Content uploaded by Paul F. Cwik
Author content
All content in this area was uploaded by Paul F. Cwik on Oct 06, 2016
Content may be subject to copyright.
ISSN 1801-0938
New Perspectives on Political Economy
Volume 1, Number 1, 2005, pp. 1 – 37
The Inverted Yield Curve and the Economic Downturn
Paul F. Cwik, Ph. D.∗
JEL Classification: B53 – Austrian Economics, E22 – Capital and Investment, E32 –
Business Cycles, G19 – Yield Curves
Abstract: This paper presents an answer to why the yield curve tends to invert one year
before a recession. The capital-based macroeconomic model used in this paper traces out
the effects of an injection of short-term working capital into the model. There are two
consequences of this injection: the Wicksell effect and the Fisher effect. The Wicksell
effect entails the downward pressure on interest rates, while the Fisher effect entails the
upward pressure on interest rates. The short-term credit can create both short- and long-
term malinvestments in the social structure of production. These malinvestments are
unsustainable and must be liquidated. The process of liquidation phase may take the form
of a credit crunch, a real resource crunch, or a combination of the two. Each scenario
culminates in an inverted yield curve approximately one year before the upper-turning
point of a recession.
∗Adjunct Scholar of the Foundation for Economic Education and of the Ludwig von Mises Institute
2 New Perspectives on Political Economy
1Overview
This paper addresses the question of why the yield curve tends to invert before a re-
cession. It does not create a model to demonstrate that such a phenomenon exists, as
this relationship has already been well established. This paper uses the capital-based
macroeconomic approach set forth by Garrison (2001) to explain that a correlation exists
between the yield curve’s spread and real output. Accordingly, the topic is examined by
disaggregating investment and capital-formation decisions. The capital-based approach of
macroeconomic theory is well-suited for the examination of this question, since it is a
theory of the upper-turning point of a business cycle.1
Macroeconomic theories attribute economic downturns to either monetary or real
factors. The capital-based approach allows for both. A disaggregated approach allows
for analysis and insights that other theories cannot provide. Unlike the capital-based ap-
proach, most macroeconomic theories that examine the upper-turning point focus on the
immediate causes of the downturn. They do not include the underlying capital structure
as a part of the theory, because this structure is viewed as an unnecessary complication to
the theory. By ruling out capital (and the malinvestments that could be built up during
the expansionary phase), the leading macroeconomic theories focus on more aggregated
causes–such as monetary or real shocks to the economy.
These models are too aggregated to properly answer the question of why the yield
curve tends to invert before a recession. Prior to the 1990-91 recession, several economists
called attention to the past performance of the spread as a predictor of a business cycle’s
upper-turning point. However many dismissed the signal, declaring it may be a false pos-
itive.2Another had argued that there would not be a recession in 1989 or 1990 because
there was “an absence of the kind of gross imbalances in the economy that have typically
preceded past recessions.”3It is possible in retrospect to see that the imbalances were in
the economy and were liquidated in the 1990-91 recession.4The current approach, how-
1See Hayek (1969) where he states,“This theory [the Austrian Business Cycle Theory or ABCT] never
claimed to do more than account for the upper turning point of the typical nineteenth-century business
cycle.” p. 282.
2See Brown and Goodman (1991) and Estrella and Mishkin (1998).
3See Keen (1989) p. 40.
4See, for example, Hughes (1997) for an empirical analysis supporting this claim.
Cwik: The Inverted Yield Curve and the Economic Downturn 3
ever, lacks the ability to see the imbalances (malinvestments) created during the “boom”
phase. It is here that Austrian theory can provide insight. This approach allows one to
analyze and draw conclusions about the problem in a manner that is superior to more
aggregated methods.
The remainder of this paper is divided into six sections. The next section reviews the
empirical relationship and surveys the current models that attempt to explain the rela-
tionship. Section 3 presents the implication of monetary expansion. Section 4 continues
the analysis by showing that monetary injections lead to a malinvestment boom. Sec-
tion 5 establishes how such malinvestments are not sustainable and inevitably lead to the
“crunch phase” of the business cycle and then demonstrates why the yield curve inverts
during the crunch phase (prior to the upper-turning point). Section 6 summarizes and
concludes.
2Presentation of the Relationship and Current Research
Economists, government officials, and businessmen have long searched for accurate busi-
ness cycle indicators. One strong predictor of the upper-turning point of a business cycle
is the inverted yield curve. Chart 1 illustrates the 10-year Treasury Bond and the 1-year
T-Bill spread and the 10-year T-Bond and the 3-month T-Bill spread between April 1953
and October 2003.5(Please see Chart 1 below.)
5The NBER dating of the recessions is used. All data for this paper were obtained from the Federal Reserve
Bank of St. Louis’ FRED II.
4 New Perspectives on Political Economy
Chart 1: Yield Curve Spreads Between 1953-2005
-5,00
-4,00
-3,00
-2,00
-1,00
0,00
1,00
2,00
3,00
4,00
5,00
1953-04
1955-04
1957-04
1959-04
1961-04
1963-04
1965-04
1967-04
1969-04
1971-04
1973-04
1975-04
1977-04
1979-04
1981-04
1983-04
1985-04
1987-04
1989-04
1991-04
1993-04
1995-04
1997-04
1999-04
2001-04
2003-04
Recession
Spread 10 year - 1 year
Spread 10 year - 3 month
Recessions are dated according to the NBER.
The data for int erest rates were obtained from FRED II.
Cwik: The Inverted Yield Curve and the Economic Downturn 5
An inverted or humped yield curve has occurred no more than 5 quarters before ev-
ery recession since the mid-1950s. Except for the Q3:1990-Q1:1991 recession, the yield
curve has inverted in every recession since the mid-1960s. Prior to the Q3:1957-Q2:1958,
Q2:1960-Q1:1961 and Q3:1990-Q1:1991 recessions, the 10-year/3-month spread did not
become negative. The lowest points for this spread were 0.24% in February 1957, 0.20%
in December 1959 and 0.13% in August 1989. Preceding these recessions, the yield curve
was technically humped and not inverted. The 10-year/1-year spread was negative in
December 1956 and from February through April 1957 and, according to McCulloch
(1990), the 15-year/6-month spread (not shown in the chart) was negative from Novem-
ber 1956 through March 1957. The 10-year/1-year spread was also negative in the period
of September 1959 through February 1960 and February through September 1989.
There is one instance where an inverted yield curve was not followed by a recession.
From September 1966 through January 1967 the yield curve inverted, but no recession
took place. Some refer to this occurrence as a false positive, but the second quarter of
1967 did experience a negative growth rate of -0.06% (real GDP). While this decline in
real output did not constitute an official recession, it does confirm the relationship under
study.6
The historical record does not show this connection to be only a post-WWII phe-
nomenon. The yield curve inverted between June 1920 and March 1921 and again be-
tween January 1928 and November 1929.7Data from the 19th century are incomplete
and do not easily lend themselves to analysis.8Nevertheless, support for the thesis of the
yield curve as a predictor of business cycles can be traced as far back as the mid-1800s.9
The current research can be separated into two basic models: the consumption-based
capital asset pricing model (CCAPM) and the Estrella models. The essential idea of
the CCAPM is that investor’s smooth income across business cycles. Criticism of the
CCAPM has led to the development of an alternative theoretical model.
6The true exception to this relationship is the Q2:1953-Q2:1954 recession, where the yield curve flattened
but did not invert.
7Cecchetti (1987) demonstrates that the observed bond market data from the 1930s and 1940s is distorted
due to heavy government intervention. As an illustration of the degree of distortion, on December 31,
1932, institutional forces caused the 31
2% US Liberty Bond’s yield to fall to a nominal rate of –1.74%.
8Davis (1971) examined U.S. capital markets from 1820 through 1930. He shows these markets were not
integrated until after World War I.
9See Keen (1989).
6 New Perspectives on Political Economy
The second type of model builds on variations of the following economic tools: the
Expectations Hypothesis, the Phillips curve, the IS curve, a monetary reaction function,
and the Fisher Equation. This category of models is motivated by a monetary shock.
The origin of these models is Estrella (1998). The Estrella model is also inadequate for
understanding why the yield curve tends to invert before a recession.10 The Estrella
models derive the relation between interest rates and real output from the Phillips curve
and the IS curve. The Phillips curve (an empirical and not a theoretical relationship)
fails to explain the connection between interest rates and real output. It only shows
that a connection exists between nominal rates and output. However, the theoretical
underpinnings needed to understand the relationship are not explained. Furthermore,
the use of the Keynesian IS curve is insufficient to create a credible model.11
Over the past 15 years, the debate on theory remains unresolved. No article during
this period examines the effects of non-neutral monetary injections through a hetero-
geneous capital structure. The use of Austrian insights can provide an alternate (and
fruitful) perspective to this debate.
