the bond premium in a dsge model with long-run real and …swanson2/pres/bankofitaly.pdf ·...
TRANSCRIPT
![Page 1: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/1.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Bond Premium in a DSGE Modelwith Long-Run Real and Nominal Risks
Glenn D. Rudebusch Eric T. Swanson
Economic ResearchFederal Reserve Bank of San Francisco
Banca D’ItaliaApril 16, 2010
![Page 2: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/2.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Outline
1 Motivation and Background
2 Epstein-Zin Preferences in a Standard NK Model
3 Long-Run Risks
4 Model Implications
5 Conclusions
![Page 3: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/3.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Bond Premium Puzzle
The equity premium puzzle: excess returns on stocks are muchlarger (and more variable) than can be explained by standardpreferences in a DSGE model (Mehra and Prescott, 1985).
The bond premium puzzle: excess returns on long-term bonds aremuch larger (and more variable) than can be explained by standardpreferences in a DSGE model (Backus, Gregory, and Zin, 1989).
Note:Since Backus, Gregory, and Zin (1989), DSGE models withnominal rigidities have advanced considerably
The UIP premium puzzle: excess returns on high-interest-rateforeign currencies are much larger (and more variable) than can beexplained by standard preferences in a DSGE model.
![Page 4: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/4.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Bond Premium Puzzle
The equity premium puzzle: excess returns on stocks are muchlarger (and more variable) than can be explained by standardpreferences in a DSGE model (Mehra and Prescott, 1985).
The bond premium puzzle: excess returns on long-term bonds aremuch larger (and more variable) than can be explained by standardpreferences in a DSGE model (Backus, Gregory, and Zin, 1989).
Note:Since Backus, Gregory, and Zin (1989), DSGE models withnominal rigidities have advanced considerably
The UIP premium puzzle: excess returns on high-interest-rateforeign currencies are much larger (and more variable) than can beexplained by standard preferences in a DSGE model.
![Page 5: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/5.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Bond Premium Puzzle
The equity premium puzzle: excess returns on stocks are muchlarger (and more variable) than can be explained by standardpreferences in a DSGE model (Mehra and Prescott, 1985).
The bond premium puzzle: excess returns on long-term bonds aremuch larger (and more variable) than can be explained by standardpreferences in a DSGE model (Backus, Gregory, and Zin, 1989).
Note:Since Backus, Gregory, and Zin (1989), DSGE models withnominal rigidities have advanced considerably
The UIP premium puzzle: excess returns on high-interest-rateforeign currencies are much larger (and more variable) than can beexplained by standard preferences in a DSGE model.
![Page 6: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/6.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Bond Premium Puzzle
The equity premium puzzle: excess returns on stocks are muchlarger (and more variable) than can be explained by standardpreferences in a DSGE model (Mehra and Prescott, 1985).
The bond premium puzzle: excess returns on long-term bonds aremuch larger (and more variable) than can be explained by standardpreferences in a DSGE model (Backus, Gregory, and Zin, 1989).
Note:Since Backus, Gregory, and Zin (1989), DSGE models withnominal rigidities have advanced considerably
The UIP premium puzzle: excess returns on high-interest-rateforeign currencies are much larger (and more variable) than can beexplained by standard preferences in a DSGE model.
![Page 7: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/7.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Kim-Wright Term Premium
61 66 71 76 81 86 91 96 01 060
2
4
6
8
10
12
14
Survey-based10-yearinflationexpectations
Fig. 1 10-year Treasury bond yield and inflation expectations Percent
Data are quarterly. The 10-year zero-coupon Treasury bond yield is the end-of-quarter yield from Gurkaynak, Sack, and Wright (2007). 10-year inflation expectations are from the Federal Reserve Board, which is from three sources: from 1991 onward, the data are inflation expectations from 5 to 10 years ahead from the Survey of Professional Forecasters; from 1981 to 1991, the data are inflation expectations from 5 to 10 years ahead from the Blue Chip Survey of forecasters; prior to1981, this series was extended backward by Federal Reserve Board staff using multiple data sources and the FRB/US model.
10-yearzero-couponyield
0
2
4
6
8
10
12
14
61 64 67 70 73 76 79 82 85 88 91 94 97 00 03 06
Fig. 2 Affine, no-arbitrage model decomposition of 10-year bond yield
10-yearzero-couponyield
Risk-neutral10-yearzero-couponyield
10-yearterm premium
Data are quarterly, sampled at the end of each quarter. Source: Kim and Wright (2005).
Percent
![Page 8: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/8.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Why Study the Term Premium?
![Page 9: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/9.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Why Study the Term Premium in a DSGE Model?
Relative to equity premium, the term premium:applies to a larger volume of securitiesis used by central banks to measure expectations of monetarypolicy, inflationonly requires modeling short-term interest rate, not dividendsor leverageprovides an additional perspective on the modeltests nominal rigidities
More generally:many empirical questions about risk premia require astructural DSGE model to provide reliable answersDSGE models widely used in macroeconomics; total failure toexplain risk premia may signal flaws in the model
![Page 10: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/10.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Why Study the Term Premium in a DSGE Model?
