Stocks and houses are the two major assets on the balance sheet of American households. Changes in the two markets have a large influence on wealth and the general economy. This thesis investigates the relationship between the stock market and the house market in the U.S from 1987 to 2013.
The sample period is unique since it includes two structural breaks, the dot-com bubble and the most recent financial crisis. A bivariate correlation analysis is applied to investigate the correlation and the Granger causality test, based on a vector autoregressive model (VAR), investigates the causality between the two variables. The causality test is performed with and without GDP and interest rates as control variables.
In order to investigate a potential dynamic causality, the full sample period is divided into two periods. The bivariate correlation analysis concludes that a large and positive correlation exists between the two markets. All causality tests indicate a unidirectional causality running from the stock market to the house market. An impulse response function (IRF) is estimated in order to investigate the size and timing of the causality.
The IRF concludes that a one percentage change in the stock market affects the house market by 0.032581 percent three years later, corresponding to a change in the value of real estates possessed by American households of 7.04 billion of dollars. The same number amounts to 47.00 billion of dollars five years later. It is concluded that the relationship has a significant influence on the wealth of American households hence a need for developing a hedging instrument which reduces the ramifications from changes in the stock market is large.
Source: Jönköping University
Authors: Andersson, Erik