View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

FRBSF

WEEKLY LETTER

Number 94-18, May 6, 1994

Just-In-Time Inventory Management:
Has It Made a Difference?
Inventory investment is a closely watched cyclical indicator, even though it is normally only a
small portion of real GOP. It is valuable as an indicator because it is so volatile in the short run
that it often accounts for much of the volatility in
real GOP. Indeed, since World War II, declines
in inventory investment have accounted for over
70 percent of declines in real GOP during recessions, on average. Therefore, an unusual buildup
of inventories may be viewed as a precursor of
an economic slowdown, as firms cut back production to eliminate unwanted inventories, while
a rapid inventory drawdown suggests just the opposite. This framework assumes that there is a
benchmark or "normal" level of inventories held
by businesses relative to sales or output, and that
an unusual deviation in the inventory-to-sales
ratio relative to its long-run value might be
viewed as indicating a change in the cyclical
phase.
Recently, the business press and management
literature have focused much attention on the
adoption of the "just-in-time" UIT) inventory
technology in the U.S. during the 1980s, which
lets firms operate with a significantly lower inventory-to-sales ratio. It is argued that the widespread adoption of JIT has lowered the normal
level of inventories in the

u.s.

This Letter examines whether there has been a
significant downward shift in the business sector's inventory holdings relative to sales in the
1980s, and finds that the evidence is not as clear
as some observers claim. In fact, the analysis
suggests that the relationship between inventory
investment and its key economic determinants
changed little during the 1980s.

The upside and downside of )IT
A number of case studies have described the adoption of jlT inventory management techniques by
firms, and two stories will serve to illustrate
its upside and its downside. On the upside, the
Journal of Purchasing and Materials Management (1987) reported the experiences of HarleyDavidson, the motorcycle firm. According to the

u.s.

article, the company was driven to the verge of
bankruptcy by its japanese competitors, who
practiced liT. Harley-Davidson lobbied for a
temporary trade barrier, and while the trade
protection was in place, it adopted jlT, which
contributed to its recovery and eventual re-emergence as a competitive business.
The downside risks of jlT are illustrated by the
experience of a California computer producer in
the mid-1980s. In this case, the firm's assembly
plant received parts on a weekly jlT basis from a
supplier in the Far East. But when one shipment of
parts turned out to be defective, the operations
of the entire plant came to a halt, and activity
throughout the chain of sales and distribution
was disrupted for a couple of weeks. To correct
the situation the company had to send executives
to the parts supplier and had a special shipment
of parts air-freighted (Ramey 1989). This example
illustrates that jlT does not necessarily always offer an unambiguous net benefit to all firms.

Measurement problems
The inventory-to-sales (IS) ratio is a commonly
observed measure of inventory behavior. The particular IS ratio considered here measures the
existing stock of inventories relative to sales of
the manufacturing and trade sectors. In considering the IS ratio, it is important to note that inventories in the manufacturing sector are made
up of goods in various stages of production, and
it is difficult to obtain a precise measurement of
inventories in the earlier stage. The inventories
consist of goods in three stages: materials and
supplies, work-in-progress, and finished goods.
As of the late 1980s, finished goods accounted
for less than 30 percent of the manufacturing sector's total inventory holdings, while materials and
supplies and work-in-progress made up about 40
and 30 percent, respectively.
The case of the automobile industry can be used
to illustrate the problem in obtaining good constant dollar estimates of inventories, which are
not affected by changes in the prices over time.
First, data become available on the nominal

FRBSF
values of the inventories in each stage of production. To arrive at the constant dollar values,
the proper price deflator must be applied to the
nominal values. This step is particularly difficult
for inventories in the earliest stages of produc~
tion-materials and supplies. For the auto industry these inventories consist of items such
as tires, metal frames, and computer chips. To
measure the constant dollar value of such diverse
items accurately, the price deflator must incorporate individual input factor prices, as well as their
share in the total stock of materials and supplies.
The necessary information on the mix of input
factors in producing a particular product is rather
limited. As a consequence, analysts use "inputoutput" tables, which show how much and what
type of input factors typically are required to produce a particular type of output.

