This dissertation is a theoretical and empirical investigation of the actual changes in Federal S&L asset portfolios following the deregulation of the 1980's which loosened the restrictions on the amount of non-housing related lending that Federal S&L's could undertake. In particular the study focuses on the effects of deregulation and the forces promoting and constraining the individual S&L's expansion into non-housing related assets.

The theoretical model provides a framework for the empirical examination of the deregulation in the DIDMCA of 1980 and Garn-St Germain Act of 1982. The theoretical model is an adaptation of the Mingo and Wolkowitz (1977) banking model. The peculiarities of the S&L industry are embodied through adaptations of the Mingo and Wolkowitz (1977) model which emphasize after-tax profit maximization (tax laws reward specialization in housing related assets), constrain diversification into non-housing related assets, and differentiate between mutual and stock associations.

Using the method of Lagrange multipliers, an expression is obtained for the effect of a change in after-tax profits for a relaxation of the constraint on diversification which becomes the focus of the analysis. By integrating the Lagrange multiplier with economic and regulatory controls, systems of regressions are developed which examine the changes in asset portfolio composition for Federal associations using balance sheet and income statement data between 1979 and 1983.

The findings and implications of the empirical analysis are summarized as follows:

1. The tax laws do not appear to have constrained the diversification.

2. Specialization effects with respect to housing related assets appear to have constrained the diversification into non-housing related assets.

3. Non-housing related assets and liquid assets appear to be substitutes.

4. Stock associations, on average, have expanded into non-housing related assets to a greater extent than mutual associations.

5. The changes in liability legislation appear to have restrained the diversification into non-housing related assets.

6. Large associations appear more able to acquire the expertise needed to diversify.

7. Profitability appears to be correlated with the expansion into "new products."