Macroeconomic Costs of in a Low -rate Environment

Jack Chen1 Daria Finocchiaro2 Jesper Lindé3 Karl Walentin4

1IMF 2Uppsala University and Riksbank 3IMF and CEPR 4Riksbank

Fourth Annual ECB Macroprudential Policy and Research Conference

Fourth Annual ECB Macroprudential Policy and Research Conference 1 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Household indebtedness have risen to unprecedented levels Raises concerns about their vulnerability to normalized interest rates...

Elevated household debt and house prices could pose risks for financial stability. Low interest rates keep the debt-service-to-income ratio at sustainable levels. But, low nominal rates also mean less room for monetary interventions. Various macroprudential policies (MPP) have been put forward to stem the elevated vulnerabilities.

Fourth Annual ECB Macroprudential Policy and Research Conference 2 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 A phenomenon that is shared among many countries...

House prices Households’ debt-to-income ratio

3.2 350

2.8 300

2.4 250

2.0 200 1.6

150 1.2

0.8 100

0.4 50 90 92 94 96 98 00 02 04 06 08 10 12 14 16 90 92 94 96 98 00 02 04 06 08 10 12 14 16

Real mortgage rate Annual rate of

20 12

10 16

8 12

6 8 4

4 2

0 0

-4 -2 90 92 94 96 98 00 02 04 06 08 10 12 14 16 90 92 94 96 98 00 02 04 06 08 10 12 14 16

Sweden Nor way Denmark UK USA Australia Canada

Residential Investment Fourth Annual ECB Macroprudential Policy and Research Conference 3 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Research question: MPP, MP and ZLB

Are some macropru instruments better than others in a low- environment?

Fourth Annual ECB Macroprudential Policy and Research Conference 4 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Research question: MPP, MP and ZLB

Are some macropru instruments better than others in a low-interest rate environment? “Better”= Less output loss for a given debt reduction (minimum “debt-sacrifice ratio”)

Fourth Annual ECB Macroprudential Policy and Research Conference 5 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 What we do

Quantify short-run and long-run macroeconomic effects of MPP in a low-real interest rate/high-debt environment. Compare contractionary effects of tightening various MPP tools when central bank can not provide accommodation: Loan-to-value (LTV) Loan-to-income (LTI) Debt-service-to-income (DSTI) Mortgage interest deductibility (MID) removal ...i.e. only borrower-based measures. Welfare analysis: MPP tightening beneficial in the long run.

Fourth Annual ECB Macroprudential Policy and Research Conference 6 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Literature review

Housing and the macroeconomy: Iacoviello (2005), Iacoviello and Neri (2010), Justiniano Primiceri and Tambalotti (2015).

Monetary policy transmission mechanism and HH debt: Garriga et al. (2017), Gelain et al. (2017), Pietrunti and Signoretti (2018), Hedlund et al. (2017), Calza et al. (2013), Cloyne et al. (2018), Flodén et al. (2018).

Interaction between MP and MPP: Lambertini et al. (2013), Angelini et al. (2014), Alpanda and Zubairy (2017), Gelain and Ilbas (2017), De Paoli and Paustian (2017).

MPP at the ZLB: Ferrero et al. (2018), Rubio and Yao (2018), Mendicino et al. (2019), Korinek and Simsek (2016), Farhi and Werning (2016).

Fourth Annual ECB Macroprudential Policy and Research Conference 7 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 The model - an extension of Iacoviello and Neri (2010)

Fourth Annual ECB Macroprudential Policy and Research Conference 8 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 The model - an extension of Iacoviello and Neri (2010)

Labor

Intermediate input Borrowers

Long-term Repo collateralized Housing Non Household sector Land debt Housing

Savers

Capital

Final Good +Housing

Fourth Annual ECB Macroprudential Policy and Research Conference 9 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Key extensions: housing transaction costs and long-term debt

1 Housing transaction costs Make housing an illiquid asset Generates reasonable consumption responses of borrowers (Target: Cloyne et⇒ al., 2018)

2 Long-term debt (Alpanda and Zubairy, 2017): Realistic Generates gradual debt responses to MP shocks Stock of debt evolution: Dt = (1 κ) Dt 1 + Lt − −