3Monetary Expansion
The capital-based approach posits that the initial disequilibrium of the business cycle is
caused by monetary injections.12 Credit is injected at the short-end of the yield curve and
spreads through the economy causing non-neutral effects. The effect of the new credit
is separated into the Wicksell effect and Fisher effect. These opposing effects distort
the ability of price signals to transmit relative scarcities to entrepreneurs. As a result,
monetary expansion lowers and alters interest rates that falsely signal entrepreneurs to
embark upon malinvestments.
The effects of monetary expansion are traced through the yield curve, which was de-
veloped in Cwik (2004), Chapter 3. The modified Preferred-Habitat Theory of the yield
curve is a combination of time-preference (in the Böhm-Bawerkian sense), expectations,
10 See Cwik (2004) Chapter 2.
11 Ibid.
12 Many factors can cause an economic downturn–war, sweeping changes in institutions, radical changes in
monetary policy, etc., but these are outside of the scope of this topic. This paper specifically focuses on
downturns caused by monetary injections.
Cwik: The Inverted Yield Curve and the Economic Downturn 7
liquidity-preference, and risk aversion (the preference for matching debt and equity).
Böhm-Bawerk’s analysis is the basis for the formation of interest rates, since it satis-
fies both the essentialist and functionalist questions regarding interest. After an initial
interest rate is established, a yield curve can be derived by adding expectations, liquidity-
preference, and risk aversion to the analysis. As shown in Cwik (2004), the modified
Preferred-Habitat Theory is consistent with the empirical observations of the yield curve.
When the monetary authority engages in a policy of monetary expansion, the new
money is injected into the monetary system at specific points.13 The effect of additional
liquidity is sometimes called the Wicksell effect.14 The Fisher effect is the change in
interest rates caused by changes in the expectations of future inflation.15 The Wicksell
effect and the Fisher effect are opposing forces. The Wicksell effect tends to lower interest
rates while the Fisher effect tends to raise them.
With a policy of monetary expansion, the Wicksell effect first dominates interest
rate movements. As money is injected into the short end of the yield curve (through
the monetary base and thus the Fed funds rate) an initial lowering of short rates and a
steepening of the slope of the yield curve results. Keeler (2002) states,
The liquidity effect of a monetary shock will lower interest rates in general and
lower short-term rates relative to long-term rates. The yield curve will shift down
and become steeper in slope. . . . 16
Although Keeler is correct about the steepening of the yield curve, empirical observation
does not support the tendency of the yield curve to shift, as long rates tend to remain
stable relative to short rates. Bernanke and Blinder (1992) argue that the short rates move
13 The Federal Open Market Committee typically adjusts monetary policy through the use of open market
operations and the use of the discount rate which change the aggregate level of depository institutions’
reserves. Changes in these reserves induce changes to the Fed funds rate. The Fed funds rate is the
interbank interest rate for short-term loans, usually overnight. See Miller and VanHoose (2001).
14 The phrase “Wicksell Effect” was first used in the Cambridge Capital Controversy of the 1960s. The
phrase was divided into a “Real Wicksell Effect” and a “Price Wicksell Effect,” describing the change
in the relationship between the rate of profit and capital intensity in real or value terms. The phrase
“Wicksell Effect” used in this paper refers to an “Interest Wicksell Effect” (or a liquidity effect) where
money added to an economic system, by expanding the supply of investable funds, initially reduces the
market rate of interest.
15 See Mishkin (2001) pp. 107-108.
16 Keeler (2002) p. 5. See also Keeler (2001) pp. 333 and 335.
8 New Perspectives on Political Economy
while the long rates remain stable. The Fisher effect increases the forward short rates,
thus applying upward pressure to long rates. However, the new money is arbitraged
across the term structure. The Wicksell effect prevents the long rates from rising. Thus
the yield curve rotates instead of shifting, as shown in Figure 1. The new yield curve is
presented as the dashed curve.
Wicksell Effect
Interest Rate
Figure 1: The Wicksell and Fisher Effects Combined
Maturity
Fisher Effect
According to Keeler (2002), the steepening of the yield curve begins during the expan-
sion phase of the business cycle. However, the yield curve is steepest at the lower turn-
ing point of the business cycle.17 These observations are not inconsistent with Keeler’s
observations if the recovery phase of the business cycle is also included as part of the
expansionary phase of the next cycle.
In sum, the analysis begins with the Böhm-Bawerkian framework to establish an
initial interest rate. Expectations, liquidity-preference and risk aversion are added to
create a modified Preferred Habitat theory of the yield curve. With monetary expansion,
the Wicksell effect shows the lowering of short rates due to an increase in the supply of
investable funds. The long rates tend to remain stable relative to the short rates due to
the interaction of the Wicksell and Fisher effects.
Since new money does not enter an economic system uniformly or at a steady rate,
the already difficult job of entrepreneurs, to correctly read market signals, becomes even
17 This empirical observation was made as early as Kessel (1965) and has since been seen as a consistent
pattern of the yield curve.
Cwik: The Inverted Yield Curve and the Economic Downturn 9
more difficult.18 Entrepreneurs need to correctly read these market signals to make prof-
its, and as a consequence, coordinate the economy.19 Since the price changes resulting
from an increase in the money supply are not uniform, entrepreneurs have difficulty de-
termining whether the price change is a result of a change in relative scarcity or whether
the price change is the result of inflationary pressures. In other words, they are not able
to distinguish between relative price changes and inflationary price changes. As a result,
the economy becomes more wasteful and less efficient.
This paper posits that the monetary authority injects money into the economy
through short-term credit markets. The addition of credit lowers short rates, while the
Fisher effect should increase long rates. However, the yield curve steepens and does not
shift. The Wicksell effect is transmitted across the term structure of interest through
the process of arbitrage and reduces the Fisher effect on long rates. Thus, monetary
injections artificially lower interest rates across the entire yield curve. These false rates
signal to entrepreneurs that consumers have shortened their time-preferences, leading to
a malinvestment boom.
4Malinvestment “Boom”
In the previous section, the idea of malinvestments was introduced. This section exam-
ines the nature of these malinvestments in the context of a capital-based macroeconomic
approach. The crisis stems from the need to liquidate the malinvestments that are built
during the boom. During this crisis, which is the upper-turning point of a business cycle,
the yield curve inverts as a consequence of the liquidation process.
To clarify terminology, a distinction must be made between an individual project’s
period of production and the social period of production. Schmitz (2003) distinguishes
between the individual and social periods of production. The individual period of pro-
duction corresponds to the length of time that an entrepreneur’s project will take until it
yields output for the next stage of production. These projects are distributed throughout
the structure of production. The social period of production corresponds to the degree of
18 Prices are packets of information that signal to entrepreneurs the relative scarcities of the various goods
and services throughout the economy. See Hayek (1945).
19 See Mises (1980).
10 New Perspectives on Political Economy
complexity of an economy. In other words, the overall degree of economy-wide round-
aboutness is the social period of production. Individual projects’ are divided into long
and short terms and correspond to the long and short rates of the yield curve.20 Both
long and short-term projects are found at every stage in the structure of production.
Furthermore, capital is divided into two forms: working capital and fixed capital.
Working capital, also known as circulating capital, refers to the funds that flow through
the structure of production. Fixed capital is the capital equipment, buildings, machines,
etc. that do not flow through the structure of production. Instead, fixed capital is embed-
ded at the different stages within the structure of production. Through the investment
process, working capital is used to purchase inputs such as labor and goods-in-process;
additionally working capital is also transformed into fixed capital.
Expansion of Short-Term Working Capital
This paper seeks to demonstrate that monetary injections (in the form of working capital)
into an economic system necessarily lead to an inverted yield curve prior to an economic
downturn. As a result, this paper assumes the extreme case where monetary expansion
initially takes the form of working capital in the short-term credit markets.21 As this
assumption is relaxed, the argument is strengthened.
As previously demonstrated, the monetary injections during an economic boom (and
also during a recovery) cause the yield curve to steepen.22 Short rates fall, while long rates
tend to remain relatively stable. With a monetary injection at the short end of the yield
curve, the modified Preferred-Habitat theory suggests that the yield curve should shift
down instead of rotate. However, the empirical evidence shows that yield curve rotates
and steepens, but shifts very little. This seeming inconsistency with the theory is due to
the Wicksell effect explored in the previous section. The impact of the Wicksell effect on
long-term and early stage malinvestments is discussed below.
20 A long-term project may be financed through a series of short-term loans. A simplifying assumption of a
hard link between the length of the project and the duration of the loan will not change the analysis.
21 This paper is additionally assuming that the traditional role of thrift institutions of “transforming matu-
rities,” by borrowing short and lending long, does not take place.
22 Cwik (2004), Chapter 3, section 3.6 provides empirical evidence that expansion ary monetary policy is
the typical case during the boom and the recovery phases of the business cycle.
