Search

The Bank of Canada uses several short-term forecasting models for the monitoring of key foreign economies - the United States, the euro area, Japan and China. The design of the forecasting models used for each region is influenced by the level of detail required, as well as the timeliness and volatility of data. Forecasts from different models are typically combined to mitigate model uncertainty, and judgment is applied to the model forecasts to incorporate information that is not directly reflected in the most recent indicators.

Equilibrium bond-pricing models rely on inflation being bad news for future growth to generate upward-sloping nominal yield curves. We develop a model that can generate upward-sloping nominal and real yield curves by instead using ambiguity about inflation and growth.

The degree to which financial constraint is binding is often not directly observable in commonly used business data sets (e.g., Compustat). In this paper, we measure and estimate the likelihood of a firm being constrained by external financing using a data set of small- and medium-sized Canadian firms.

The Bank of Canada’s current suite of models faces challenges in addressing network effects that integrate household and firm-level heterogeneity and their behaviours. We develop CANVAS, a Canadian behavioural agent-based model to contribute to the Bank’s next-generation modelling effort. CANVAS improves forecasting performance and expands capacity for model-based scenario analysis.

With increasing levels of household debt in recent years, the number of households that may be vulnerable to a negative economic shock is rising as well. Decisions made by both the debtor and the creditor can contribute to insolvency. This article presents some stylized facts about insolvency in Canada’s household sector and analyzes the role of creditors in insolvencies. The average debt of an individual filing for bankruptcy is more than 1.5 times that of an average Canadian household; bankruptcy filers tend to be unemployed or in low-wage jobs, and are typically renters. The article reports that banks that approve more loans per branch, which is interpreted as less-intensive use of soft information (such as the loan officer’s assessment of the applicant’s character), experience more client bankruptcies. This finding has important policy implications, because financial institutions that do not use soft information risk further deterioration in their lending portfolios.

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.

Conventional models imply that central banks aiming to raise inflation should lower nominal rates and thus stimulate aggregate demand. However, several economists have recently challenged this conventional wisdom in favour of an alternative “neo-Fisherian’’ view under which higher nominal rates might in fact lead to higher inflation.

Business cycles are substantially correlated across countries. Yet most existing models are not able to generate substantial transmission through international trade. We show that the nature of such transmission depends fundamentally on the features determining the responsiveness of labor supply and labor demand to international relative prices.

We develop a model to analyze the link between financial leverage, worker pay structure and the risk of job termination. Contrary to the conventional view, we show that even in the absence of any agency problem among workers, variable pay can be optimal despite workers being risk averse and firms risk neutral.

We introduce a new class of time-varying parameter vector autoregressions (TVP-VARs) where the identified structural innovations are allowed to influence — contemporaneously and with a lag — the dynamics of the intercept and autoregressive coefficients in these models.