ECON1220H&I * Intro Macro Chi-Wa Yuen

5 Business cycles

5.1 Facts and theories

Decomposition of a cycle into 4 phases: prosperity, crisis, depression, and revival.

Lucas’s de…nition of cycles as deviations from trend of aggregate real output and
of business-cycle regularities as co-movements of detrended macro time series—
“Business cycles are all alike!!!”

04/03/2023 1/15
ECON1220H&I * Intro Macro Chi-Wa Yuen

Business cycle facts: by characterizing the statistical patterns of detrended/…ltered

macro time series in terms of their

– amplitudes: volatilities as measured by standard deviations

– directions: co-movements as re‡ected by the signs of correlation coe¢ cients

with real GDP — pro-/counter-/a-cyclicality

– timing : persistence and phase shifts as indicated by …rst- and higher-order

auto- and cross-correlations — leading/lagging/coincident relative to GDP

04/03/2023 2/15
ECON1220H&I * Intro Macro Chi-Wa Yuen

Construction of DSGE models: by embedding into the deterministic dynamic general

equilibrium model

– impulses: di¤erent types of shocks, e.g., productivity, endowment, preference,

policy shocks

– propagation mechanisms: transmission of aggregate shocks over time (and of

imperfectly insurable sector-speci…c shocks across sectors?), say, through capital
accumulation

“Explaining” the business cycle facts: by parameterizing and calibrating the model
(by matching the statistical moments of data simulated from the theoretical model)
to real-world data.

Using the calibrated model for prediction: by conducting impulse-response and

innovation-accounting exercises.

04/03/2023 3/15
ECON1220H&I * Intro Macro Chi-Wa Yuen

5.2 Persistence of business-cycle movements & turning points

Why even a transitory shock could produce persistent movements in major macro
variables via the multiplier-accelerator process.

Whether an economy could automatically turn around in the absence of stabilization

policies and shock reversals

– falling o¤ a peak — due to bottlenecks and capacity constraints

– rising from a trough — due to subsistence consumption (and thus non-negativity

of output and capital)

Read section 3.3 of my “Keynesian Economics” in the SAGE Encyclopedia of Po-

litical Behavior.

04/03/2023 4/15
 ECON1220H&I * Intro Macro Chi-Wa Yuen
to
'<T

o -~~t\'6'~~ ~#
p...~"
~ IX)
.
I"')