Cwik: The Inverted Yield Curve and the Economic Downturn 11
The monetary injections, which rotate the yield curve, send opposite signals to en-
trepreneurs and consumers. As new short-term working capital is injected into the econ-
omy, the economy embarks upon a malinvestment “boom.” As the cost of borrowing
decreases, the marginal borrowers (those previously excluded from the market) will now
be able to obtain the wherewithal to fund their individual projects. Real resources are
transferred to these borrowers and the working capital is converted into fixed capital as
distinct production processes are added to the economy. Machlup (1932) illustrates this
process,
The fresh borrowers employ the fresh capital–either for a new enterprise or for
the expansion of an already existing one–by demanding means of production, partly
original factors of production, partly intermediate goods. This increased demand
will raise the price of production goods. Therefore the borrowers who are in the
best position to compete are those who are less affected by the increased cost of
intermediate goods than by the lowering of the rate of interest. This is not the
case with investment in raw materials and goods in process, but it is the case with
investments in fixed capital since in calculating the prospects of such investments
the interest rate is of much greater importance than the price of the goods used.23
(italics in the original)
After debating with Haberler, Machlup demonstrates “that the investment of fresh capital
for an increase of production and output which might be technically possible without
expanding fixed capital, is economically impossible.”24 Machlup’s point is that in order
to achieve an expansion of output, working capital must be transformed into fixed capital.
In a later work, Machlup (1935) reinforces his conclusion that a decrease in interest rates
leads to the formation of new investment in fixed capital.25 While Machlup argues that
the short-term funds will eventually be transformed into longer-term fixed capital, at this
part of the analysis, the point to be emphasized is that working capital is transformed
23 Machlup (1932) pp. 276-277. With expansionary monetary policy and an increase in output, Machlup
concludes that, even with additional short-term funds, “the short-term use of capital is theoretically
impossible.” p. 277.
24 Machlup (1932) pp. 277.
25 Machlup (1935) states, “As a cost factor, the interest rate has real significance only as it applies to new
investment in fixed equipment.” p. 462. (italics in the original) “A decrease in the interest rate changes the
comparative cost-calculation in favor of those methods of production which make the heavier demands
on capital.” p. 462.
12 New Perspectives on Political Economy
into fixed capital. These short-term projects are malinvestments and must be liquidated
at a future date unless additional real savings are supplied.
In the meantime, the short-term projects in the late-stages of production (those stages
closest to the consumers) are justified through increased profitability due to the increase
in the demand for consumer goods. With a decline in short rates, the cost of financing
short-term consumer purchases (through the use of credit instruments such as credit
cards) falls. Thus, an immediate result of the monetary injection is an increase in the
demand for consumer goods. The effect is seen in Figure 2.
Expansion of
Consumer Goods
(Output)
Value of Goods
Production Process:
Early Stages
Middle Stages
Late Stages
Time
Figure 2: The First Effect from Additional Working Capital
The dashed line in Figure 2 represents the additional projects accumulating toward
the late stages of production.
By restricting monetary injections to the addition of working capital in the short-
term credit markets, the analysis leads to the conclusion that short-term projects at the
late stages of production are built up. Machlup demonstrates that, over time, the working
capital will be transformed into fixed capital. The fixed capital is combined with addi-
tional inputs to create consumer goods. These consumer goods are purchased with the
new credit extended to consumers.
Cwik: The Inverted Yield Curve and the Economic Downturn 13
Long-Term Malinvestment
In this analysis, monetary injections have been restricted to the form of additions of
working capital to the short end of the yield curve. As this working capital is transformed
into distinct production processes to supply consumer goods, a portion of the working
capital is applied to the purchase of inputs while the remainder is transformed into fixed
capital. While the analysis has focused on an expansion of the late-stages of production,
empirical observation of a boom is that early-stage markets experience larger swings rel-
ative to intermediate and consumer goods markets.26 Such empirical evidence begs two
particular questions: “If short-term rates fall relative to long rates thus increasing the
amount of short-term projects, where is the long-term malinvestment?” and “Where is
the early-stage malinvestment?”27 To find answers the analysis will use Bastiat’s “unseen.”
When the monetary authority engages in monetary expansion, entrepreneurs in-
crease their expectation of future inflation. Within the model of the modified Preferred-
Habitat theory, the middle and long rates should rise in accordance with the Fisher effect,
but empirically they do not. As previously observed, short rates fall relative to the long
rates and the long rates tend to remain stable.
Nevertheless, long-term malinvestments emerge from the injection of short-term
working capital. The arbitrage process from the shorter rates prevents the long rates
from rising. In other words, credit is flooding into the long bond markets, keeping their
yield from rising. The Wicksell effect counters the Fisher effect. The “unseen fact” is that
many long-term projects would have been curtailed with an increase in long-term rates,
but the Wicksell effect discourages their liquidation. These non-liquidated projects are
now in a state of disequilibrium. They, too, are malinvestments. The degree to which the
Wicksell effect inhibits long-term rates from rising corresponds with the extent to which
of long-term malinvestment.
Machlup (1935) presents the rate of interest as a cost and capitalization factor in the
production process. As interest rates (the short-term rates in particular) decline, the capi-
26 See Skousen (1990) pp. 303-305.
27 Long-term malinvestment and early-stage malinvestment are not the same. Long-term malinvestment
refers to individual projects with a long-term planning horizon. Such projects may be found at any stage
in the structure of production. Early-stage malinvestment refers to the projects at the higher-order stages
of production. They may employ both long and short-term individual projects.
14 New Perspectives on Political Economy
tal values of all fixed capital increases. The capitalization effect yields a greater return for
the longer-lived capital equipment.28 Such a change in the relative value of long-lived fixed
capital encourages the expansion of long-term investments despite the increase in only
short-term working capital. Thus, not only are some projects that should be liquidated
not discontinued, but new long-term projects are started. As the newly expanded short-
term malinvestments are added to the long-term malinvestments, the economy moves
beyond the production possibility frontier curve–it has an unsustainable capital struc-
ture.
The argument so far is that the monetary authority has injected working capital into
short-term credit markets. The addition of credit lowers the short rates, the yield curve
steepens, and the arbitrage process prevents increases in the long-term rates; the Wicksell
effect counters the Fisher effect. Short-term projects (malinvestments) are commenced
due to the lowering of the short rates. These short-term projects embed various degrees of
fixed capital into the structure of production. Up to this point of the analysis, the short-
term projects have been located toward the late-stages of production, but the addition of
short-term projects is not necessarily an addition of projects to a particular stage. Short-
term projects may be added throughout the structure of production.
Simultaneously, the rotation and steepening of the yield curve is evidence that long
rates are artificially held down. The prevention of increasing long rates delays the liqui-
dation of some long-term projects, and new long-term projects are started because of the
relative change in the return of long-lived fixed capital. These long-term projects are not
supported by a foundation of real savings and will need to be liquidated at some future
date. These long-term projects are malinvestments. Thus malinvestments (in both long
and short-terms) emerge throughout the structure of production.
Despite demonstrating that long- and short-term malinvestments arise from mone-
tary injections, one may be led to ask the second question posed above, “Where is the
early-stage malinvestment?” While a hard link between long-term investments and the
early-stages in the social structure of production cannot be technically proven, it is rea-
sonable to assume that such an association exists. The case can be argued that individual
long-term investment projects necessarily affect the social period of production to the
28 See Machlup (1935) p. 465.
Cwik: The Inverted Yield Curve and the Economic Downturn 15
extent these long-term projects prompt capital lengthening. Nevertheless, to avoid using
a conjecture that cannot be proven, this paper adopts the extreme case that long- and
short-term projects may occur at any stage in the production process. To the extent that
long-term investments lengthen the social period of production, the overall argument of
this paper is bolstered.
Early-Stage Malinvestment
Despite all of the stringent assumptions about the monetary authority injecting working
capital into the short end of the yield curve and that long and short-term projects may
occur at any stage in the structure of production, a malinvestment boom in the early
stages can be demonstrated.
In his seminal paper, Hayek (1945) demonstrates that entrepreneurs have only price
signals to guide them to meet consumers’ demands and make profits. Prices are informa-
tion packets that not only signal to entrepreneurs the quantity and quality of the goods
they are to produce, but also allow entrepreneurs to calculate which types of inputs and
production processes are the most efficient. It is in this manner that the economy is co-
ordinated. A network of prices ties the structure of production together. For a single
interest rate model in equilibrium, the rate of interest equals the rate of profit. When the
model is expanded to include a term structure of interest rates, the same principle applies,
but the rate of profit for each individual project corresponds to the matching instrument
in the yield curve. For example, a two-year project’s rate of return should correspond
to the yield of a two-year bond.29,30 Thus when the rates across the yield curve fall (or
are held down by the Wicksell effect), the cumulative effect is a change in the social pe-
riod of production. The decrease in the short-term interest rates and the artificially low
long rates signal to the entrepreneurs that the normal rate of economic profit has been
29 The modified Preferred-Habitat Theory is able to accommodate the segmentation of the yield curve.
30 While long-term projects may be financed through a series of short-term loans, the entrepreneur will
use the maturity that matches the individual project as the relevant yield. With a positively sloped yield
curve, the yield of rolling over short-term instruments is below that of the longer maturity instrument.
