a bank as a coalition of depositors. Instead, our bank is a self-interested agent with a technological advantage in record-keeping. We examine the implications of the resulting agency problem for the design of bank contracts and the possibility of bank-run equilibria. For a special case, we discover that the agency problem may or may not simplify the qualitative structure of bank liabilities. We also find that the uniqueness result in Green and Lin [Implementing efficient allocations in a model of financial intermediation, J. Econ. Theory 109 (2003) 1-23] is robust to our form of agency, but that the non-uniqueness result in Peck and Shell [Equilibrium bank runs, J. Polit. Econ. 111 (2003) 103-123] is not.     

An ascending double auction

We show why the failure of the affiliation assumption prevents the double auction from achieving efficient outcomes when values are interdependent. This motivates the study of an ascending price version of the double auction. It is shown that when there is a sufficiently large, but still finite, number of sellers, this mechanism has an approximate perfect Bayesian equilibrium in which traders continue bidding if and only if their true estimates of the ‘value’ of the object being traded exceed the current price. This equilibrium is ex post efficient and has a rational expectations property in the sense that along the equilibrium path traders appear to have made the best possible trades conditional on information revealed by the trading process. We thank two anonymous referees and Dan Kovenock, the Editor, whose detailed comments and suggestions have allowed us to substantially improve the paper. We also thank seminar participants at University of Toronto, University of Wisconsin-Madison, Summer 2003 North American Meetings of the Econometric Society, 2004 NSF Decentralization Conference for their comments.     

Information revelation in competitive markets

This paper analyzes a market with multiple sellers and horizontally differentiated products. We investigate the sellers’ incentives to reveal product relevant information that affects the buyer’s private valuation. The main finding is that if the number of sellers is sufficiently large, there is a unique symmetric equilibrium with full information disclosure. Thus, unlike the results by Lewis and Sappington (Int Econ Rev 35:309–327, 1994) and Johnson and Myatt (Am Econ Rev 93:756–784, 2006) for monopoly, which state that the monopolist reveals either full information or no information, intense competition results in a single extreme with respect to information disclosure. We show that the market is always inefficient, but the magnitude of inefficiency converges to zero at a high rate as competition intensifies.     

On mixed markets with asymmetric information

In asymmetric information exchange economies involving both non-negligible and negligible agents, one should expect the failure of the private Core-Walras Equivalence Theorem. This paper shows that if “large” traders are similar to each other, then they lose their market power and hence the Equivalence Theorem can be restored. We also investigate on weaker equivalences among Walrasian expectations allocations, Aubin private core and private core allocations of the original mixed economy and the atomless one associated to it, without the assumption that all atoms are of the same type. Furthermore, extensions of Hervés-Moreno-Yannelis and Schmeidler Theorems (compare Hervés et al. in J Math Econ 41:844–856, 2005a; Schemidler in Econometrica 40:579–580, 1972) are given for differential information economies in which the feasibility constraints are imposed with an equality (exact feasibility).     

Conservative traders,natural selection and market efficiency

This paper examines the impact of conservative traders on market efficiency in an evolutionary model of a commodity futures market. This paper shows that the long-run market outcome is informationally efficient, as long as in every period there is a positive probability that entering traders are more conservative than their predecessors. Conservative traders are those who correctly predict the spot price with a positive probability, and more importantly, who in their mistakes err on the side of caution, and rarely overpredict the spot price as buyers, and underpredict the spot price as sellers. This result does not require entry of traders with better information than their predecessors.     

Bilateral trade with apositional traders

We study the standard model of bilateral trade under incomplete information dropping the assumption that traders know on which side of the market they are. We consider two mechanisms that differ only in the number of offers that an agent can submit. These mechanisms are realistic and they are ex post individually rational (i.e. regret free), while the usual mechanisms proposed in the literature satisfy the weaker requirement of interim individual rationality. Properties of the Bayesian equilibria are described for the general case. For the case where valuations are uniformly distributed in the unit square, two types of equilibria are derived for each mechanism and their efficiency properties are analyzed. As expected, the equilibria under the double offer mechanism are less inefficient than those under the single offer mechanism.     

Pluralism(s) in economics: lessons from complexity and innovation. A review paper

My analysis focuses on two main observations. First, many competing schools of thoughts are currently present in economics with no predominant paradigm. We are experiencing an era of pluralism (Davis J Econ Methodol 14(3):275–290, 2007, Camb J Econ 32:249–366, 2008; Colander 2000; Colander et al. J Polit Econ 16(4):485–499, 2004). The term ‘pluralism’ is extremely interesting since, as I will show, it has different dimensions to it. These offer insights into interpreting the tangled universe of the economics. Second, there is a progressive intertwining of innovation economics with complexity economics, which I argue provides an instance of the above-described shift toward pluralism.     

How Communication Improves Efficiency in Bargaining Games

We study a double auction with two-sided private information and preplay communication, for which Myerson and Satterthwaite (1983, J. Econ. Theory28, 265–281) showed that all equilibria are inefficient and the Chatterjee–Samuleson linear equilibrium is most efficient. Like several others, we find that players use communication to surpass equilibrium levels of efficiency, especially when the communication is face-to-face. Our main contribution is an analysis of how communication helps the parties achieve such high levels of efficiency. We find that when preplay communication is allowed, efficiency above equilibrium levels is a result of what we call “dyadic” strategies that allow the parties to coordinate on a single price that reflects both parties' valuations. Journal of Economic Literature Classification Numbers: C78, D82.     

The veil of public ignorance

A theory of cooperative choice under incomplete information is developed in which agents possess private information at the time of contracting and have agreed on a utilitarian “standard of evaluation” governing choices under complete information. The task is to extend this standard to situations of incomplete information. Our first main result generalizes Harsanyi's (J. Polit. Econ. 63 (1955) 309) classical result to situations of incomplete information, assuming that group preferences satisfy Bayesian Coherence and Interim Pareto Dominance. These axioms are mutually compatible if and only if a common prior exists. We argue that this result partly resolves the impossibility of Bayesian preference aggregation under complete information.