Predicting a credit crisis: House price influence on the real estate mortgage default decision

Abstract: The thesis evaluates how house price development can indicate risk of mortgage defaults. This leads to conclusions about the capability that the method of the thesis has in predicting mortgage default related crises. The questions asked regard which states in the U.S. had the most volatile house prices, which states had unexpectedly high house prices at the end of 2006, and how house prices are affected by changes in interest rates and gross domestic product (GDP). Examination and analysis of the questions above should answer the basic question of this thesis which is: Could house price risk and the expected house price level in relation to interest rate- and GDP development have been used to predict the U.S. credit crisis of 2007? The methodology has three main parts. To find out which states that have the highest house price risk, a variance comparison is applied. The states that had the highest unexpected house prices in late 2006 are found by the use of discrete-time simulations. Last, the effect of interest rates and GDP on house prices is examined through an ordinary least squares regression. The conclusion is that house price development work as an indicator of risk of mortgage defaults. However, the method applied does not give a precise picture. The method of this thesis applied on the housing market of the U.S. around 2006 gives examples of two possible reasons for risk of mortgage defaults. These two reasons have to do with housing market characteristics.