Fixed and Variable-Rate Mortgages, Business Cycles and Monetary Policy

The aim of this paper is twofold. First, I study how the proportion of …xed and variable-rate mortgages in an economy can a¤ect the way shocks are propagated. Second, I analyze optimal implementable simple monetary policy rules and the welfare implications of this proportion. I develop and solve a New Keynesian dynamic stochastic general equilibrium model that features a housing market and a group of constrained individuals who need housing collateral to obtain loans. A given proportion of constrained households borrows at a variable rate, while the rest borrows at a …xed rate. The model predicts that in an economy with mostly variable-rate mortgages, an exogenous interest rate shock has larger e¤ects on borrowers than in a …xed-rate economy. Aggregate e¤ects are also larger for the variable-rate economy. For plausible parametrizations, di¤erences are muted by wealth e¤ects on labor supply and by the presence of savers. More persistent shocks, such as in‡ation target and technology shocks, cause larger aggregate di¤erences. From a normative perspective I …nd that, in the presence of collateral constraints, the optimal Taylor rule is less aggressive against in‡ation than in the standard sticky-price model. Furthermore, for given monetary policy, a high proportion of …xed-rate mortgages is welfare enhancing. Keywords: Fixed/Variable-rate mortgages, monetary policy, housing market, collateral constraint Bank of Spain, C/Alcalá 48, 28014 Madrid, Spain. [email protected]. Part of this project was developed during my dissertation internships at the St. Louis Fed and the Federal Reserve Board, whose hospitality I gratefully acknowledge. I am highly indebted to Fabio Ghironi, Matteo Iacoviello and Peter Ireland for their help and advice. Thanks to seminar participants of the Dissertation Workshop at BC, R@BC, University of Valencia, University of Barcelona and WEAI for useful suggestions. And special thanks to Susanto Basu, Bill Dupor, Simon Gilchrist, Michael Kiley, Andreas Lehnert, Antonio Miralles, Ed Nelson and Michael Palumbo. All errors are mine. This paper was previously circulated under the name "Fixed and Variable-Rate Mortgages and the Monetary Transmission Mechanism". Usual disclaimer applies. 1 " [:::]the structure of mortgage contracts may matter for consumption behavior. In countries like the United Kingdom, for example, where most mortgages have adjustable rates, changes in short-term interest rates have an almost immediate e¤ect on household cash ‡ows. [:::] In an economy where most mortgages carry …xed rates, such as the United States, that channel of e¤ect may be more muted. I do not think we know at this point whether, in the case of households, these e¤ects are quantitatively signi…cant in the aggregate. Certainly, these issues seem worthy of further study". Ben Bernanke, June 15, 2007. 1 Introduction Mortgage contracts in an economy can be …xed or variable rate. The proportion of variable-rate mortgages varies from country to country. In countries such as the United States, Germany and France, the majority of mortgages are …xed rate. However, the predominant type of mortgages in countries such as the United Kingdom, Australia and Spain is variable. Mortgage rate changes a¤ect the amount of mortgage interest payments, causing a direct cash‡ow e¤ect on consumption. Interest rate changes also a¤ect housing demand and housing prices. If households are using housing as a collateral, the value of this collateral changes, inducing a wealth e¤ect on household behavior and indirectly a¤ecting consumption (ECB (2003), HM Treasury (2003)). Interest rate shocks a¤ect mortgage rates di¤erently depending on whether the mortgage is …xed or variable rate. Variable-rate mortgages are mortgage loans for which the interest rate is adjusted periodically, typically in line with some measured short-term interest rate. Hence, interest rate shocks directly a¤ect variable rates. In contrast, …xed-rate mortgages are mortgage loans for which the interest rate remains constant through the term of the loan. The …xed interest rate is tied to a longer-term interest rate and is less sensitive to changes in the policy rate. This raises important questions: How does the mortgage rate structure a¤ect the way macroeconomic shocks are propagated? What are the implications in terms of monetary policy and welfare? These questions are of academic and policy interest. To give an illustrative example, the United Kingdom Treasury explicitly mentions the di¤erence in mortgage structures as an important reason not to join the euro area. In the UK, the vast majority of borrowers have variable-rate mortgages, as opposed to the large countries of the euro area. According to the UK Treasury, British households are more exposed to monetary policy changes than, say, German households (HM Treasury (2003), Miles (2004)).