Uncertainty in a model with credit frictions Ambrogio Cesa-Bianchi1 1 Bank
Emilio Fernandez-Corugedo2 of England Monetary Fund
3 International
T2M Conference March 25, 2016
Uncertainty in a model with credit frictions
1/ 29
Disclaimer
The views expressed in this paper are solely those of the authors and should not be taken to represent those of the Bank of England or the International Monetary Fund.
Uncertainty in a model with credit frictions
2/ 29
“A large black cloud of uncertainty” I
“A large black cloud of uncertainty, [...] with consumers and businesses holding back from commitments to spending [...]” Mervyn King, June 2012
Uncertainty in a model with credit frictions
—
Introduction
3/ 29
“A large black cloud of uncertainty” I
“A large black cloud of uncertainty, [...] with consumers and businesses holding back from commitments to spending [...]” Mervyn King, June 2012
I
Policy makers and researchers alike have taken the stance that widespread and heightened uncertainty was one of the key factors behind the unusual depth and duration of the “Great Recession”
Uncertainty in a model with credit frictions
—
Introduction
3/ 29
“A large black cloud of uncertainty” I
“A large black cloud of uncertainty, [...] with consumers and businesses holding back from commitments to spending [...]” Mervyn King, June 2012
I
Policy makers and researchers alike have taken the stance that widespread and heightened uncertainty was one of the key factors behind the unusual depth and duration of the “Great Recession”
I
Exactly what role uncertainty plays in explaining macroeconomic fluctuations is hard to gauge but is of outmost policy relevance
Uncertainty in a model with credit frictions
—
Introduction
3/ 29
“A large black cloud of uncertainty” I
“A large black cloud of uncertainty, [...] with consumers and businesses holding back from commitments to spending [...]” Mervyn King, June 2012
I
Policy makers and researchers alike have taken the stance that widespread and heightened uncertainty was one of the key factors behind the unusual depth and duration of the “Great Recession”
I
Exactly what role uncertainty plays in explaining macroeconomic fluctuations is hard to gauge but is of outmost policy relevance
I
This paper A quantitative assessment of whether and how uncertainty shocks can drive business cycle fluctuations
Uncertainty in a model with credit frictions
—
Introduction
3/ 29
How is uncertainty transmitted to the real economy?
I
Real options, Hartman-Abel effects, precautionary savings
Uncertainty in a model with credit frictions
—
Introduction
details
4/ 29
How is uncertainty transmitted to the real economy?
I
Real options, Hartman-Abel effects, precautionary savings
I
Risk premia
details
• With incomplete financial markets uncertainty leads to an increase in
the cost of external finance Arellano et al. (2012), Gilchrist et al. (2014), Christiano et al. (2014)
Uncertainty in a model with credit frictions
—
Introduction
4/ 29
How is uncertainty transmitted to the real economy?
I
Real options, Hartman-Abel effects, precautionary savings
I
Risk premia
details
• With incomplete financial markets uncertainty leads to an increase in
the cost of external finance Arellano et al. (2012), Gilchrist et al. (2014), Christiano et al. (2014)
I
Sticky prices • In a closed economy model with flexible prices uncertainty generates a
lack of comovement between consumption and investment • Sticky prices can fix this Basu and Bundick (2012)
Uncertainty in a model with credit frictions
—
Introduction
4/ 29
[Clarification #1] What do we mean by “uncertainty”?
