1/18

HOUSING MARKET CYCLES, PRODUCTIVITY GROWTH,AND HOUSEHOLD DEBT

Dmitry Brizhatyuk

June 3, 2021

2/18

Slow recoveries from financial crises

1985 1990 1995 2000 2005 2010 2015 202030

35

40

45

50

55

60

65

Tho

usan

ds 2

012

US

D

US Real GDP & CBO growth projections

Real per capita GDPPotential GDP, CBO Aug. 2007Potential GDP, CBO Aug. 2010

Recoveries from financial crises tend tobe slow and incomplete

(e.g. Cerra and Saxena 2008; Reinhart andRogoff 2009; Romer and Romer 2017)

A growing literature on hysteresis

(e.g. Benigno and Fornaro 2018; Comin andGertler 2006; Queralto 2019)

3/18

Slow recoveries from financial crises

1985 1990 1995 2000 2005 2010 2015 202030

35

40

45

50

55

60

65

Tho

usan

ds 2

012

US

D

US Real GDP & CBO growth projections

Real per capita GDPPotential GDP, CBO Aug. 2007Potential GDP, CBO Aug. 2010

Big question: under what conditionshysteresis effects are most prominent?

Focus of this paper: the role of housingmarket and household debt cycles

Empirical evidence and a dynamic generalequilibrium model

4/18

Intuition

Negative house price shock→ household deleveraging

- AD-driven contraction in the short-run

- Endogenous fall in growth and a persistently lower TFP level in the long run

- Amplification through a feedback loop b/w deleveraging, house price, and growth

- Sensitivity to the initial level of household debt

- Asymmetry (positive vs negative house price shocks)

5/18

Empirical evidence

Housing market boom-and-bust cycles predict lower future productivity growth

(A) Unbalanced panel of 50 countries, 1950 - 2018:

- House price indexes

- Household debt

- Real economy indicators

- Utilization-adjusted TFP (constructed using the Imbs (1999) correction )

Two experiments:

- House price shock in a panel VAR

- Event study of housing market crashes by local projections

(B) Cross-section of US MSAs since the Great Recession

6/18

House price shock in a panel VAR

Panel VAR in levels, Cholesky identification, house price ordered last:

Utilization-adjusted TFP

5 10 15 20-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

% d

evia

tion

Household debt to GDP

5 10 15 200

0.05

0.1

0.15

0.2

0.25

0.3

0.35

, %

Real house price

5 10 15 20

years

0

0.5

1

1.5

% d

evia

tion

A rise in house prices and household debt predicts lower TFP growth in the medium run

7/18

Event study of housing market crashes

63 housing market boom-and-bust events List

Elasticities of macroeconomic variables to the house price decline during the crash:

∆hyi,t+h = αhi + αht + β

h∆pcrashi,t + X′i,tΓ

h + εhit

∆hyi,t+h = log(Yi,t+h) − log(Yi,t ), country i

∆pcrashi,t – housing crash measure (3-year price decline from the peak)

αhi , αht – country and year fixed effects

Xi,t – vector of controls List

H-period response: {βh}h=1:H

8/18

Event study of housing market crashes

Real per-capita GDP

0 2 4 6 8 10 12 14-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Real per-capita consumption

0 2 4 6 8 10 12 14-0.4

-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Real per-capita investment

0 2 4 6 8 10 12 14-1

-0.5

0

0.5

% d

evia

tion

Real per-capita capital

0 2 4 6 8 10 12 14-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Employment-to-pop. ratio

0 2 4 6 8 10 12 14-0.2

-0.1

0

0.1

0.2

% d

evia

tion

Utilization-adjusted TFP

0 2 4 6 8 10 12 14-0.2

-0.15

-0.1

-0.05

0%

dev

iatio

nHousehold debt to GDP

0 2 4 6 8 10 12 14-0.4

-0.3

-0.2

-0.1

0

0.1

, %

Full sample (90% CI)

Real house price

0 2 4 6 8 10 12 14

years

-2

-1.5

-1

-0.5

0

% d

evia

tion

Deleveraging→ persistent decrease in TFP and capital driving persistence Pre-2007 sample