Relative to equity premium, the term premium:applies to a larger volume of securitiesis used by central banks to measure expectations of monetarypolicy, inflationonly requires modeling short-term interest rate, not dividendsor leverageprovides an additional perspective on the modeltests nominal rigidities
More generally:many empirical questions about risk premia require astructural DSGE model to provide reliable answersDSGE models widely used in macroeconomics; total failure toexplain risk premia may signal flaws in the model
![Page 11: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/11.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Why Study the Term Premium in a DSGE Model?
Relative to equity premium, the term premium:applies to a larger volume of securitiesis used by central banks to measure expectations of monetarypolicy, inflationonly requires modeling short-term interest rate, not dividendsor leverageprovides an additional perspective on the modeltests nominal rigidities
More generally:many empirical questions about risk premia require astructural DSGE model to provide reliable answersDSGE models widely used in macroeconomics; total failure toexplain risk premia may signal flaws in the model
![Page 12: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/12.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Some Recent Studies of the Bond Premium Puzzle
Wachter (2005)can resolve bond premium puzzle using Campbell-Cochranepreferences in endowment economy
Rudebusch and Swanson (2008)the term premium is far too small in a standard New Keynesianmodel, even with Campbell-Cochrane habitssimilar finding by Jermann (1998), Lettau and Uhlig (2000) forequity premium in an RBC model
Piazzesi-Schneider (2006)can resolve bond premium puzzle using Epstein-Zinpreferences in endowment economy
We examine to what extent the Piazzesi-Schneider resultsgeneralize to the DSGE case
![Page 13: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/13.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Some Recent Studies of the Bond Premium Puzzle
Wachter (2005)can resolve bond premium puzzle using Campbell-Cochranepreferences in endowment economy
Rudebusch and Swanson (2008)the term premium is far too small in a standard New Keynesianmodel, even with Campbell-Cochrane habitssimilar finding by Jermann (1998), Lettau and Uhlig (2000) forequity premium in an RBC model
Piazzesi-Schneider (2006)can resolve bond premium puzzle using Epstein-Zinpreferences in endowment economy
We examine to what extent the Piazzesi-Schneider resultsgeneralize to the DSGE case
![Page 14: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/14.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Some Recent Studies of the Bond Premium Puzzle
Wachter (2005)can resolve bond premium puzzle using Campbell-Cochranepreferences in endowment economy
Rudebusch and Swanson (2008)the term premium is far too small in a standard New Keynesianmodel, even with Campbell-Cochrane habitssimilar finding by Jermann (1998), Lettau and Uhlig (2000) forequity premium in an RBC model
Piazzesi-Schneider (2006)can resolve bond premium puzzle using Epstein-Zinpreferences in endowment economy
We examine to what extent the Piazzesi-Schneider resultsgeneralize to the DSGE case
![Page 15: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/15.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Some Recent Studies of the Bond Premium Puzzle
Wachter (2005)can resolve bond premium puzzle using Campbell-Cochranepreferences in endowment economy
Rudebusch and Swanson (2008)the term premium is far too small in a standard New Keynesianmodel, even with Campbell-Cochrane habitssimilar finding by Jermann (1998), Lettau and Uhlig (2000) forequity premium in an RBC model
Piazzesi-Schneider (2006)can resolve bond premium puzzle using Epstein-Zinpreferences in endowment economy
We examine to what extent the Piazzesi-Schneider resultsgeneralize to the DSGE case
![Page 16: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/16.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Related Strands of the Literature
The Bond Premium in a DSGE Model:Backus-Gregory-Zin (1989), Donaldson-Johnson-Mehra (1990),Den Haan (1995), Doh (2006), Rudebusch-Swanson (2008)
Epstein-Zin Preferences and the Bond Premium in an EndowmentEconomy:
Piazzesi-Schneider (2006), Colacito-Croce (2007), Backus-Routledge-Zin (2007), Gallmeyer-Hollifield-Palomino-Zin (2007),Bansal-Shaliastovich (2008), Doh (2008)
Epstein-Zin Preferences in a DSGE Model:Tallarini (2000), Croce (2007), Levin-Lopez-Salido-Nelson-Yun(2008)
Epstein-Zin Preferences and the Bond Premium in a DSGE Model:van Binsbergen-Fernandez-Villaverde-Koijen-Rubio-Ramirez (2008)
![Page 17: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/17.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences in a Standard DSGE Model
