Incomplete Market Models and The Housing Market

Abstract: This thesis investigates the characteristics of the market for housing derivatives. We fit two incomplete market models to data using futures contracts on the S&P/Case-Shiller Composite 10 Index. First the index is assumed to follow a Geometric Brownian Motion, second we introduce seasonality through a Geometric Ornstien Uhlenbeck process with drift. Further good deal bounds are studied, and their numerical properties evaluated, for each specification. The GBM model explain 93.4% of the variation within the sample period with an average negative pricing bias of 1.2%. The GOU model explain 80% of the variation within the sample period. Through out the whole thesis we assume a constant market price of risk. For both models this turns out to be a reasonable assumption, even tough we find evidence in favor of a non constant lambda for each model. However due to the poor data quality the results are uncertain.