Uncertainty in a model with credit frictions
—
Introduction
5/ 29
[Clarification #1] What do we mean by “uncertainty”? At = ρA At−1 + εt
εt ∼ N (0, σ 2t )
4
εt
2
0
−2
−4 t=−1
Uncertainty in a model with credit frictions
t=0
—
Introduction
t=1
t=2
5/ 29
[Clarification #2] Two different notions of uncertainty
Uncertainty in a model with credit frictions
—
Introduction
6/ 29
[Clarification #2] Two different notions of uncertainty I
Macro uncertainty • Time-varying variance of aggregate shocks in representative agent
models • Example: technology, preference, monetary policy, fiscal policy,... Fernandez-Villaverde and Rubio-Ramirez (2007), Justiniano and Primiceri (2008), Fernandez-Villaverde et al. (2011,2014), Basu and Bundick (2012), Born and Pfeifer (2014)
Uncertainty in a model with credit frictions
—
Introduction
6/ 29
[Clarification #2] Two different notions of uncertainty I
Macro uncertainty • Time-varying variance of aggregate shocks in representative agent
models • Example: technology, preference, monetary policy, fiscal policy,... Fernandez-Villaverde and Rubio-Ramirez (2007), Justiniano and Primiceri (2008), Fernandez-Villaverde et al. (2011,2014), Basu and Bundick (2012), Born and Pfeifer (2014)
I
Micro uncertainty • Time-varying cross-sectional dispersion of idiosyncratic shocks in
models that feature some kind of heterogeneity • Example: firm-level productivity shocks Bloom (2009), Bloom et al. (2012), Bachmann and Bayer (2013), Dorofeenko et al. (2008), Christiano et al. (2003,2014), Gilchrist et al. (2013), Chugh (2015)
Uncertainty in a model with credit frictions
—
Introduction
6/ 29
This paper I
Question Are uncertainty shocks a major source of business cycle fluctuations?
Uncertainty in a model with credit frictions
—
Introduction
7/ 29
This paper I
Question Are uncertainty shocks a major source of business cycle fluctuations?
I
How • Set up a DSGE model with sticky prices and financial frictions
(Bernanke, Gertler and Gilchrist, 1999) • Infer the time series properties of uncertainty from the data • Quantify the role of uncertainty in driving business cycle fluctuations
Uncertainty in a model with credit frictions
—
Introduction
7/ 29
This paper I
Question Are uncertainty shocks a major source of business cycle fluctuations?
I
How • Set up a DSGE model with sticky prices and financial frictions
(Bernanke, Gertler and Gilchrist, 1999) • Infer the time series properties of uncertainty from the data • Quantify the role of uncertainty in driving business cycle fluctuations I
Contribution • Consider macro and micro uncertainty shocks in a unified framework • Investigate (quantitatively) the role of imperfect financial markets and
nominal rigidities • Contribute to the debate on how to parametrize micro uncertainty
processes in this class of models Uncertainty in a model with credit frictions
—
Introduction
7/ 29
What we find
I
Micro uncertainty matters more than macro uncertainty • According to our model, the impact of micro uncertainty shocks on
output is one order of magnitude larger than macro uncertainty shocks
Uncertainty in a model with credit frictions
—
Introduction
8/ 29
What we find
I
Micro uncertainty matters more than macro uncertainty • According to our model, the impact of micro uncertainty shocks on
output is one order of magnitude larger than macro uncertainty shocks I