2 Epstein-Zin Preferences in a Standard NK ModelEpstein-Zin PreferencesStandard New Keynesian ModelPrice Assets in the ModelSolve the ModelResults
![Page 18: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/18.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Standard preferences:
Vt ≡ u(ct , lt ) + βEtVt+1
Epstein-Zin preferences:
Vt ≡ u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Note:need to impose u ≥ 0
or u ≤ 0 and Vt ≡ u(ct , lt )− β(Et (−Vt+1)1−α)1/(1−α)
We’ll use standard NK utility kernel:
u(ct , lt ) ≡c1−γ
t1− γ
− χ0l1+χt
1 + χ,
![Page 19: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/19.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Standard preferences:
Vt ≡ u(ct , lt ) + βEtVt+1
Epstein-Zin preferences:
Vt ≡ u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Note:need to impose u ≥ 0
or u ≤ 0 and Vt ≡ u(ct , lt )− β(Et (−Vt+1)1−α)1/(1−α)
We’ll use standard NK utility kernel:
u(ct , lt ) ≡c1−γ
t1− γ
− χ0l1+χt
1 + χ,
![Page 20: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/20.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Standard preferences:
Vt ≡ u(ct , lt ) + βEtVt+1
Epstein-Zin preferences:
Vt ≡ u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Note:need to impose u ≥ 0
or u ≤ 0 and Vt ≡ u(ct , lt )− β(Et (−Vt+1)1−α)1/(1−α)
We’ll use standard NK utility kernel:
u(ct , lt ) ≡c1−γ
t1− γ
− χ0l1+χt
1 + χ,
![Page 21: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/21.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Standard preferences:
Vt ≡ u(ct , lt ) + βEtVt+1
Epstein-Zin preferences:
Vt ≡ u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Note:need to impose u ≥ 0
or u ≤ 0 and Vt ≡ u(ct , lt )− β(Et (−Vt+1)1−α)1/(1−α)
We’ll use standard NK utility kernel:
u(ct , lt ) ≡c1−γ
t1− γ
− χ0l1+χt
1 + χ,
![Page 22: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/22.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Standard preferences:
Vt ≡ u(ct , lt ) + βEtVt+1
Epstein-Zin preferences:
Vt ≡ u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Note:need to impose u ≥ 0
or u ≤ 0 and Vt ≡ u(ct , lt )− β(Et (−Vt+1)1−α)1/(1−α)
We’ll use standard NK utility kernel:
u(ct , lt ) ≡c1−γ
t1− γ
− χ0l1+χt
1 + χ,
![Page 23: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/23.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Household optimality conditions with EZ preferences:
µt u1∣∣(ct ,lt )
= Ptλt
−µt u2∣∣(ct ,lt )
= wtλt
λt = βEtλt+1(1 + rt+1)
µt = µt−1(Et−1V 1−α
t)α/(1−α)V−αt , µ0 = 1
Recall: Vt = u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Stochastic discount factor:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
![Page 24: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/24.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Household optimality conditions with EZ preferences:
µt u1∣∣(ct ,lt )
= Ptλt
−µt u2∣∣(ct ,lt )
= wtλt
λt = βEtλt+1(1 + rt+1)
µt = µt−1(Et−1V 1−α
t)α/(1−α)V−αt , µ0 = 1
Recall: Vt = u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Stochastic discount factor:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
![Page 25: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/25.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Epstein-Zin Preferences
Household optimality conditions with EZ preferences:
µt u1∣∣(ct ,lt )
= Ptλt
−µt u2∣∣(ct ,lt )
= wtλt
λt = βEtλt+1(1 + rt+1)
µt = µt−1(Et−1V 1−α
t)α/(1−α)V−αt , µ0 = 1
Recall: Vt = u(ct , lt ) + β(EtV 1−α
t+1
)1/(1−α)
Stochastic discount factor:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
![Page 26: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/26.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
New Keynesian Model (Very Standard)
Continuum of differentiated firms:face Dixit-Stiglitz demand with elasticity 1+θ
θ , markup θset prices in Calvo contracts with avg. duration 4 quartersidentical production functions yt = At k1−η lηthave firm-specific capital stocksface aggregate technology log At = ρA log At−1 + εA
t
Parameters θ = .2, ρA = .9, σ2A = .012
Perfectly competitive goods aggregation sector
![Page 27: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/27.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
New Keynesian Model (Very Standard)
Government:imposes lump-sum taxes Gt on householdsdestroys the resources it collectslog Gt = ρG log Gt−1 + (1− ρg) log G + εG
t
Parameters G = .17Y , ρG = .9, σ2G = .0042
Monetary Authority:
it = ρi it−1 + (1− ρi) [1/β + πt + gy (yt − y) + gπ(πt − π∗)] + εit
Parameters ρi = .73, gy = .53, gπ = .93, π∗ = 0, σ2i = .0042
![Page 28: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/28.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
New Keynesian Model (Very Standard)
Government:imposes lump-sum taxes Gt on householdsdestroys the resources it collectslog Gt = ρG log Gt−1 + (1− ρg) log G + εG
t