Fourth Annual ECB Macroprudential Policy and Research Conference 10 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 LTI 2 LTI : Lt θ wt Nt + HEWt ≤ t Labor income DSTI wt Nt 3 DSTI L | {z } HEW : t θt F + t ≤ (1 τt )r + κ − t Labor income/Debt Service | {z }

The various borrowing constraints

LTV N 1 LTV : Lt θ q IHt +HEWt ≤ t t New loans Housing collateral |{z} | {z }

N HEWt = γ[qt (1 δh)Ht 1 (1 κ)Dt 1] − − − − −

Fourth Annual ECB Macroprudential Policy and Research Conference 11 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 DSTI wt Nt 3 DSTI L HEW : t θt F + t ≤ (1 τt )r + κ − t Labor income/Debt Service | {z }

The various borrowing constraints

LTV N 1 LTV : Lt θ q IHt +HEWt ≤ t t New loans Housing collateral LTI 2 LTI : |{z}Lt θ w|t N{zt +} HEWt ≤ t Labor income | {z }

N HEWt = γ[qt (1 δh)Ht 1 (1 κ)Dt 1] − − − − −

Fourth Annual ECB Macroprudential Policy and Research Conference 11 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 The various borrowing constraints

LTV N 1 LTV : Lt θ q IHt +HEWt ≤ t t New loans Housing collateral LTI 2 LTI : |{z}Lt θ w|t N{zt +} HEWt ≤ t Labor income DSTI wt Nt 3 DSTI L | {z } HEW : t θt F + t ≤ (1 τt )r + κ − t Labor income/Debt Service

| N{z } HEWt = γ[qt (1 δh)Ht 1 (1 κ)Dt 1] − − − − −

Fourth Annual ECB Macroprudential Policy and Research Conference 11 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Monetary and macroprudential policy

Monetary Policy Central bank reacts to inflation, the output gap and output growth.

rπ (1 ρ) 1 + πt rY r∆Y − ρ Rˆ t = R¯ Y ∆Y R exp (εr ,t ) 1 + π GAP,t t t 1     − Subject to a ZLB constraint

Rt = max 1, Rˆ t 

Macroprudential Policy Tools: LTV, LTI, DSTI and MID Interest rate deductions financed with lump-sum paid by borrowers, period-by-period

M Dt 1 Tt = τt rt 1 − − Pt

Fourth Annual ECB Macroprudential Policy and Research Conference 12 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Calibration: LTV vs. LTI

Data source: Swedish FSA Mortgage Survey, average 2015-2017

Fourth Annual ECB Macroprudential Policy and Research Conference 13 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Drivers of increase in household debt

Low debt vs. High debt 1990’s 2010’s Resulting LTI increase Realrate 3% 0.5% 54% LTV 75% 85% 24% HEWfraction 0.015 0.021 12% Inflation 2.0% 1.5% 9%

Long-run equilibrium 1990’s 2010’s LTI borrowers 217% 433% DSTI(aftertax)borrowers 14.2% 19.1% Interest (after )/income of borrowers 7.67% 6.08% Non-residential investment /GDP 17.1% 20.9% Residential investment /GDP 3.0% 5.2% House prices (∆ from 1990’s to 2010’) 36.5%

More LTV

Fourth Annual ECB Macroprudential Policy and Research Conference 14 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Results: Long-term effects of MPP instruments (% change)

2010’s Constraint LTV LTI DSTI Tightening LTV MID LTI DSTI LTI aggregate -10.2 -10.2 -10.2 -10.2 Output -0.32 -0.37 -0.56 -0.56 Consumption -0.07 -0.08 -0.56 -0.56 Non-residential investment -0.20 -0.24 -0.56 -0.56 Residential investment -3.05 -3.56 -0.58 -0.58 House prices -1.21 -1.41 -0.07 -0.07 DSTI (after tax) of borrowers -10.2 2.09 -10.2 -10.2 Consumption of borrowers 0.98 1.11 -0.33 -0.33 Housing of borrowers -7.94 -9.30 -0.27 -0.27 Hours worked of borrowers -1.11 -1.27 -1.79 -1.79 Income of borrowers -0.38 -0.44 -0.56 -0.56

Fourth Annual ECB Macroprudential Policy and Research Conference 15 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 MPP in the short-run: simulations set-up

We assume that the economy is driven to the ZLB by “a mix of adverse shocks”: Perturbation setting: no need to specify which shocks, only the path of the matters (Erceg and Lindé, 2014). Assume that MPP actions not large enough to impact ZLB duration.