However, the entrepreneur will not view the short-term instruments as a relevant substitute for the
project. Instead, if he is looking to engage in a long-term project, he will look to the yield of the longer-
term instrument as the opportunity cost of such an investment. For example, if the yield of a one-year
bond is 4% and the yield of a two-year bond is 5%, the entrepreneur will regard the 5% yield as the
opportunity cost of embarking upon a two-year project.
16 New Perspectives on Political Economy
lowered.
To illustrate this process, the analysis begins with a single interest rate model. Suppose
that all individual projects have a length of 3 months and there is a corresponding single
interest rate for 3-month instruments. The monetary injections falsify the price signals
to the entrepreneurs.31 The effect of the additional credit lowers the interest rate and also
lowers the normal rate of return for these projects. In other words, the opportunity cost
of each project is lowered.
As the rate of interest changes, so does the rate of return necessary to obtain nor-
mal economic profits.32 As the monetary injections lower the interest rate, two effects
emerge. The first effect is that consumers, experiencing a decrease in their return on their
savings, shift their wealth into consumer goods. Garrison (2001) refers to this situation
as over-consumption.33 As consumers dedicate less resources to their savings, retailers
face an increase in the consumers’ demand curves in their markets. As a result, retail-
ers increase their demand of wholesaler goods in order to take advantage of the profit
opportunities. The cumulative effect of the entrepreneurs’ actions at the late-stages of
production is to reduce the degree of roundaboutness in the economy. The effect is illus-
trated in Figure 2 above.
The second effect is that the injected money is lent simultaneously to entrepreneurs,
thereby increasing the amount of investment throughout the structure of production. In
this phase, the amount of investment is no longer equal to the amount of savings. A
“tug-of-war,” to use Garrison’s phrase, arises between saving and investment. Garrison
argues that “the conflict is resolved initially in favor of investment spending–because the
investment community has more to pull with, namely, the new money that was lent into
existence at an attractive rate of interest.”34 As a result of this conflict, the economy is
pulled in the direction of more roundabout production processes by the investors and
consumers pull the economy in the direction of less roundabout production processes.
31 The importance of Hayek’s observation is that entrepreneurs only have price signals to guide them in
their conduct. Evans (1987) argues that a “mis-assessment of risk” by investors can occur when the “true”
risk structure of the economy is uncertain. Since the true risk structure is never precisely known, en-
trepreneurial error can occur even under the best conditions. When the monetary authorities manipulate
price and interest rate signal, these errors are intensified.
32 The following analysis follows the analysis of Cwik (1998).
33 See Garrison (2001) pp. 69-70.
34 Ibid. p. 71.
Cwik: The Inverted Yield Curve and the Economic Downturn 17
A “dueling production structure”35 emerges. Figure 3 illustrates this result.
Expansion of
Consumer Goods
(Output)
Value of Goods
Production Process:
Early Stages
Middle Stages
Late Stages
Time
Figure 3: The Dueling Production Structure
Expansion of Early
Stages
Garrison is able to achieve this result by arguing that lower interest rates make longer-
term investments more profitable. To the extent that longer-term investments are capital
deepening, this result is correct. However, there is another, more fundamental reason
why the economy becomes more roundabout: as the normal rate of profit falls, the effect
of the decrease in the interest rate is compounded through the structure of production
(via relative price changes) and yields its largest impact at the earliest stages of production.
With an increase in investable funds, the normal rate of profit for all businesses de-
creases. As firms react to compete for the new profits in the late-stages of production,
they bid up input prices until they establish this new rate of profit. The cumulative ef-
fect of bidding up input prices creates windfall returns for the firms in the early stages.
These economic profits attract new investment into the early stages creating the dueling
production structure.
To illustrate this idea, suppose that there is a simple production process through
which there are four stages. Each stage is 3-months long and the initial rate of inter-
est is 10%. As a starting point, assume that the initial price of inputs is $100. Under
equilibrium, each stage meets the normal rate of profit of 10%. Using the discounted
35 This phrase was coined by John Cochran (2001) p. 19.
18 New Perspectives on Political Economy
present value formula, the price of the final output (one year later) is $146.10. Figure 4
demonstrates this relationship.
Consumer Goods
(Output)
Value of Goods
Figure 4: Prices for Each Stage Production
Production Process:
0
3-months 6-months 9-months 12-months
$100 $110 $121 $133.10 $146.41
Prices at each stage:
Holding, for the moment, the price of the final output constant, when the rate of
interest falls to 8%, the price of 9-month goods will be bid up to $135.56, thus yielding
an economic profit of 1.85% to those firms operating at the wholesale stage of the pro-
duction process. Table 1 shows the economic profit is highest for the owners of natural
resources.
Table 1: The Effects of Relative Price Changes on Economic Profit
Stage: 0-months 3-months 6-months 9-months 12-months
(Natural (Final
Resource Output)
Owner)
Prices with 10% $100 $110 $121 $133.10 $146.41
interest rate.
Prices with 8% $107.61 $116.22 $125.52 $135.56 $146.41
interest rate.
Rate of 7.61% 5.66% 3.74% 1.85% 0%
Economic Profit
Cwik: The Inverted Yield Curve and the Economic Downturn 19
The rate of economic profit is expressed in Equation 1.
( )
( )
1Equation 100.1
r1
r1
Profit Economic of Rate
2
1∗
−
−
+
−
+
=xn
xn
tt
tt
where r1is the initial interest rate,
r2is the interest rate after the monetary injection,
tn is the maximum number of stages in the production process, and
tx is the stage under examination.
Equation 1 shows that rate of economic profit is larger for the earlier-stages of pro-
duction. This process demonstrates how the early-stage markets are able to expand while
consumer market’s demand curves are rising. While it may look as though the expan-
sion of the early-stages depends upon longer-term investment (through the use of the
discounted present value formula), the assumption early-stages having long-term invest-
ments is not necessary. The large swings in early-stage production processes result from
changes in relative input/output prices.
The example provided above makes the unrealistic assumption of a specific number
of stages that follow a precise order. In the real world, there is no method by which to
determine where a firm is located in the structure of production. Additionally, there are
many recursive loops in the structure of production, where a portion of a firm’s output
may be sold in both the consumer and early-stage markets. An example of such a product
is the desktop computer. They are sold to research and development institutions and to
consumers. Nevertheless, the principle illustrated above holds true when the economy is
examined from the perspective of the social period of production.
When the assumption of holding the output price constant is relaxed and the output
price is allowed to rise in accordance with the increased demand for consumer goods, the
effect upon the level of economic profit is magnified. Furthermore, when the assumption
of a single rate of interest is relaxed, the same formula and analysis can be applied and
extended across the entire term structure of interest rates. The early-stage firms are able
to derive economic profits from engaging in both long and short-term projects. The
major difference for the long-term interest rates is that r1 becomes the rate of interest
20 New Perspectives on Political Economy
that would have materialized on the market if the Wicksell effect did not affect it.36
Responding to the compounding effects of relative price and interest rates changes,
entrepreneurs act as if the social period of production has changed and build more round-
about processes. Keeler (2002) empirically demonstrates that relative prices are the key
component of the propagation mechanism during the malinvestment boom phase of the
business cycle. Furthermore, he establishes that investment reallocation and capacity
utilization are expanded toward the early stages of production.37
Additionally, Machlup (1932) points out that even short-term investment in working
capital requires an array of higher-order capital (a superstructure) to support its produc-
tion. Thus even if a short-term project is transformed into fully integrated fixed capital,
it requires an additional array of higher-order capital to maintain its output. Machlup
further argues that the effect on capital is compounded the “more distant” an individ-
ual production project is from consumers. Thus, a malinvestment boom in the early
production stages occurs even when only short-term working capital is expanded.
A conclusion from this analysis is that the extent of the Wicksell effect corresponds
to the degree of malinvestments. In other words, to the degree that the new credit is able
to prevent an upward shift (or even cause a downward shift) in yields across maturities,
one will see maintenance (and expansion) of disequilibrated capital projects. It is impor-
tant to note that only a disaggregated approach can examine the capital structure in this
manner. The more aggregated theories are unable to draw these conclusions. Thus the
disaggregated, capital-based approach explains the malinvestment boom that Keen (1989),
Brown and Goodman (1991) and Estrella and Mishkin (1998) missed prior to the 1990-1
recession.
A key aspect of the malinvestment boom is not the boom, but the malinvestment.