Parameters G = .17Y , ρG = .9, σ2G = .0042
Monetary Authority:
it = ρi it−1 + (1− ρi) [1/β + πt + gy (yt − y) + gπ(πt − π∗)] + εit
Parameters ρi = .73, gy = .53, gπ = .93, π∗ = 0, σ2i = .0042
![Page 29: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/29.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Asset Pricing
Asset pricing:
pt = dt + Et [mt+1pt+1]
Zero-coupon bond pricing:
p(n)t = Et [mt+1p(n−1)
t+1 ]
i(n)t = −1
nlog p(n)
t
Notation: let it ≡ i(1)t
![Page 30: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/30.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Asset Pricing
Asset pricing:
pt = dt + Et [mt+1pt+1]
Zero-coupon bond pricing:
p(n)t = Et [mt+1p(n−1)
t+1 ]
i(n)t = −1
nlog p(n)
t
Notation: let it ≡ i(1)t
![Page 31: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/31.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,a perpetuity that pays $1, δc , δ2
c , δ3c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 32: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/32.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,
a perpetuity that pays $1, δc , δ2c , δ3
c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 33: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/33.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,a perpetuity that pays $1, δc , δ2
c , δ3c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 34: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/34.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,a perpetuity that pays $1, δc , δ2
c , δ3c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 35: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/35.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,a perpetuity that pays $1, δc , δ2
c , δ3c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 36: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/36.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
The Term Premium in the Standard NK Model
In DSGE framework, convenient to work with a default-free consol,a perpetuity that pays $1, δc , δ2
c , δ3c , . . . (nominal)
Price of the consol:
p(n)t = 1 + δc Etmt+1p(n)
t+1
Risk-neutral consol price:
p(n)t = 1 + δc e−it Et p
(n)t+1
Term premium:
ψ(n)t ≡ log
(δc p(n)
t
p(n)t − 1
)− log
(δc p(n)
t
p(n)t − 1
)
![Page 37: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/37.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Solving the Model
The standard NK model above has a relatively large numer of statevariables: At−1, Gt−1, it−1, ∆t−1, πt−1, εA
t , εGt , εi
t
We solve the model by approximation around the nonstochasticsteady state (perturbation methods)
In a first-order approximation, term premium is zeroIn a second-order approximation, term premium is a constant(sum of variances)So we compute a third-order approximation of the solutionaround nonstochastic steady statePerturbation AIM algorithm in Swanson, Anderson, Levin(2006) quickly computes nth order approximations
![Page 38: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/38.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Solving the Model
The standard NK model above has a relatively large numer of statevariables: At−1, Gt−1, it−1, ∆t−1, πt−1, εA
t , εGt , εi
t
We solve the model by approximation around the nonstochasticsteady state (perturbation methods)
In a first-order approximation, term premium is zeroIn a second-order approximation, term premium is a constant(sum of variances)So we compute a third-order approximation of the solutionaround nonstochastic steady statePerturbation AIM algorithm in Swanson, Anderson, Levin(2006) quickly computes nth order approximations
![Page 39: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/39.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Solving the Model
The standard NK model above has a relatively large numer of statevariables: At−1, Gt−1, it−1, ∆t−1, πt−1, εA
t , εGt , εi
t
We solve the model by approximation around the nonstochasticsteady state (perturbation methods)
In a first-order approximation, term premium is zeroIn a second-order approximation, term premium is a constant(sum of variances)So we compute a third-order approximation of the solutionaround nonstochastic steady statePerturbation AIM algorithm in Swanson, Anderson, Levin(2006) quickly computes nth order approximations
![Page 40: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/40.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Empirical and Model-Based Unconditional Moments
EU EZ “best fit” EZVariable U.S. Data Preferences Preferences Preferences
sd[C] 1.19 1.40 1.46 2.12sd[L] 1.71 2.48 2.50 1.89
sd[w r ] 0.82 2.02 2.02 2.02sd[π] 2.52 2.22 2.30 2.96sd[i ] 2.71 1.86 1.93 2.65
sd[i (40)] 2.41 0.52 0.57 1.17
mean[ψ(40)] 1.06 .010 .438 1.06sd[ψ(40)] 0.54 .000 .053 .162
mean[i (40) − i ] 1.43 −.038 .390 0.95sd[i (40) − i ] 1.33 1.41 1.43 1.59mean[x (40)] 1.76 .010 .431 1.04
