Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Objective Findings Comments
Discussion on “Stagnation Traps”
Jang-Ting Guo
Department of EconomicsUniversity of California, Riverside
May 15, 2015
Jang-Ting Guo Discussion on “Stagnation Traps” 1 / 12
Objective Findings Comments
Existence and Persistence of Stagnation Trap in a MonetaryEndogenous Growth Model with Quality Ladders
⇒ Coexistence of Positive Unemployment, Low Growth, andLiquidity Trap
The Key Mechanism
(1) Unemployment and Weak Aggregate Demand ⇒ ReducesFirms’ Investment in Innovation ⇒ Low Growth
(2) Low Growth ⇒ Reduces Real Interest Rate ⇒ PushesNominal Interest Rate to Zero
Jang-Ting Guo Discussion on “Stagnation Traps” 2 / 12
Objective Findings Comments
Existence and Persistence of Stagnation Trap in a MonetaryEndogenous Growth Model with Quality Ladders
⇒ Coexistence of Positive Unemployment, Low Growth, andLiquidity Trap
The Key Mechanism
(1) Unemployment and Weak Aggregate Demand ⇒ ReducesFirms’ Investment in Innovation ⇒ Low Growth
(2) Low Growth ⇒ Reduces Real Interest Rate ⇒ PushesNominal Interest Rate to Zero
Jang-Ting Guo Discussion on “Stagnation Traps” 2 / 12
Objective Findings Comments
Two Steady States in Baseline Model
(1) Full Employment y f = 1, High Growth g f , PositiveNominal Interest Rate i f > 0, and Positive/Negative InflationRate πf ≷ 1
(2) Unemployment yu < 1, Low Growth gu < g f , ZeroNominal Interest Rate iu = 0, and Negative Inflation Rateπu < 1
Two Extensions: Precautionary Savings and Time-VaryingInflation Rate
Constant or Countercyclical Subsidy to Firms’ Investment inInnovation ⇒ Removal of Low-Growth Steady State
Jang-Ting Guo Discussion on “Stagnation Traps” 3 / 12
Objective Findings Comments
Two Steady States in Baseline Model
(1) Full Employment y f = 1, High Growth g f , PositiveNominal Interest Rate i f > 0, and Positive/Negative InflationRate πf ≷ 1
(2) Unemployment yu < 1, Low Growth gu < g f , ZeroNominal Interest Rate iu = 0, and Negative Inflation Rateπu < 1
Two Extensions: Precautionary Savings and Time-VaryingInflation Rate
Constant or Countercyclical Subsidy to Firms’ Investment inInnovation ⇒ Removal of Low-Growth Steady State
Jang-Ting Guo Discussion on “Stagnation Traps” 3 / 12
Objective Findings Comments
Two Steady States: y f = 1 and yu < 1
⇒ y Denotes the Level of Actual Output
⇒ 1− y = Output Gap
Figure 1 ⇒ Local Stability Property of Each Steady State:Saddle, Sink or Source
Possibility of Global Indeterminacy ⇒ Various Forms ofBifurcations
Jang-Ting Guo Discussion on “Stagnation Traps” 4 / 12
Objective Findings Comments
Two Steady States: y f = 1 and yu < 1
⇒ y Denotes the Level of Actual Output
⇒ 1− y = Output Gap
Figure 1 ⇒ Local Stability Property of Each Steady State:Saddle, Sink or Source
Possibility of Global Indeterminacy ⇒ Various Forms ofBifurcations
