Abstract. The WALRAS algorithm calculates competitive equilibria via a distributed tatonnementlike process, in which agents submit single-good demand functions to market-clearing auctions. The algorithm is asynchronous and decentralized with respect to both agents and markets, making it suitable for distributed implementation. We present a formal description of this algorithm, and prove that it converges under the standard assumption of gross substitutability. We relate our results to the literature on general equilibrium stability and some more recent work on decentralized algorithms. We present some experimental results as well, particularly for cases where the assumptions required to guarantee convergence do not hold. Finally, we consider some extensions and generalizations to the WALRAS algorithm.

In this paper, I argue that applied monetary policy analysis is hard. In particular, all of our models are grossly deficient relative to the ideal, and this cannot be corrected in the medium term. This view has important implications for answering the Simsian question of whether any given change in policy analysis methods is progress or regress. As an application of these ideas, I assess the potential role of dynamic stochastic general equilibrium (DSGE) models to policy analysis. If DSGE models are modified so they can play all the functions as old-style models, progress or regress may be unclear. If these models are used in the roles for which they are best suited, they can make a vital positive contribution.

entities are used during the analysis stage to conceptu- alize the system, while concrete entities are used within the design process and typically have direct counterparts in the run-time system.

A fundamental question that arises in derivative pricing is why investors trade at a "fair" price. A common opinion attributes trading to dierences in the beliefs that market participants have about the future development of market prices. We develop a model that accounts for investors' pre-existing liability structures and enables us to show, through a series of experiments, that investors trade even when their belief structures are identical and accurate. More generally, we show that multi-agent simulation of a nancial market provides a mechanism for conducting experiments that shed light on fundamental properties of the market. As all processes in nancial markets (including decision making) become automated, it becomes crucial to have a mechanism by which we can observe the patterns that emerge from a variety of possible investor behaviors. Our simulator provodes this mechanism.

stor profiles in a single market. A theoretical study of such a model is virtually impossible. The strength of multi-agent approach is the ability to (experimentally) study large heterogeneous populations. Exploring financial market models with heterogeneous agents will help better understand security pricing and develop novel pricing techniques. Further, we introduce a particular microeconomic model of trading in an incomplete market that pre-supposes a heterogeneous population and study it experimentally. 2 Private Valuation of Contingent Claims Pricing a derivative security against an established market is one of the key problems in financial engineering. Arbitrage pricing theory (APT), while supplying pricing techniques based on fairness of the market, does not give a comprehensive answer to the question why investors trade in derivative securities. We believe that the answer lies in heterogeneity of the investor community. The ability to model and analize such communities will