The present paper is a progress report and will be only one
contribution to a series of Korean macro-models. The ancestors
of this model have been used to make a number of annual
forecasts, providing helpful tools in formulating the national
economic plans. Since it was out first experience in building a
quarterly model of the Korean economy, it will set the stage for
more work and further improvement to come.
The limitation of our quarterly data is one of the most
serious problems in developing short-term forecasting models.
Disaggregated production functions, for manufacturing, social
overhead and service industries, can be estimated when quarterly
dates on capital stock are available. Inventory might have to be
subdivided into agricultural and non-agricultural groups, and
imported raw materials and other imports. Consumption
expenditures might also be needed to disaggregate into durable,
non-durable, and service consumption expenditures.
Investment equations are especially unsatisfactory and we
need to develop better ways to estimate those equations.
Prediction of the inflow of long-term foreign loans is also very
difficult, yet it is a very significant determinant of fixed capital
formation over the sample period. The inflow of long-term
foreign loans was depended largely upon government policy
decisions and international economic and political conditions.
However, it should be eventually treated as endogenous variables
because inflow of foreign loan is determined by investment
demand and availability of domestic resources.
It would be desirable to introduce the supply of money
function, since the role of monetary policy is very important in
economic fluctuations, growth, stabilization. The exogenous
treatment of the government sector (government consumption,
investments and loan, and tax revenues), is not desirable. It will
be sensible to distinguish between induced government
expenditures and autonomous categories. Then some equations
also can be developed for the induced part of the expenditures.
All these improvements require substantial research work,
but they are all feasible and can be added to the basic system