Download Monetary Policy Rules in the Presence of an Punnoose Jacob Christie Smith

Monetary Policy Rules in the Presence of an
Occasionally Binding Borrowing Constraint
Punnoose Jacob
Christie Smith
Fang Yao
Oct 2014, Wellington
Reserve Bank of New Zealand.
Research Question
How does an occasionally-binding Loan-to-Value Ratio (LVR) constraint
a¤ect the conduct of monetary policy in terms of an interest rate rule?
Local Context
New Zealand’s LVR restrictions were introduced on 1 October 2013,
responding to
Local Context
New Zealand’s LVR restrictions were introduced on 1 October 2013,
responding to
I
Annual NZ house price in‡ation reached 10 percent, December 2013 (16% in
Auckland).
Local Context
New Zealand’s LVR restrictions were introduced on 1 October 2013,
responding to
I
Annual NZ house price in‡ation reached 10 percent, December 2013 (16% in
Auckland).
I
The proportion of high LVR lending exceeded 30% in early 2013.
Local Context
New Zealand’s LVR restrictions were introduced on 1 October 2013,
responding to
I
Annual NZ house price in‡ation reached 10 percent, December 2013 (16% in
Auckland).
I
The proportion of high LVR lending exceeded 30% in early 2013.
Housing market led to …nancial stability concerns
Local Context
New Zealand’s LVR restrictions were introduced on 1 October 2013,
responding to
I
Annual NZ house price in‡ation reached 10 percent, December 2013 (16% in
Auckland).
I
The proportion of high LVR lending exceeded 30% in early 2013.
Housing market led to …nancial stability concerns
Reluctance to use the interest rates: concerns about low in‡ation and
elevated exchange rate.
Contribution
We start from Iacoviello (2005)
Contribution
We start from Iacoviello (2005)
I
Housing market
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
I
Borrowing constraint on loans
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
I
Borrowing constraint on loans
Our extensions
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
I
Borrowing constraint on loans
Our extensions
I
Open economy DSGE model for NZ
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
I
Borrowing constraint on loans
Our extensions
I
Open economy DSGE model for NZ
I
Occasionally-binding borrowing constraint
Contribution
We start from Iacoviello (2005)
I
Housing market
I
Patient and impatient households
I
Borrowing constraint on loans
Our extensions
I
Open economy DSGE model for NZ
I
Occasionally-binding borrowing constraint
I
We study optimal monetary policy rules
Main Findings
Imposing an occasionally-binding LVR makes the economy respond
asymmetrically to positive and negative shocks.
Main Findings
Imposing an occasionally-binding LVR makes the economy respond
asymmetrically to positive and negative shocks.
The LVR a¤ects macro volatilities and hence changes monetary policy.
Main Findings
Imposing an occasionally-binding LVR makes the economy respond
asymmetrically to positive and negative shocks.
The LVR a¤ects macro volatilities and hence changes monetary policy.
The optimal monetary policy rule under an LVR constraint transfers welfare
from savers to borrowers.
Main Findings
Imposing an occasionally-binding LVR makes the economy respond
asymmetrically to positive and negative shocks.
The LVR a¤ects macro volatilities and hence changes monetary policy.
The optimal monetary policy rule under an LVR constraint transfers welfare
from savers to borrowers.
Removing the LVR results in gradual adjustment.
THE MODEL
Households’problem
Maximise expected utility subject to
1
Budget constraint
Households’problem
Maximise expected utility subject to
1
Budget constraint
2
Collateral constraint
Rl ,t Lt0
µ Et qh,t +1 Pc ,t +1 ht0
The Rest of the Model
Banks channel savings from domestic and foreign savers to borrowers.
Home good produced with capital and Labour
Sold at home and abroad
Foreign output, in‡ation and the interest rate: New Keynesian closed
economy model
Monetary and LVR Policy
Interest rate rule
Rt
=
R̄
Rt 1 r r
R̄
π c ,t
π̄ c
rπ
yt
yt 1
r ∆y 1 rr
exp ω r ,t
Monetary and LVR Policy
Interest rate rule
Rt
=
R̄
Rt 1 r r
R̄
π c ,t
π̄ c
rπ
yt
r ∆y 1 rr
yt 1
LVR policy
Rl ,t Lt0
µLVR Et qh,t +1 Pc ,t +1 ht0
exp ω r ,t
Parameter Values
Most of structural parameters are calibrated to match NZ data.
