Empirical studies of financial asset returns

Sammanfattning: This dissertation aims at understanding differences in rates of returns on financial assets, both in the cross-section and over time. It contains four chapters. The first paper is "Non-separable preferences and risk aversion: Results from the UK". Due to poor empirical support of the Consumption Capital Asset Pricing Model it has been suggested that the model should be specified as being non-separable over durable and non-durable consumption, by using a Constant Elasticity of Substitution (CES) utility function. I test this model on UK data with an emphasis on analyzing if it can resolve the "Equity Premium Puzzle" in the UK, as documented by earlier studies. This analysis reveals that the non-separable specification does not offer any improvement compared to a standard separable model. In fact, the analysis reveals that the non-separable CES model is ill-suited for resolving the "Equity Premium Puzzle" in the UK market. The second paper is "Investment-based CAPM in the UK". Cochrane (1991, 1996) suggests that equity returns can be fruitfully modeled by looking from the production side of an economy – leading to an Investment-based Capital Asset Pricing Model. I evaluate the model on UK data, and find that, once a lead-lag relationship between investments and stocks are taken into consideration, the model does live up to one of it's primarily prediction: average returns do line up with correlations between investment growth and returns. This is an encouraging result, since it indicates a rational interaction between the real economy and the financial market. The result has an important implication for future research on understanding the link between investments and financial markets in the UK: It is essential to allow for lags in the investment process. The third paper is "Evaluating a non-linear asset pricing model on international data". If international financial markets are integrated, then only global systematic risk variables should affect expected returns. A central, but still unsolved problem is the identification of these global risk variables. The paper analyses the ability of a non-linear asset pricing model, suggested by Dittmar (2002), to explain the returns on international value and growth portfolios. We find support for the model of Dittmar (2002). Evaluated conditionally, this model successfully passes all the different diagnostic tests performed in the analysis. The fourth paper is "An empirical analysis of factors driving the swap spread". Due to its importance for financial markets, the swap spread, i.e. the difference between the swap rate and the government bond yield, is one of the most closely monitored financial indicators. In this paper, we perform a robust analysis of the determinants of US swap spreads using a wide range of theoretically motivated candidate factors. Among other things, we find that Treasury- and stock market volatility as well as the activity of the Mortgage Backed Security holders have strong impacts on the US swap spread.