View original document

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

ESSAYS ON ISSUES
	

	 THE FEDERAL RESERVE BANK 	
OF CHICAGO

NOVEMBER 2009
	 NUMBER 268

Chicag­ Fed Letter
o
The Chicago Fed National Activity Index and business cycles
by Scott Brave, business economist

This article discusses how the Chicago Fed National Activity Index—a monthly
index designed to gauge economic activity and related inflationary pressures—
can be used as an indicator of business cycle turning points.

The Chicago Fed National Activity
Index (CFNAI) is a monthly index of
U.S. economic activity constructed to
summarize variation in 85 data series
classified into four groups: production
and income; employment, unem1. CFNAI-MA3 and business cycles
ployment, and
index
hours; personal
2
consumption and
housing; and sales,
1
orders, and inven+0.2
tories.1 The index
0
is designed as a co–0.7
incident indicator
–1
of national eco–2
nomic activity and
serves as a leading
–3
indicator of activityrelated inflationary
–4
pressures. In this
1967
’73
’79
’85
’91
’97
2003
’09
Chicago Fed Letter,
Notes: CFNAI-MA3 is the Chicago Fed National Activity Index’s three-month moving
I reexamine the reaverage. Shading indicates official periods of recession as identified by the National
Bureau of Economic Research; the dashed vertical line indicates the most recent busilationship between
ness cycle peak. A CFNAI-MA3 value below –0.70 following a period of economic expansion indicates an increasing likelihood that a recession has begun. A CFNAI-MA3
the CFNAI and
value above +0.20 following a period of economic contraction indicates a significant
likelihood that a recession has ended.
business cycle turning points.

The CFNAI is an example of a “Goldilocks” index. In essence, this means
that the information in various data
series on national economic activity is
combined in a way to reflect deviations
around a trend rate of economic growth.
Accordingly, the CFNAI is normalized
to have a mean of zero and a standard
deviation of one. In the Goldilocks
terminology, this means that a zero

value of the index is “just right,” in that
the economy is proceeding along its
historical growth path. A negative value
of the index is “cold,” in that growth is
below average, while a positive value is
“hot,” in that it is above average.
The CFNAI can be very volatile, since
many of the monthly series that make
up the index vary significantly from
month to month. For this reason, the
focus is often given to the three-month
moving average of the index, i.e., the
CFNAI-MA3 (figure 1), which smoothes
the month-to-month variations over
time in order to provide a more consistent picture of variations in economic
growth around trend. When the values
of this index reach certain levels that
have been identified in previous research as “too hot,” the likelihood of
an inflationary period rises; when it
gets “too cold,” the likelihood of a
recession rises.2
Historical performance

Here, I focus on those levels of the
three-month moving average index
that are typical of economic activity
during recessions as identified by the
National Bureau of Economic Research
(NBER). These “threshold values” identify large deviations of the CFNAI-MA3
from its mean of zero given its standard deviation of one. For instance,
a CFNAI-MA3 value below –0.7, or
greater than two-thirds of one standard
deviation from its mean, is defined as

2. Four categories of the CFNAI-MA3
A. Production and income

B. Employment, unemployment, and hours

index

index

2

2

1

1

0

0

–1

–1

–2

–2

–3

–3

–4
1967

’73

’79

’85

’91

’97

2003

’09

–4
1967

’73

’79

’85

’91

’97

2003

’09

’91

’97

2003

’09

D. Sales, orders, and inventories

C. Personal consumption and housing

index

index

2

2

1

1

0

0

–1

–1

–2

–2

–3

–3

–4
1967

’73

’79

’85

’91

’97

2003

’09

CFNAI-MA3 category

–4
1967

’73

’79

’85

CFNAI-MA3

Notes: Notes: CFNAI-MA3Chicago Fed National ActivityActivity Index’s three-month moving average. Shading indicates official periods of recession as identified National BureauBureau of Economic
CFNAI-MA3 is the is the Chicago Fed National Index’s three-month moving average. Shading indicates official periods of recession as identified by the by the National of Economic Research;
Research; the dashed vertical line business cycle peak.
the vertical line indicates the most recent indicates the most recent business cycle peak.

indicating a significant likelihood that
a recession has begun. Similarly, a recession is indicated to have likely ended based on the CFNAI-MA3 returning
to a level of +0.2 after having crossed
the –0.7 threshold.
Historically, both thresholds have performed fairly well with respect to the
recession timing as determined by the
NBER (figure 1). The –0.7 threshold
has correctly predicted a recession
month with 86% accuracy since 1967,
identifying the beginning of a recession within three months on average;
the +0.2 threshold has correctly predicted the end of each recession since

