Browse research

I show how to combine large numbers of forecasts using several approaches within the framework of a Bayesian predictive synthesis. I find techniques that choose and combine a handful of forecasts, known as global-local shrinkage priors, perform best.

We introduce generalized autoregressive gamma (GARG) processes, a class of autoregressive and moving-average processes in which each conditional moment dynamic is driven by a different and identifiable moving average of the variable of interest. We show that using GARG processes reduces pricing errors by substantially more than using existing autoregressive gamma processes does.

While Government of Canada bond transactions of hedge funds are typically in the opposite direction to those of other market participants, during the peak period of market turmoil in March 2020, hedge funds sold these bonds, just as other market participants did. This shows that hedge funds can at times contribute to one-sided markets and amplify declines in market liquidity.

Since 2014, the Bank of Canada (BoC) has maintained a comprehensive database of sovereign defaults to systematically measure and aggregate the nominal value of the different types of sovereign government debt in default. The database is posted on the BoC’s website and is updated annually in partnership with the Bank of England (BoE).

Valuation ratios in the Canadian stock market can help reveal investors’ expectations about future economic growth because the impact of economic growth on valuation ratios can vary across industries. We find that this variation helps produce accurate forecasts of future growth of real gross domestic product in Canada. The forecasts from our model declined by just over 3 percentage points between January 2022 and February 2023—a period when the Bank of Canada rapidly increased the overnight rate. As well, we find that interest-rate-sensitive industries had an outsized contribution to this expected slowdown in growth. 

We study whether the credit derivatives of firms reflect the risk from climate transition. We find that climate transition risk has asymmetric and significant economic impacts on the credit risk of more vulnerable firms, and negligible effects on other firms.

This paper quantifies global demand, supply and uncertainty shocks and compares two major global recessions: the 2008–09 Great Recession and the COVID-19 pandemic. We use two alternate approaches to decompose economic shocks: text mining techniques on earnings calls transcripts and a structural Bayesian vector autoregression model.

We construct new indicators of demand and supply for the Canadian economy by using natural language processing techniques to analyze earnings calls of publicly listed firms. Our results indicate that the new indicators could help central banks identify inflationary pressures in real time.

Since 2022, consumer inflation expectations have shifted, with a significant increase in those expecting high inflation in the coming year and a surge in those expecting deflation further in the future. Using data from the Canadian Survey of Consumer Expectations, this paper seeks to assess the factors that influence people to expect high inflation, moderate inflation or deflation.

We propose a model to understand low observed migration rates by considering the interaction between location and wealth decisions. We look at different policies and find that temporary moving vouchers only slightly increase welfare, while lower housing regulations can decrease the welfare gap by lowering house prices nationwide.