sd[x (40)] 23.43 6.52 6.87 10.77
memo: IES .5 .5 .5quasi-CRRA 2 75 90
![Page 41: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/41.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Empirical and Model-Based Unconditional Moments
EU EZ “best fit” EZVariable U.S. Data Preferences Preferences Preferences
sd[C] 1.19 1.40 1.46 2.12sd[L] 1.71 2.48 2.50 1.89
sd[w r ] 0.82 2.02 2.02 2.02sd[π] 2.52 2.22 2.30 2.96sd[i ] 2.71 1.86 1.93 2.65
sd[i (40)] 2.41 0.52 0.57 1.17
mean[ψ(40)] 1.06 .010 .438 1.06sd[ψ(40)] 0.54 .000 .053 .162
mean[i (40) − i ] 1.43 −.038 .390 0.95sd[i (40) − i ] 1.33 1.41 1.43 1.59mean[x (40)] 1.76 .010 .431 1.04
sd[x (40)] 23.43 6.52 6.87 10.77
memo: IES .5 .5 .5quasi-CRRA 2 75 90
![Page 42: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/42.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Arrow-Pratt:−C u′′(C)
u′(C)
Here:
Vt =c1−γ
t1− γ
− χ0l1+χt
1 + χ+ β
(EtV 1−α
t+1
)1/(1−α)
CRRA =−W V ′′(W )
V ′(W )+ α
W V ′(W )
V (W )
=−u11 + λu12
u1
c1 + wλ
+ αc u1
u
see “Risk Aversion, the Labor Margin, and Asset Pricing in a DSGE Model”
![Page 43: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/43.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Arrow-Pratt:−C u′′(C)
u′(C)
Here:
Vt =c1−γ
t1− γ
− χ0l1+χt
1 + χ+ β
(EtV 1−α
t+1
)1/(1−α)
CRRA =−W V ′′(W )
V ′(W )+ α
W V ′(W )
V (W )
=−u11 + λu12
u1
c1 + wλ
+ αc u1
u
see “Risk Aversion, the Labor Margin, and Asset Pricing in a DSGE Model”
![Page 44: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/44.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Arrow-Pratt:−C u′′(C)
u′(C)
Here:
Vt =c1−γ
t1− γ
− χ0l1+χt
1 + χ+ β
(EtV 1−α
t+1
)1/(1−α)
CRRA =−W V ′′(W )
V ′(W )+ α
W V ′(W )
V (W )
=−u11 + λu12
u1
c1 + wλ
+ αc u1
u
see “Risk Aversion, the Labor Margin, and Asset Pricing in a DSGE Model”
![Page 45: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/45.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Arrow-Pratt:−C u′′(C)
u′(C)
Here:
Vt =c1−γ
t1− γ
− χ0l1+χt
1 + χ+ β
(EtV 1−α
t+1
)1/(1−α)
CRRA =−W V ′′(W )
V ′(W )+ α
W V ′(W )
V (W )
=−u11 + λu12
u1
c1 + wλ
+ αc u1
u
see “Risk Aversion, the Labor Margin, and Asset Pricing in a DSGE Model”
![Page 46: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/46.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Arrow-Pratt:−C u′′(C)
u′(C)
Here:
Vt =c1−γ
t1− γ
− χ0l1+χt
1 + χ+ β
(EtV 1−α
t+1
)1/(1−α)
CRRA =−W V ′′(W )
V ′(W )+ α
W V ′(W )
V (W )
=−u11 + λu12
u1
c1 + wλ
+ αc u1
u
see “Risk Aversion, the Labor Margin, and Asset Pricing in a DSGE Model”
![Page 47: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/47.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Epstein-Zin preferences:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
Barillas-Hansen-Sargent (2008):
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
ψt+1
ψt
Pt
Pt+1
Guvenen (2006), Moskowitz-Vissing-Jorgensen (2009):heterogeneous agents
![Page 48: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/48.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Epstein-Zin preferences:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
Barillas-Hansen-Sargent (2008):
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
ψt+1
ψt
Pt
Pt+1
Guvenen (2006), Moskowitz-Vissing-Jorgensen (2009):heterogeneous agents
![Page 49: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/49.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Coefficient of Relative Risk Aversion
Epstein-Zin preferences:
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
(Vt+1(
EtV 1−αt+1
)1/(1−α)
)−αPt
Pt+1
Barillas-Hansen-Sargent (2008):
mt ,t+1 ≡βu1
∣∣(ct+1,lt+1)
u1∣∣(ct ,lt )
ψt+1
ψt
Pt
Pt+1
Guvenen (2006), Moskowitz-Vissing-Jorgensen (2009):heterogeneous agents
![Page 50: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/50.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Risk Aversion and the Term Premium
40
60
80
100
120
140
erm
pre
miu
m (
basi
s po
ints
)
-20
0
20
0 10 20 30 40 50 60 70 80 90 100
mea
n te
quasi-CRRA
Expected Utility Preferences,no long‐run risk
![Page 51: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/51.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Risk Aversion and the Term Premium
40
60
80
100
120
140
erm
pre
miu
m (
basi
s po
ints
)
Epstein‐Zin Preferences,no long‐run risk
-20
0
20
0 10 20 30 40 50 60 70 80 90 100
mea
n te
quasi-CRRA
Expected Utility Preferences,no long‐run risk
![Page 52: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/52.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Risk Aversion and the Term Premium
40
60
80
100
120
140
erm
pre
miu
m (
basi
s po
ints
)
Epstein‐Zin Preferences,no long‐run risk
Epstein‐Zin Preferences,with long‐run inflation risk
-20
0
20
0 10 20 30 40 50 60 70 80 90 100
mea
n te
quasi-CRRA
Expected Utility Preferences,no long‐run risk
