Discussion of
\Balance Sheet Recessions"
by Zhen Huo and Jose-Victor Rios-Rull
Dirk Krueger
University of Pennsylvania, CEPR, and NBER
MACROECONOMIC DYNAMICS WITH HETEROGENEOUS AGENTS
WORKSHOP IN LONDON
June 11, 2013
Objective and Approach of the Paper
Model-based, quantitatively accurate picture of the great recession.
View taken here: great recession
{ is caused by an exogenous nancial shock, modeled as sudden tightening of housing collateral constraints
{ is ampli ed through an endogenous fall in house prices (negative
wealth e ect, further tightening of collateral constraints)
{ is propagated through decline in aggregate demand in frictional
consumption and labor markets (potentially with sticky wages).
Motivating Macro Facts
Big increase in debt/income ratio prior to great recession. Dramatic
de-leveraging thereafter.
Big increase in house prices (and house price-rent-ratio) prior to great
recession. Big fall thereafter.
UE rate rose * 4% to 10% in great recession, then \jobless recovery".
Signi cant decline in (durable) consumption during great recession.
Total Debt Balance and its Composition
Trillions of Dollars
15
Mortgage
HE Revolving
Auto Loan
Credit Card
Trillions of Dollars
15
Student Loan
Other
2012Q1 Total: $11.44 Trillion
2011Q4Total: $11.53 Trillion
12
(3%)
12
(8%)
(6%)
(6%)
9
(5%)
9
(72%)
6
6
3
3
0
0
03:Q1
04:Q1
05:Q1
06:Q1
07:Q1
08:Q1
Source: FRBNY Consumer Credit Panel/Equifax
4
09:Q1
10:Q1
11:Q1
12:Q1
Figure 1: Price-Rent Ratios in the Data
The figure compares three measures of the price-rent ratio. The first measure (“Flow of Funds”) is the ratio of residential real estate
wealth of the household sector from the Flow of Funds to aggregate housing services consumption from NIPA. The second measure
(“Freddie”) is the ratio of the Freddie Mac Conventional Mortgage Home Price Index for purchases to the Bureau of Labor Statistics’s
price index of shelter (which measures rent of renters and imputed rent of owners). The third series (“Case-Shiller”) is the ratio of the
Case-Shiller national house price index to the Bureau of Labor Statistics’s price index of shelter. All indices are normalized to a value
of 100 in 2000.Q4. The data are quarterly from 1970.Q1 until 2008.Q4. The REITs series starts in 1972.Q4 and the Case-Shiller series
in 1987.Q1.
150
Freddie Mac
140
Case−Shiller
130
Flow of Funds
Index
120
110
100
90
80
1970
1975
1980
1985
1990
Year
1995
2000
2005
Facts on the last recession: I
6
10
4
9
2
8
0
7
−2
6
−4
5
−6
−8
2004
2006
2008
2010
2012
4
2004
Real output
2006
2008
2010
2012
Unemployment
6
30
4
20
2
10
0
0
−2
−10
−4
−20
−6
−30
−8
−10
2004
2006
2008
2010
2012
−40
2004
Consumption
Huo & Rı́os-Rull (UMN, Mpls Fed, CAERP)
2006
2008
2010
Balance Sheet Recessions
2012
. . . . . . .
.. .. .. .. ,.. .. ..
Macro Dyn with HA, LBS, 6/11/13
2 / 58
Investment
The Model in a Nutshell
Krusell-Smith economy with aggregate risk ( ; ( 0j )); frictionless
housing, and goods and labor markets with search/matching frictions.
Household problem: aggregate state ( ; S ), individual state s = ( ; e; a):
Dynamic programming problem:
V (s; ; S ) =
max
cN ;cT ;h;d;b;k;IN 0
n
u(cN ; IN ; cT ; h; d) +
o
0
0
0
E ;e; V (s ; ; S )
p(S )cN IN + cT + ph(S )h + k + & ( )b = a + 11w(S ) + 10w + q b
a0 = ph(S 0)h + R(S; S 0)k b
b
( )ph(S )h=q ( ; b)
IN = d d (D)
Thought Experiments (I Think!?)
One time (unexpected?):
{ tightening of credit constraint: from
= 0:85 to
{ increase in mortgage rate: from & = 0 to & = 30bp
Immediate vs. 4 period gradual change
Flexible vs.
xed wages
= 0:75
Key Results
De-leveraging?
Movement in house prices?
(Un-)employment dynamics?
Real output and consumption?
Dynamics of trade balance?
Comments I:
Right thought experiment? Want to tell the "story" of great recession:
{ tightening of credit constraints: Yes
{ increase in interest rates? Not so much
Right results? Mostly promising, but
{ I don't see a strong di erence in dynamics of TFP and Y =L in
data. Paper wants to get Y =L " and succeeds in model. But it's
not in their data!