9/18

General equilibrium model

Productionsector

Wholesalers

RetailersFinal good

basket

Borrowerhouseholds

Innovators

Intermediatesector

Saverhouseholds

Monetarypolicy

Equityfinancing

Labor, Capital

Labor

Blueprints

Intermediategoods

Differentiation

Packaging

Collateralizedborrowing

- Borrower-saver NK model

- Housing as collateral (Iacoviello 2005)

- Borrowing subject to an occasionallybinding constraint

- Endogenous growth through productcreation (Romer 1990)

- Experiment: a housing market crashtriggered by negative housing demandshocks (Liu et al. 2013)

10/18

Endogenous growth through innovation

Aggregate production function: Yt = F(

Kt , LtRival

factors

,

∫ Nt0

xt (ω)dω

Non-rival“ideas”

)

New “ideas” through innovation (S): ÛNt = φtSρt

Positive externality in innovation: φt = φNt (generates growth)

Monopolistic competition: xt (ω) are imperfectly substitutable →positive profit → entry subject to a sunk cost

Connection to business cycles: Entry incentives depend on cyclical conditions

11/18

Housing as collateral

max Åt∑∞

j=t βj−t [u(Cj , Lj ) + ηjg(hBj )

Utility fromhousing

]

Budget constraint: Ct + Pht (ht − ht−1) + (1 + rt−1)Bt−1Pt

= BtPt + other terms

Occasionally binding collateral constraint: Bt ≤ mPht htFraction of

housing value

Åt(β

u′ct+1u′ct

1+rtΠt+1

)= 1 − χt

Intertemporaldistortion

χt ≥ 0 ≡ Lagrange multiplier w.r.t. the collateral constraint

The rest of the model includes standard quantitative NK features: nominal rigidities,capital accumulation subject to adjustment costs, varying capital utilization, etc.

12/18

IRF matching

Crisis experiment: a sequence of negative housing preference shocks to mimic theempirical housing price decline

The resulting theoretical IRFs are used to estimate a set of quantitative parameters P

IRF matching estimator: choose P to minimize the weighted distance between empirical(ΣLP ) and theoretical (ΣDSGE ) impulse responses:

minP(ΣDSGE (P) − ΣLP ) Ω−1 (ΣDSGE (P) − ΣLP )′

Quantitative parameters: Capital adjustment costs (ψK ); R&D adjustment costs (ψN );Borrowing limit inertia (ρb); Labor disutility inertia (γ), Capital utilization parameter (c2)

13/18

Housing market crash: model vs evidence

Y

0 2 4 6 8 10 12 14

-0.4

-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

95% CI

C

0 2 4 6 8 10 12 14

-0.4

-0.3

-0.2

-0.1

0

0.1I

0 2 4 6 8 10 12 14-1

-0.5

0

0.5K

0 2 4 6 8 10 12 14-0.3

-0.2

-0.1

0

0.1

L

0 2 4 6 8 10 12 14-0.4

-0.2

0

0.2Util.-adjusted TFP

0 2 4 6 8 10 12 14-0.2

-0.15

-0.1

-0.05

0

0.05House price

0 2 4 6 8 10 12 14-2

-1.5

-1

-0.5

0

Model Data

0 2 4 6 8 10 12 14

years

-8

-6

-4

-2

0Housing pref. shock

More IRFs

14/18

Model-based decomposition of output and TFP dynamics

0 10 20 30 40 50 60quarters

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0%

dev

iatio

n

Growth accouting decomposition

LaborCapitalTFPTotal GDP response

0 10 20 30 40 50 60quarters

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0Measured TFP decompisition

Innovation effectMarkup effectUtilization effectTotal TFP response

∆GDPt = ∆TFPt + α∆KtCapital

+ (1 − α)∆LtLabor

∆TFPt = ∆ΩtMarkup

+ α∆utUtilization

+ (1 − α)∆NtInnovation

15/18

Asymmetric belief-driven boom and bust cycle

- Housing cycles driven by beliefs about future demand (Kaplan, Mitman, Violante 2020)

- Asymmetry is driven by occasionally binding collateral constants that amplify negativebut not positive shocks