![Page 53: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/53.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Risks
3 Long-Run RisksLong-Run Inflation RiskLong-Run Real Risk
![Page 54: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/54.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Long-run inflation risk makes long-term bonds more risky:same idea as Bansal-Yaron (2004), but with nominal riskrather than real risklong-term inflation expectations more observable thanlong-term consumption growth
![Page 55: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/55.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Long-run inflation risk makes long-term bonds more risky:same idea as Bansal-Yaron (2004), but with nominal riskrather than real risklong-term inflation expectations more observable thanlong-term consumption growth
![Page 56: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/56.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
6
8
10
12
14
Per
cent
10-year zero-coupon Treasury yield
0
2
4
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
10-year inflation expectationsfrom SPF, Blue Chip
![Page 57: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/57.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Long-run inflation risk makes long-term bonds more risky:same idea as Bansal-Yaron (2004), but with nominal riskrather than real risklong-term inflation expectations more observable thanlong-term consumption growth
other evidence (Kozicki-Tinsley, 2003, Gürkaynak, Sack,Swanson, 2005) that long-term inflation expectations in theU.S. vary
![Page 58: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/58.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Long-run inflation risk makes long-term bonds more risky:same idea as Bansal-Yaron (2004), but with nominal riskrather than real risklong-term inflation expectations more observable thanlong-term consumption growthother evidence (Kozicki-Tinsley, 2003, Gürkaynak, Sack,Swanson, 2005) that long-term inflation expectations in theU.S. vary
![Page 59: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/59.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Suppose:π∗t = ρ∗ππ
∗t−1 + επ
∗t
Then:inflation is volatile, but not riskyin fact, long-term bonds act like insurance:when π∗ ↑, then C ↑ and p(40) ↓result: term premium is negative
![Page 60: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/60.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Suppose:π∗t = ρ∗ππ
∗t−1 + επ
∗t
Then:inflation is volatile, but not riskyin fact, long-term bonds act like insurance:when π∗ ↑, then C ↑ and p(40) ↓result: term premium is negative
![Page 61: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/61.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Consider instead:
π∗t = ρ∗ππ∗t−1 + (1− ρ∗π)θπ∗(πt − π∗t ) + επ
∗t
θπ∗ describes pass-through from current π to long-term π∗
Gürkaynak, Sack, and Swanson (2005) found evidence forθπ∗> 0 in U.S. bond response to macro data releasesmakes long-term bonds act less like insurance:when technology/supply shock, then π ↑, C ↓, and p(40) ↓supply shocks become very costlyThe term premium is positive, closely associated with θπ∗
![Page 62: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/62.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Inflation Risk
Consider instead:
π∗t = ρ∗ππ∗t−1 + (1− ρ∗π)θπ∗(πt − π∗t ) + επ
∗t
θπ∗ describes pass-through from current π to long-term π∗
Gürkaynak, Sack, and Swanson (2005) found evidence forθπ∗> 0 in U.S. bond response to macro data releasesmakes long-term bonds act less like insurance:when technology/supply shock, then π ↑, C ↓, and p(40) ↓supply shocks become very costlyThe term premium is positive, closely associated with θπ∗
![Page 63: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/63.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model-Based Moments with Long-Run Inflation Risk
EU Preferences EZ PreferencesVariable U.S. Data & LR π∗ Risk & LR π∗ Risk
sd[C] 1.19 1.70 2.01sd[L] 1.71 3.02 1.37
sd[w r ] 0.82 2.40 1.52sd[π] 2.52 3.65 3.25sd[i ] 2.71 3.32 2.94
sd[i (40)] 2.41 1.71 1.89
mean[ψ(40)] 1.06 .003 1.05sd[ψ(40)] 0.54 .001 .51
mean[i (40) − i ] 1.43 −.10 .96sd[i (40) − i ] 1.33 1.73 1.10mean[x (40)] 1.76 .003 1.04
sd[x (40)] 23.43 13.07 11.64
memo: IES .5 1.1quasi-CRRA 2 90
![Page 64: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/64.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Long-Run Productivity Risk
Following Bansal and Yaron (2004), introduce long-run real risk tomake the economy more risky:
Assume productivity follows:
log At = log A∗t + εAt
log A∗t = ρA∗ log A∗t−1 + εA∗t
makes the economy much riskier to agentsincreases volatility of stochastic discount factor
![Page 65: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/65.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model-Based Moments w/Long-Run Productivity Risk
