RDP 7902: Financial Modelling in Australia 5. The Way Ahead

At this stage the hard working reviewer is rewarded with a little space for his suggestions. In my view, the important directions which Australian financial sector studies should take are:

i) as market forces become more important in determining interest rates, efforts to explain interest rates in this way should be stepped up.

ii) it must be recognised that financial asset demands are functions not only of current interest rates but of expected rates. Further efforts to measure and to explain interest rate expectations should be made.

iii) the evidence from the general sectoral studies for links between the real and financial sectors is sufficient to encourage further work in this area. However, the statistical problems with very general specifications, which have been addressed in the past by imposing highly restricted structures (separable utility functions; specific decision hierarchies; monetary disequilibrium; simple stock adjustment models) must be overcome using more general models, based on empirically testable and theoretically reasonable restrictions.

iv) further attention must be given to the role of foreign financial markets in affecting asset demands both through expectations and as direct competing assets. Specific attention here needs to be placed on the role of exchange rate expectations.

v) the role of the equity market, which is neglected in the full model studies, must be further considered. There is conflicting evidence from the sectoral studies, which appears to stem partly from data problems. The first step would be to settle these data issues, and then to disaggregate the corporate and household sectors.

vi) the issue of parameter instability must be assessed further. Elasticity and speed of adjustment estimates suggest that parameter stability could be a real problem for financial modellers who plan to use data from both the 1960's and 1970's.

vii) more attention must be given to the dynamic simulation properties of financial models.

viii) the role of wealth must be investigated further. Simulation evidence from RBA1/74 suggests that the links between wealth and real activity are not strong. However the volatile behaviour of wealth in these experiments suggests that alternative ways of measuring and explaining the links between wealth and the real sector may be needed.