Impulse responses are constructed as follows Baseline: MP is constrained by the ZLB for 8 quarters. Scenario: Add a macroprudential shock to the system. The IRFs we plot are:

IRF = Scenario-Baseline

Fourth Annual ECB Macroprudential Policy and Research Conference 16 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Dynamic effects of MPP at the ZLB Aggregate effects of LTV tightening in a liquidity trap

1. Nom Policy Rate (APR, dev from baseline) 2. Inflation (APR) 3. Output (Dev from Trend)

0.1 0

0

•0.5 0

•0.2 •1

•0.1

•0.4 •1.5 t t t n n n e e e c c c

r •0.2 r r

e e e •2 P P •0.6 P

•2.5 •0.3 •0.8

•3

•0.4 •1 •3.5

Low Initial Debt High Initial Debt •0.5 •1.2 •4 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 17 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Dynamic effects of MPP at the ZLB Effects of LTV tightening: dissecting the mechanism

Effects of LTV tightening at the ZLB notably larger when debt is high Two mechanisms: An MPP tightening in this setting is more contractionary and therefore requires a larger dose of monetary accommodation. Monetary policy more potent and hence monetary constraints have larger adverse effects

Fourth Annual ECB Macroprudential Policy and Research Conference 18 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Short-run effects of MPP tools: LTV, LTI and DSTI tightening

1. Nom Policy Rate (APR, dev from baseline) 2. Inflation (APR) 3. Output (Dev from Trend) 0.1

0 0

0 •0.5

•0.2

•1 •0.1

•0.4 •1.5 t t t n n n e e e c c c

r •0.2 r r e e e P P •0.6 P •2

•2.5 •0.3 •0.8

•3

•0.4 •1 •3.5 LTV tightening LTI tightening DSTI tightening •0.5 •1.2 •4 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 19 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Short-run effects of MPP tools: LTV, LTI and DSTI tightening

1. Aggegate Consumption 2. Savers Consumption 3. Borrowers Consumption 0 0.8 0 •0.5 0.6 •2

t t 0.4 t n •1 n n e e e

c c c •4 r r 0.2 r e •1.5 e e P P P 0 •6 •2 •0.2 •8 •2.5 •0.4 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter 4. Non•residential Investment 5. Residential Investment 6. Housing Price 2 0 0

1 •5 •1 t t t

n n n •2 e e •10 e c c c r r r

e 0 e e •3 P P •15 LTV tightening P LTI tightening •4 •1 •20 DSTI tightening •5

0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter 7. Savers Housing 8. Borrowers Housing 9. LTI of borrowers (Perc. Points) 0 20 1.5 •1 0 t t •2 t n n n

e 1 e e c c c r r r

e e •3 e

P P P •20 0.5 •4 •40 •5 0 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 20 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Short-run effects of various MPP tools

In the short-run, LTV tightening more contractionary than LTI/DSTI: Due to feedback through house prices: The fall in house prices amplifies the contraction in borrowing capacity, which then further reduce borrowers’demand for both housing and other consumption

Fourth Annual ECB Macroprudential Policy and Research Conference 21 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 MID Removal: Aggregate effects - similar to LTV

1. Nom Policy Rate (APR, dev from baseline) 2. Inflation (APR) 3. Output (Dev from Trend) 0.1

0 0

0 •0.5

•0.2

•1 •0.1

•0.4 •1.5 t t t n n n e e e c c c

r •0.2 r r e e e P P •0.6 P •2

•2.5 •0.3 •0.8

•3

•0.4 •1 •3.5

•0.5 •1.2 •4 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 22 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Welfare, LTV and the ZLB

Total Welfare Savers Welfare Borrowers Welfare

•620 •76.2 •2.59 •76.3 •622 •76.4 •2.595 •76.5 •624 •76.6 •2.6

•626 •76.7

•2.605 No ZLB •76.8 ZLB •628 •76.9 0.65 0.7 0.75 0.8 0.85 0.9 0.65 0.7 0.75 0.8 0.85 0.9 0.65 0.7 0.75 0.8 0.85 0.9 LTV LTV LTV