@ cl,.:tN!~a!r,-tlNit
~~-~rf~ - t
o
~ " ~,t~ I"')
~~~ ~. ~~
(1) ~ ~~I~C.~
. ~~ ~ tta.~..~ ~
~ N
N

IX)

GNP 1982$
Trend GNP
'<T

'-1955195819611964196719701973197619791982198519881991

Figure 1.1 Log of Real GNP and ItsGrowth Component

Trillionsof 1982 dollars(log scale). tu. ~~ ~

04/03/2023 5/15
ECON1220H&I * Intro Macro Chi-Wa Yuen

04/03/2023 6/15
 ECON1220H&I * Intro Macro Chi-Wa Yuen
Business Cycle Facts
[from Thomas C. Cooley and Edward C. Prescott, "Economic Growth and Business Cycles,"
in Cooley (1995) ed, Frontiers of Business Cyc!!LResearch. Princeton.]

The following are some salient features of the business cycle based on US time
. series data:

1) The magnitude of fluctuations in output and aggregate hours of work

are nearly equal. It is well known that the business cycle is most clearly
manifested in the labor market and this observation confirms thaL 4"'")-
2) Employment fluctuates almost as much as output and total hours of work,
while average weekly hours fluctuate.considerably less. This suggests that
most fluctuations in total hours represent movements into and out of the
work force rather than adjustments in average hours of work.
3) Consumption of nondurables and services (CNDS) is smooth, fluctuating
much less than OutpUL a:"'''$ <tr; -
~~
4) Investmentin both producers' and consutners' durablesfluctuatesmuch
more than OUtpUL d7, '7 d) I CS; ? ~
5) The capital stock fluctuates much less than output and i~ largely
uncorrelated with OUtpUL tr.Ie ~ (1'; / ~ ( t<. 1- ) "'" 0
~ less than
6) Productivityis slightlyprocyclical~ut variesconsideraly
OUtpUL (JArN < ~ I C4wlAIN, ~) 7 ()
7) Wagesvary less than productivity. o-w < dit,1tI
8) The correlation between average hourly compensation and output is
essentiallyzero.20 ~(~, 0) ~ \)

9) Government expenditures are essentially uncorrelated with output. ~ ( q , ~) ~ 0

10) Imports are more strongly procyclical than exports.

C#w- (J.M,.~) ;> ~ (~, j> -;> 0

04/03/2023 7/15

---
 - - - ----
4..k, '...

ECON1220H&I * Intro Macro

Table 1.1 Chi-Wa Yuen
Cyclical Behavior of the U.5. Ecanomy: Deviations from Trend of Key Variables, 1954:1-1991:11
Cross-Corrclationof Output with:

;r{-4) x(-2) x( -1) x x(+1) x(+2) .1:(+3) .t(+4) X(+5)

Variable SD% x( -5) x(-3)

.38 .63 .85 1.0 .85 .63 .38 .16 -.02

GNP component \-,@- .02 .16
Output
Consumption expenditureS
.72 .82 t"\'1) .67c:..t:Ul.\.46
!, .u...,..t -.01
.22 -.20
.25 .42 .57
CONS 0.14 .68 .78 .77 .64 .47 .27 .06 -.11
S . .1- .40 .55
.65 .75 .78 .61 .38 . .11 . -.13 -.31
CD 4.96 .u .37 .49
Investment .76 .50 .22 -.04 -.24
.04 .19 .38 .59 .79
INV 4.11 .82 .81 .60 .35 .09 -.12
INVF 5.34 .08 .25 .43 .63
.30 .57 .79 .88 .83 .60 .46 .24
INVN 5.11 -.26 -.12 .05
.63 .39 .11 -.14 -.33 -.43
.55 .65
INVR
Ch.INV
Government purchases
10.7
@>
2.04
.42
-.03
.03
.07
-.Ol
.22
-.03
.
:.:

-.0 1
@ .51

.08
.27

.11
.04

.16
1'
-.15

.25
-.

!3
GOVT
Exponsand impons .15 .50 .54 .54 .52 .
-.29 -.10
EXP
IMP 4.88
5.53 -.48
.11
-.42
.19 .31 .45 .62 a1 7" .71 .52 .28 .04 -.18

Laborinputbasedon hourvey .82 .69 .52 .32 .11

HSHOURS0,' - -.06 .09 .30 .53 .74 @ .52 .37 .23 .09 -.05
.16 .34 .48 .63 .62
HSAVGHRS .04
0.;7
___.RIa

.04 .23 .46 .69 .76 .59 .40 .18

frF N GNP/HSHOURS'- .14 .20 .851) .00
HSEMPLMT ".g: . II> .06
-.10 .30 .33 @ -.18 -.25 -.24
Laborinputbasedon 0 ' .r 1-
establishment survey
ESHOURS 1.69 -.12 .07 .38 .54 .78 .92 .90 .78 .63 .42 .21
ESAVGHRS 0.48 .14 .26 .42 .58 .68 .62 045 .22 .05 -.15 -.30
ESMPLMT 1.41 -.19 -.01 .22 .47 .72 .89 .92 .86 .73 .55 .34
GNPIESHOURS 0.73 .35 .44 .44 .45 .34 .34 .10 -.09 -.30 -.38 -.42
Average hourly earnings
based on establishment
survey
WAGE
Average hourly compen-
€I .20 .35 .47 .58 .66 .68 .59 .46 .29 .12 -.03

sation based on nation-

al income accounts
COMP 0.55 .24 .25 .21 .14 .09 -.07 -.09 -.09 -.09 -.10

Notes: GNP-reaJ GNP. 1982$: CONS-persona1 consumption cxpendilUre. 1982$: CNDS-consumption of nondurables and services. 1982$: CD-conswnption of
durables. 1981$: INV-gross private domestic invesanent. 1981$: 1NVf-fixed invesanent. 1981$: INVN-nonresidential fixed invesanent. 1982$; INVR-residential
