On aggregation and representative agent equilibria (Q684175)

According to the authors' conclusion: ``A key goal of this paper is to obtain existence of representative agent equilibria in very general settings and under weak assumptions. It is therefore natural to ask whether further extensions are nonetheless possible.'' This goal in the general microeconomic setting can be reformulated as the goal to find characteristics of the neoclassical economy created on the individualistic (reductionistic) basis with many rational and independent actors (consumers, producers, traders), and conditions on the environment and information characteristics of the market -- conditions which provide rationality properties of aggregate characteristics (the aggregate market's demand and supply), similar to the corresponding characteristics of individual actors. These aggregate characteristics are attributed to the representative agent representing all individual agents of the corresponding type. It is well known that this goal had appeared unattainable for individualistic models of economy with multicommodity consumer markets of Walrasian type in the setting of Arrow and Debreu, and its generalizations. Such an insolvency is being discussed, particularly, in the fourth chapter of the widely known text book by \textit{A. Mas-Colell} et al. [Microeconomic theory. New York, NY: Oxford University Press (1995; Zbl 1256.91002)], in the papers by \textit{A. Kirman} [J. Economic Perspectives. V. 6. No 2, 117--136 (1992); CESifo Economic Studies. V. 56, No. 4, 498--535 (2010)], by \textit{V. Polterovich} [Ekonomicheskaya Nauka Sovremennoy Rossii, No. 1, 46--66 (1998)], and by \textit{V. K. Gorbunov} [Zh. Èkon. Teor. 2009, No. 1, 85--94 (2009; Zbl 1183.91087); Ècon. Nauka Sovrem. Ross. 2013, No. 4, 19--36 (2013; Zbl 1310.91089)]. However, the authors investigate this generally hard problem for financial markets which are similar to markets of production factors where preferences of buyers are primitive: they are only profit or expenses in money measure. In the financial markets the traders' preferences are also primitive. Here consumption has also only money measure, and it is the only economic argument in the individual utility functions which represent these preferences. This circumstance allows advancing in different directions of this aggregation problem of modern economic theory created on the reductionistic basis. The authors consider a financial market of \(m\) assets with \(n\) traders heterogeneous with respect to their preferences. The market can be incomplete, with trading constraints, heterogeneous beliefs, and differential information. The process of assets' price is assumed as \(m\)-dimensional semimartingale. A generic trader chooses a feasible trading strategy compatible with the initial holdings, an endowment stream, and a consumption plan (in money measure) according her/his preferences which are presented through a `stochastic utility function' (the definition is presented). The stochastic endowment streams and dividend processes may be not absolutely continuous with respect to Lebesgue measure, and they can have jumps. Each trader maximizes her utility functional which is the expectation of integral of the individual utility function over a consumption stream. An equilibrium is a price process along with trading strategies and consumption plans such that individual optimality holds, and markets clear with respect to trading and consumption. The traders' aggregation problem consists in clarification of conditions under which total indicators of micro-equilibrium coincide with the corresponding optimum macro-indicators of one `representative trader' (RT) which maximizes a single (aggregate) stochastic utility function depending on the total consumption stream. The regular stochastic individual utility functions can be aggregated into one stochastic utility function by means of the sup-convolution construction, which was first used by \textit{T. Negishi} [Metroeconomica 12, 92--97 (1960; Zbl 0104.38403)]. There are three key theorems in this paper. The first one proves existence of RT under some sufficient conditions, and the second proves existence of a RT equilibrium under another sufficient conditions which relate to the existence of a stochastic process satisfying the first order conditions for the solution to the RT's optimal consumption problem. The third theorem connects existence and characteristics of an equilibrium of the RT and those of the origin market with many traders. These theorems generalize earlier results in the literature mentioned in the Introduction.