- Example: unrealized positive housing demand news shock about t=12:

0 10 20 30 40 50 60-0.5

0

0.5

% d

evia

tion

TFP, innovation component

0 10 20 30 40 50 60-5

0

5

10

15, %

Household debt to GDP

0 10 20 30 40 50 60-10

0

10

20

30

40

50

% d

evia

tion

House price

0 10 20 30 40 50 60Quarters

-5

-2.5

0

2.5

% d

evia

tion

GDP

16/18

Housing market crash: main channels

Falling individualspending

Falling expectedaggregate growth

Declining aggregatehousing wealth

Tighter creditconstraint

(1) AD channel Flex price IRFs Binding ZLB IRFs

Demand effects of deleveraging

(2) Productivity growth channel No growth IRFs

Endogenous slowdown in TFP growthprolonging the crisis

(3) Fisherian debt deflation channel Housing IRFs

Negative feedback loop betweendeleveraging and the collateral price

(4) Expected income growth channelNegative feedback loop betweenexpected growth and consumption

17/18

Monetary policy and the the welfare cost of the crisis

- Counterfactual simulations under various parameters of the Taylor rule

- Welfare cost in % of the steady-state consumption

0.1 0.2 0.3 0.4

Taylor rule response to output

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

2.8

3

Tay

lor

rule

res

pons

e to

infla

tion

Baseline model

-0.2-0.18

-0.16-0.14

-0.12

-0.1

-0.1

-0.08

-0.0

8

-0.06

-0.06

-0.06

-0.04

-0.0

4

-0.04 -

0.02

-0.24

-0.22

-0.2

-0.18

-0.16

-0.14

-0.12

-0.1

-0.08

-0.06

-0.04

0.1 0.2 0.3 0.4

Taylor rule response to output

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

2.8

3

Tay

lor

rule

res

pons

e to

infla

tion

No endogenous growth

-0.0

7-0

.06

-0.0

5

-0.04-0.03

-0.0

3

-0.02

-0.0

2

-0.02

-0.24

-0.22

-0.2

-0.18

-0.16

-0.14

-0.12

-0.1

-0.08

-0.06

-0.04

18/18

Conclusion

Housing market crashes are transitory events but they can leave long-lasting scars oneconomic activity...

- ...especially in the economy with a high household debt burden

- ...especially when monetary policy focuses on inflation stabilization relative to outputstabilization and/or is constrained by the zero lower bound

- occasionally binding collateral constants make these effects asymmetric: housingmarket booms do not induce comparable increases in productivity growth

1/0

APPENDIX

2/0

Utilization-adjusted TFP

Utilization adjustment approach of Imbs (1999) based on a partial-equilibrium version of amodel from Burnside and Eichenbaum (1996)

Firms problem:

maxKt ,ut ,et

[Zt (utKt )α (etLt )1−α − w(et )Lt − (rt + δuφt )Kt

]Households problem:

max{Ct+j ,Lt+j ,et+j }∞j=0

Åt

∞∑j=0

β j

(ln(Ct ) −

L1+�t1 + �

−e1+ψt1 +ψ

)s.t. Ct ≤ w(et )Lt

Capital utilization: ut =(

Yt/KtY /K

) δr+δ Labor effort: et =

(Yt/CtY /C

) 11+ψ

Back

3/0

US factor utilization

1960 1970 1980 1990 2000 2010-4

-3

-2

-1

0

1

2

3

4

5

util

izat

ion,

%Basu, Fernald, and KimballAuthor`s calculations

Back

4/0

Event study: sample of housing market crashes

Peak Trough 3 years Total Peak Trough 3 years Total Peak Trough 3 years Total

BEL 1979 1985 -26% -38% GBR 1973 1977 -24% -29% NLD 1978 1985 -34% -48%BGR 1996 2002 -40% -52% GBR 1989 1996 -22% -30% NLD 2008 2013 -11% -26%BGR 2008 2013 -39% -44% GBR 2007 2012 -16% -23% NOR 1987 1992 -29% -43%BRA 2014 2017 -16% -16% GRC 2007 2017 -15% -45% NZL 1974 1980 -18% -36%CAN 1981 1985 -26% -30% HKG 1981 1984 -47% -47% NZL 2007 2009 -11% -11%CHE 1973 1976 -20% -20% HKG 1997 2003 -42% -57% PER 1999 2003 -15% -29%CHE 1990 2000 -20% -33% HRV 1999 2002 -14% -14% PHL 1996 2004 -36% -53%CHE 1959 1961 -12% -12% HRV 2009 2015 -19% -24% POL 2010 2013 -16% -16%COL 1989 1992 -13% -13% HUN 2006 2013 -17% -37% PRT 1992 1996 -11% -12%COL 1995 2003 -14% -35% IRL 2006 2012 -30% -46% RUS 2008 2011 -33% -33%CZE 2008 2013 -15% -19% ISL 2007 2010 -32% -32% SGP 1983 1986 -31% -31%DEU 1981 1987 -11% -14% ITA 1981 1986 -21% -31% SGP 1996 1998 -32% -34%DNK 1979 1982 -34% -34% ITA 1992 1997 -14% -26% SRB 2010 2013 -29% -29%DNK 1986 1993 -18% -31% JPN 1974 1977 -23% -23% SVK 2008 2012 -21% -26%DNK 2007 2012 -19% -28% JPN 1991 2012 -13% -51% SVN 2011 2014 -21% -21%ESP 1991 1996 -13% -15% KOR 1991 1998 -25% -43% SWE 1979 1985 -26% -35%ESP 2007 2014 -15% -36% LTU 2007 2010 -43% -43% SWE 1990 1993 -30% -30%EST 2007 2009 -51% -52% LUX 1980 1984 -22% -23% THA 2006 2009 -30% -30%FIN 1974 1979 -25% -31% LVA 2007 2010 -47% -47% USA 2006 2012 -14% -26%FIN 1989 1993 -42% -47% MYS 1997 1999 -15% -18% ZAF 1984 1987 -39% -39%FRA 1980 1985 -11% -16% NLD 1964 1966 -27% -29% ZAF 2007 2012 -16% -19%

- 63 events in total, 39 before 2006,

- Median duration: 5 years peak to though, -30.6% price decline Back

5/0

Local projections, control variables

Value at the peak and one lag:

- Growth rate of the response variable

- Real per-capita investment growth

- GDP-deflator inflation rate

- Real house price growth rate

- Net exports to GDP

Value at the peak:

- Investment to GDP

- Exchange rate regime indicator(Ilzetzki et al. 2019)

- Systematic banking & currency crisesindicator (Laeven and Valencia 2012)

Back

6/0

Event study of housing market crashes, pre-2007 sample

Real per-capita GDP

0 2 4 6 8 10 12 14-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Real per-capita consumption

0 2 4 6 8 10 12 14-0.4

-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Real per-capita investment

0 2 4 6 8 10 12 14-1

-0.5

0

0.5

% d

evia

tion

Real per-capita capital

0 2 4 6 8 10 12 14-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Employment-to-pop. ratio