30-year Fixed Rate Conventional Mortgages
11
10
9
8
7
6
5
4
3
http://www.economagic.com/ Jun 6 2013
19
90
19
92
19
94
19
96
19
98
20
00
20
02
20
04
20
06
20
08
20
10
20
12
20
14
1 Sudden change of λ,
Flex. w
0.5
0
0
−2
−0.5
Fixed w
−4
−1
−6
−1.5
−8
−2
−10
−2.5
−12
−3
−3.5
0
1
2
3
4
5
6
7
8
9
−14
0
10
1
2
3
Wealth
4
5
6
7
8
9
10
Debt
0
−1
−2
−3
−4
−5
−6
−7
−8
−9
0
1
2
3
4
5
6
7
Housing price
Huo & Rı́os-Rull (UMN, Mpls Fed, CAERP)
Balance Sheet Recessions
8
9
10
. . . . . . .
.. .. .. .. ,.. .. ..
Macro Dyn with HA, LBS, 6/11/13
35 / 58
1: Sudden change of λ.
Flex. w
Fixed w
11
0.5
0
10
−0.5
9
−1
8
−1.5
−2
7
−2.5
6
−3
−3.5
0
1
2
3
4
5
6
7
8
9
10
5
0
1
2
Real output
3
4
5
6
7
8
9
10
8
9
10
Unemployment
1
5
0
0
−5
−1
−10
−2
−15
−20
−3
−25
−4
−5
0
−30
1
2
3
4
5
6
7
8
9
10
−35
0
1
Huo & Rı́os-Rull (UMN, Mpls Fed, CAERP)
2
3
4
5
6
7
Investment
Consumption
Balance Sheet Recessions
. . . . . . .
.. .. .. .. ,.. .. ..
Macro Dyn with HA, LBS, 6/11/13
34 / 58
1. Sudden change of λ,
Flex. w
0.4
Fixed w
1.5
0.2
1
0
0.5
−0.2
−0.4
0
−0.6
−0.5
−0.8
−1
0
1
2
3
4
5
6
7
8
9
10
−1
0
1
TFP with total hours
2
3
4
5
6
7
8
9
10
9
10
Labor Productivity
1.6
0.2
1.4
0
1.2
−0.2
1
0.8
−0.4
0.6
−0.6
0.4
−0.8
0.2
−1
0
−0.2
0
1
2
3
4
5
6
7
Labor quality
Huo & Rı́os-Rull (UMN, Mpls Fed, CAERP)
8
9
10
−1.2
0
1
2
3
4
5
6
7
8
TFP with total labor inputs
Balance Sheet Recessions
. . . . . . .
.. .. .. .. ,.. .. ..
Macro Dyn with HA, LBS, 6/11/13
36 / 58
Facts on the last recession: III
3
4
2
3
1
2
0
1
−1
0
−2
−1
−3
−2
−4
2004
2006
2008
2010
2012
−3
2004
TFP with total hours
0.8
4
0.6
3
0.4
2
0.2
1
0
0
−0.2
−1
−0.4
−2
−0.6
−3
−0.8
−1
2006
2008
2010
2012
Labor productivity
−4
2004
2006
2008
2010
Labor quality
Huo & Rı́os-Rull (UMN, Mpls Fed, CAERP)
2012
−5
2004
2006
2008
2010
2012
. . . . . . .
.. .. .. .. ,.. .. ..
Macro Dyn with HA, LBS, 6/11/13
4 / 58
TFP with total labor inputs
Balance Sheet Recessions
Comments II: Modeling of Housing
Note: h is not a state variable.
In presence of rental markets complete separation of housing as nancial asset and housing as consumption good.
Convenient formulation (avoids nonconvexity in controls, additional
state variable). But it is plausible?
Also: houses in xed supply! Helps making Ph move a lot (and it does
not move nearly as much as in data).
The Model in a Nutshell
Krusell-Smith economy with aggregate risk ( ; ( 0j )); frictionless
housing, and goods and labor markets with search/matching frictions.
Household problem: aggregate state ( ; S ), individual state s = ( ; e; a):
Dynamic programming problem:
V (s; ; S ) =
max
cN ;cT ;h;d;b;k;IN 0
n
u(cN ; IN ; cT ; h; d) +
o
0
0
0
E ;e; V (s ; ; S )
p(S )cN IN + cT + ph(S )h + k + & ( )b = a + 11w(S ) + 10w + q b
a0 = ph(S 0)h + R(S; S 0)k b
b
( )ph(S )h=q ( ; b)
IN = d d (D)
Comments III: Modeling Collateral Constraint
Collateral constraint can be written as
q ( ; b)b
b
( )ph(S )h
(1 + r ) ( )ph(S )h
:
1+r
&( )
Seems right for new mortgages, not quite right for households with
existing mortgages. "Margin call" if ph #?
Makes credit (and thus consumption smoothing possibilities) too sensitive to nancial shocks ( ); & ( ): Seems crucial for the potency of
the mechanism.
Concluding Remarks
Physics: in search of the great unifying theory of everything.
A theory of everything (ToE) or nal theory is a putative theory of theoretical physics that fully explains and links together
all known physical phenomena, and predicts the outcome of any
experiment that could be carried out in principle...The central issue is how to combine general relativity and quantum mechanics.
[Wikipedia]
Concluding Remarks
This paper is equally ambitious and has the same avor.
Lots (!) of facts to be explained within uni ed theory (that combines
RBC theory with demand-driven business cycles).
Lots (!) of moving pieces. All necessary qualitatively? All useful
quantitatively?
Success? Given the results so far perhaps!?