The rest are estimated using Bayesian methods
Sample period:1993 Q4 to 2013 Q3 before the LVR restriction was
introduced.
The estimated model does not have the borrowing constraint.
9 data series :
I
GDP growth, Consumption growth, Residential investment growth, Business
investment growth, Housing loan growth, 90-day rate, CPI in‡ation, House
price In‡ation, Mortgage spread.
Occasionally-Binding Solution
Occasionally-binding Solution
We use the "OccBin" Toolbox developed by Guerrieri and Iacoviello (2014)
A piecewise-linear approximation of occasionally binding constraints
It is able to deal with large models with many predetermined variables.
Asymmetric IRFs: Monetary Policy Shock
Contractionary
% from S.S.
90-Day Rate
House Price
2
1
1
0
Loan
O utput
0
CPI Inflation
0
0
-0.5
-1
-1
0
-1
-1
0
4q
8q
-2
-2
0
4q
8q
-3
-0.5
-1.5
0
4q
8q
-2
0
4q
8q
-1
0
4q
8q
Occasionally-binding
P erpetually-binding
Asymmetric IRFs: Monetary Policy Shock
C o n tr a c tio n a r y
90-Day Rate
House Price
% fr o m S.S.
2
Loan
1
0
1
0
-1
0
-1
-2
-1
-2
0
4q
8q
O utput
-3
0
4q
8q
CPI Inflation
0
0
-1
- 0 .5
-2
0
4q
8q
-1
0
4q
8q
0
4q
8q
O c c a s io n a lly - b in d in g
P e r p e tu a lly - b in d in g
E x p a n s io n a r y
% fr o m S.S.
90-Day Rate
House Price
Loan
1
2
3
0
1
2
-1
0
1
-2
-1
0
4q
8q
O utput
0
0
4q
8q
CPI Inflation
2
1
1
0 .5
0
0
4q
8q
0
0
4q
8q
0
4q
8q
Stochastic Simulation
Stochastic Simulation
Stochastic Simulation
Comparing Moments from the Perpetually- and Occasionally-binding Models
LVR
0.90
0.70
Binding Frequency
Occasional Perpetual
10.4%
100%
12%
100%
Output S.D. (%)
Occasional Perpetual
0.77
1.13
0.75
0.87
CPI In‡ation S.D. (%)
Occasional Perpetual
0.19
0.21
0.18
0.19
Optimal Policy
Optimal Monetary Policy Rules
Taylor Rules
Estimated:
Occ.binding optimal:
Always binding optimal:
R̂t = 0.80R̂ t 1 +0.2 (1.89π̂ c ,t + 0.32∆ŷ t )
R̂t = 0.80R̂ t 1 +0.2 (1.1π̂ c ,t
R̂t = 0.80R̂ t 1 +0.2 (3π̂ c ,t
∆ŷ t )
∆ŷ t )
Optimal Monetary Policy Rules
Taylor Rules
Estimated:
Occ.binding optimal:
Always binding optimal:
R̂t = 0.80R̂ t 1 +0.2 (1.89π̂ c ,t + 0.32∆ŷ t )
R̂t = 0.80R̂ t 1 +0.2 (1.1π̂ c ,t
R̂t = 0.80R̂ t 1 +0.2 (3π̂ c ,t
∆ŷ t )
∆ŷ t )
Extend Taylor rule to include house price in‡ation and credit growth
Welfare Evaluation
Welfare Level (Gain in terms of consumption)
Saver
Borrower
Social
Taylor Rules
Estimated:
-84.83
-101.42
-1.912
Occ.binding:
-85.88 ( 1.04%)
-100.71 (0.71%)
-1.910 (0.002%)
Always binding:
-75.4 (9.8%)
-113.12 ( 11.6%)
-2.01 ( 0.2%)
Temporary LVR Tightening
Loan-to-Value Ratio
0.92
Occasionally-binding
Perpetually-binding
0.91
Level
0.9
0.89
0.88
0.87
0
8q
16q
24q
32q
40q
48q
Temporary LVR Tightening
Loan-to-Value Ratio
Loans
% from S.S.