0 2 4 6 8 10 12 14-0.2

-0.1

0

0.1

0.2

% d

evia

tion

Utilization-adjusted TFP

0 2 4 6 8 10 12 14-0.25

-0.2

-0.15

-0.1

-0.05

0%

dev

iatio

nHousehold debt to GDP

0 2 4 6 8 10 12 14-0.4

-0.3

-0.2

-0.1

0

0.1

% d

evia

tion

Real house price

0 2 4 6 8 10 12 14

years

-2

-1.5

-1

-0.5

0

% d

evia

tion

Pre-2007 sample (90% CI) Full sample

Baseline results are not driven by the GFC Back

7/0

Housing market crash and productivity growth across US MSAs

-50 -40 -30 -20 -10 0 10

House price %, Q1 2007 - Q1 2010

-10

-5

0

5

10

15

20 L

abor

pro

duct

ivity

%, 2

007

- 20

17US states

-60 -50 -40 -30 -20 -10 0 10

House price %, Q1 2007 - Q1 2010

-20

-10

0

10

20

30

Lab

or p

rodu

ctiv

ity %

, 200

7 -

2017

US MSAs

Higher exposure to the crash, slower post-crisis labor productivity growth

8/0

Housing market crash and productivity growth across US MSAs

∆ 20072017

log (Y /L)i = α + η ∆ 20072010

logPHi + X′i Γ + εi

-0.1 0 0.1 0.2 0.3 0.4

OLS + controls

IV 1 + controls

IV 1

OLS

IV 2

IV 2 + controls

95% CI

Higher exposure to the crash,slower labor productivity growth

Can explain >40% of the US GDPgap relative to the pre-GFC trend

Identification

IV 1: housing supply elasticityIV 2: regional sensitivity

9/0

Housing market boom and productivity growth across US MSAs

0 20 40 60 80 100 120

House price %, Q1 2002 - Q1 2007

0

5

10

15

20

25

30 L

abor

pro

duct

ivity

%, 2

002

- 20

07US states

0 20 40 60 80 100 120 140

House price %, Q1 2002 - Q1 2007

-20

-10

0

10

20

30

Lab

or p

rodu

ctiv

ity %

, 200

2 -

2007

US MSAs

No relation between the house price growth and productivity growth during the boom

10/0

Production sector, full problem

Production function: Ft = Zt(K̃ αt L

1−αt

)1−ξ (∫ Nt0 xt (ω)

1ν dω

)νξmax

{xt+j (ω),Lt+j ,Kt+j }∞j=0Åt

∞∑j=0

ΛBt,t+j

[pFt Ft+j − RKt+j K̃t+j −Wt+jLt+j −

∫ Nt0

pxt+j (ω)xt+j (ω)dω]

Labor demand: Wt = pFt (1 − α)(1 − ξ)FtLt

Capital demand: RKt = pFt α(1 − ξ)

FtK̃t

Intermediate-good demand: pxt (ω) = pFt ξFtXt

xt (ω)1−νν

11/0

Intermediate sector, full problem

maxpxt (ω)

[(pxt (ω) − A−1)xt (ω)

]s.t. pxt (ω) = pFt ξ

FtXt

xt (ω)1−νν

Optimal relative price: pxt = ν A−1

Optimal quantity: xt =(Aξν

) 11−ξ(pFt Zt )

11−ξ N

νξ−11−ξ

t K̃αt L

1−αt

Real profit: dt =ν − 1ν

pxt xt =ν − 1

Axt

12/0

Innovators, full problem

Individual production function: N iet = φit S

it Aggregate productivity: φt = φ

NtNρt S

1−ρt

max{Sit+j }

∞j=0

Åt

∞∑j=0

ΛBt,t+j

(pi,bt+jφ

it+jS

it+j − (1 + ACS,t+j )S

it+j

)

Optimal blueprint price: pi,bt =1φ it

(1 + ACS,t + AC ′S,tS

it − Åt

(ΛBt,t+1AC

′S,t+1S

it+1

))

13/0

Downstream sectors: retailers and wholesalers, full problem

max{P(j)t+k }∞k=0

Åt

∞∑k=0

Λt,t+k

[Pt+k (j)

PtYt+k (j) −

PFt+kPt

Ft+k (j) − ACp,k (j) − Γ], s.t

Production fucntion: Yt (j) = Ft (j)

Retailers demand: Yt (j) =(Pt (j)Pt

)−ηYt

Price adjustment cost: ACp,t (j) =ψp

2

(Pt (j)

Pt−1(j)Π− 1

)2Yt

Pt (j) = µtPFt µt =η

(η − 1) +ψp ΠtΠ(ΠtΠ − 1

)−ψpÅtΛt,t+1

(Πt+1Π − 1

)Πt+1Π

Yt+1Yt

14/0

Households: savers

max{CSj ,L

Sj ,h

Sj ,B

Sj+1 }

∞j=t

Åt

∞∑j=t

βj−tS

(u(CSj , L

Sj ) + g(h

Sj )

)s. t.