Output Gap Savers Consumption (in levels) Borrowers Consumption (in levels) 0.3 Accounting for ZLB10 moves optimal LTV from15 0.84 to 0.7

0.25 Fraction of periods at the ZLB = 0.06, for LTV = 0.7 8

0.2 10 6

0.15 Fourth Annual ECB Macroprudential Policy and Research Conference 23 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36

4

Frequency Frequency Frequency 0.1 5

2 0.05 No ZLB ZLB 0 0 0 •40 •30 •20 •10 0 10 1.3 1.4 1.5 1.6 1.7 1.8 0.2 0.3 0.4 0.5 0.6 0.7 Outcome Outcome Outcome Total Welfare Savers Welfare Borrowers Welfare

•620 •76.2 •2.59 •76.3 •622 •76.4 •2.595 ZLB yields skewed distribution •76.5 •624 •76.6 Simulation•2.6 for LTV=0.7

•626 •76.7

•2.605 No ZLB •76.8 ZLB •628 •76.9 0.65 0.7 0.75 0.8 0.85 0.9 0.65 0.7 0.75 0.8 0.85 0.9 0.65 0.7 0.75 0.8 0.85 0.9 LTV LTV LTV

Output Gap Savers Consumption (in levels) Borrowers Consumption (in levels) 0.3 10 15

0.25 8

0.2 10 6

0.15

4

Frequency Frequency Frequency 0.1 5

2 0.05 No ZLB ZLB 0 0 0 •40 •30 •20 •10 0 10 1.3 1.4 1.5 1.6 1.7 1.8 0.2 0.3 0.4 0.5 0.6 0.7 Outcome Outcome Outcome

Fourth Annual ECB Macroprudential Policy and Research Conference 24 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 MPP may be welfare enhancing Restricting debt ex-ante is optimal in this class of models

In our model, two externalities motivate the use of MPP to curb household debt:

1 Pecuniary (Lorenzoni, 2008) Effects of house prices on collateral constraints not internalized by atomistic agents. Relies on the existence of a borrowing constraint.

2 Aggregate Demand (Farhi and Werning, 2016, Korinek and Simsek, 2016) Households take financial decisions based on private rather than social marginal utilities. Changes in the distribution of wealth affect demand but this is not internalized by atomistic agents. The key friction is a constraint on monetary policy (ZLB).

Fourth Annual ECB Macroprudential Policy and Research Conference 25 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Conclusions

Established plausible factors that doubled LTI ratio in recent decades Mainly the fall in the real mortgage rate and the increase in maximum LTV The current high household indebtedness and low interest rate makes macroeconomy sensitive to tightening of MPP In the long-run, similar small output effects of various MPP tools. In the short-run, contractionary effects from MPP can be substantial when the ZLB is binding LTV or MID tightening is associated with large output costs: Financial accelerator: negative feedback through house prices. MID removal has similar effects as LTV even when nominal rates are low. LTI/DSTI tightening is less contractionary: Crucial mechanism: avoids negative feedback effects of housing prices via the collateral constraint.

Fourth Annual ECB Macroprudential Policy and Research Conference 26 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Conclusions (part II)

Key policy messages: Need to think carefully about monetary constraints and initial debt levels when designing MPP. The macroeconomic costs of MPP tools could differ substantially in the short-run. Welfare analysis is in favor of restricting debt The presence of the ZLB matters for optimal LTV High household debt and ZLB a toxic combination

Fourth Annual ECB Macroprudential Policy and Research Conference 27 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Thank you!

