Project topics and materials
Get Complete Project Material File(s) Now! »
A market perspective
The first approach constitutes the first chapter of this dissertation. It is a contribution to the market microstructure literature, which is joint work with Edouard Challe, and has been published in the Journal of Economic Theory. The central question adressed in this paper is the one of the relationship of the choice of orders (limit orders, market orders) made by imperfectly informed traders trading an asset on a market, and the informationnal content of the price.
Modelling approach and related literature
Market microstructure is a financial subfield dealing with the impact of financial market decision and design on financial outcomes. The National Bureau of Economic Research Market Microstructure Working Group defines his role as follows:
The market microstructure research group is devoted to theoretical, empirical, and experimental research on the economics of securities markets, including the role of information in the price discovery process, the definition, measure-ment, control, and determinants of liquidity and transactions costs, and their implications for the efficiency, welfare, and regulation of alternative trading mechanisms and market structures. 1
The approach developped in this chapter is theoretical, and studies a stylised, static financial market with three types of participants: informed traders who participate in the market according to the private information they own on the security’s future payoff, noise traders who participate in the market only for liquidity consumption smoothing reasons, unrelated to the information they own about the security, and market makers who close the deals, setting prices up. While the early market microstructure literature has been mainly concerned with the impact of market makers in the information aggregation process (see e.g. Stoll (19), Glosten (8), Glosten (9)), a later branch has emphasized the role of privately informed traders. This branch follows pioneering work of Kyle (15), Glosten and Milgrom (10), Easley and Maureen (7), among many others. In this vein, we study competitive rational expectations model with asymmetric information in a static setup closely related to developments by Hellwig (14), Grossman and Stiglitz (11), Diamond and Verrecchia (6), Admati (1), who considers a multiasset market, and Vives (21). Our focus is on the types of orders informed traders are allowed to post in the market. The market microstructure literature studies essentially two types of orders: market orders and full demand schedule (in the rational expectations tradition). Informed traders who choose to set market orders place only one order in the market, while full demand schedule consists of an infinite number of orders, each being associated with a price at which the trade takes place. The question this chapter adresses is the type of orders allocation, and the way information is aggregated into prices when you let traders choose their order types.
Main results
The article examines the joint determination of the informational content of asset prices and market composition by order type on a deep and competitive financial market in which information on economic fundamentals is dispersed among investors. Specifically, we consider a market structure in which investors receive a noisy signal on the value of the dividend, following which they can invest in the asset, and, in order to do so, use two types of orders. The first type is the one which is conventionally studied in the literature on the aggregation of information: each investor is supposed to post a complete demand curve, that is to say an order that specifies the quantity of acquired/sold for any possible value of the equilibrium price. Such an order is tantamount to placing a continuum of limit orders, and we assume that such complexity requires the payment of a fixed (but potentially low) cost. The interest of this type of order is that it covers the investor against the risk of the execution price, that is to say the uncertainty about the effective exchange price resulting from the fact that the orders are placed before being aggregated by the market to give the equilibrium price. Alternatively, investors can place market orders, which are conditional purchase orders to private information but not to the equilibrium price. These orders are much simpler than complete demand curves, and we normalize their cost to zero. Investors using this type of order are exposed to the risk of price execution (in addition to the dividend risk), and this additional risk generally leads them to limit the size of their orders, with a given private signal. In this framework, we ask the basic question: what is the equilibrium composition of the market by type of order, i.e. the proportion of investors choosing each of these orders? Our first result is that there is a strategic substitutability in the choice of order types when the private information of investors is sufficiently precise. In other words, the fact that some investors place demand curves is likely to dissuade other investors from doing so.
The explanation for this result is simple: investors placing demand curves are protected against the run-price risk, they respond more aggressively to their private information than investors placing market orders. When the information is very precise, this aggressiveness leads to efficient alignment of the asset price to the fundamental (in other words, the price becomes more informative), which reduces the risk of execution price and therefore the incentive to hedge against this risk. We show that as the accuracy of the signals increases, so that the asset price becomes more and more informative, the fraction of investors placing demand curves is reduced to become asymptotically residual. It thus appears that the market microstructure most commonly assumed in the literature is not robust to the introduction of a much simpler and marginally less costly alternative order type. Introducing endogenous order’s choice in a standard model of market microstructure yields to distortions in the information aggregation by prices, compared to the one usually found in this literature: as people tend to choose less costly orders, the information aggregation is less efficient than in the standard approach with full demand schedule, implying that observed prices in markets can in fact be a much more blurry signal about asset’s future dividends than in a situation where agents cannot choose their orders. When considering markets where severe pricing adjustments occur in times of crises (e.g. sovereign debt market), the individual reluctance to pay for costlier types of order might be a driver of bad information aggregation into prices.
Markets and runs
The second chapter adopts a broader view, by embedding the mechanism developped in the first chapter into a two-stage global game. This chapter, which is also joint work with Edouard Challe, has been submitted. The question we wish to adress is the impact of market microstructure in a financial market on the equilibrium selection in a coordination game which considers the price outcome observed in the financial market as a public information.