Micro uncertainty shocks can account for a non-trivial share of output volatility
Uncertainty in a model with credit frictions
—
Introduction
8/ 29
What we find
I
Micro uncertainty matters more than macro uncertainty • According to our model, the impact of micro uncertainty shocks on
output is one order of magnitude larger than macro uncertainty shocks I
Micro uncertainty shocks can account for a non-trivial share of output volatility
I
Both the degree of price stickiness and the severity of the credit friction are important amplifiers of micro uncertainty shocks
Uncertainty in a model with credit frictions
—
Introduction
8/ 29
A DSGE model with sticky prices and credit frictions
CENTRAL BANK Policy Rate
BANKS Deposits
Loans
HOUSEHOLDS
ENTREPRENEURS Capital
Labour
MON. FIRMS
Uncertainty in a model with credit frictions
—
The model
9/ 29
The model – Aggregate productivity
CENTRAL BANK Policy Rate
BANKS Deposits
Loans
HOUSEHOLDS
ENTREPRENEURS Capital
Labour
MON. FIRMS
Uncertainty in a model with credit frictions
—
The model
- TFP:
𝑨𝒕 = 𝝆𝑨𝒕−𝟏 + 𝜺𝒕
- Aggreg. productivity shocks:
𝜺𝒕 ~𝑵(𝟎, 𝝈𝟐 𝑨,𝒕 )
10/ 29
The model – Idiosyncratic productivity
- Entrepreneurial returns: 𝝎(𝟏 + 𝑹𝒌,𝒕 )
CENTRAL BANK Policy Rate
- Idiosyn. productivity shocks :
BANKS Deposits
Loans
HOUSEHOLDS
ENTREPRENEURS
𝝎~𝒍𝒐𝒈𝑵(𝟏, 𝝈𝟐 𝝎,𝒕 )
Capital
Labour
MON. FIRMS
Uncertainty in a model with credit frictions
—
The model
- TFP:
𝑨𝒕 = 𝝆𝑨𝒕−𝟏 + 𝜺𝒕
- Aggreg. productivity shocks:
𝜺𝒕 ~𝑵(𝟎, 𝝈𝟐 𝑨,𝒕 )
11/ 29
The model – Macro uncertainty shocks
- Entrepreneurial returns: 𝝎(𝟏 + 𝑹𝒌,𝒕 )
CENTRAL BANK Policy Rate
- Idiosyn. productivity shocks :
BANKS Deposits
Loans
HOUSEHOLDS
ENTREPRENEURS
𝝎~𝒍𝒐𝒈𝑵(𝟏, 𝝈𝟐 𝝎,𝒕 )
Capital
Labour
MON. FIRMS
Uncertainty in a model with credit frictions
—
The model
- TFP:
𝑨𝒕 = 𝝆𝑨𝒕−𝟏 + 𝜺𝒕
- Aggreg. productivity shocks:
𝜺𝒕 ~𝑵(𝟎, 𝝈𝟐 𝑨,𝒕 )
Macro Uncertainty
12/ 29
The model – Micro uncertainty shocks
- Entrepreneurial returns: 𝝎(𝟏 + 𝑹𝒌,𝒕 )
CENTRAL BANK Policy Rate
- Idiosyn. productivity shocks :
BANKS Deposits
Loans
HOUSEHOLDS
ENTREPRENEURS
𝝎~𝒍𝒐𝒈𝑵(𝟏, 𝝈𝟐 𝝎,𝒕 )
Micro Uncertainty
Capital
Labour
MON. FIRMS
Uncertainty in a model with credit frictions
—
The model
- TFP:
𝑨𝒕 = 𝝆𝑨𝒕−𝟏 + 𝜺𝒕
- Aggreg. productivity shocks:
𝜺𝒕 ~𝑵(𝟎, 𝝈𝟐 𝑨,𝒕 )
Macro Uncertainty
13/ 29
Solution method
I
DSGE models are normally solved by taking a linear (i.e., first-order Taylor series) approximation around the non–stochastic steady state equilibrium • However, when using the traditional linear approximation, uncertainty
shocks do not play any role by construction
Uncertainty in a model with credit frictions
—
The model
14/ 29
Solution method
I
DSGE models are normally solved by taking a linear (i.e., first-order Taylor series) approximation around the non–stochastic steady state equilibrium • However, when using the traditional linear approximation, uncertainty
shocks do not play any role by construction I
Higher order approximation is needed (at least 3rd order) ⇒ obtained using Dynare • The “pruned state-space system” (Andreasen, Fernandez-Villaverde,
Rubio-Ramirez, 2013 NBER) is then used to simulate the model and to compute non-linear impulse responses
Uncertainty in a model with credit frictions
—
The model
14/ 29
Parameters in our model
I
Standard values for most deep parameters • Great ratios, steady state inflation, monetary policy,...