Jang-Ting Guo Discussion on “Stagnation Traps” 4 / 12
(yu, gu)
(1, gf )
AD
GG
gro
wth
g
output gap y
Objective Findings Comments
This Paper
max∞∑t=0
βtC 1−σt − 1
1− σ, 0 < β < 1
Ct = exp
(∫ 1
0ln qjtcjtdj
)and Qt = exp
(∫ 1
0lnqjtdj
)(ct+1
ct
)σ
= β (1 + rt) g1−σt+1 , where gt+1 =
Qt+1
Qt
Need σ > 1 such that
(1) Positive Relationship between Present Consumption andInnovation Growth
(2) Existence of Unemployment Steady State
(3) i f > 0 at Full-Employment Steady State
Jang-Ting Guo Discussion on “Stagnation Traps” 5 / 12
Objective Findings Comments
This Paper
max∞∑t=0
βtC 1−σt − 1
1− σ, 0 < β < 1
Ct = exp
(∫ 1
0ln qjtcjtdj
)and Qt = exp
(∫ 1
0lnqjtdj
)(ct+1
ct
)σ
= β (1 + rt) g1−σt+1 , where gt+1 =
Qt+1
Qt
Need σ > 1 such that
(1) Positive Relationship between Present Consumption andInnovation Growth
(2) Existence of Unemployment Steady State
(3) i f > 0 at Full-Employment Steady State
Jang-Ting Guo Discussion on “Stagnation Traps” 5 / 12
Objective Findings Comments
Alternative Specification (Footnote 14)
max∞∑t=0
βtc1−σt − 1
1− σ, 0 < β < 1
yt = f
(∫ 1
0qjtXjtdj
)= f (Qt)
(ct+1
ct
)σ
= β (1 + rt)
(ct+1
ct
)σ
= β (1 + rt) g1−σt+1 , where gt+1 =
Qt+1
Qt
⇒ Isomorphic Formulations Only When σ = 1
Jang-Ting Guo Discussion on “Stagnation Traps” 6 / 12
Objective Findings Comments
Alternative Specification (Footnote 14)
max∞∑t=0
βtc1−σt − 1
1− σ, 0 < β < 1
yt = f
(∫ 1
0qjtXjtdj
)= f (Qt)
(ct+1
ct
)σ
= β (1 + rt)
(ct+1
ct
)σ
= β (1 + rt) g1−σt+1 , where gt+1 =
Qt+1
Qt
⇒ Isomorphic Formulations Only When σ = 1
Jang-Ting Guo Discussion on “Stagnation Traps” 6 / 12
Objective Findings Comments
This Paper
Euler:
(ct+1
ct
)σ
= β(1 + it)
πg1−σt+1
Growth : 1 = β
[(ctct+1
)σ
g1−σt+1 (χ
γ − 1
γyt+1 + 1− ln gt+2
ln γ)
]When σ > 1⇒ Positive Relationship between yt+1 and gt+1
Market Clearing: ct +ln gt+1
χ ln γ= yt
Monetary Policy: 1 + it = max{
(1 + ı) yφt , 1}
Jang-Ting Guo Discussion on “Stagnation Traps” 7 / 12
Objective Findings Comments
This Paper
Euler:
(ct+1
ct
)σ
= β(1 + it)
πg1−σt+1
Growth : 1 = β
[(ctct+1
)σ
g1−σt+1 (χ
γ − 1
γyt+1 + 1− ln gt+2
ln γ)
]When σ > 1⇒ Positive Relationship between yt+1 and gt+1
Market Clearing: ct +ln gt+1
χ ln γ= yt
Monetary Policy: 1 + it = max{
(1 + ı) yφt , 1}
Jang-Ting Guo Discussion on “Stagnation Traps” 7 / 12
Objective Findings Comments
Alternative Specification
Period Utility:c1−σt − 1
1− σ
Final Good: Yt = A
∫ 1
0(qjtXjt)
αdj , A > 0, 0 < α < 1
Demand for Xjt : Xjt =
(AαqαjtPjt
) 11−α
Supply for Xjt : Xjt = Ljt , where
∫ 1
0Ljtdj + LRDt + Ut = L
R&D Firms’ Profits: πjt = (Pjt −Wt)Xjt ,Wt
Wt−1= π
Jang-Ting Guo Discussion on “Stagnation Traps” 8 / 12
Objective Findings Comments
Alternative Specification
Period Utility:c1−σt − 1
1− σ
Final Good: Yt = A
∫ 1
0(qjtXjt)
αdj , A > 0, 0 < α < 1
Demand for Xjt : Xjt =
(AαqαjtPjt
) 11−α
Supply for Xjt : Xjt = Ljt , where
∫ 1
0Ljtdj + LRDt + Ut = L