EZ Preferences EZ PreferencesVariable U.S. Data & LR π∗ Risk & LR A∗ risk
sd[C] 1.19 2.01 2.37sd[L] 1.71 1.37 2.13
sd[w r ] 0.82 1.52 1.81sd[π] 2.52 3.25 2.95sd[i ] 2.71 2.94 2.86
sd[i (40)] 2.41 1.89 1.66
mean[ψ(40)] 1.06 1.05 0.98sd[ψ(40)] 0.54 0.51 0.28
mean[i (40) − i ] 1.43 0.96 0.89sd[i (40) − i ] 1.33 1.10 1.36mean[x (40)] 1.76 1.04 0.96
sd[x (40)] 23.43 11.64 12.20
memo: IES 1.1 .5quasi-CRRA 90 90
![Page 66: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/66.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Implications
4 Model ImplicationsNominal Yield Curve is Upward-SlopingTerm Premium is CountercylicalModel Is Nonhomothetic, Heteroskedastic
![Page 67: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/67.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Nominal Yield Curve is Upward-Sloping
Backus-Gregory-Zin (1989), Den Haan (1995)if interest rates are low in recessionsthen bond prices rise in recessions=⇒ the term premium should be negativethe yield curve slopes downward
This paper:technology shocks imply that inflation is high in recessionsthen nominal bond prices fall in recessions=⇒ the nominal yield curve slopes upward
Note: Backus et. al intuition still applies to real yield curve
![Page 68: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/68.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Nominal Yield Curve is Upward-Sloping
Backus-Gregory-Zin (1989), Den Haan (1995)if interest rates are low in recessionsthen bond prices rise in recessions=⇒ the term premium should be negativethe yield curve slopes downward
This paper:technology shocks imply that inflation is high in recessionsthen nominal bond prices fall in recessions=⇒ the nominal yield curve slopes upward
Note: Backus et. al intuition still applies to real yield curve
![Page 69: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/69.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Nominal Yield Curve is Upward-Sloping
Backus-Gregory-Zin (1989), Den Haan (1995)if interest rates are low in recessionsthen bond prices rise in recessions=⇒ the term premium should be negativethe yield curve slopes downward
This paper:technology shocks imply that inflation is high in recessionsthen nominal bond prices fall in recessions=⇒ the nominal yield curve slopes upward
Note: Backus et. al intuition still applies to real yield curve
![Page 70: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/70.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Nominal Yield Curve is Upward-Sloping
3.0
3.1
3.2
3.3
3.4
3.5
3.6
4.6
4.7
4.8
4.9
5
5.1
5.2
Yield (p
ercent)
Model‐Implied Nominal Yield Curve (left axis)
2.6
2.7
2.8
2.9
4.2
4.3
4.4
4.5
0 4 8 12 16 20 24 28 32 36 40
Maturity (months)
Model‐Implied Real Yield Curve (right axis)
![Page 71: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/71.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
US Yield Curve, 1994–2007
4.5
4.7
4.9
5.1
5.3
5.5
5.7
Yield (p
ercent)
US Nominal Yield Curve, 1994‐2007 (left axis)
3.7
3.9
4.1
4.3
0 4 8 12 16 20 24 28 32 36 40
Maturity (months)
![Page 72: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/72.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
US Yield Curve, 1994–2007
2.0
2.2
2.4
2.6
2.8
3.0
3.2
4.5
4.7
4.9
5.1
5.3
5.5
5.7
Yield (p
ercent)
US Nominal Yield Curve, 1994‐2007 (left axis)
1.2
1.4
1.6
1.8
3.7
3.9
4.1
4.3
0 4 8 12 16 20 24 28 32 36 40
Maturity (months)
US Real Yield Curve, 2004‐2007 (right axis)
![Page 73: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/73.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
UK Yield Curve, 1994–2007
2.7
2.8
2.9
3.0
3.1
3.2
3.3
5.3
5.4
5.5
5.6
5.7
5.8
5.9
Yield (p
ercent)
UK Nominal Yield Curve, 1994‐2007 (left axis)
2.3
2.4
2.5
2.6
4.9
5.0
5.1
5.2
0 4 8 12 16 20 24 28 32 36 40
Maturity (months)
UK Real Yield Curve, 1994‐2007 (right axis)
![Page 74: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/74.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Term Premium is Countercylical
0 5 10 15 20-.2
0
.2
.4
.6Consumption
Per
cent
0 5 10 15 20-.06
-.04
-.02
.00Consumption
0 5 10 15 20-.2
-.1
.0
Consumption
0 5 10 15 20-1.2
-1.0
-.8
-.6
-.4
-.2
0Inflation
Per
cent
0 5 10 15 20-.004
-.002
.000
0
.004
Inflation
0 5 10 15 20-.12
-.1
-.08
-.06
-.04
-.02
0
.02Inflation
0 5 10 15 20.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4Long-Term Bond Price
Per
cent
0 5 10 15 20-.07
-.06
-.05
-.04
-.03
-.02
-.01
0Long-Term Bond Price
0 5 10 15 20-.25
-.2
-.15
-.1
-.05
0
.05Long-Term Bond Price
Response to
Technology Shock
Response to Government
Spending Shock
Response to Monetary
Policy Shock
Figure 2. Impulse responses to structural shocks.
Impulse responses of consumption, inflation, long-term bond prices, and term
premiums to positive one standard deviation shocks to technology, government
spending, and monetary policy.