I
Financial friction parameters: monitoring cost (µ), steady state level ¯ of micro uncertainty (S) • External finance premium, default probability
I
Time series properties of micro and macro uncertainty inferred from the data
Uncertainty in a model with credit frictions
—
Model parametrization
15/ 29
Parameter values
General µ γ α δ β % ν τ θ φk π ss S
Monitoring Cost Survivial Prob. Capital Share Depr. Rate Discount Factor Risk Aversion Inv. Frish Elasticity GHH Scaling Factor Demand elasticity Rotemberg Inv. Adj. Cost Inflation (SS) Micro Uncert. (SS)
Uncertainty in a model with credit frictions
Macro uncertainty shock 0.25 0.985 0.33 0.025 0.994 2 1 2.5 10 105 1.5 2% 0.225
—
ρA σA ρW σW
TFP TFP Persistence Macro Uncert. St. Dev. Macro Uncert.
? ? ? ?
ρS σS
Micro uncertainty shock Persistence Micro Uncert. St. Dev. Micro Uncert.
? ?
ρr ρy ρπ
Monetary policy Int. Rate Smoothing Output Inflation
0.25 0.5 1.5
Model parametrization
16/ 29
Micro uncertainty I
Annual data from the Census panel of manufacturing establishments over the sample period 1972–2009
Uncertainty in a model with credit frictions
—
Model parametrization
17/ 29
Micro uncertainty I
Annual data from the Census panel of manufacturing establishments over the sample period 1972–2009
I
Compute establishment-level TFP and its cross-sectional standard deviation as a proxy for micro uncertainty (St ) • Data available from Bloom et al. (2012) at
http://www.stanford.edu/˜nbloom/index files/Page315.htm
Uncertainty in a model with credit frictions
—
Model parametrization
17/ 29
Micro uncertainty I
Annual data from the Census panel of manufacturing establishments over the sample period 1972–2009
I
Compute establishment-level TFP and its cross-sectional standard deviation as a proxy for micro uncertainty (St ) • Data available from Bloom et al. (2012) at
http://www.stanford.edu/˜nbloom/index files/Page315.htm I
Fit an AR(1) process on the deviations of St from an HP trend to get ρS and σ S
Uncertainty in a model with credit frictions
—
Model parametrization
17/ 29
Micro uncertainty I
Annual data from the Census panel of manufacturing establishments over the sample period 1972–2009
I
Compute establishment-level TFP and its cross-sectional standard deviation as a proxy for micro uncertainty (St ) • Data available from Bloom et al. (2012) at
http://www.stanford.edu/˜nbloom/index files/Page315.htm I
Fit an AR(1) process on the deviations of St from an HP trend to get ρS and σ S
I
Recover quarterly values for ρS and σ S using simulations form the quarterly model and time-aggregating the simulated data
Uncertainty in a model with credit frictions
—
Model parametrization
17/ 29
Micro uncertainty (a) St. dev. of establishment−level TFP shocks 0.65
(b) Cyclical component 0.08 0.06
0.6 0.04 0.55
0.02 0
0.5
−0.02 0.45 −0.04 0.4 1970
1980
1990
2000
2010
−0.06 1970
1980
1990
2000
2010
Figure Micro Uncertainty: Cross-sectional Dispersion Of TFP Shocks. The left panel plots the cross-sectional standard deviation of establishment-level TFP ) from Bloom et al (2012). The right panel plots the deviation of shocks (sigmamicro t sigmamicro from an HP trend with smoothing parameter equal to 100, i.e. our proxy t for St .
Uncertainty in a model with credit frictions
—
Model parametrization
18/ 29
Parameter values – Micro uncertainty
General µ γ α δ β % ν τ θ φk π ss S
Monitoring Cost Survivial Prob. Capital Share Depr. Rate Discount Factor Risk Aversion Inv. Frish Elasticity GHH Scaling Factor Demand elasticity Rotemberg Inv. Adj. Cost Inflation (SS) Micro Uncert. (SS)
Uncertainty in a model with credit frictions
Macro uncertainty shock 0.25 0.985 0.33 0.025 0.994 2 1 2.5 10 105 1.5 2% 0.225
—
ρA σA ρW σW
TFP TFP Persistence Macro Uncert. St. Dev. Macro Uncert.
? ? ? ?