fixed invesancut. 1982$: Ch. JNV-change in inventories. 1982S: OOVf-govemment purchases of goods and services. 1982S: EXP--expons of goods and services. 1982$:
IMP-impons of goods and services. 1982S: HSHOURS-total bours of wOrk(Household SW'Ve)'):HSAVGHRS-avcragc weekly hours of work (Household Surveo
HSEMPLMT--cmployment (Household SW'Ve)');ESHOURS-tolll hours of worts (Establishment Survcy): ESAVGHRS-average weekly hours of work (Establish'"
SW'Ve)'):ESEMPLMT-employment {Establishment SW'Ve)'):WACiE-aYerage hourly earning. 1981$ (Estab1ishmcnt SW'Ve)');COMP-evmgc toIIl compenSAt<- ,.,..
holti'. 1982S(Nalionallncomc Accounts). The Establishment Survey sample is for 1964:I-1991:JL

Q -.v6\o.~ l: W ~
(Y ~~~
04/03/2023
~ ~ ~ I r~"'L- 8/15
 n. ~.,. _., _. .. .

ECON1220H&I
Ch. 14: * Intro
Resuscitating Real Business Cycles Macro Chi-Wa Yuen
935

8 Nondurables Consumption and Output

I:
.. 0Utput
-8
--- CanoumpIIan (NO)

.-8
47 &2 87 82 87 . 72 77 82 87 82
Dat8

30 Durab'" Consumption and Output

25
20 -OUtput
18 - - - CanoumpIIan
(D)

11 10
6
J 0
-Ii
-10
-18
40
47 82 87 82 87 72 77 82 87 82
Dat8

16 Inv8stment and Output

10

1:11
-10

.18
82
40
f 82 87. 82 87 72 77 82 87

18 Government Spending and Output

10

)-8
-10
-OUtput
-15 --- Gav'I SpendInIJ

82 87 82 87 72 77 82 87 82

Fig. 2. Cyclicalcomponentsof US expenditures.Sampleperiod is 1947:1-1996:4.All variablesare

04/03/2023 detrended using the Hodrick-Prescott filter. 9/15

-- - ---
936 R.G. King and S.T. Rebelo
ECON1220H&I * Intro Macro Chi-Wa Yuen
8 Total Hounl and output

I
..-2
0

04

.e
.e
47 52 57 82 57 72 77 82 ff1 82
DaI8

8 Productivity (Solow Idual) and Output

1! 0

1-2

04

.e

.-8
47 52 57 82 72 77 82 57 82
DaI8

Fig. 3. Cyclical component of US factors of production. Sampleperiod is 1947:1-1996:4. All variables

04/03/2023 are detrended using the Hodrick-Prescott filter. 10/15
 ECON1220H&I * Intro Macro Chi-Wa Yuen
Ch. 14: Resuscitating Real Business Cycles 937

Hours per Workerand Output

8T
4
2

j (iii .c c:r,.
04

.e
.e
47 62 fiT 62 87 72 71 62 87 82
D8I8

\ .
8.. Employment and Ouput

4
2

J o- Z Cia
04

.e

.e
47 62 51 112 87 72 71 8Z 17 82
DII8

8 Average PnxIuc:t and Output

1! 0

102

04

.e
.e
47 62 1ST 112 71 8Z 17 82

8 Real Wall" and Output

4
2

I
04

.e
.e
47 62 57 112 87 72 71 8Z 87 82
D8I8

Fig. 4. Cyclical component of US labor market measures. Sampleperiod is 1947:1-1996:4. All variables
are detrended using the Hodrick-Prescott filter.
04/03/2023 11/15
 - - -- - --

ECON1220H&I * Intro Macro Chi-Wa Yuen

SUMMARY10 Variable Direction Timing

Production.
.TheCyclical
Industrial production Procyclical Coincident
Behavioro.fKey volatiletlum nondurable
DurablegoodsindustriesaremOTe goodsandseroices
Maaoeconomic
Variables(The" Expenditure
Business
Cycle Corummpnoo Procyclical Coincident
Business fixed investment Proeyclical Coincident
Facts) Residential investment . Procyclical Leading
Inventory investment. Procyclical Leading
A_.I 1-
Gove~ent purchas~_ . Procyclical
Investment is more volatile than. consumption . JJ--~a '.

. .
Labor Market Variables
Employment Procyclical Coincident
Unemployment Countercyclical Unclassi.6.edb
Average labor producnvity Procyclical Leadinga
Real..wage Procyclical _I

Money Growth and Inflation

Money growth Procy~cal Leadingc
Inflation .
Procyclical Lagging
..
. Financial Variables
Stock prices Procyclical Leading
Nominal interest rates Procyclical Lagging
Real interest rates - Acyclical ---
· Tuning is not officially designated by Bureau of Economic:Analysis.
b Officially designated as "undassified" by the Bureau of Economic: Analysis.
c M~ money growth is a leading variable; M2 money growth leads the cycle at peaks but is coincident at
troughs and is officially "un...I"-i6~" overalL

Source: S/lTf1eY of Current BusWsS, May 1993. Industrial production: series 41 (total industrial production);
c;onswnption: series S1 (manulacturing and trade sales, 1981 dollars); business fixeci investment: series 86
