Welcome to my website. I’m a Senior Economist at the Bank of Canada as part of both the Model Development & Research Division of the Financial Stability Department and the Economics and Financial Research Department. My research focuses on household finance and macroeconomics, with an emphasis on the aggregate implications of household borrowing and consuming. I have a PhD in Economics from Duke University.
All views on this website are my own.
PhD in Economics, 2016–2021
Duke University
MA in Economics, 2019
Duke University
BSc in Financial Economics, 2015
University of Toronto
In the United States, student debt currently represents the second largest component of consumer debt, just after mortgage loans. Repayment of those loans reduces disposable income early in their life cycle when marginal utility is particularly high, and limits households' ability to build a buffer stock of wealth to insure against background risks. In this paper we study alternative student debt contracts, which offer a 10-year deferral period. During this period individuals either make interest payments only (“Principal Payment Deferral”, PPD) or make no payments at all (“Full Payment Deferral”, FPD) with the missed interest payments added to the value of the debt outstanding. We first calibrate an equilibrium with the current contracts, and then solve for counterfactual equilibria with the PPD or FPD contracts. We find that both alternatives generate economically large welfare gains, which are robust to different assumptions about the behavior of the lenders and borrower preferences. We decompose the gains into the percentages resulting from loan repricing and from the deferral of debt repayments.
How do households respond to unanticipated income shocks? Reflecting the fact that households are not perfect planners, I build and estimate a quantitative model of bounded rationality in which reoptimization is costly. Households respond to windfall income shocks by choosing a finite planning horizon over which to reoptimize. The optimal horizon is increasing in income, wealth, and the magnitude of the income shock. In the estimated model, the distribution of consumption responses is consistent with two motivating facts: highly liquid households have large consumption responses out of income shocks that cannot be driven by borrowing constraints, and larger income shocks induce smaller consumption responses.
We study the role that the joint distribution of liquid assets and credit card debt plans in the aggregate marginal propensity to consume. We show that grouping households across the distribution of liquid wealth, as is typically done, confounds two very different types of households: true hand-to-mouth households with low liquid wealth due to low liquid assets, and households with low liquid wealth due to high debt. This latter type displays a high marginal propensity to repay debt and a low marginal propensity to consume. To quantify the impact of unsecured credit card debt on stimulative fiscal transfers, we add a cash-in-advance constraint to a standard consumption-savings model. The model generates the co-holding of liquid assets and debt observed in the data and matches the empirically observed marginal propensities to consume and repay debt.
Rent control, intended to benefit renters by capping rent increases, may disincentivize landlords from lowering rents during temporary negative demand shocks because they are unable to quickly increase rent afterward. To test this prediction, I use a unique combination of exogenous variation in rent control policy in Toronto and a negative demand shock induced by the COVID-19 pandemic. In line with theory, rent per square foot decreased by 3.6% for rent controlled units and 6.5% for exempt units. Using a model of differentiated demand, I construct a counterfactual exercise and estimate that in the absence of rent control, rent would have decreased by 9.1% for rent-controlled units and 9.9% for exempt units.
This paper analyzes the tradeoff between targeted versus timely fiscal stimulus payments in a quantitative two-sector HANK model. In response to a negative sectoral shock, fiscal policy is specified as the total size of transfers, the degree of targeting towards households in the affected sector, and the length of periods until the policy can be implemented. The key trade-off in the model is that the degree of targeting is increasing in the delay until policy can be implemented. In the baseline calibration of symmetric equilibrium with one household wholly employed in each sector, the key result is that fully targeting the stimulus program to the household in the affected sector yields less total welfare than intermediate levels of targeting.
Using 14,800 forecasts of one-year S&P 500 returns made by Chief Financial Officers over a 12-year period, we track the individual executives who provide multiple forecasts to study how their beliefs evolve dynamically. While CFOs' return forecasts are systematically unbiased, their confidence intervals are far too narrow, implying significant miscalibration. We find that when return realizations fall outside of ex-ante confidence intervals, CFOs' subsequent confidence intervals widen considerably. These results are consistent with a model of Bayesian learning which suggests that the evolution of beliefs should be impacted by return realizations. However, the magnitude of the updating is dampened by the strong conviction in beliefs inherent in the initial miscalibration and, as a result, miscalibration persists.
As part of the CARES Act, the IRS distributed $300 billion in Economic Impact Payments (EIPs) directly to US households. In the Census Bureau’s Household Pulse Survey (HPS), almost 75% of households receiving an EIP reported mostly spending it. Conditioning on labor status, 63% of employed households reported mostly spending their EIPs, compared to 84% of unemployed households. Both types of households reported spending largely on consumption goods, but unemployed households tended to pay regular bills while employed households paid down debt or increased savings. The evidence suggests that in designing an untargeted stimulus program and trading off potential efficacy for timeliness, Economic Impact Payments were overall very effective in supporting consumer spending.
To analyze monetary policy implementation in a negative rate environment, we add the option to exchange central bank reserves for cash to the standard workhorse model of monetary policy implementation (Poole 1968). Importantly, we show that monetary policy can be constrained when the target overnight rate is below the yield on cash. At this point, the overnight rate equals the yield on cash instead of the target rate. Modifications to the implementation framework, such as a reserve requirement that varies with cash withdrawals, can help restore the implementation of monetary policy such that the overnight rate equals the target rate.
Formerly Bank of Canada Staff Working Paper 2017-25.