The predictive power of price dynamics: - The causality between house prices and stock prices in Sweden 1992-2009

Abstract: We examine whether a causal relationship exists between the prices of households’ two major assets; stocks and houses. In particular we study the causal impact and the direction of causality between these two asset types during the period 1992q1-2009q1 in Stockholm and in Sweden as a whole. We argue that a causal relationship should be present since households are financially affected by changes in these price series and thereby choose consumption and investment accordingly. By conducting a granger-casualty test controlling for the GDP and the interest rate, we find a positive unidirectional causal relationship running from the stock market to the housing market with a time lag of one year both in Sweden and in Stockholm. Moreover, an in-depth analysis shows that the causal effect is larger and more persistent in Stockholm than in Sweden as a whole. A 10 percent increase in stock prices has a positive impact on house prices after one year with approximately 1.3 and 2.8 percent for Sweden and Stockholm respectively. We conclude that these results are not only important for households but also for policy makers concerned with financial stability. Our findings stress the importance for creating a tool for households to hedge against house price fluctuations.