New vs. old measures of the IS ratio
In early 1993 the Bureau of Economic Analysis
released a benchmark revision to the inventory
and sales data. The revision had an important effect on estimates of the IS ratio data, as illustrated
in Figure 1. The revised is ratio is measured in
1987 dollars, and it reflects the input-output table
as of 1982, whereas the pre-revision IS ratio was
based on the input-output table of 1967, and is
measured here in 1982 dollars. Both the new and
the old measures show similar cyclical patterns.
They rise markedly just before business cycle
peaks; they continue to rise during subsequent
contractions (shaded periods in Figure 1), and
then they tend to fall in expansions.

Figure 1
Manufacturing IS Ratios

1.: 1
1.8

During the second half of the 1980s, however, the
two measures exhibit different patterns with regard to the secular trend. The ratio measured in
1982 dollars shows a sharp decline in the late
1980s, suggesting shift in the levels of inventories. This would support the view that JIT is
playing a big role in inventory management.
However, a downward trend is less obvious in
the revised data.
Part of the explanation for the upward revision in
the IS ratio in the 1980s is that, since the revised
input-output table is more detailed, it better reflects the technology components in the production mix. Integrated computer chips, for example,
common in today's automobiles, were not used in
cars produced in the late 1960s. Furthermore, the
new IS estimates also reflect the significant price
declines in computing equipment during the
1980s. Under the old measure, the constant dollar value of the inventories of computer chips for
1987 were underestimated because they were deflated by a higher 1982 price.
Based on these considerations, the 1987 constant
dollar IS ratio is more reliable than the one measured in terms of th~ 1982 constant dollar. The IS
ratio measured in current dollars is a potentially
useful alternative. However, the amplitude of
fluctuations in the current dollar IS ratio around
recessions is very big and hence it is hard to discern a change in the long term trend.
The IS ratio examined so far is limited to the
manufacturing and trade sectors. For a more
complete analysis, we need to look at aggregate
inventory behavior. Also, a more systematic examination of the dynamic relationship between
inventory investment and related variables is
needed to see whether there has been a shift in
the responsiveness of inventories to changes
in their economic determinants.

Behavior of aggregate inventory investment

1.7
1.6
1.5
1.4
1.3
1.2
77

79

81

83

85

87

89

91

93

Is there a stable long-term relationship between
aggregate inventory investment and changes in
aggregate outputover time? Such a relationship
offers a frame of reference for comparing the pattern of aggregate inventory investment since
1980. Indeed, a simple relationship between inventory investment and output growth has been
observed for a number of years. Namely, the ratio
between the aggregate inventory investment and
the annual change in real CDP remains stable
over time (see, for example, Hall and Taylor
1991). Based on this relationship, it will be useful

to see whether the observations in the 1980s, the
period when JIT would have had the most impact, align with the general historical pattern
between inventory investment and changes in
output. Figure 2 presents the scatter plots of the
two variables where the observations from 1948
to 1980 are marked by triangles and the observations since 1980 are marked by squares. The
results do not support the contention that there
has been a significant downward shift in inventory due to an improvement in inventory management techniques, because the post-1980
observations, as a group, are not different from
the pattern seen in the observations for the earlier period.

Figure 2
Inventory Investment and Change in Real GDP

period would have been different from the first
one estimated using data from 1959-1980: However, the dynamic behavior of the two models in
response to an unforeseen change in output, or
in the interest rate, is not substantially different
across the two models.
Next, the dynamic simulations from the two
models were examined to see whether there was
a significant divergence. Both models were used
to forecast inventory investment for 1994 through
1996 conditioned on the same set of information.
The difference in the forecasts from the two models is quite small and statistically insignificant. in
other words, the model estimated from the sample that includes the additional10-year period
from the 1980s is not materially different from
the model estimated from the shorter sample.

Conclusion
100
80

-

60

l/I

40

c:

Q)

E
Q)

>
E

~

.8
c:

The relationship between aggregate inventory
investment and its key determinants does not appear to have shifted significantly since 1980. This
finding does not deny that there have been technological developments in the area of inventory
management, especially in manufacturing. However, the effect of any improvement in inventory
management is not discernible in either the manufacturing sector or the aggregate real inventory
investment data to this point.

