Essays on portfolio behavior and asset pricing

Sammanfattning: This thesis consists of four self-contained essays on portfolio behavior and asset pricing.The first Essay employs portfolio theory to study the balance sheet composition of theSwedish Manufacturing Industry 1965-90. During this period there have been quite dramaticchanges in firms' balance sheets, and in particular, there appears to have been a substitution offinancial assets for tangible assets (real capital and inventories). I try to analytically explain theobserved development as a rational adjustment to changes in relative returns. Even thoughtests-for the mean-variance restrictions of symmetry and homogeneity are rejected, I find thatto some extent changes in relative returns can explain the observed development.In Essay two I study asset substitutability and debt management in Sweden 1980-1992.Including a foreign security in investors' asset menus, I find that lengthening the maturitycomposition of government debt drives up the return on long-term bonds and corporateshares. However, in quantitative terms the effects are small. The investors' demand for theforeign asset appears to be sensitive to the maturity composition of government debt.In the third Essay, written jointly with Hans Dillén, the Market Model - estimated for 20stocks on the Stockholm Stock Exchange - is examined under different assumptions regardingthe distribution of the residuals. We find strong evidence that the residuals have a leptokurticdistribution and our results suggest that much of the leptokurticness can be attributed to ajump component in the distribution. Even though the OLS estimator is consistent it is sensitiveto outliers in the distribution. The alternative estimators we propose are more robust toextreme observations. In a Monte-Carlo experiment, based on a bootstrapping technique, wefind that the probability of obtaining estimates that deviate more than ten percent from the truevalue is reduced by our alternative estimators.In the fourth Essay I estimate the stock market's relative valuation of dividends andcapital gains for the period 1983-96, using a version of the After-tax CAPM. I find thatinvestors require a higher risk-adjusted pre-tax return to hold high-yield shares compared tolow-yield shares. This result cannot be attributed to any information effect, as has beensuggested in the international literature, due to the specific characteristics of Swedish data.Instead our interpretation is that investors require a tax-premium to hold high-yield sharescompared to low-yield shares. After the Tax Reform Act of 1991 the tax-premium hasdecreased. Presumably, its an effect of the tax reform or, alternatively, a consequence of theincreased influence of foreign investors on the Swedish stock market.