Several points should be made about the nature of aggregate (and individual
firm) investment behavior implicit in equations (4) and (5). First, high profitability
in any given period, by generating increases in firm equity levels (for
non-credit-constrained firms) and increased cash flow (for credit-constrained
firms), will lead to increased future investment. Thus, the model suggests the
kind of significant relationship between current operating cash flow and investment
found by Hubbard, Fazzari, and Petersen (1988), among others.
Also, if high profitability in any period is related to increases in demand in
that period, the model will exhibit the kind of accelerator behavior that has
been so successful in explaining actual investment behavior.18 The model can
be usefully thought of, therefore, as providing a microeconomic rationale for
both the cash flow and accelerator aspects of investment behavior that appear
to play such a significant role in practice.19 Second, firms wishing to borrow
pay firm-specific rates of interest, not some average rate on all assets. With
imperfectly informed lenders, changes in general market rates do not necessarily
lead to changes in the rates charged to borrowing firms.20 Some part of
the shift in loan supply is absorbed by increased credit rationing since charging
higher interest rates has an adverse effect on the quality and riskiness of
the borrower pool (see below). Thus, the rate, rt (and the associated expected
return to lenders ft), which enters the investment model above may vary significantly
less than widely observed market rates (such as Treasury-bill rates),
which would be available for use in any empirically estimated investment
equation. Second, the impact of interest rates tends to be small relative to the
impact of changes in a firm's financial position,21 and real interest-rate series
have, until the very recent past, been observably quite stable. Thus, the variability
in the financial positions of firms and the perceived riskiness of the
environment they face over the business cycle can be responsible for a far
greater share of the variation in investment over time than market interest
rates. For both reasons, the model provides an explanation for the relatively
small and elusive role that interest rates play in empirical investment equations
and suggests that interest rates themselves do not play a primary role in
macroeconomic stabilization.22
Finally, as will be noted extensively in later sections of this paper, investment,
although defined for explanatory purposes as investment in physical
capital of the usual sort, need not and should not be interpreted so narrowly.
Part of investment takes the form of working capital and the hiring and training
of workers, and a rise in the cost of investment (because, e.g., of a deterioration
in a firm's equity position) will be reflected as a reduction in working
capital, in
