Natural interest rates in the U.S., Canada and Mexico

The natural interest rate, or r-star, has been a key determinant of monetary policy normalization in the U.S. and other countries. Since December 2015, the U.S. Federal Reserve (Fed) has increased the federal funds target rate by 100 basis points and, policymakers and market participants expect further interest rate increases. Central banks in other countries will have to balance the spillover effects from the Fed’s normalization with their own internal dynamics. In countries with close ties to the U.S. economy, the influence could be more profound. This is the case of Canada and Mexico, two traditional small-open economies that are highly integrated with the U.S. through international trade and capital markets. All of our data are from Haver Analytics.

In this paper, we first apply the methodology developed by Laubach and Williams (2003) to estimate r-stars for the U.S., Canada and Mexico. With this methodology we also decompose the natural interest rate into a growth and non-growth components. The growth component is the growth rate of potential output while the non-growth component captures the aggregate effects of other factors such as the scarcity of safe assets (Caballero et al., 2017), income inequality (Summers, 2014), and the global saving “glut” (Bernanke et al., 2011). We further discuss the co-movement of the three estimated natural interest rates using a dynamic factor model to extract a common factor for all three countries.

In our country-by-country analysis, we find that a declining trend in potential output growth explains Canada's decreasing natural interest rate. Meanwhile, Mexico's natural interest rate is dominated by its volatile non-growth component, which may result from higher vulnerability to financial conditions and income inequality. In our study of the natural interest rate co-movement, we show that the U.S. has a strong influence within the region, as its r-star significantly correlates with the common factor. However, this connection decreased after the global financial crisis.

Our findings suggest that the natural interest rate in the U.S. would disproportionately affect the components of r-star of both Canada and Mexico through different channels. Therefore, we simulate the path for short-term interest rates in these countries conditional on growth and non-growth factors and likely paths of the federal funds rate in the U.S. Our simulations highlight the distinctive challenges faced by the central bank in each country.