(gross private nonresidential fixed investment); residential investment: series 89 (gross private residential
fixed investment); inventoty investment: series 30 (change in business inventories); employment: series 41
(employees on nonagric:ultural payroUs); unemployment: series 43 (c:ivilian unemployment rate); money
supply: series 85 (percent change in money supply, Ml) and series 102 (pe:cent change in money supply,
M2); inflation: series 120 (CPr for services, dwtge from previous month. smoothecI); stoc:1c: prices: series 19
(index of stoc:1c:
prices, 500 c:oUunon stoclcs); nominal interest rates: series 119 (Federal funds rate), series 114
(discount rate on new issues of 91~y Treasury bills), series 109 (average prime rate c:harged by banks)

04/03/2023 12/15

I
 938 L
ECON1220H&I * Intro Macro
R.G. Killg and S.T. Rebelo
Chi-Wa Yuen
Table 1
Busines$, cycle statistics for the US Economy

Standard dviation Relative standard . First-order Contemporaneous

deviation autocorrelation correlation with output

Y "1.81 -' 1.00 0.84 1.00

C 1.35 0.74 0.80 0.88
J 5.30 2.93 0.87 ....0.80
N 1.79 0.99 0.88 0.88
YIN. 1,02 0.56 0.74. 0.s5
.
w 0.68 0.38 0.66 0.12 . .
r 0.30 0.16 0.60 -0.35
A 0.98 0.54 0.74 0.78

a All variables are in logarithms (with the eXceptionof the real interest rate) and have been detrended
with the HP filter. ~ta sources are described in Stock and Watson (1999), who create<lthe real rate.
using VAR inflation expectations. Our notation in this table corresponds to that in the text, so that Y is
per capita outp~t. C is per capita consumption. J is per capita investment, N is per capita hours, w is.
the real wage (compens2tion per hour), r is the real interest rate, and A is total fact!>rproductivity.

. .

)C ~' vt It... ~~1":S

..
14: Rt's'lScilaling Real Busine.f." C.vclles
/
I ,
~~'\n1 " ,..~~
957

, Table 3
Business cyclc.statistics for basic RBC modela.b

Standard deviatiGn Relative standard First-order Contemporaneous

deviation autocorrelation correlation witlr10utput

;.!f 1.39 1.00 0.72 1.00

{C 0.61 0.44 ox 0.79 0.94.
{.J 4.09 2.95 0.71 . 0.99
. .,
N 0.67 0.48 x 0.71 0.97
.
YRI 0;75 0.54 0.76 0.98
:!
W': 0.75 0.54 0.76 0.98
r 0.05 0.04 0.71 0.95
.4 0.94 0.68 0.72 1.00

r-'" A,IIvariables have been logged (with the exception of the real interest rate) and detrended with the
:' HPfilter. .
" b The moments.in this table are population moments computed ftom the solution of the model. Prescott
(1986) produced multiple simulations, each with the same number of obServations available in the data,
04/03/2023
and reported the average HP-filtered moments across these simulat:ons. 13/15
-._--

- - ----
 .-------.. "0 __ ~._. ..___.

Ch. 14: Resuscitating Real Business Cycles 959

ECON1220H&I * Intro Macro Chi-Wa Yuen
8,. Output
,
I
4
2

I
-4 ---
-D818
Model

-8

-8
51 58 81 ee 71 78 81 ee 91 88
Data

..I
-4 --- Model
-D818
-8

-8 51 58 81 88 71 78 81 88 81 88
Data

Investment
20
15
10
5

I
..
0
oS
-10
-15

51 88 81 ee 71 78 81 88 81 88
Data

8- Consumption
5
---
-Data
Model

II 1\ 1\ f"\., +ra:
I jt
.1

-3
-4
51 58 81 ee 71 78 81 88 91 88
Data

Fig. 7. Basic model: simulated business cycles. Sample period is 1947:2-1996:4. All variables are
04/03/2023 detrended using the Hodrick-Prescott filter. 14/15
 ECON1220H&I * Intro Macro Chi-Wa Yuen

~
- -

Table1.2
CyclicalBehaviorof the ArtificialEconomy:DeviationsfromTrendof KeyVariables,150Observations
-.- - ~ H__'.~...(i

Cross-Correlation of Output with:

Variable SD% x( -5) x (-4) x( -3) x( -2) x( -1) x x(+I) x(+2) x( +3) x( +4) x( +5)

Output I j.JS-r -.049 .071 .232 .441 .698 1.0 .6983 .441 .232 .071 -.049
(0.148)
I Consumption .232 .340 .460 .592 .725 .843 .502 .229 .022 -.128 -.234
O.'Uf't (0.041)
Investment ..5J)S4- -.112 -.007 .171 .389 .664 .992 .713 .470 .270 .115 -.003
4.401 (0.646)
-.130 -.012 .152 .373 .652 .986 .715 .478 .281 .127 .010
1 Hours
, 0 'rb1 (0.083)
Productivity .055 .175 .325 .512 .732 .978 .649 .376 .160 -.002 -.122

o'o«{-, (0.068)

Note: Values in parentheses are standard deviations across simulations.

~~ fVV)e'~l~

2.~'t r~~1< 11Pt+J

~.z..
,.
04/03/2023 15/15