I am a Senior Economist at the Bank of Canada in the International Studies Division in the Department of International Economic Analysis.
My research interests are in Macroeconomics and International Macroeconomics.
Work in Progress:
This paper assesses the prospects of a 2021 time bomb in SME failures triggered by the generous support policies enacted during the 2020 COVID-19 crisis. Policies implemented in 2020, on their own, do not create a 2021 “time-bomb” for SMEs. Rather, business failures and policy costs remain modest. By contrast, credit contraction poses a significant risk. Such a contraction would disproportionately impact firms that could survive COVID-19 in 2020 without any fiscal support. Even in that scenario, most business failures would not arise from excessively generous 2020 policies, but rather from the contraction of credit to the corporate sector.
We estimate the impact of the COVID-19 crisis on business failures among small and medium size enterprises (SMEs) in seventeen countries using a large representative firm-level database. We use a simple model of firm cost-minimization and measure each firm’s liquidity shortfall during and after COVID-19. Our framework allows for a rich combination of sectoral and aggregate supply, productivity, and demand shocks. We estimate a large increase in the failure rate of SMEs under COVID-19 of nearly 9 percentage points, absent government support. Accommodation & Food Services, Arts, Entertainment & Recreation, Education, and Other Services are among the most affected sectors. The jobs at risk due to COVID-19 related SME business failures represent 3.1 percent of private sector employment. Despite the large impact on business failures and employment, we estimate only moderate effects on the financial sector: the share of Non Performing Loans on bank balance sheets would increase by up to 11 percentage points, representing 0.3 percent of banks’ assets and resulting in a 0.75 percentage point decline in the common equity Tier-1 capital ratio. We evaluate the cost and effectiveness of various policy interventions. The fiscal cost of an intervention that narrowly targets at risk firms can be modest (0.54% of GDP). However, at a similar level of effectiveness, non-targeted subsidies can be substantially more expensive (1.82% of GDP). Our results have important implications for the severity of the COVID-19 recession, the design of policies, and the speed of the recovery.
Policymakers are often concerned about large and volatile capital flows potentially amplifying business cycle fluctuations. The purpose of this paper is to estimate how business cycle variables respond to exogenous capital inflows. I construct a shift-share instrument for capital inflows using mutual fund level micro-data and estimate a series of local projection regressions to trace out the dynamic response of business cycle variables to capital inflows. The key results are that GDP rises; equity inflows lead to an investment boom and debt inflows lead to a consumption boom. While I estimate that capital inflows tend to reverse, there is little evidence that these reversals lead GDP to fall even after the initial inflow disappears. I show that modest adjustments to the canonical setup of international macroeconomic models can capture the qualitative business cycle responses of both equity and debt inflows.
“Estimating the effects of central bank communications.” (with Chaewon Baek)
Recent work suggests that Central Bank communications alter the overall impact of monetary policy shocks. While lowering rates should stimulate demand, lower rates also communicates that the central bank believes that demand needs stimulating. If strong enough, the second force could lead to looser monetary policy lowering demand. The purpose of this paper is to separate the two channels. We exploit the fact that the Federal Reserve typically releases its minutes 3 weeks after FOMC announcements to separate the effect of information from the announcement of policy. We use a high-frequency event study design to estimate the impact of minutes releases and FOMC announcements. Further we use narrative methods to classify the types of information released in each FOMC announcement and in each set of minutes. Finally, with these classifications, we use the Rigabon heteroskedasiticy identification method to separate the economic effects of monetary policy from the economic effects of information .
“The Effects of Monetary Policy on Bubbles” (with Yuriy Gorodnichenko)
Since the financial crisis it has been suggested that monetary policy could be used more actively to prevent excessive bubbles from forming and the associated misallocation of investment that follows. Work by Gali (2014) suggests that when bubbles are “rational” tighter monetary policy could exacerbate bubble growth rather than limit it. The purpose of this paper is to empirically test the effects of monetary policy on bubbles. We use the bubble classification algorithm of Phillips, Shi and Yu (2015) to classify bubbles in each S&P500 stock from 1994 to the present. We then use a high-frequency identification of monetary policy shocks to see if stocks classified as in bubbles respond differently to monetary policy changes than stocks not classified as in bubbles. Our results suggest that a 100 basis point tightening of monetary policy leads to 2 percentage point decline in stocks classified as in a bubble relative to those not classified as in a bubble. This implies that monetary policy can be used to pop bubbles in contrast to Gali’s paper.
“Estimating Markups allowing for Capital-Augmenting Technical Change” (with Pierre-Olivier Gourinchas)
De-Loecker & Eeckhout (2018) argued that the US has experienced a large secular rise in market power by estimating rising markups among US firms since 1980. However, their estimation procedure does not allow for capital-augmenting technical change in the production function. When capital becomes more efficient and the elasticity of substitution between labor and capital is above 1, the labor share could fall without markups changing. We re-estimate the change in markups allowing for capital-augmenting technical change and our preliminary results suggest a modest fall in markups in the median 3-digit SIC industry in the US.