Budget constraint: CSt + Pht ∆h

St + (1 + rt−1)

BStPt

= WtLSt +BSt+1Pt

15/0

Households: borrowers

max{CBj ,L

Bj ,h

Bj ,B

Bj+1,Ij ,Kj+1,ιj+1,uj }

∞j=t

Åt

∞∑j=t

βj−tB

(u(CBj , L

Bj ) + g(h

Bj )

)s. t.

Budget constraint: CBt + It + Pht ∆h

Bt + (1 + rt−1)

BBtPt

+ ιt+1vt (Nt + Net ) =

= ιt (vt + dt )Nt +WtLBt + RKt Kt +BBt+1Pt

Capital accumulation: Kt+1 = (It − ACI,t ) + (1 − δK (ut ))Kt

Collateral constraint: BBt ≤ρBBBt−1Πt

+ (1 − ρB)mPht hBt

Capital utilization: δK (ut ) = δK + c1(ut − 1) + (c2/2)(ut − 1)2

16/0

Calibration summary

Calibrated parameters Value Source / target

βS Savers discount factor 0.9968 4% annual real interest rateβB Borrowers discount factor 0.9918 βB = βS − 0.005σ Relative risk aversion 2 Conventional

1/�L Elasticity of labor supply 1 Conventionalν/(ν − 1) Intermediate-good elasticity of subst. 1.6 BGP requirement ξ(ν − 1)/(1 − ξ) = 1 − αη Retail-good elasticity of subst. 11 10% steady-state markup

1/A Intermediate sector marginal cost 1 Normalizationρ R&D output elasticity 0.8 Comin and Gerlter (2006)

δN Intermediate sector exit rate 0.025 Bilbiie et al. (2012)φy ; φπ ; ρr Taylor rule: output; inflation: inertia 0.5/4; 1.5; 0.7 Conventional

Z̄ Final sector productivity 1.74 Normalization, YGDP = 1ψp Price adjustment cost 120 4-quarter average Calvo price ridigity equivalent

−1/�h Elasticity of housing demand -0.2 Hanushek and Quigley (1980)m Max leverage 0.75 Warnock and Warnock (2008)α Capital share 0.4 Data median, PWT 9.1

δK Steady state capital depreciation 0.025 Conventionalφ R&D productivity 0.11 Annual per-capita TFP growth = 0.8% (data median, PWT 9.1)

κ Share of housing in utility 0.03 Mortgage debt to GDP = 0.55

ξ Intermediate good share 0.5 Comin and Gertler (2006)

17/0

Utility function

GHH preference: u(CHt , LHt ) =((

CHt − Υt (LHt )1+�L/(1 + �L))1−σ

− 1)/(1 − σ)

Housing utility: g(hHt ) = (hHt )1−�h/(1 − �h)

Labor supply: Wt = Υt (LHt )�L Υt = Υγt−1N

1−γt

Time-varying disutility of labor (Queralto 2019; Jaimovich and Rebelo 2009)

BGP with constant hours exists but the short-run effect of growth on labor supply is limited