The malinvestments maintained and created during the boom phase are malinvestments
because real savings does not support them.38 As a result, they must be liquidated at a
future date. These malinvestments are revealed during the crunch phase of the business
36 The rate, r2, is still the observed rate of interest after the monetary injection has had its effect.
37 See Keeler (2002) p. 15.
38 Additional savings (entering the economy from abroad or through a tax cut on savings) will transform
some (or maybe even all) malinvestments back into stable investments. However, additions to the money
supply changes the international price of the currency and reduces the incentive of foreigners to invest in
economy under examination.
Cwik: The Inverted Yield Curve and the Economic Downturn 21
cycle.
5The Crunch Phase and an Inverted Yield Curve
As noted above, the theories of the upper-turning point of the business cycle center on
either monetary or real factors as the primary cause of the downturn.39 While the capital-
based macroeconomic theory of the upper-turning point is not unique in describing the
upper-turning point, the significance of this approach is that it is able to account for
multiple factors. Robertson (1959) presents a classic observation on the phenomenon:
How is this cumulative upward process [of the economy] stopped and reversed?
It seems to me unlikely that there is a single answer applicable to all occasions; there
is a great variety of reasons why, in Haberler’s language, the system may become
more and more sensitive to “deflationary shocks” as expansion proceeds. Some in-
terpreters have laid stress on purelymonetary factors–the fact that the banks, finding
their reserves slipping away through withdrawals of legal tender money to pay the
enhanced wage-bills, etc., ultimately draw in their horns with a jerk. Others lay
stress on the emergence of what they call a “shortage of saving,” which no liberality
on the part of the banks could remedy. According to this picture, windfall profits
are eaten into by rising wages and interest rates, which at this stage no longer lag
appreciably behind the rise in prices, and with the disappearance of windfall profits
the main source of demand for instrumental goods is dried up. There turns out to be
an overproduction of such goods in the sense, as Cassel puts it, of “an overestimate
of the. . .amount of savings available for taking over the real capital produced.”40
Within the above passage, Robertson identifies two causes of the onset of a recession
as either a “deflationary shock” or a shortage of savings. The capital-based approach
identifies each of these factors as a potential and immediate cause of a recession, but the
underlying factor in each case is the malinvestment built up during the boom phase.
Monetary injections create disequilibria that cannot be maintained forever. The crunch
phase of the business cycle, where the scramble to prevent the liquidation of malinvest-
ments takes place, can come about in two ways–the credit crunch or the real resource
39 See fn. 12 above.
40 Robertson (1959) pp. 96-97. Robertson does not provide the specific cite for the Cassel quotation.
22 New Perspectives on Political Economy
crunch. While each scenario may cause the economy to turn from boom to bust, they
often occur together. However, the capital-based approach, by not over-aggregating the
effects of monetary injections, shows that each of these “causes” have the same root–
malinvestments.
Since the mid-1960s, there have been six official recessions.41 Except for the 1990-1
recession, monetary policy was tightened in each instance. However, when tightening
occurred after the recession started, it cannot be concluded that the recession was caused
by a credit crunch.42 In five instances, 1966, 1969-70, 1973-5, 1981-2, and 2001, a credit
crunch preceded an economic downturn. The recessions that are not preceded by a
policy of credit tightening are: 1980, and 1990-1. These recessions were caused by a
real resource crunch where economic pressure increased input prices which led to an
economic downturn.
Credit Crunch
The credit crunch occurs when the monetary authority determines inflation (or expected
inflation) is too high and “slams on the monetary brake.” The monetary authority’s
actions force short-term rates to rise. The yield curve rotates instead of shifts because the
rate of future inflation is expected to fall. The Wicksell effect dominates the Fisher effect
at the short-end of the yield curve and they negate each other at the long-end. Thus the
yield curve tends to invert prior to a recession, as seen in Figure 5.
41 In addition to the six recessions, the second quarter of 1967 experienced a negative growth rate. It was
preceded by an inverted yield curve and a credit crunch in 1966.
42 Owens and Schreft (1995) argue that there was tight credit in 1966, 1969-70, 1973-74, first half of 1980,
1981-2, and early 1990-2. However, they state that the 1990 crunch was market induced, while the others
stem from actual policies of monetary tightening or threats of increased regulatory oversight.
Cwik: The Inverted Yield Curve and the Economic Downturn 23
Figure 5: Inverting the Yield Curve with the Wicksell Effect
and the Fisher Effect
Interest Rate
Maturity
Fisher Effect
Wicksell Effect
Hayek (1969) states that in order to maintain the level of malinvestment, the rate
of money supply increases must be accelerated even when expectations of future prices
remain constant. If there is an expectation of future inflation, the rate of monetary
expansion must also outpace expectations of inflation. During periods of increasing price
levels, expectations of future inflation are not constant. In the neo-classical model of
the Long-Run/Short-Run Phillips curves, the economy is on a point to the left of the
Long-Run curve. Such a point is inherently unstable and the only manner in which the
economy can maintain its level of output is through accelerating rates of inflation.
With monetary expansion, price levels increase for two reasons: the previous expan-
sions of the money supply drive prices higher in an uneven manner and the instability
of the malinvestments induces entrepreneurs to bid up input prices. Malinvestments are
projects for which there is not enough savings to support them. During the monetary
expansion phase of the boom, new investments are encouraged and consumers increase
their levels of consumption and decrease their level of savings.43 As a result, there is
a shortage of real resources at existing prices. Assuming that prices are not sticky up-
wards, the consequence is an increasing input-price level at an accelerating rate. The
43 Again the significance of using the modified Preferred-Habitat theory becomes important. The fact
that there is a divergence between entrepreneur’s plans and that of consumers is based upon the in-
clusion of Böhm-Bawerk’s formulation of time-preference. Since consumers have not changed their
time-preferences, when new money is injected into the economy, they decrease their rate of savings.
A Preferred-Habitat theory using Fisher as its foundation could not support such a claim.
24 New Perspectives on Political Economy
inflationary effects of the earlier monetary expansion are compounded due to the need
of entrepreneurs to finance their malinvestments. Only with a disaggregated approach,
such as with the capital-based approach, can it be shown that inflation must not only
accelerate to account for expectations, but that it must expand at a higher multiple to
accommodate increasing input prices in order to maintain output levels.44
Kashyap, Stein and Wilcox (1993) demonstrate that when the monetary authorities
engage in a policy of monetary contraction, there is an immediate effect on the money
stock. The first consequence is a reduction of new loans made to entrepreneurs. As in-
put prices increase, there is a need for new financial capital to complete or maintain the
malinvestments. The firms with investment-grade bonds have access to credit markets,
but firms without this rating scramble for financial capital. They drive up short-term
rates in order to finance their projects. Cantor and Wenninger (1993) illustrate how, in a
time of credit tightness, funds flow away from low-grade investment instruments (in their
analysis, away from the junk bond market) and into bonds with at least a grade of Baa.
Long-term rates do not change due to two factors: expectations for future inflation have
not changed and firms with investment-grade bonds are able to borrow long-term by tap-
ping the funds flowing out of the low-grade investment instruments. Romer and Romer
(1993) show that “[S]mall firms are particularly dependent on banks for finance.. . .”45
They also conclude that during the periods of monetary contraction where the Federal
Reserve is able to increase short-term interest rates, banks are able to maintain the levels
of loans. However, banks do not increase their loan levels that would be required to
maintain the malinvestments.
When the monetary authorities believe that the current rate or future rate of inflation
is too high, they engage in a policy of monetary tightening. Christiano, Eichenbaum and
Evans (1996) show that a contractionary monetary policy increases the federal funds rate.
The short-term rates increase relative to the long-term rates. Kashyap, Stein and Wilcox
(1993) show that the issuance of commercial paper and bonds jumps relative to bank loans
after monetary tightening. Bernanke and Blinder (1992) argue that a tight monetary
policy leads to a short-run sell-off of the banks’ security holdings (with little effect on
44 If the monetary authorities adopt a policy of accelerating inflation, the consequence is hyperinflation.
However, a real resource crunch will usually come about before that point is reached, e.g., the 1980
recession.
45 See Romer and Romer (1993) p. 39.
Cwik: The Inverted Yield Curve and the Economic Downturn 25
loans), therefore reducing the capital value of these assets and making it more difficult for
firms to borrow against their assets. They demonstrate that, over time, banks terminate
old loans and refuse to make new ones. In other words, the monetary shock hits securities
first. After the securities are sold off, banks rebuild their portfolios and loans start to fall
off. After an average period of 2 years, securities return to their previous levels and loans
reflect the entire decline. When there is a monetary contraction, according to the results
of Kashyap, Stein and Wilcox (1993), a reduction of the supply of loans and the effects on
production will not begin to materialize until 6-9 months later. Furthermore, they find
evidence that output corresponds with loans. Christiano, Eichenbaum and Evans (1996)
argue that households do not adjust their financial assets and liabilities for several quarters
after a contractionary monetary shock. The authors also support the findings that the
net funds raised by businesses are able to increase for up to one year after the policy
shock, after that period, these funds decline. This delay explains the timing issue–the fact
that the yield curve tends to invert approximately one year before a recession.
