An EightQuarter
Forecast of
Gross
Private Domestic Investment
RODNEY W. DAVIS
College of Business
Western Carolina University
Abstract
The U.S. real Gross Private Domestic Investment
is projected to decrease by approximately 16% by yearend 2002.This
bleak outlook of investment expenditures could suggest a recession.Because
inflation is relatively low and unemployment is also below the natural
rate of 6%, it might be in the best interest of the Federal Reserve to
loosen the current monetary policy.The
basis of this forecast is the Keynesian aggregate investment function.This
forecast uses the 3month Tbill rate (secondary market) and the 20year
Treasury constant maturity rate as the explanatory variables.(JEL:
E22).
Part 1. Introduction
This paper forecast U.S. real gross private domestic
investment (GDPI) for the years 2001 and 2002.The
explanatory variables used are the 20year Treasury constant maturity rate
and the 3month Treasury bill interest rate.The
approach for this forecast paper is based on the Keynesian investment function.
The target of this forecast, real gross private
domestic investment, was selected because GPDI has a strong correlation
to U.S. economic growth, described as GDP.As
investment expenditures increase so does possible GDP.This
paper forecast eight quarters through yearend 2002.This
shorter time period enables a more precise forecast of gross private domestic
investment.
The remainder of the paper is organized as follows:part
2. presents the data used to forecast GPDI; part 3. explains the theoretical
basis for the approach adopted in forecasting GDPI; part 4. presents the
regressions and the forecasting of GPDI for years 2001 and 2002; part 5.
explains the significance of the forecasting as it relates to the economy;
and part 6. discusses the conclusions for monetary policy.
Part 2. Data
All variables are taken from the Federal Reserve
Bank of St. Louis Federal Reserve Economic Data (FRED).Gross
Private Domestic Investment is FRED variable GPDIC1 and is given in billions
of chained 1996 dollars at seasonally adjusted annual rates (SAAR).The
U.S. Department of Commerce Bureau of Economic Analysis is the primary
source of data.
The interest rate data given are the 3month Treasury
bill rate (secondary market) and the 20year Treasury constant maturity
rate.These are given as FRED variables
TB3MS and GS20.Both are given as
a percent discount.The data is taken
from a sample period beginning in 1994 and ending with yearend 2000.The
interest rate data are both monthly variables.Gross
private domestic investment, however, is a quarterly variable.The
value taken for each quarter is the value given for the first month of
each quarter (January, April, July, and October).The
primary source for this data is the Board of Governors, U.S. Federal Reserve
System.
The right hand side variables, the interest rates,
were lagged eight quarters to the GPDI.This
was done because the forecast target was two years into the future.A
regression analysis of this data was run.
Part 3. Forecasting with
the Keynesian Investment Function
Investment spending is very volatile and is subject
to change primarily based on interest rate change.Interest
rate data is used to predict the investment expenditures for the next two
years.The forecasting of real gross
private domestic investment spending will be based on the Keynesian Investment
Function which is:
I = f (i)
This is a simple linear model that will be used to
predict investment spending.Dollars
spent on investment is I.When
the short and long term interest rate variable are added the Keynesian
Investment Function becomes:
I = a + b*I_{(LR)} + c*I_{(SR)}
Part 4. An Eight Quarter Projection
of Real Gross Private Domestic Investment
The investment function was estimated with 1994.1
– 2000.1 was estimated with quarterly data.The
regression estimate is (tstatistics in parentheses):
I_{t} = 2277.43(5.57) – 245.67(4.71)i_{(LR)t2} + 155.23(2.80)i_{(SR)t2}
In this estimate I is investment spending, the short
term interest rate is the 90day Treasury Bill rate, and the long term
interest rate is the 20year Treasury constant maturity rate.
The adjusted Rsquare of the estimate was .4789,
which indicates that 47.89% of the forecast investment spending data can
be explained by the interest rate variables in a linear equation.The
Fstatistic for this regression is 13.41, which indicates a strong rejection
of the null hypothesis that the coefficients of the slope parameters equal
zero.All tstatistics have a significance
level less than 5%. This shows a strong correlation between interest rates
and investment spending.
The forecasting of GPDI using a fourquarter mean
for 2001 is 1490.61 and is 1619.32 for the year 2002.Investment
expenditures were forecast quarterly and are reported in table 1.All
numbers are reported in billions of chained 1996 dollars.
Table 1
Forecast Real Gross Private Domestic Investment 2001.1 – 2002.4

Quarter

Investment Forecast

2001.1

1612.24

2001.2

1513.58

2001.3

1440.93

2001.4

1395.69

2002.1

1417.97

2002.2

1637.80

2002.3

1679.46

2002.4

1742.05

Part 5. The Bears are Here!
This forecast shows a considerable decrease in investment
expenditures for the next two years. Whether or not this will spawn a recession
is undeterminable.The likelihood
of recession is higher based on the data showing a choking of investment
spending which translates into less capital available for real gross domestic
product.GDP being the primary measure
of growth, it would appear from this forecast to indicate a slowing of
growth in the U.S. economy.
This forecast, bases on interest rates, appears to
indicate a 19% decrease in investment spending from the year 2000 to the
yearend 2001.However, 2002 will
see an 8.5% increase in GPDI.Though
a twoyear average of expenditures for the years 2001 and 2002 will show
an overall decrease, this may only indicate a oneyear setback for 2001.It
appears investmentspending will began to grow again beginning with the
first quarter of 2002.
Part 6. Policy Conclusions
Because of the predicted large decrease in real GPDI
in 2001, it would be wise for the Federal Reserve to loosen the monetary
policy and lower interest rates in the short term.This
could stimulate investment and because unemployment has remained below
the natural rate (6%), this could also translate into growth.However
this forecast only predicts a decrease in GPDI for the year 2001.As
GPDI increases in 2002, the Federal Reserve should begin to raise interest
rates to control growth and maintain low inflation.
Another alternative to interest rate reduction for
economic stimulation is the proposed tax cut.This
could provide a boost to the economy within a time period of 6 months,
however the long term benefits and repercussions are undeterminable because
this cut is based on tenyear government budget projections.
Reference
Federal Reserve Bank of St. Louis, Federal Reserve
Economic Data (FRED), (February
25, 2001), http://www.stls.frb.org/fred/ (February
25, 2001).