ρS σS
Micro uncertainty shock Persistence Micro Uncert. St. Dev. Micro Uncert.
0.79 0.025
ρr ρy ρπ
Monetary policy Int. Rate Smoothing Output Inflation
0.25 0.5 1.5
Model parametrization
19/ 29
Macro uncertainty
I
Quarterly TFP for the U.S. business sector over the 1972:Q1-2009:Q4 • Data from Basu, Fernald, and Kimball (2006) available at
http://www.frbsf.org/economic-research/total-factor-productivity-tfp/
Uncertainty in a model with credit frictions
—
Model parametrization
20/ 29
Macro uncertainty
I
Quarterly TFP for the U.S. business sector over the 1972:Q1-2009:Q4 • Data from Basu, Fernald, and Kimball (2006) available at
http://www.frbsf.org/economic-research/total-factor-productivity-tfp/ I
Compute conditional heteroskedasticity of TFP growth with a GARCH(1,1) model as a proxy for micro uncertainty (Wt )
Uncertainty in a model with credit frictions
—
Model parametrization
20/ 29
Macro uncertainty
I
Quarterly TFP for the U.S. business sector over the 1972:Q1-2009:Q4 • Data from Basu, Fernald, and Kimball (2006) available at
http://www.frbsf.org/economic-research/total-factor-productivity-tfp/ I
Compute conditional heteroskedasticity of TFP growth with a GARCH(1,1) model as a proxy for micro uncertainty (Wt )
I
Fit an AR(1) process on the deviations of Wt from an HP trend to get ρW and σ W
Uncertainty in a model with credit frictions
—
Model parametrization
20/ 29
Macro uncertainty (a) Conditional heteroscedasticity aggregate TFP 1.3
(b) Cyclical component 0.25
1.2
0.2
1.1
0.15
1
0.1
0.9
0.05
0.8
0
0.7
−0.05
0.6 0.5 1972
−0.1 1980
1988
1996
2004
−0.15 1972
1980
1988
1996
2004
Figure Macro Uncertainty: Conditional Heteroskedasticity Of Aggregate Productivity (TFP). The left panel plots the conditional heteroscedasticity of the growth rate of quarterly TFP (σ macro ) estimated using a t GARCH(1,1) model using the TFP data from Basu, Fernald, and Kimball (2006). The right panel plots the deviation of σ macro from an HP trend with smoothing parameter t equal to 1600, i.e. our proxy for Wt .
Uncertainty in a model with credit frictions
—
Model parametrization
21/ 29
Parameter values – Macro uncertainty
General µ γ α δ β % ν τ θ φk π ss S
Monitoring Cost Survivial Prob. Capital Share Depr. Rate Discount Factor Risk Aversion Inv. Frish Elasticity GHH Scaling Factor Demand elasticity Rotemberg Inv. Adj. Cost Inflation (SS) Micro Uncert. (SS)
Uncertainty in a model with credit frictions
Macro uncertainty shock 0.25 0.985 0.33 0.025 0.994 2 1 2.5 10 105 1.5 2% 0.225
—
ρA σA ρW σW
TFP TFP Persistence Macro Uncert. St. Dev. Macro Uncert.
? ? 0.63 0.048
ρS σS
Micro uncertainty shock Persistence Micro Uncert. St. Dev. Micro Uncert.
0.79 0.025
ρr ρy ρπ
Monetary policy Int. Rate Smoothing Output Inflation
0.25 0.5 1.5
Model parametrization
22/ 29
Parameter values – TFP
General µ γ α δ β % ν τ θ φk π ss S
Monitoring Cost Survivial Prob. Capital Share Depr. Rate Discount Factor Risk Aversion Inv. Frish Elasticity GHH Scaling Factor Demand elasticity Rotemberg Inv. Adj. Cost Inflation (SS) Micro Uncert. (SS)
Uncertainty in a model with credit frictions
Macro uncertainty shock 0.25 0.985 0.33 0.025 0.994 2 1 2.5 10 105 1.5 2% 0.225
—
ρA σA ρW σW
TFP TFP Persistence Macro Uncert. St. Dev. Macro Uncert.