Bernanke and Gertler (1995) argue that interest rates initially spike after monetary
contraction and return to their trends after 9 months. This evidence corresponds to the
data that show that the yield curve tends to return to its normally positive slope just prior
to a recession. This phenomenon was observed in 1957, 1960, 1967, 1989-90 and 2001.
Furthermore, Bernanke and Gertler (1995) argue that with a monetary contraction, final
demand falls off before production does, implying that inventories rise in the short-run.46
According to their results, durable spending displays the largest response to monetary
policy shocks, which corresponds to the arguments presented in section 4.
Owens and Schreft (1995) argue that the recessions of 1953-4, 1957-8 and 1960-1 were
caused by credit crunches. They report that a credit crunch occurred in the spring of
1953 and the recession began in Q2:1953. While the yield curve did not become invert
or humped, it flattened throughout the preceding period.
The next US recession began in Q3:1957. Based on the Minutes of the FOMC Meet-
ings, Romer and Romer (1993) identify a contractionary monetary shock in September
1955. The yield curve displayed the effects of a credit crunch when it became humped in
46 Dimelis (2001) demonstrates that business inventories are procyclical and are more volatile than sales.
She also points out that EU swings are larger than US swings. She attributes this characteristic to better
inventory practices in the US.
26 New Perspectives on Political Economy
December 1956, but it did not invert over the course of the recession.47
Owens and Schreft (1995) find evidence of a credit crunch in last third of 1959. The
recession began in Q2:1960, and while the yield curve became humped in September
1959, it did not invert.
As noted above, the economic downturns of 1966, 1969-70, 1973-5, 1981-2 and 2001
were also caused by a credit crunch. In February 1966, President Johnson publicly stated
that he feared an approaching inflation and said that he was counting on the Federal
Reserve to prevent it. Owens and Schreft (1995) report that the Federal Reserve met
with bankers and imposed quantitative limits on certain types of lending. The yield curve
inverted in September 1966 and the economy experienced negative growth for Q2:1967.
In late 1968, the fear of inflation arose again. Romer and Romer (1993), using the
Minutes of the FOMC Meetings, identify the contractionary monetary shock in Decem-
ber 1968. Owens and Schreft (1995) identify the January 14, 1969 meeting of the FOMC
where a shift toward tighter monetary policy took place. The recession began in Q4:1969
and lasted through Q4:1970. In any case, the yield curve became humped in November
1968 and inverted in briefly in January 1969 and then again inverted between July 1969
and August 1969. It inverted once more between November 1969 and January 1970.48
For the 1973-5 recession (which began in Q4:1973), the fear of inflation led the Federal
Reserve to raise discount rate 4.5% to 5% on January 15th, 1973. The yield curve became
humped in February 1973, inverted in June of the same year, and remained inverted
until September 1974. However Romer and Romer (1993), again using the Minutes of
the FOMC Meetings, identify the contractionary monetary shock in April 1974.
Preceding the recession of 1981-2, which began in Q3:1981, Owens and Schreft (1995)
argue that the Fed maintained tight credit policy throughout 1981. The yield curve
became humped in September 1980 and inverted in November 1980.
On December 5th, 1996, Chairman Greenspan used the phrase “irrational exuber-
ance,” sending the first warning that inflationary pressures were on the horizon. How-
47 However, Owens and Schreft (1995) do not find evidence of a credit crunch until the fall of 1957.
48 It is interesting to note that the yield curve became humped before the contractionary monetary policy
was put into place. A possible reason for this is that there was a real resource crunch just staring to take
effect at this time as well. The real resource crunch is explained below.
Cwik: The Inverted Yield Curve and the Economic Downturn 27
ever, after a series of rate cuts (cutting the discount rate by 50 basis points to 4.50% by
December 1998) the Federal Reserve did not increase the discount rate until August 1999.
Beginning in that month, the Federal Reserve began a series of discount rate increases,
which culminated in a discount rate of 6.00% in June 2000. The stated reason for the
change in policy is found in the FOMC Press Release August 24th, 1999: “Today’s in-
crease in the federal funds rate, together with the policy action in June and the firming of
conditions more generally in U.S. financial markets over recent months, should markedly
diminish the risk of rising inflation going forward.” The November 16th, 1999 FOMC
Press Release stated that the Federal Reserve was increasing the fed funds rate and the
discount rate to check “inflationary imbalances.” However the annualized rate of infla-
tion, according to the Consumer Price Index (CPI), for August and November 1999 were
merely 1.48% and 1.47% respectively.
As a consequence of the policy of monetary tightening, the yield curve became
humped in April 2000 and inverted in August 2000. It and remained inverted through
December 2000. The NBER dates the beginning of the recession in March 2001.
While the cause of the 2001 recession may be claimed to be the monetary policy, the
Federal Reserve was actually reacting to significant inflationary pressures in the producers
markets.49 Between April 2000 and January 2001, the Producer Price Index (PPI) for
industrial commodities increased over 8.09% and the PPI for all commodities increased
over 7.11%. Also during this period, the CPI increased at an approximate rate of only
2.21%. These inflationary pressures are accounted for by a real resource crunch where
malinvestments that have built up in the economy can no longer be supported without
an accelerating rate of inflation. In other words, if the Federal Reserve had not intervened
with a contractionary monetary policy, the economy would have experienced an inverted
yield curve and recession because of the impending real resource crunch. The action of
the Federal Reserve only hastened the outcome, but did not substantively change the
result.
In the surveyed downturns, the monetary authority actively followed a policy of
monetary contraction. However, not all recessions are caused by such policies. The
recessions of 1980 and 1990-1 were caused by a real resource crunch. The existence of
49 This case is the opposite scenario of the 1980 recession, which was a recession caused by a real resource
crunch and enhanced by a credit crunch.
28 New Perspectives on Political Economy
malinvestments can be analytically identified only with a capital-based approach. The
subsequent need to liquidate these malinvestments is significant because this need causes
the yield curve to invert and sets the economy down a path toward recession even without
a policy of monetary contraction.
Real Resource Crunch
Unlike more aggregated models, the capital-based approach can also account for the
upper-turning point of a business cycle even when the monetary authorities do not en-
gage in monetary tightening. During the malinvestment boom, entrepreneurs are given
false signals to undertake malinvestments. Also during the boom phase, consumers rebal-
ance their portfolios so that they increase their spending on consumer goods and reduce
their level of savings. These malinvestments are unstable because there are not enough re-
sources to complete and maintain each of these projects. As Robertson described above,
windfall profits disappear, wages and input prices rise and “no longer lag appreciably be-
hind the rise in prices. . . .”50 Consequently, there is a scramble for financial capital by
entrepreneurs to prevent the liquidation of their projects. They bid up short-term rates
and the yield curve inverts due to a real resource crunch.
As described above, the price level is driven upward during the malinvestment boom
because of two factors: the expectation of future inflation and the bidding up of in-
put prices by entrepreneurs to prevent the liquidation of their projects. Even when the
monetary authorities engage in a policy of monetary expansion to meet the increasing
expectations of inflation, the total amount of stable investments is limited at any one
point in time by the level of savings in the economy. Savings provide the wherewithal
for entrepreneurs to build, complete and maintain their projects.
Monetary injections falsify the price signals to the entrepreneurs, causing them to
begin more projects than are actually tenable at that point in time. Additionally, the
steepening of the yield curve signals to consumers that short-term credit for consumer
purchases are less expensive. As described in section 4, a decrease in short-term rates
indicates that the cost of financing short-term purchases falls. Furthermore, with ever
increasing rates of inflation, consumers will rebalance their portfolios away from savings
50 See Robertson (1959) pp. 97 and fn. 41 above.
Cwik: The Inverted Yield Curve and the Economic Downturn 29
and checking accounts (and cash holdings) and into tangible assets. In the aggregate, the
demand for consumer goods increases and consumers save less. With fewer savings in the
economy, the total amount of possible stable investment projects diminishes.
The initial effect of the monetary expansion is that entrepreneurs are able to bid
resources to their projects because they are able to cheaply borrow and use the new
credit. However, the act of embarking on these investment projects bids up these resource
prices. At first the effects of the misalignment of the social structure of production is not
apparent and the dueling structure of production emerges (as seen in Figure 2 above).
To better illustrate this process, the following example is provided.51 Suppose that
a builder exists who has enough bricks to finish four houses, yet he starts to build five.
With a decrease in the normal rate of profit, he sees the additional house as a potential
windfall (economic) profit. He figures that he will be able to purchase the bricks neces-
sary for the completion of the project in the future when he needs them. Suppose further
that he borrows $100,000 to finance the project. Competing entrepreneurs also follow
this pattern due to the false market signals. As the entrepreneur runs out of bricks, he
starts to purchase more to complete the project. However, other entrepreneurs are also
bidding for more bricks. The price of the bricks increases with the increasing demand.
