Research

Working Papers

"Heterogeneous Real Estate Agents and the Housing Cycle" (with Paul Goldsmith-Pinkham, (Accepted at the Review of Financial Studies)

The real estate market is highly intermediated, with 90% of buyers and sellers hiring an agent to help them transact a house. However, formal training to become an agent is short, and agents primarily learn on the job. Low entry barriers and fixed commission rates result in a market where inexperienced intermediaries have a large market share, especially during and after boom periods. Using rich micro-level data on listings and deeds, we first show that agents’ experience affects clients’ outcomes and is particularly important in the busts. We then study the aggregate implications of the experience distribution on the efficiency of the real estate market by building a theoretical entry and exit model of real estate agents with aggregate shocks and considering several policies that raise the cost of entry and thus favorably alter the equilibrium distribution of experience. We find that low entry barriers amplify the cycles in the housing market.

“New Listing Alert: Alternative Theory of Housing Search"

This paper proposes an alternative approach to modeling search in real estate. Instead of search frictions resulting from buyers’ inability to sample the entire distribution of houses, all successful matches between buyers and sellers realize immediately. The current stock of buyers and sellers are therefore mismatched and both sides are waiting for the new arrivals in the market to find a successful counter party. Thus, houses on the market sell to newly arrived buyers only, and, similarly, buyers who have been looking for a while only buy newly listed properties. Using data of 3.3 million listings from one of the biggest Multiple Listing Service platform in the United States, I test several predictions of this model that are at odds with the more commonly used search models in this literature. I find that, consistent with the model, 1) sales are much more correlated with new listings than they are with the market inventory, 2) sale hazard rates decline after the initial period, and 3) houses that sell in the initial period are likely to sell at a premium while those that spend a long time on the market sell at a discount.

Publications 

"The Price to Rent Ratio: A Macroprudential Application" (with Andrew Haughwout, Rebecca J. Landau, and Joe Tracy, Real Estate Economics, 51(2), 503-532, 2023)

We examine the potential for the price-to-rent ratio to be used as a macroprudential tool. In addition to using standardized appraisal methods, appraisers could estimate the current market rent for a property. The resulting price-to-rent ratio would provide a useful signal for speculative pressures. We show this by estimating price-to-rent ratios using the American Housing Survey. The distribution of price-to-rent ratios shifted up dramatically during the housing boom with the 75th and 95th percentile reaching a historic peak in 2006 at 22 and 44, respectively. We propose a lending policy that incorporates the price-to-rent ratio to generate countercyclical loan-to-value ratios.

"Rational Buyers Search When Prices Increase*" (with Luís Cabral, Journal of Economic Theory, Vol 187, 104998, 2020.)

We develop a dynamic pricing model motivated by observed patterns in business-to-business (and some business-to-customer) transactions. Seller costs are perfectly correlated and evolve according to a Markov process. In every period, each buyer observes (for free) the price set by their current supplier, but not the other sellers' prices or the sellers' (common) cost level. By paying a cost s the buyer becomes “active” and benefits from (Bertrand) competition among sellers. We show that there exists a semi-separating equilibrium whereby sellers increase price immediately when costs increase and otherwise decrease price gradually. Moreover, buyers become active when prices increase but not otherwise. In sum, we deliver a theory whereby buyers become active (“search”) if and only if their supplier increases price.

"Firm Dynamics and Pricing under Customer Capital Accumulation"  Supplementary Material (with Pau Roldan, Journal of Monetary Economics, 2020)

In a search model of firm dynamics, customer accumulation is shown to affect the lifecycle of firms and the dynamics of markups in response to aggregate demand shocks. In the model, sellers of different sizes and productivities post dynamic pricing contracts to strike a balance between attracting new customers and exploiting pre-existing ones. Calibrated using establishment-level and micro-pricing data from the U.S. retail sector, the model provides a quantitatively good fit to the lifecycle and cross-sectional properties of retail establishments. Further, the model predicts that markups are procyclical to aggregate demand shocks, with a larger contribution by small sellers.

"Deficits, Public Debt Dynamics and Tax and Spending Multipliers", with Gauti Eggertsson and Matthew Denes, Economic Journal, 123 (566), 133-163, 2013.

Cutting government spending can increase the budget deficit at zero interest rates according to a standard New Keynesian model, calibrated with Bayesian methods. Similarly, increasing sales taxes can increase the budget deficit rather than reducing it. Both results suggest limitations of ‘austerity measures’. At zero interest rates, running budget deficits can be either expansionary or contractionary depending on how they interact with expectations about long-run taxes and spending. The effect of fiscal policy action is thus highly dependent on the policy regime. A successful stimulus, therefore, needs to specify how the budget is managed not only in the short but also medium and long run.

Research in Progress

"Urgency to Buy: Amplification of House Price Fluctuations" 

In the bust, home buyers take their time in choosing a house, unafraid that the property they liked will be snatched by another buyer while they search for other potential options. In the boom, the rivalry grows as the number of buyers increases relative to the inventory of available houses.  This urgency-to-buy shifts the bargaining power from sellers to buyers across aggregate states and amplifies the magnitude of the house price cycle.  To quantify this channel, I build a dynamic search model where buyers are able to recall their past searches, unless the previously viewed houses are purchased by another rival buyer. I examine two counterfactuals to assess the contribution of this channel to the magnitude of the business cycle. First, I consider a model of naive buyers who make search and purchase decisions as if they can always recall previously viewed properties. Second, I estimate the model with no recall.

"Estimating search costs from price distribution of heterogeneous goods" 

I extend the Hong and Shum (2006) framework for estimating search costs using only the price distribution. Instead of a single search good, I allow for multiple substitute goods of varying quality. While an additional step of demand estimation is needed to tease out quality, this allows for richer settings where firms are competing within all substitutable goods, rather than identical products only.