18/0

Baseline simulation, extended set of impulse responses

0 10 20 30 40 50 60-0.4

-0.2

0%

dev

iatio

nY

0 10 20 30 40 50 60-0.3

-0.2

-0.1

0C

0 10 20 30 40 50 60-0.4

-0.2

0

0.2L

0 10 20 30 40 50 60-3

-2

-1

0Debt

0 10 20 30 40 50 600.018

0.02

0.022

0.024

leve

l

OBC multiplier

0 10 20 30 40 50 60-0.5

0

0.5Profit

0 10 20 30 40 50 60-0.6

-0.4

-0.2

0Innovation spending

0 10 20 30 40 50 60-0.15

-0.1

-0.05

0Util.-adjusted TFP

0 10 20 30 40 50 60-1

-0.5

0I

0 10 20 30 40 50 60-0.2

-0.1

0K

0 10 20 30 40 50 60-3

-2

-1

0Borrowers housing wealth

0 10 20 30 40 50 60-2

-1

0Housing price

0 10 20 30 40 50 601.018

1.019

1.02

1.021

leve

l

Inflation

0 10 20 30 40 50 601.056

1.058

1.06

1.062

leve

l

Policy rate

0 10 20 30 40 50 60

Quarters

-10

-5

0Housing pref. shock

Back

19/0

Aggregate demand channel: baseline vs flexible price economy

0 10 20 30 40 50 60-0.4

-0.2

0

% d

evia

tion

Y

Baseline Flexible prices

0 10 20 30 40 50 60-0.4

-0.2

0

0.2C

0 10 20 30 40 50 60-0.4

-0.2

0

0.2L

0 10 20 30 40 50 60-3

-2

-1

0Debt

0 10 20 30 40 50 600.018

0.02

0.022

0.024

leve

l

OBC multiplier

0 10 20 30 40 50 60-0.6

-0.4

-0.2

0Innovation spending

0 10 20 30 40 50 60-0.2

-0.1

0K

0 10 20 30 40 50 601.015

1.02

1.025

leve

l

Inflation

0 10 20 30 40 50 60-0.15

-0.1

-0.05

0Util.-adjusted TFP

0 10 20 30 40 50 60

Quarters

-10

-5

0Housing pref. shock

Nominal frictions matter Back

20/0

Aggregate demand channel: baseline vs binding ZLB

0 10 20 30 40 50 60-1

-0.5

0

% d

evia

tion

Y

Baseline Binding ZLB for the first 8 quarters

0 10 20 30 40 50 60-1

-0.5

0C

0 10 20 30 40 50 60-1

-0.5

0

0.5L

0 10 20 30 40 50 60-3

-2

-1

0Debt

0 10 20 30 40 50 600.018

0.02

0.022

0.024

leve

l

OBC multiplier

0 10 20 30 40 50 60-1.5

-1

-0.5

0Innovation spending

0 10 20 30 40 50 60-0.3

-0.2

-0.1

0K

0 10 20 30 40 50 601.015

1.02

1.025

leve

l

Inflation

0 10 20 30 40 50 60-0.3

-0.2

-0.1

0Util.-adjusted TFP

0 10 20 30 40 50 60

Quarters

-10

-5

0Housing pref. shock

The amplification role of the binding zero lower bound constraint Back

21/0

Productivity growth channel: baseline vs no growth response

0 10 20 30 40 50 60-0.4

-0.2

0

% d

evia

tion

Y

Baseline No response of endogenous growth

0 10 20 30 40 50 60-0.3

-0.2

-0.1

0C

0 10 20 30 40 50 60-0.4

-0.2

0

0.2L

0 10 20 30 40 50 60-3

-2

-1

0Debt

0 10 20 30 40 50 600.018

0.02

0.022

0.024

leve

l

OBC multiplier

0 10 20 30 40 50 60-0.6

-0.4

-0.2

0Innovation spending

0 10 20 30 40 50 60-0.2

-0.1

0K

0 10 20 30 40 50 601.018

1.02

1.022

leve

l

Inflation

0 10 20 30 40 50 60-0.2

-0.1

0

0.1Util.-adjusted TFP

0 10 20 30 40 50 60

Quarters

-10

-5

0Housing pref. shock

Endogenous productivity growth is key for generating the empirically-relevant persistentresponse of TFP, consumption, and output Back

22/0

Fisherian debt deflation: details of the housing mrkt dynamics

0 10 20 30 40 50 60-0.6

-0.4

-0.2

0

0.2

% d

evia

tion

Housing units,borrowers

0 10 20 30 40 50 60-2.5

-2

-1.5

-1

-0.5

0

Housing wealth,borrowers

0 10 20 30 40 50 60-2.5

-2

-1.5

-1

-0.5

0

Housing wealth,savers

0 10 20 30 40 50 60-2

-1.5

-1

-0.5

0Housing price

0 10 20 30 40 50 600.019

0.02

0.021

0.022

0.023

leve

l

OBC multiplier

0 10 20 30 40 50 60

Quarters

-8

-6

-4

-2

0Housing pref. shock

- The aggregate shock has an asymmetric effect across borrowers and savers

- Credit-contained borrowers reduce their housing demand by more than savers

- GE effects amplify the fall in borrowers housing wealth and exacerbate deleveraging

Back

AppendixAppendix