The $100,000 initially borrowed to complete the project is no longer enough. The en-
trepreneur must find additional financial capital to complete all five houses.
To state the situation more generally, the amount of funds previously borrowed to
complete projects (across all lengths) is now insufficient. There is an immediate need for
funds to complete and maintain the malinvestments. The scramble for additional funds
may be more intense with short-term projects. All entrepreneurs are faced with the
alternative of liquidation or of finding supplementary funds to complete their project,
but the intensity in demand for funds may be much higher for projects that are almost
complete. In other words, an entrepreneur may be more highly motivated to secure funds
to finish a project that will produce output next month than he would be to secure funds
for a project that will not yield a return until next year.
As a result of these actions, short-term rates are bid up quickly, inverting the yield
51 This example is similar to one presented in Mises (1966) pp. 559-560.
30 New Perspectives on Political Economy
curve.52 Tribó (2001) argues that smaller firms must look for short-term credit when
faced with output problems.53 He further finds that larger firms are able to tap into
the long-term credit markets. However, as noted above, those firms without investment-
grade ratings scramble for financial capital.
Over the course of a business cycle, long rates tend to remain relatively stable. The
larger firms with investment-grade ratings are able to attract funds for long-term invest-
ments from the low-grade investment sectors. The effect from the increases in the short
rates is diminished across the yield curve, because long-term lenders take on less risk since
they tend to be the mortgage holders, etc. They are the first to collect if the firm enters
bankruptcy. There is a liquidity premium to lending long, yet long-term instruments
have an inherent hedge against business cycle risk. Thus, the yield curve inverts instead
of shifting.
Three of the recessions since the mid-1950s were not caused by a credit crunch. While
there is evidence that there was monetary tightening in every recession except for the
1990-1 recession, the tightening for the 1980 downturn did not take place until after the
recession was under way.
As shown above, the 1969-70 recession has elements of both a credit crunch and a
real resource crunch. The yield curve became humped in November 1968, a month or
two before the policy of monetary contraction was agreed upon by the Federal Reserve.
These two causes are not mutually exclusive and may work simultaneously, thus this
evidence is not surprising.
The recession of 1980 is an example of a recession caused by a real resource crunch
and enhanced by a credit crunch.54 The recession began in Q1:1980. The yield curve
became humped in September 1978, inverted in December 1978, and remained so until
52 Summarizing his empirical findings, Keeler (2002) states, “As the aggregate economy expands and firms
progress in building capital and expanding output, shortages of resources occur which raise resource
prices. Short term interest rates are increased toward long term rates and the yield curve flattens or may
invert. The primary mechanism in this endogenous market process is the intertemporal disequilibrium
between sources and uses of income; at low interest rates, consumers and investors increase spending,
and need to finance an increase in both consumption and investment, but savers decrease the quantity
supplied of funds.” p. 5.
53 See also Romer and Romer (1993).
54 As noted above, the 1980 and 2001 recession are opposite scenarios but both contain the same compo-
nents: a policy of monetary contraction and a real resource crunch.
Cwik: The Inverted Yield Curve and the Economic Downturn 31
April 1980. While Romer and Romer (1993) identify contractionary monetary shocks in
August 1978 and October 1979, Owens and Schreft (1995) argue that the credit crunch did
not occur until the first half of 1980. Their position is that in order to regain control over
inflation and the expectation of high rates of future inflation, the Carter Administration
imposed credit controls on March 14th, 1980. Furthermore, the Federal Reserve did
not fully enforce these regulations until the Federal Reserve officials met on May 17th,
where Chairman Volcker warned the banks that the Federal Reserve would enforce the
program.
Producer prices grew at an accelerating rate between September 1978 and January
1980 (the dates where the yield curve became humped and the beginning of the recession).
Over this period, the PPI for industrial commodities grew at a rate over 22.64%, and the
PPI for all commodities at a rate over 17.19%. The CPI, over the same period, grew at
a rate of 16.99%. This evidence corresponds with the scramble by entrepreneurs to find
funding to complete their projects.
The first post-war recession that did not experience a contractionary monetary pol-
icy was the 1990-1 recession, which began in Q3:1990. Although the yield curve never
inverted, it became humped in February 1989 and lasted in this state through September
1989. While Romer and Romer (1993) identify a contractionary monetary shock in De-
cember 1988, most analysts doubt that such a shock occurred. Cantor and Wenninger
(1993) state that, “One of the most striking features of the recent credit cycle [of the
1990-1 recession] has been the [credit] crisis that never happened.” They argue that there
was a boom in the credit markets between 1986 and 1991. Bernanke (1993) interprets
the 1990-1 recession as the result of a credit crunch without a contractionary monetary
policy.
A real resource crunch is evidenced by increasing rates of input prices while output
prices fail to keep pace. Unfortunately, the aggregated data from the Federal Reserve Bank
of St. Louis is not specific enough to capture this relationship.55 Hughes (1997) provides
some evidence that corresponds to the real resource crunch scenario. He shows that
a malinvestment boom took place, with long-term bank borrowing by manufacturing
55 Nevertheless, the data from FRED II is as follows: from the date when the yield curve began to change,
January 1989, through the beginning of the recession, June 1990, the PPI for industrial commodities
increased at a rate greater than 3.28%, and the PPI for all commodities increased faster than 3.43%.
32 New Perspectives on Political Economy
industries increasing from $60.5 billion in 1981 to $197.2 billion in 1991 (in unadjusted
dollars). He also demonstrates that early-stage firms (such as primary metals and Iron
and Steel industries) greatly expanded their capacity from 1981 to 1985, but their output
prices collapsed in 1986. While his arguments tend to support the Austrian Business
Cycle theory, he seems to argue that, at least for the early-stage industries, the 1990-1
recession really began in 1986. Thus to find evidence of a real resource crunch for the
1990-1 recession, one must look at the market which most analysts identify as the one
which set off the recession–the commercial real estate market.
When viewed from the perspective of the commercial real estate market, one sees
that the 1990-1 recession was caused by a real resource crunch. Cantor and Wenninger
(1993) demonstrate that there was a large increase in the value of real estate prices prior to
the late 1980s. Additionally, the authors argue that deregulation forced non-bank thrift
institutions (like insurance companies) to look for ways to increase their rates of return.
Thus these institutions extended credit to more risky ventures (like those in the real estate
market), but the capital requirements for these thrift institutions remained low and many
weak firms were exempted from tough regulatory scrutiny.
Cantor and Wenninger (1993) point out that the real estate market collapsed in the
late 1980s, after which regulators increased scrutiny of these types of loans, making it
very difficult to obtain funding for real estate ventures. Owens and Schreft (1995), in
their paper which also observed decreasing real estate values in the late 1980s, state that
there were complaints that regulators were too closely scrutinizing real estate portfolios,
making real estate lending extremely difficult. Many new buildings came on the market
at the same time, depressing rental and sales prices. Additionally in many cases, the
tax breaks that made commercial building profitable were removed. Bernanke (1993)
observes that large losses in the real estate market reduced the amount of bank capital,
which he labeled a “capital crunch.” However, Bernanke downplays the supply of funds
as a major cause of the recession, because as bank loans fell in 1989, commercial paper
increased. Cantor and Wenninger (1993) state that during the period between 1986 and
1991, “nondepository credit growth continued to exceed GDP growth by a wide margin
(4.5 to 5.5 percentage points). Depository credit, on the other hand, decelerated sharply
as thrift credit went into an outright decline in the 1989-1991 period.”56 There was a
56 See Cantor and Wenninger (1993) p. 5.
Cwik: The Inverted Yield Curve and the Economic Downturn 33
shift from financing through banks, etc. to self-financing in the commercial paper and
commercial bond markets. Those firms with investment-grade securities were able to
obtain financing, while those without such a rating were not. Cantor and Wenninger
(1993) further argue that just prior to the recession, (1989-90) money stopped flowing into
“junk” bonds and instead went into investment-grade corporate bonds. Despite an easy
monetary policy,57 those borrowers without direct access to the financial markets (i.e.,
those without investment-grade ratings) did not benefit from this policy. Their scramble
for financial capital caused the yield curve to become humped. Below investment-grade
borrowers were shut out of the long and short-term money markets and eventually were
forced to liquidate their projects, while those with investment-grade ratings benefited
from the easy credit policy.