1967 within eight months on average,
with no false positive signals. However,
the performance of the latter threshold has varied significantly over time,
with the 1990–91 and 2001 recessions
ending 18 months on average before
+0.2 was reached.
Given the subpar performance of the
index in identifying the end of the two
most recent recessions, it may be useful to consider alternative definitions
of the likelihood of a recession’s end.
I propose here two such alternatives.
These alternatives seek to make use
of auxiliary information on U.S. recessions since 1967 and unify information

contained in the CFNAI-MA3 on business cycle turning points around a single
threshold value of –0.7.
The first alternative relies on the fact
that since 1967 all recessions have
lasted for a minimum of six months.
Using this information, a new threshold can be constructed where the end
of a recession is signaled when the
three-month moving average index
attains a value greater than –0.7 at least
seven months after an NBER-defined
recession has begun. Using this definition, the CFNAI correctly predicts the
end of each recession since 1967 within
one month on average, and all six of

the false positive signals from this definition are contained in the two recessions of the 1970s.
Another alternative is to base the definition of the likelihood of a recession’s
end on the CFNAI-MA3 attaining a value
greater than –0.7 following the point
in time during an NBER recession when
it first fell below this threshold. This alternative removes any potential error
in identifying a recession’s likely ending point from incorrectly identifying

August of this year, the CFNAI-MA3 has
increased by nearly 2.5 standard deviations. Despite this steady improvement,
at –1.09 as of the September 28, 2009,
release, the index continues to be below
the –0.7 threshold. Substantial economic
slack exists in the U.S. economy as evidenced by the fact that the three-month
moving average remains just slightly more
than one standard deviation below its
mean. However, the index now indicates
a level of economic activity above the

Taking a closer look at the categories of indicators that
make up the index can be helpful in explaining the ongoing
weakness in the economy in relation to past recessions.
its beginning. By this definition, the
CFNAI also correctly predicts the end
of each recession within one month on
average, with four false positive signals
that all occur during the 1969–70 recession. Thus, both of these alternatives
more accurately describe the timing of
recessions since 1980.
The monthly release of the CFNAI is
a benefit that can be exploited in real
time to determine business cycle turning points. For instance, with regard
to the current recession, the index
correctly identified December 2007
as the start date in its March 24, 2008,
release, nearly eight months before the
official declaration by the NBER. However, using the index as an indicator of
a recession’s end in real time is more
difficult. The real-time use of both my
alternative threshold definitions presupposes that the recession has already
been identified by the NBER. Given the
length of a typical recession and the
performance of the index in identifying the beginning of a recession, this
concern should be minimal.
CFNAI indicator categories

The history of the index shown in
figure 1 demonstrates that the current
recession marks the steepest decline in
economic activity since the 1973–75
recession, and possibly the longest since
the 1981–82 recession, depending on
its official end. From January through

trough of the two most recent recessions prior to the current one.
Taking a closer look at the categories
of indicators3 that make up the index
can be helpful in explaining the ongoing weakness in the economy in relation to past recessions. It is well known
that the recoveries from the 1990–91
and 2001 recessions were muted in
part because of the sluggish response
of the labor market. This points out a
potential weakness in the relationship
between the CFNAI-MA3 and business
cycle turning points.4 Recessions are
typically defined based on widespread
changes in economic activity. However,
when individual sectors of the economy diverge sharply from the rest, this
information can provide further insight
into the nature of a recession and the
subsequent recovery. Therefore, looking
just at the CFNAI-MA3 may obscure
potentially useful information.
Consider figure 2, which charts the contributions to the CFNAI-MA3 from each
of its four categories of indicators. One
can see notable differences in them over
the business cycle. For instance, the contribution of the employment, unemployment, and hours category (panel B)
tends to lag the CFNAI-MA3 as a whole.
This category’s greatest negative contribution to the index is generally made
near the end of a recession. Also, in the
case of the 1990–91 and 2001 recessions,
the impact of the jobless recovery on