0 5 10 15 20-4.5
-4
-3.5
-3
-2.5
-2
-1.5
Quarters
Term Premium
Bas
is P
oin
ts
0 5 10 15 200
.01
.02
.03
.04
.05
Quarters
Term Premium
0 5 10 15 20-.1
0
.1
.2
Quarters
Term Premium
0 5 10 15 20-.2
0
.2
.4
.6Consumption
Per
cent
0 5 10 15 20-.06
-.04
-.02
.00Consumption
0 5 10 15 20-.2
-.1
.0
Consumption
0 5 10 15 20-1.2
-1.0
-.8
-.6
-.4
-.2
0Inflation
Per
cent
0 5 10 15 20-.004
-.002
.000
0
.004
Inflation
0 5 10 15 20-.12
-.1
-.08
-.06
-.04
-.02
0
.02Inflation
0 5 10 15 20.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4Long-Term Bond Price
Per
cent
0 5 10 15 20-.07
-.06
-.05
-.04
-.03
-.02
-.01
0Long-Term Bond Price
0 5 10 15 20-.25
-.2
-.15
-.1
-.05
0
.05Long-Term Bond Price
Response to
Technology Shock
Response to Government
Spending Shock
Response to Monetary
Policy Shock
Figure 2. Impulse responses to structural shocks.
Impulse responses of consumption, inflation, long-term bond prices, and term
premiums to positive one standard deviation shocks to technology, government
spending, and monetary policy.
0 5 10 15 20-4.5
-4
-3.5
-3
-2.5
-2
-1.5
Quarters
Term Premium
Bas
is P
oin
ts
0 5 10 15 200
.01
.02
.03
.04
.05
Quarters
Term Premium
0 5 10 15 20-.1
0
.1
.2
Quarters
Term Premium
![Page 75: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/75.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
![Page 76: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/76.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
![Page 77: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/77.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
![Page 78: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/78.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
![Page 79: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/79.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
term premium term premiumModel mean (bp) std dev (bp)
baseline model 86.5 11.0
![Page 80: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/80.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Model Is Nonhomothetic, Heteroskedastic
p(2)t − p(2)
t = Etmt+1p(1)t+1 − Etmt+1Etp
(1)t+1 = Covt (mt+1,p
(1)t+1)
time-varying term premium ⇐⇒ conditional heteroskedasticity
Second-order solution:
xt = µx +∑
αxdxt−1 +∑
αεεt
+∑
αxxdxt−1dxt−1 +∑
αxεdxt−1εt +∑
αεεεtεt + . . .
term premium term premiumModel mean (bp) std dev (bp)
baseline model 86.5 11.0log-linear log-normal 86.5 0.0
![Page 81: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/81.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Conclusions
1 The term premium in standard NK DSGE models is verysmall, even more stable
2 Habit-based preferences can solve bond premium puzzle inendowment economy, but fail in NK DSGE framework:although agents are risk-averse, they can offset that risk
3 Epstein-Zin preferences can solve bond premium puzzle inendowment economy, are much more promising in NK DSGEframework:agents are risk-averse and cannot offset long-run real ornominal risks
4 Long-run risks reduce the required quasi-CRRA, increasevolatility of risk premia, help fit financial moments
![Page 82: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/82.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Conclusions
1 The term premium in standard NK DSGE models is verysmall, even more stable
2 Habit-based preferences can solve bond premium puzzle inendowment economy, but fail in NK DSGE framework:although agents are risk-averse, they can offset that risk
3 Epstein-Zin preferences can solve bond premium puzzle inendowment economy, are much more promising in NK DSGEframework:agents are risk-averse and cannot offset long-run real ornominal risks
4 Long-run risks reduce the required quasi-CRRA, increasevolatility of risk premia, help fit financial moments
![Page 83: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/83.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Conclusions
1 The term premium in standard NK DSGE models is verysmall, even more stable
2 Habit-based preferences can solve bond premium puzzle inendowment economy, but fail in NK DSGE framework:although agents are risk-averse, they can offset that risk
3 Epstein-Zin preferences can solve bond premium puzzle inendowment economy, are much more promising in NK DSGEframework:agents are risk-averse and cannot offset long-run real ornominal risks
4 Long-run risks reduce the required quasi-CRRA, increasevolatility of risk premia, help fit financial moments
![Page 84: The Bond Premium in a DSGE Model with Long-Run Real and …swanson2/pres/bankofitaly.pdf · 2012-07-30 · Motivation Basic Model Long-Run Risks Model Implications Conclusions The](https://reader036.vdocuments.us/reader036/viewer/2022081612/5f7f0d892b0df757cf0a4198/html5/thumbnails/84.jpg)
Motivation Basic Model Long-Run Risks Model Implications Conclusions
Conclusions
1 The term premium in standard NK DSGE models is verysmall, even more stable
2 Habit-based preferences can solve bond premium puzzle inendowment economy, but fail in NK DSGE framework:although agents are risk-averse, they can offset that risk
3 Epstein-Zin preferences can solve bond premium puzzle inendowment economy, are much more promising in NK DSGEframework:agents are risk-averse and cannot offset long-run real ornominal risks
4 Long-run risks reduce the required quasi-CRRA, increasevolatility of risk premia, help fit financial moments