Bernanke and Lown (1991) support the conclusion that the decrease in bank loans
was not supply-side related. Thus, they are skeptical that a credit crunch caused the reces-
sion. The lack of contractionary monetary policy explains the appearance of a humped
yield curve instead of an inverted yield curve. While Bernanke and Lown argue that
the demand for loans was a more important cause of the 1990-1 slowdown, the evidence
they provide is that all forms of credit (including commercial paper) decreased during the
1990-1 recession–indicating a decrease in demand for credit. There seems to be a timing
discrepancy in their analysis. In 1989 commercial paper increased and then, when the
recession began, all forms of credit decreased. Their evidence supports the real resource
crunch scenario, instead of the scenario where a fall in the demand for credit caused the
recession. There was a scramble for credit in the late 1980s, which is seen in the increase
in commercial paper issuances and the change in the shape of the yield curve. When
businesses started to fail in 1990, the demand for all credit fell and the recession was
underway.
6Summary
The capital-based approach compares favorably with CCAPM and Estrella models be-
cause it is able to explain why the yield curve tends to invert before a recession. Unlike
these other models, this approach centers its focus on the malinvestments built up during
57 Owens and Schreft (1995) report that the Fed eased monetary policy in spring of 1990.
34 New Perspectives on Political Economy
the malinvestment boom. These other models, by over aggregating, are unable recognize
that the root cause of the inversion of the yield curve is the malinvestments.
In this paper, the assumption was made that the initial monetary injection was short-
term working capital. It has been shown that this capital is transformed into fixed capital,
long-term projects and early-stage malinvestments. To the extent that these projects are
inconvertible, the liquidation process becomes more severe.
The modified Preferred-Habitat theory is an essential component to the model used,
because it is able to illustrate how monetary injections lead to a disequilibrium be-
tween consumption/savings horizons of households and the investment projects of en-
trepreneurs. Monetary injections cause the yield curve to steepen which falsely signals
entrepreneurs to begin new investments and encourages households to increase their de-
mand for final goods and services.
The unstable malinvestments force input prices to rise and lead to a credit crunch,
a real resource crunch, or a combination of both. In their attempts to prevent their
individual projects from being liquidated, entrepreneurs will cause the yield curve to
flatten, become humped or even invert as they scramble for financial capital (even when
the monetary authority adopts a policy of easy credit). Thus in every recession since the
mid-1950s, an inverted or humped yield curve occurred no more than 5 quarters prior to
the upper-turning point of the business cycle.
Bibliography
[1] Bernanke, Ben, 1993. “How Important is the Credit Channel in the Transmis-
sion of Monetary Policy?: A Comment,” Carnegie-Rochester Conference Series
on Public Policy, 39, 47-52.
[2] Bernanke, Ben, and Alan Blinder, 1992. “The Federal Funds Rate and the Channels
of Monetary Transmission,” American Economic Review, 82(4), 901-921.
[3] Bernanke, Ben, and Mark Gertler, 1995. “Inside the Black Box: The Credit Chan-
nel of Monetary Policy Transmission,” The Journal of Economic Perspectives, 9(4),
27-48.
Cwik: The Inverted Yield Curve and the Economic Downturn 35
[4] Bernanke, Ben, and Cara Lown, 1991. “The Credit Crunch,” Brookings Papers on
Economic Activity, 2, 205-247.
[5] Brown, William, and Douglas Goodman, 1991. “A Yield Curve Model for Predict-
ing Turning Points in Industrial Production,” Business Economics, July, 55-58.
[6] Cantor, Richard, and John Wenninger, 1993. “Perspective on the Credit Slow-
down,” Federal Reserve Bank of New York, Quarterly Review, 18(1), 3-36.
[7] Cecchetti, Stephan, 1987. “The Case of the Negative Nominal Interest Rates: New
Estimates of the Term Structure of Interest Rates During the Great Depression,”
NBER, Working Paper no. 2472.
[8] Christiano, Lawrence, Martin Eichenbaum, and Charles Evans 1996. “The Effects
of Monetary Policy Shocks: Evidence from the Flow of Funds,” The Review of
Economics and Statistics, 78(1), 16-34.
[9] Cochran, John, 2001. “Capital-Based Macroeconomics,” Quarterly Journal of Aus-
trian Economics, 4(3), 17-25.
[10] Cwik, Paul, 2004. “An Investigation of Inverted Yield Curves and Economic
Downturns,” forthcoming Dissertation, Auburn University.
[11] Cwik, Paul, 1998. “The Recession of 1990: A Comment,” Quarterly Journal of
Austrian Economics, 1(2), 85-88.
[12] Davis, Lance, 1971. “Capital Mobility and American Growth,” in: Fogel, Robert,
and Stanley Engerman (Eds.), The Reinterpretation of American Economic His-
tory, New York, NY: Harper & Row, pp. 285-300.
[13] Dimelis, Sophia, 2001. “Inventory Investment over the Business Cycle in the EU
and the US,” International Journal of Production Economics, 71(1-3), 1-8.
[14] Estrella, Arturo, 1998. “Monetary Policy and the Predictive Power of the Term
Structure of Interest Rates,” Federal Reserve Bank of New York, Working Paper,
November.
36 New Perspectives on Political Economy
[15] Estrella, Arturo, and Frederic Mishkin, 1998. “Predicting U.S. Recessions: Finan-
cial Variables as Leading Indicators,” The Review of Economics and Statistics,
80(1), 45-61.
[16] Evans, Martin, 1987. “A Macroeconomic Model of the Term Structure of Interest
Rates,” New York University Salomon Brothers Center, Working Paper no. 445,
29.
[17] Garrison, Roger, 2001. Time and Money: The Macroeconomics of Capital Struc-
ture, New York, NY: Routledge.
[18] Hayek, Friedrich von, 1969. “Three Elucidations of the Ricardo Effect,” Journal of
Political Economy, 77(2), 274-285.
[19] Hayek, Friedrich von, 1945. “The Use of Knowledge in Society,” American Eco-
nomic Review, 35(4), 519-530.
[20] Hughes, Arthur, 1997. The Recession of 1990: An Austrian Explanation, Review
of Austrian Economics, 10(1), 107-123.
[21] Kashyap, Anil, Jeremy Stein and David Wilcox, 1993. “Monetary Policy and Credit
Conditions: Evidence from the Composition of External Finance,” American Eco-
nomic Review, 83(1), 79-98.
[22] Keen, Howard, 1989. “The Yield Curve as a Predictor of Business Cycle Turning
Points,” Business Economics, 24(4), 37-43.
[23] Keeler, James, 2002. “Relative Prices and the Business Cycle,” paper presented at
the Southern Economic Association Annual Meetings in Tampa, the Society for
the Development of Austrian Economics, Session 42C: Research in Empirical Aus-
trian Economics.
[24] Keeler, James, 2001. “Empirical Evidence on the Austrian Business Cycle Theory,”
Review of Austrian Economics, 14(4), 331-351.
[25] Kessel, Reuben A, 1965. “The Cyclical Behavior of the Term Structure of Interest
Rates,” NBER, Occasional Paper no. 91.
Cwik: The Inverted Yield Curve and the Economic Downturn 37
[26] Machlup, Fritz, 1935. “The Rate of Interest as Cost Factor and Capitalization Fac-
tor,” American Economic Review, 25(3), 459-465.
[27] Machlup, Fritz, 1932. “The Liquidity of Short-Term Capital,” Economica, 12(35-
38), 271-284.
[28] McCulloch, J. Huston, 1990. “Appendix B: U.S. Term Structure Data, 1946-87,”
in: Friedman, B.M., and F.H. Hahn (Eds.), Handbook of Monetary Economics,
North-Holland, Amsterdam, pp. 672–687.
[29] Miller, Roger LeRoy and David VanHoose, 2001. Money, Banking, & Financial
Markets, 2001 edition, Australia: South-Western / Thomson Learning.
[30] Mises, Ludwig von, 1980. “Profit and Loss,” in Planning for Freedom, Fourth Edi-
tion, South Holland, IL: Libertarian Press, 108-150.
[31] Mises, Ludwig von, 1966. Human Action: A Treatise on Economics, Third Revised
Edition, Chicago, IL: Contemporary Books, Inc.
[32] Mishkin, Frederic, 2001. The Economics of Money, Banking, and Financial Mar-
kets, Sixth edition, New York, NY: Addison Wesley Longman.
[33] Owens, Raymond, and Stacey Schreft, 1995. “Identifying Credit Crunches,” Con-
temporary Economic Policy, 13(2), 63-76.
[34] Robertson, Dennis, 1959. Lectures On Economic Principles, Volume 3, London:
Staples Press.
[35] Romer, Christina and David Romer, 1993. “Credit Channels or Credit Actions?”
NBER, Working Paper no. 4485.
[36] Schmitz, Stefan, 2003. “Uncertainty in the Austrian Theory of Capital,” unpub-
lished manuscript.
[37] Skousen, Mark, 1990. The Structure of Production, New York, NY: New York
University Press.
[38] Tribó, Josep, 2001. “Inventories, Financial Structure and Market Structure,” Inter-
national Journal of Production Economics, 71(1-3), 79-89.