0.95 0.009 0.63 0.048
ρS σS
Micro uncertainty shock Persistence Micro Uncert. St. Dev. Micro Uncert.
0.79 0.025
ρr ρy ρπ
Monetary policy Int. Rate Smoothing Output Inflation
0.25 0.5 1.5
Model parametrization
23/ 29
Quantitative analysis
I
4 sets of IRFs • Macro uncertainty shock
[baseline & price stickiness]
[baseline & fin. frictions]
• Micro uncertainty shock
[baseline & price stickiness] I
[baseline & fin. frictions]
Simulated business cycle statistics
Uncertainty in a model with credit frictions
—
Quantitative analysis
24/ 29
x 10
x 10
0
0
The impact of a macro uncertainty shock 0.6
−1
0.4
−5 −2
I
−10
The role of price stickiness 5 10 15 20
0.2
−3 5
Consumption (%)
10
15
20
−3 x 10 Investment (%)
0
0
0
0
−0.005
5
10
15
20
−3 x 10 Total Output (%)
−5 −5
−0.01
−10 −0.015
−10 −15
−0.02 5
10
15
20
Baseline (4 quarters)
Uncertainty in a model with credit frictions
—
5
10
15
20
High Stickiness (5 quarters)
Quantitative analysis
5
10
15
20
Flex. Prices
25/ 29
x 10
x 10
0
0
The impact of a macro uncertainty shock 0.6
−1
0.4
−5
Hours (%)
I
0 −10
The role−0.005 of price stickiness 5 10 15 20 Consumption (%)
−0.01 0 −0.005 −0.015 −0.01
5 −3
−0.015 x 10 0 −0.02 −1
−2
5
10
15
20
Net worth (%)
10
−3
15
−2 −3 x 10Inflation (annual, %) 0 −3 −2 5 10 15 −4 −3 x 10 Investment (%) −6 0 −8 −5
5
10
15
0.2 0 0 20−0.005
−0.5
10
15
The role of financial frictions 5 10 15 20 0
20
10
15
20
Risk. Premium (annual, bp)
0.3
5
10
15
20
15
20
Flex. Prices
0.1 5
Consumption (%)
15
−15 0.4 20
−1
0
5 −10
−5
I
10
−5
20
High Stickiness (5 quarters) 0.2
Baseline (4 quarters)
−4
5
5
−0.01 −3 x 10 Total Output (%) 0 −0.015
−3 x 10Price of Capital (%) −10 0
20
Policy Rate (annual, %)
10
15
20
−3 x 10 Investment (%)
0
5
10
Total Output (%) 0
−1 −2
−0.005
−0.005
−3 −4
−0.01
−5 5
10
15
20
Baseline (µ=0.25)
Uncertainty in a model with credit frictions
—
−0.01 5
10
15
High Credit Frict. (µ=0.45)
Quantitative analysis
20
5
10
15
20
Low Credit Frict. (µ=0.05)
25/ 29
0
40
The impact of a micro uncertainty shock 0
30
−0.05
20
−0.2
10
−0.1
I
The role−0.4of price 5 10stickiness 15 20
0 5
Consumption (%)
10
15
20
5
Investment (%)
10
15
20
Total Output (%)
0 −0.05
0
−0.1
−0.2 −0.1
−0.15 −0.2
−0.4 −0.2
−0.25 5
10
15
20
Baseline (4 quarters)
Uncertainty in a model with credit frictions
—
5
10
15
20
High Stickiness (5 quarters)
Quantitative analysis
5
10
15
20
Flex. Prices
26/ 29
0
40
The impact of a micro uncertainty shock 0
30
−0.05
20
−0.2
Hours (%)
−0.1 0
0
I
The role−0.4 of price −0.1 5 10stickiness 15 20 −0.2
10
15
20
Policy Rate (annual, %)
5 −0.5
10
15
20
Total Output (%)
0 −0.4 5
10
15
20 −0.2
5
10
15
20
−0.05 −1 −0.1
5
10
15
20
−0.15 Net worth (%)
−0.4 0
−0.20 −0.1
5
10
15
20
Price of Capital (%)
5
10
15
−0.2 −0.25 40 20
Risk. Premium (annual, bp)
5
10
15
20
15
20
−0.05
−0.2
Baseline (4 quarters)
−0.3
I
5
10 0 0
Investment (%)
−0.3 0
−0.1
−0.2
Consumption (%)
Inflation (annual, %)
High Stickiness (5 quarters) 20
