There has been much recent interest in the effect of the level of wealth holdings on consumed expenditure. Numerous empirical studies have been reported in the literature including Jonson (1976) ; Townend (1976) ; Davidson, Hendry, Srba and Ye0 (1978); Bean (1978); Hendry and von Ungern-Sternberg (1981), etc. Although these studies look at the infiuence of different aspects of the portfolio on expenditure (with particular emphasis on liquid assets), the approaches adopted have little to say on the determinants of saving or portfolio behaviour. This tendency of the literature on consumers' expenditure to treat savings as a residual is mirrored by the tendency of the literature on portfolio allocation to take savings and, hence, wealth as given and concentrate on the allocation of net wealth between different assets and liabilities, e.g., Brainard and Tobin (1968); Backus, Brainard, Smith and Tobin (1980) ; Kopcke (1977) ; Saito (1977), etc. A more reasonable approach is to treat the decision to save and decisions on the magnitude and form that such savings take aa inextricably linked with expenditure decisions.Hence, if wealth or liquid asset holdings are believed to be important in explaining expenditure, an appropriate framework in which to analyse this issue is one which embodies conscious savinga and portfolio allocation decisions and allows for interrelationships between expenditure and portfolio composition. It is also important to note that emphasis of certain wealth components and neglect of others is a simplifying abstraction. Disequilibrium in consumed w e t s and liabilities other than liquid assets can affect their *Manuscript received 6.9.83; final version weived 16.11.84. tl a m grateful to Michael Artis and two referees for helpful comment8 on an earlier veraion of this paper.
65