the index is clear, since it took this category a substantially longer period of
time to return to trend.
In contrast, the contributions of the
production and income (panel A) and
sales, orders, and inventories (panel D)
categories tend to turn negative more
quickly during a recession and turn
positive once the recovery begins. That
said, the contribution of the personal
consumption and housing category
(panel C) demonstrates substantial variability in its behavior during recessions
(sometimes leading, sometimes lagging);
and in the case of the 2001 recession,
it shows no discernible impact.
These observations are in line with welldocumented business cycle facts. For
instance, it is well known that the labor
market tends to be a lagging indicator
of the business cycle. Furthermore, the
casual definition of a recession as two
quarters of negative real gross domestic
product growth conforms well to the
fact that production and income indicators tend to decline at the onset of
recessions. Finally, the behavior of the
sales, orders, and inventories category reflects the fact that declining sales and
orders that accompany a recession are
typically followed by liquidations of
inventories, only to have firms build
them up again when demand improves.
Charles L. Evans, President; Daniel G. Sullivan, Senior
Vice President and Director of Research; Douglas D. Evanoff,
Vice President, financial studies; Jonas D. M. Fisher,
Vice President, macroeconomic policy research; Daniel
Aaronson, Vice President, microeconomic policy research;
William A. Testa, Vice President, regional programs,
and Economics Editor; Helen O’D. Koshy and
Han Y. Choi, Editors; Rita Molloy and Julia Baker,
Production Editors.
Chicago Fed Letter is published by the Economic
Research Department of the Federal Reserve Bank
of Chicago. The views expressed are the authors’
and do not necessarily reflect the views of the
Federal Reserve Bank of Chicago or the Federal
Reserve System.
© 2009 Federal Reserve Bank of Chicago
Chicago Fed Letter articles may be reproduced in
whole or in part, provided the articles are not
reproduced or distributed for commercial gain
and provided the source is appropriately credited.
Prior written permission must be obtained for
any other reproduction, distribution, republication, or creation of derivative works of Chicago Fed
Letter articles. To request permission, please contact
Helen Koshy, senior editor, at 312-322-5830 or
email Helen.Koshy@chi.frb.org. Chicago Fed
Letter and other Bank publications are available
at www.chicagofed.org.
ISSN 0895-0164

Behavior that deviates from these norms
can, thus, be illustrative of the type of
recession and recovery to be expected.
For instance, the contribution from
the personal consumption and housing category reached an all-time low
during the current recession, with
the beginning of the decline in this
category preceding the recession by
over a year. The extent of the decline,
reflecting the struggles of the residential housing market after home prices
began to fall sharply in 2006, is unprecedented in the index’s history.
As of the September 28, 2009, release,
despite some recent improvement in
residential housing indicators, the
contribution to the CFNAI-MA3 from
this category was still –0.50. Given the
persistence of this component, it is
likely to be a drag on economic activity
for some time to come.
That said, looking at the employment,
unemployment, and hours category
offers some hope for the beginning
of a recovery. The negative contribution of this category bottomed out in
February 2009. Based on its behavior
during past recessions, this often has
served as an early signal of a recovery.
However, such a signal has not always
been accurate. For instance, during

the 1981–82 recession this category exhibited a W-like pattern, rising and falling more than once before the end of
the recession. In addition, the amount
of resource slack indicated by the category’s August 2009 contribution of –0.58
also suggests further skepticism about
a full recovery is warranted.
Still, in recent months, contributions
from both the production and income
and the sales, orders, and inventories
categories have also improved considerably. The contributions of both groups
bottomed out at the CFNAI-MA3’s most
recent trough of January 2009. In the
past, a neutral contribution of these
groups during a recession has been
an early sign of a recovery. As of the
September 28, 2009, release, the contribution of the production and income
category had turned slightly positive,
while the sales, orders, and inventories
category remained slightly negative. Thus,
it appears likely that recovery from the
current recession had not yet reached
full swing as of August 2009, but early
signs of recovery are apparent.

the ends of recent recessions has been
subpar. Adjusting the way the threemonth moving average index is interpreted based on alternative definitions
of the likelihood of a recession’s end
can help to improve its performance
in this regard. Furthermore, individual
inspection of the four categories of indicators that make up the index can
provide additional information on the
nature of a recession and the subsequent
recovery. Applying these methods to
the current recession offers some signs
of a recovery.
1	 Additional information on the construction of the CFNAI can be found at
www.chicagofed.org/cfnai.

The CFNAI has historically performed
well in relation to NBER recession dates.
However, its accuracy in determining

See Charles L. Evans, Chin Te Liu, and
Genevieve Pham-Kanter, 2002, “The 2001
recession and the Chicago Fed National
Activity Index: Identifying business cycle
turning points,” Economic Perspectives,
Federal Reserve Bank of Chicago, Vol. 26,
No. 3, Third Quarter, pp. 26–43.

3	

The four indicator categories discussed
here were made publicly available with
the September 28, 2009, release at www.
chicagofed.org/cfnai.

4	

Conclusion

2	

See Mary Daly, Bart Hobijn, and Joyce
Kwok, 2009, “Jobless recovery redux?,”
FRBSF Economic Letter, Federal Reserve
Bank of San Francisco, No. 2009-18, June 5,
available at www.frbsf.org/publications/
economics/letter/2009/el2009-18.html.


Federal Reserve Bank of St. Louis, One Federal Reserve Bank Plaza, St. Louis, MO 63102