Modelling approach and related literature
Coordination games with strategic complementarities (i.e such that the decisions and actions of the different agents mutually reinforce one another) tend to generate multi-ple equilibria, as emphasized by Cooper and John (5). Given the variety of economic phenomena associated to such games, in particular in the fields of macroeconomics and finance (speculative attacks, debt crises,..), and the positive ambiguity of multiple equilib-ria outcomes, a strand of the game theoretic literature has seeking to develop equilibrium selection techniques. Following Carlsson and van Damme (4), global games technique have been developed. They introduce refinement through perturbation techniques, which help pinning down unique equilibrium. The fundamental idea is that perturbations generate enough differentiations between the agents to allow for differentiated behaviors between the agents. One concern with the robustness of this result is that the introduction of a theory of prices in global coordination games may reintroduce multiplicity of equilibria. The concern about the robustness of the global game result has been raised by Atkeson (2) in his comment toMorris and Shin (16) . This point has been formalized and addressed in Werning and Angeletos (22) and Tsyvinski, Mukherji, and Hellwig (20). They show that equilibrium multiplicity may be restored by the existence of prices acting as an endoge-nous public signal, provided that private information is sufficiently precise. The question we are after in this paper is the robustness of this result to the modelling of the financial market itself.
Main results
Speculators contemplating an attack (e.g., on a currency peg) must guess the beliefs of other speculators, which they can do by looking at the stock market. This paper examines whether this information-gathering process is stabilizing by better anchoring expectations ñor destabilizing by creating multiple self-fulfilling equilibria. To do so, we study the outcome of a two-stage global game wherein an asset price determined at the trading stage of the game provides an endogenous public signal about the fundamental that a§ects tradersí decision to attack in the coordination stage of the game. In the trading stage, placing a full demand schedule (i.e., a continuum of limit orders) is costly, but traders may use riskier (and cheaper) market orders, i.e., order to sell or buy a fixed quantity of assets unconditional on the execution price. Price execution risk reduces traders aggressiveness and hence slows down information aggregation, which ultimately makes multiple equilibria in the coordination stage less likely. In this sense, microstructure frictions that lead to greater individual exposure (to price execution risk) may reduce aggregate uncertainty (by pinning down a unique equilibrium outcome). The outcome of this chapter is that microstructure frictions in the financial market may impact the information embedded into a public signal used for coordination purpose, even if in our simple setup, this friction does not eliminate all possibilities for self-fulfilling equilibria, it is able to reduce the multiple-equilibria zone. A potential reason for this is that the setup we propose offers a binary option with respect to price execution risk: agents are allowed to choose at a fixed cost to eliminate completely their price-execution risk, which is likely to favorize information aggregation into prices, hence dampening our equilibrium selection effect.
Crises and financial intermediation
The point of view of the last chapter is more oriented towards the macro-financial litera-ture. The issue at the heart of this work is the coexistence and interaction between the traditional banking sector and the shadow banking sector, which has developed since the 1970s to the current size whoseorder of magnitude is similar to the size of the conventional banking sector. The analysis carried out here considers the interaction between these two sectors in times of crisis as a prism of analysis and draws a series of conclusions about the possible channels of interactions between them. Joint work with Victor Lyonnet, it has been awarded the Prize of the K2 Circle in Finance, and beneftied from a research grant from the Institut Louis Bachelier and the Europlace Institute of Finance.
Modelling approach and related literature
This chapter is embedded into the literature on shadow banking, which broadly defines a parallel banking system, consisting in whole chains of intermediation performed outside the traditional banking system, which taken together provide similar services as the traditional banking system without access to central bank liquidity or public sector credit guarantees. Part of this literature aims at measuring, and defining closely the Shadow Banking system, as well as the way the 2007-2008 crisis affected it. References include Pozsar, Adrian, Ashcraft, and Boesky (18), He, Khang, and Krishnamurthy (13), Begenau, Bigio, and Majerovitz (3). Another strand of the literature has studied the way shadow banks interact with banks. An emphasis has been made on the regulatory arbitrage view of shadow banks (see for instance Plantin (17)). Finally, a strand of the literature has focused on the way these two types of institutions coexist. Hanson, Shleifer, Stein, and Vishny (12) develop a model where these two types of intermediaries coexist by investing in different types of assets. Our contribution falls at the intersection between these two last strands.
Table of contents :
1 Résumé
1.1 Une perspective de marché
1.2 Marchés et paniques
1.3 Crises et intermédiation financière
1.4 Suite de la dissertation
Bibliographie
2 General introduction
2.1 A market perspective
2.2 Markets and runs
2.3 Crises and financial intermediation
2.4 Rest of the dissertation
Bibliography
3 Market composition and price informativeness in a large market with endogenous order types
3.1 Introduction
3.2 The model
3.3 Exogenous trader types
3.4 Endogenous trader types
3.5 Concluding remarks
3.6 Appendix
Technical Appendix
3.A Proof of Proposition 2
3.B Generalising the information structure
3.C Proof of Proposition 2b
Bibliography
4 Market microstructure, information aggregation and equilibrium uniqueness in a global game
4.1 Introduction
4.2 The model
4.3 Equilibrium uniqueness versus multiplicity
4.4 Endogenous sorting
4.5 Concluding remarks
4.6 Appendix
Bibliography
5 Traditional and shadow banks during the crisis
5.1 Introduction
5.2 Motivating evidence
5.3 Model setting
5.4 Model analysis
5.5 Discussion
5.6 Conclusion
Technical Appendix
5.A The data
5.B Shadow bank’s program
5.C Traditional bank’s program
5.D Market equilibria
Bibliography
6 Conclusion
6.1 Conclusive remarks
6.2 Future research