Fourth Annual ECB Macroprudential Policy and Research Conference 28 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Calibration: LTV Ratio

Back

Fourth Annual ECB Macroprudential Policy and Research Conference 29 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Calibration

Take most parameters from Iacoviello-Neri Four parameters set to match moments:

Parameterexplanation Value Momentmatched Housing preference weight, savers jP 0.1235 Residential investm. / GDP Housing preference weight, borrowers jI 0.2316 LTI of borrowers 1990’s Housing adjustment costs φh 10 C-response of borrowers to MP shock HEW fraction 2010’s γ 0.021 Doubling of LTI 1990’sto 2010’s

Fourth Annual ECB Macroprudential Policy and Research Conference 30 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Calibration - remaining housing parameters

Description Symbol Value Amortization rate on HH loans κ 0.0075 Share refinancing every quarter Φ 0.3 Sources: Swedish credit registry data, Swedish FSA Mortgage Survey

Back

Fourth Annual ECB Macroprudential Policy and Research Conference 31 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Steady state very similar across models

2010’s LTV LTI DSTI LTI borrowers 433 433 433 DSTI (after tax) borrowers 19.1 19.1 19.1 Interest (after tax)/income of borrowers 6.08 6.08 6.07 Non-residential investment /GDP 20.9 20.9 21.0 Residential investment /GDP 5.2 4.4 4.4 House prices (%∆ from 1990’s to 2010’) 36.5 34.4 34.4

Fourth Annual ECB Macroprudential Policy and Research Conference 32 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Dynamic effects of MPP at the ZLB Effects of permanent LTV tightening: digging deeper

1. Aggegate Consumption 2. Savers Consumption 3. Borrowers Consumption 0 0.1 0 •0.5 0 •2 t t t

n •1 n •0.1 n e e e c c c

r r r •4 e •1.5 e •0.2 e P P P •6 •2 •0.3 •0.4 •8 •2.5 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter 4. Non•residential Investment 5. Residential Investment 6. Housing Price

0.5 •5 •2

t 0 t t

n n •10 n •3 e e e c c c r •0.5 r r e e e

P P •15 P •4 •1 Low Initial Debt •20 High Initial Debt •1.5 •5

0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter 7. Savers Housing 8. Borrowers Housing 9. DTI of borrowers (Perc. Points) 1.5 •1 20 10 •2 t t t

n 1 n n

e e e 0 c c c r r •3 r e e e

P P P •10 0.5 •4 •20

•5 •30 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 33 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Explanation 1: LTV tightening needs more accommodation Aggregate effects of an LTV tightening in alternative debt environments when monetary policy is unconstrained

1. Nom Policy Rate (APR, dev from baseline) 2. Inflation (APR) 3. Output (Dev from Trend) 0 0

0.08

•0.1 •0.1

0.07

•0.2 •0.2

0.06

•0.3 •0.3

0.05 t t t n n n

e e e •0.4 c c c r r r

e •0.4 e e P P 0.04 P

•0.5 •0.5 0.03

•0.6 •0.6 0.02

•0.7 •0.7 0.01

Low Initial Debt High Initial Debt •0.8 •0.8 0 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 0 4 8 12 16 20 24 28 32 36 Quarter Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 34 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Explanation 2: Strength of MP depends on debt level Effects of a contractionary MP shock

1. Nominal Policy Rate (APR) 2. Yearly Inflation (4q•change) 0.25 0

s 0.2 s t t n n i i •0.01 o o

P 0.15 P e e g g a a

t 0.1 t •0.02 n n e e c c r r

e 0.05 e Low Indebtedness P P •0.03 High Indebtedness 0 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 Quarter Quarter 3. GDP 4. House Price 0 0

•0.05 t t •0.05 n n e e c c r r e e

P •0.1 P •0.1

•0.15 •0.15 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 Quarter Quarter 5. Aggegate Consumption (GDP Contrib.) 6. Borrowers Consumption 0 0

•0.02 •0.1

t t •0.2 n •0.04 n e e c c r r

e e •0.3

P •0.06 P •0.4 •0.08 •0.5 •0.1 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 Quarter Quarter 7. Non•residential10•3 Investment (GDP Contrib.) 8. Residential Investment (GDP Contrib.)

0 0 data1 data2 •0.01 data3 t •5 t n n e e

c c •0.02 r r e e P •10 P •0.03

•0.04 •15

0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 Quarter Quarter

Fourth Annual ECB Macroprudential Policy and Research Conference 35 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36 Residential Investment

Fourth Annual ECB Macroprudential Policy and Research Conference 36 Chenetal.(2019)Costs of Deleveraging xxx in a LIRE () / 36