Research Statement

My research program in macroeconomics and finance focuses on three areas: interactions between asset markets and the real economy, the business cycle, and corporate finance. My dissertation explores this agenda with computational, econometric, and analytic tools.

The travails of the global economy in the last five years are deeply intertwined with the gyrations of housing, stock, and bond markets. This has hammered home the necessity of incorporating financial markets with macroeconomics. This motivation was firmly in mind when writing my job market paper, “The Effects of Housing Adjustment Costs on Consumption Dynamics”. Homes are the largest asset on most households' balance sheets. Unfortunately, most work in macroeconomics ignores housing's unique features: differing expected returns, costly adjustment, minimal correlation with stock market returns, and that it is also a source of consumption services. Canonical frictionless models predict that consumption and portfolio choices are constant fractions of wealth. In the presence of a realistic treatment of housing, I find consumption and portfolio choices depend not just on wealth but also on the price and quantity of housing. These features allow the model to match (in a S.S.E. and standard deviation sense) NIPA non-durable consumption changes better than frictionless models. Further, the model predicts the timing and magnitude of housing adjustments in response to the Great Recession.

In future work, I plan to add new features to this model including: human capital to the household portfolio, life-cycle housing adjustment motives, and intensive adjustment to housing. Another planned extension is to test whether the differential responses of house-poor and house-rich households to wealth shocks predicted by the model are borne out in panel data. A third extension will exploit regional home price dynamics to assess the contribution of housing to economic dynamics. Related work that is underway (coauthored with Jess Diamond of The Bank of Japan) explores how segmentation of housing markets provides additional power for estimating the stochastic discount factor and forming a more plausible market portfolio for CAPM analyses.

The second chapter of my dissertation, “Employment and the Great Moderation: International Evidence”  (coauthored with Thomas Daula of UCSD) considers international business cycle dynamics. We study the reasons for the prolonged period of stability that marks the Great Moderation and the apparent exit from that regime with the 2008 financial crisis. There are three explanations: i) good luck, ii) good policy and iii) structural change. Though the Great Moderation was a global phenomenon, the literature has focused on explanations based on American experiences of monetary policy and structural change. There remains significant disagreement over the results of these investigations. Galí and van Rens (2010) accounts for the decline in output volatility and a number of recently documented employment regularities with a single cause, the decline in American labor market frictions. We consider employment and output across 14 OECD countries and find that under a variety of econometric techniques the US experience is atypical. Under Galí and van Rens, this would be caused by differences in labor frictions. We estimate labor frictions with data from the OECD and the LSE Centre for Economic Performance. We find international differences in labor frictions do not explain differences in business cycle dynamics and therefore it is unlikely that structural change in the labor market can account for the Great Moderation. This narrows the field of possible alternatives to the null hypothesis of good luck.

Since others have argued that better monetary policy is an important cause of the Great Moderation, future work will extend the model to have a monetary policy channel. This project alerted us to the widespread use of univariate filtering in studying macroeconomic dynamics. Since univariate filtering may induce spurious relationships and needlessly reduce statistical power, future work will revisit those analyses with multivariate filtering.

The third chapter of my dissertation, “How Do Firms Switch Among Tools Used to Monitor the Agency Problem?” (coauthored with Cindy Vojtech of the Federal Reserve Board) explores the relationships between the mechanisms that mitigate agency problems. Since firms face differing monitoring costs and levels of agency problems, the portfolio of monitoring tools selected is endogenous to firm characteristics. The minimum requirements on board composition established by the Sarbanes–Oxley Act and contemporaneous changes in NASDAQ and NYSE rules provide exogenous variation in monitoring. We study how treated firms adjust their choice and magnitude of monitoring methods in response to this natural experiment using a difference-in-differences estimation strategy. Depending on the rule change, we find that treated firms decreased dividends and on average lowered CEO ownership by 1.9% and lowered leverage by 1.6%. This provides evidence that independent board members are substitutes for other techniques to address the principal-agent dilemma such as dividends, CEO ownership, and debt.

Currently, we are extending the paper by examining measures of good governance before the law change. Our goal is to establish if there was variation in the degree of substitution and complementarity with pre-law change governance. We hope to continue this line of research by examining other natural experiments in governance and managerial independence.