Amy Sun, the 2014 Queen’s Economics Department Faculty Research Prize recipient, uses Search Theory to give insight into economic policy. The following article was originally published in the QED alumni newsletter. 

By Amy Sun, Queen’s University

Classical economic theory treats the transaction process as an instantaneous step. Experience from the real world, however, suggests that in many contexts this simplification is too much of an abstraction from reality. In particular, many interesting and important questions in economics are left aside: How and where does one find a trading partner? How does one reach an agreement upon price? How long does it take for one to locate a trading partner and for the transaction to go through? What are the consequences given that one may have to wait a period before finally getting to trade? In particular, how does the length of the expected-time-before-transaction affect one’s trading strategies, e.g., the choice of transaction price? Finally, what are the macroeconomic implications of the fact that trades can take a non-negligible amount of time to be fulfilled?

Search Theory has been developed to model non-trivial trading processes and to provide insights into the aforementioned questions. The theoretical framework takes seriously the frictions that prevent transactions from being conducted instantaneously. Namely, individuals must “search” for a trading opportunity in such models. Search theory has been influential in many fields. My recent research aims to extend and apply search theory to broad theoretical frameworks that provide a serious micro-foundation for the notion of endogenous liquidity. My coauthors and I are using these frameworks to study issues associated with monetary theory, policy, and housing economics.

My collaboration with Guido Menzio from The University of Pennsylvania and Shouyong Shi from The University of Toronto resulted in an article entitled “A Monetary Theory with Non-Degenerate Distributions,” which was published in the Journal of Economic Theory. In this article, we contribute to monetary theory by constructing and analyzing a tractable money-search model with a non-degenerate distribution of money holdings.

It has long been a plague that search models of money (i.e., models with trading frictions in product markets) tend to be intractable due to the large dimension of the endogenous money distribution across agents. Economists have come up with theoretical shortcuts to make these models tractable by keeping the money distribution in a trivial (i.e., degenerate) form. This, however, comes with a significant cost because the endogenous distribution of liquid assets in itself is interesting and important from the perspective of both theoretical and policy work.

My paper with Menzio and Shi provides a solution to the above dilemma. We show that analytical tractability can come from modeling decentralized exchange as directed search as opposed to non-directed, random search commonly adopted in the previous literature. The key difference is that in the former, individuals are informed about and “directed” by the terms of trade before setting out to search for a trading partner. Moreover, we demonstrate the advanced techniques used to characterize the equilibrium of our environment with a non-degenerate money distribution. Later in a single-authored paper, I extend this theoretical framework to address the distributional effects of monetary and fiscal policies.

I have also collaborated with my colleagues, Allen Head and Huw Lloyd-Ellis, on applying the search theory to examine the functioning of the housing market. Our joint paper, entitled “Search, Liquidity and the Dynamics of House Prices and Construction,” was published in the American Economic Review. In this article, we study the consequences of a time-consuming trading process for the dynamics of house prices, sales, and construction at the city level. We characterize the short-run dynamics of average house prices, home sales, construction and population growth for a panel of U.S. cities. We construct a model of house search, in which the entry of new buyers and the construction of new houses in response to shocks are endogenously determined.

Our model generates serial correlation in the growth rates of house prices and construction, even if income is strictly mean-reverting following shocks. This had been proven difficult to reconcile with traditional housing models without search frictions. Finally, we show that our theory achieves much success in quantitatively accounting for the dynamics of housing price as opposed to non-search models. In my latest work with Chenggang Zhou from Queen’s University, we extend the framework into a dynamic model of housing and lending to analyze the effect of mortgage debt on house-selling decisions, as well as the impact of housing market liquidity on mortgage standards. Calibration of our model generates interesting results that are consistent with empirical features of housing and mortgage markets, at both the micro and the macro levels.