Flex. Prices
−0.1
The role −0.4 of financial frictions 5 10 15 20 Consumption (%)
0 5
10
15
20
5
Investment (%)
0
10
Total Output (%)
0 −0.05
−0.05
−0.1
−0.2
−0.1
−0.15
−0.15
−0.2
−0.4
−0.2
−0.25 5
10
15
20
Baseline (µ=0.25)
Uncertainty in a model with credit frictions
—
5
10
15
High Credit Frict. (µ=0.45)
Quantitative analysis
20
5
10
15
20
Low Credit Frict. (µ=0.05)
26/ 29
Unconditional business cycle statistics
Volatility Output
(1)
(2)
(3)
(4)
(5)
(6)
(7)
Data
Baseline
Macro
Micro
Micro: large shock
Micro: high fin. frictions
Micro: high stickiness
1.55
1.58
0.00
0.11
0.30
0.15
0.15
Volatility of aggregate variables relative to output volatility Consumption 0.82 0.81 2.36 Investment 2.92 2.35 8.66
0.78 4.56
0.77 4.85
0.77 4.05
0.84 3.63
First-order autocorrelation Output 0.88 Consumption 0.88 Investment 0.91
0.64 0.71 0.53
0.64 0.70 0.46
0.64 0.71 0.52
0.71 0.79 0.54
0.98 0.94
0.98 0.79
0.98 0.93
0.98 0.89
0.92 0.91 0.79
0.99 1.00 1.00
Contemporaneous correlation with aggregate consumption Consumption 0.88 0.94 0.95 Investment 0.95 0.83 0.91
Uncertainty in a model with credit frictions
—
Quantitative analysis
27/ 29
Conclusions
I
Micro uncertainty has a larger impact on output than macro uncertainty
I
A 1 standard deviation shock to micro uncertainty decreases output by 0.25 percent
I
Shocks to micro uncertainty can drive about 10 percent of total volatility of output
I
The above estimates are increasing in the degree of price stickiness and the severity of the credit friction
Uncertainty in a model with credit frictions
—
Conclusions
28/ 29
How is uncertainty transmitted to the real economy? I
Real option delay effects • Under (partial) irreversibility uncertainty can delay
investment/consumption decision Bernanke (1983), Dixit and Pindyck (1996), Bertola and Caballero (1990)
Uncertainty in a model with credit frictions
—
Appendix
29/ 29
How is uncertainty transmitted to the real economy? I
Real option delay effects • Under (partial) irreversibility uncertainty can delay
investment/consumption decision Bernanke (1983), Dixit and Pindyck (1996), Bertola and Caballero (1990)
I
Hartman-Abel effects • When marginal revenue product of capital is convex in TFP
uncertainty increases expected profits Hartman (1976), Abel (1983)
Uncertainty in a model with credit frictions
—
Appendix
29/ 29
How is uncertainty transmitted to the real economy? I
Real option delay effects • Under (partial) irreversibility uncertainty can delay
investment/consumption decision Bernanke (1983), Dixit and Pindyck (1996), Bertola and Caballero (1990)
I
Hartman-Abel effects • When marginal revenue product of capital is convex in TFP
uncertainty increases expected profits Hartman (1976), Abel (1983)
I
Risk aversion • Uncertainty induces prudent consumers to increase their precautionary
savings Leland (1968), Kimball (1990) ... go back Uncertainty in a model with credit frictions
—
Appendix
29/ 29