Level
0.91
0.9
0.89
0.1
0.5
0
0
-0.5
0.88
0.87
House Price
1
% from S.S.
0.92
0
8q
16q
24q
32q
40q
-1
48q
-0.2
0
8q
Output
16q
24q
32q
40q
48q
-0.5
0.1
0
Occasionally-binding
Perpetually-binding
8q
16q
24q
8q
32q
40q
48q
-0.1
16q
24q
32q
40q
48q
40q
48q
5.4
Annualised Level in %
Annualised %
0
0
0
Policy rate
0.3
0.2
% from S.S.
-0.3
Inflation
0.5
-1
-0.1
0
8q
16q
24q
32q
40q
48q
5.2
5
4.8
4.6
0
8q
16q
24q
32q
Conclusion
We study macro dynamics under an occasionally-binding LVR .
The LVR makes the economy respond asymmetrically to positive and
negative shocks and hence changes macro volatilities and monetary policy.
Future Research:
I
Extend monetary policy rule to include house price in‡ation and credit growth
I
Endogenise LVR policy
I
Open economy dimensions
Bank’s problem
Bank dividend
∞
max Et
∑ βbτ
τ =0
λt +τ Db,t +τ (j )
λt
Pc ,t +τ
Budget constraint
Db,t (j ) Rt 1 St 1 (j ) Φt 1 Rt 1 St 1 (j ) Lt (j )
+
+
+
Pc ,t
Pc ,t
et Pc ,t
Pc ,t
R
L
(
j
)
St ( j )
S (j )
t 1
+ t
+ l ,t 1
Pc ,t
et Pc ,t
Pc ,t
The bank is subject to a capital requirement constraint
Lt ( j )
St (j ) St (j )/et
=1
Lt ( j )
b
where : µb,t = µbb,t
1
<–
Lt
Pd ,t yd ,t
µb,t
b l (1 b b )
, bl > 0, bb 2 [0, 1)
(1)
IRFs to Monetary Policy Shock
S hoc k P roc es s
P o l i c y R a te
1
O u tp u t
C P I In fl a ti o n
0 .2
0
9 7 .5 % i l e
M edian
% from S.S.
0 .2
2 .5 % i l e
0 .1
- 0 .5
0
4q
8q
12q 16q 20q
- 0 .4
- 0 .4
- 0 .6
- 0 .8
- 0 .6
0
C o n s . S a ve r
0 .5
4q
8q
12q 16q 20q
0
C ons. B orrower
0
4q
8q
0
12q 16q 20q
4q
8q
12q 16q 20q
H ous e P r ic e
H o u s i n g In v.
0 .5
0
0
% from S.S.
- 0 .2
0
0
- 0 .2
0
- 0 .4
- 0 .5
- 0 .5
- 0 .5
-1
-1
- 1 .5
- 1 .5
0
4q
8q
12q 16q 20q
0
Loans
% from S.S.
0
- 0 .2
0 .5
0 .0 3
- 0 .2
0 .0 2
- 0 .4
0 .0 1
- 0 .6
0
- 0 .8
- 0 .0 1
4q
8q
12q 16q 20q
8q
-1
- 0 .8
- 1 .5
12q 16q 20q
0
L e n d in g S p r e a d
0
0
4q
- 0 .6
4q
8q
0
12q 16q 20q
4q
B u s i n e s s In v.
8q
12q 16q 20q
R ER
0
1 .5
- 0 .5
0 .5
-1
- 0 .5
1
0
0
4q
8q
12q 16q 20q
0
4q
8q
12q 16q 20q
0
4q
8q
12q 16q 20q
Optimal Monetary Policy Rules under LVR
We evaluate Taylor type operational rules based on unconditional
expectations of social welfare.
Due to the occasionally binding constraint, we apply a simulation-based
welfare measure.
We simulate the model based on estimated parameters and driving forces for
500 periods, repeating it for 200 times.
Averaging across 200 replications yields a sample approximation to the
expected values of variables for the welfare measure.
We repeat this exercise for each candidate policy rule on a grid of Taylor rule
coe¢ cients.