全球智库信息集成服务系统(G2TT): Global convergence of the stochastic tatonnement process.

G2TT

来源类型	Article
规范类型	其他
DOI	10.1016/S0304-4068(00)00037-9
	Global convergence of the stochastic tatonnement process.
	Ermoliev YM; Keyzer MA; Norkin VI
发表日期	2000
出处	Journal of Mathematical Economics 34 (2): 173-190
出版年	2000
语种	英语
摘要	The paper introduces stochastic elements into the Walrasian tâtonnement process, both to make it more realistic and to ensure its global convergence, with probability 1. It is assumed that the aggregate excess demand satisfies standard assumptions but is subject to measurement error. We distinguish two cases. First, the true aggregate excess demand is assumed to satisfy the Weak Axiom of Revealed Preference. This condition will be met if the underlying economy has a single consumer, several consumers with identical homothetic utility functions, or if it maximizes a social welfare function. We prove that after two minor modifications, under a fairly general specification of the measurement error and by imposing a certain consistency property on the estimator of excess demand, the tâtonnement process converges with probability 1 to an equilibrium. Second, we consider the case that the Weak Axiom only holds around some of the equilibria. The procedure proposed imposes a random shock at time intervals of increasing duration. We also discuss how the procedure could be extended to determine all equilibria, to deal with jumps in excess demand including those that do not satisfy the Weak Axiom, and to represent agents who only gradually learn how to find an optimum.
主题	Risk, Modeling and Society (RMS)
关键词	Stochastic tâtonnement Path dependence Stochastic optimization Martingales Simulated annealing
URL	http://pure.iiasa.ac.at/id/eprint/5989/
来源智库	International Institute for Applied Systems Analysis (Austria)
引用统计
资源类型	智库出版物
条目标识符	http://119.78.100.153/handle/2XGU8XDN/127913
推荐引用方式 GB/T 7714	Ermoliev YM,Keyzer MA,Norkin VI. Global convergence of the stochastic tatonnement process.. 2000.