/),1948-1979
.1980-1992

•
/),

/), l'~t.~/::"

20
0

~
c: -20

·40

/::,.

•

•

·60
-100

"!t
0

~

t:>

rio/),

•
•

100

Chan Huh
Economist
200

300

Change in Real GOP
This informal test is confirmed by econometric
analysis (of a vector autoregression), which models the change in real GOP, inventory investment
(both measured in 1987 dollars), consumer price
inflation, and the 6-month commercial paper
rate (the results are available from the author
upon request). The latter two variables are included to capture the effects of inventory financing costs. Two identically specified econometric
models were estimated for the periods 1959.Q21980.Q4 and 1959.Q2-1989.Q4. If there had
been a significant shift in the relationship during
the 1980s, the model covering the whole sample

References
Blinder, A.S., and L.J. Maccini. 1991. "Taking Stock: A
Critical Assessment of Recent Research on Inventories." Journal of Economic Perspectives (Winter).
Hall, R.E., and J.B. Taylor. 1991. Macroeconomics, 3rd
edition. W.W. Norton & Company.
Raira, Ernest. 1987. "Just-In-Time USA: Journey to
World Class:' Journal of Purchasing and Materials
Management.
Ramey, V.A.1989. "Inventories as Factors of Production and Economic Fluctuations:' American
Economic Review Oune}.
Survey of Current Business. 1993. U.S. Department of
Commerce Oune).

Opinions expressed in this newsletter do not necessarily reflect the views of the management of the Federal Reserve Bank of
San Francisco, or of the Board of Governors of the Federal Reserve System.
Editorial comments may be addressed to the editor or to the author.... Free copies of Federal Reserve publications can be
obtained from the Public Information Department, Federal Reserve Bank of San Francisco, P.O. Box 7702, San Francisco 94120.
Phone (415) 974-2246, Fax (415) 974-3341.

~ ~

·S)jU! ueaqAos L1l!M

'V '0' Jaded papkJaJ uo palulJd

01:Lt'6 V:J 'O:>SPUl'J:I Ul'S

WLL xog 'O"d

O)SPUOJ:J UOS

JO

~U08

aAJaSa~

IOJapa:J

~uaw~Jodaa

4)JOaSa8

Index to Recent Issues of FRBSF Weekly Letter

DATE NUMBER TITLE
11/5
11/12
11/19
11/26
12/3
12/17
12/31
1/7
1/14
1/21
1/28
2/4
2/11
2/18
2/25
3/4
3/11

3/18

3/25
4/1
4/8
4/15
4/21
4/29

93-38
93-39
93-40
93-41
93-42
93-43
93-44
94-01
94-02
94-03
94-04
94-05
94-06
94-07
94-08
94-09
94-10
94-11
94-12
94-13
94-14
94-15
94-16
94-17

Real Interest Rates
A Pacific Economic Bloc: Is There Such an Animal?
NAFTA and the Western Economy
Are World Incomes Converging?
Monetary Policy and Long-Term Real Interest Rates
Banks and Mutual Funds
inflation and Growth
Market Risk and Bank Capital: Part 1
Market Risk and Bank Capital: Part 2
The Real Effects of Exchange Rates
Banking Market Structure in the West
Is There a Cost to Having an Independent Central Bank?
. Stock Prices and Bank Lending Behavior in Japan
Taiwan at the Crossroads
1994 District Agricultural Outlook
Monetary Policy in the 1990s
The IPO Underpricing Puzzle
New Measures of the Work Force
Industry Effects: Stock Returns of Banks and Nonfinancial Firms
Monetary Policy in a Low Inflation Regime
Measuring the Gains from International Portfolio Diversification
Interstate Banking in the West
California Banks Playing Catch-up
California Recession and Recovery

AUTHOR
Trehan
Frankel/Wei
Schmidt/Sherwood-Call
Moreno
Cogley
Laderman
Motley
Levonian
Levonian
Throop
Laderman
Walsh
Kim/Moreno
Cheng.
Dean
Parry
Booth
Motley
Neuberger
Cogley
Kasa
Furlong
Furlong/Soller
Cromwell

The FRBSF Weekly Letter appears on an abbreviated schedule in June, July, August, and December.