R&D Firms’ Profits: πjt = (Pjt −Wt)Xjt ,Wt
Wt−1= π
Jang-Ting Guo Discussion on “Stagnation Traps” 8 / 12
Objective Findings Comments
Alternative Specification
Period Utility:c1−σt − 1
1− σ
Final Good: Yt = A
∫ 1
0(qjtXjt)
αdj , A > 0, 0 < α < 1
Demand for Xjt : Xjt =
(AαqαjtPjt
) 11−α
Supply for Xjt : Xjt = Ljt , where
∫ 1
0Ljtdj + LRDt + Ut = L
R&D Firms’ Profits: πjt = (Pjt −Wt)Xjt ,Wt
Wt−1= π
Jang-Ting Guo Discussion on “Stagnation Traps” 8 / 12
Objective Findings Comments
Monopoly Pricing: Pjt =Wt
α
Equilibrium Quantity: Xjt =
(Aα2qαjtWt
) 11−α
Aggregate Output: Yt = A1
1−αα2α1−αW
−α1−αt Qt ,
where Qt =
∫ 1
0q
α1−α
jt dj
Equilibrium Profit: πjt = α(1− α)qα
1−α
jt
Yt
Qt
Jang-Ting Guo Discussion on “Stagnation Traps” 9 / 12
Objective Findings Comments
Monopoly Pricing: Pjt =Wt
α
Equilibrium Quantity: Xjt =
(Aα2qαjtWt
) 11−α
Aggregate Output: Yt = A1
1−αα2α1−αW
−α1−αt Qt ,
where Qt =
∫ 1
0q
α1−α
jt dj
Equilibrium Profit: πjt = α(1− α)qα
1−α
jt
Yt
Qt
Jang-Ting Guo Discussion on “Stagnation Traps” 9 / 12
Objective Findings Comments
Probability of Innovating =χLRDtL
= χµt
Value Function: Vt = β
(ct+1
ct
)−σ[πjt+1 + (1− χµt+1)Vt+1]
Free Entry: LRDt Wt = χµtVt ⇒ LWt = χVt
Innovation Growth: gt+1 =Qt+1
Qt= χµtγ
α1−α ⇒ Yt+1
Yt= gt+1π
−α1−α
Growth: 1 =(βπ
σα1−α
)g−σt+1
[α(1− α)q
α1−α
j(t+1)
χYt+1
LWtQt+1+ π(1− gt+2
γα
1−α
)]
When σ > 0⇒ Positive Relationship betweenYt+1
Qt+1and gt+1
Jang-Ting Guo Discussion on “Stagnation Traps” 10 / 12
Objective Findings Comments
Probability of Innovating =χLRDtL
= χµt
Value Function: Vt = β
(ct+1
ct
)−σ[πjt+1 + (1− χµt+1)Vt+1]
Free Entry: LRDt Wt = χµtVt ⇒ LWt = χVt
Innovation Growth: gt+1 =Qt+1
Qt= χµtγ
α1−α ⇒ Yt+1
Yt= gt+1π
−α1−α
Growth: 1 =(βπ
σα1−α
)g−σt+1
[α(1− α)q
α1−α
j(t+1)
χYt+1
LWtQt+1+ π(1− gt+2
γα
1−α
)]
When σ > 0⇒ Positive Relationship betweenYt+1
Qt+1and gt+1
Jang-Ting Guo Discussion on “Stagnation Traps” 10 / 12
Objective Findings Comments
Probability of Innovating =χLRDtL
= χµt
Value Function: Vt = β
(ct+1
ct
)−σ[πjt+1 + (1− χµt+1)Vt+1]
Free Entry: LRDt Wt = χµtVt ⇒ LWt = χVt
Innovation Growth: gt+1 =Qt+1
Qt= χµtγ
α1−α ⇒ Yt+1
Yt= gt+1π
−α1−α
Growth: 1 =(βπ
σα1−α
)g−σt+1
[α(1− α)q
α1−α
j(t+1)
χYt+1
LWtQt+1+ π(1− gt+2
γα
1−α
)]
When σ > 0⇒ Positive Relationship betweenYt+1
Qt+1and gt+1
Jang-Ting Guo Discussion on “Stagnation Traps” 10 / 12
Objective Findings Comments
Alternative Specification
Euler:
(ct+1
ct
)σ
= β(1 + it)
π
Growth: 1 =(βπ
σα1−α
)g−σt+1
[α(1− α)q
α1−α
j(t+1)
χYt+1
LWtQt+1+ π(1− gt+2
γα
1−α
)]
When σ > 0⇒ Positive Relationship betweenYt+1
Qt+1and gt+1
Market Clearing: ct = Yt ⇒ct+1
ct=
Yt+1
Yt= gt+1π
−α1−α
Monetary Policy: 1 + it = max{(1 + i)Yt
Qt, 1}
Jang-Ting Guo Discussion on “Stagnation Traps” 11 / 12
Objective Findings Comments
Alternative Specification
Euler:
(ct+1
ct
)σ
= β(1 + it)
π
Growth: 1 =(βπ
σα1−α
)g−σt+1
[α(1− α)q
α1−α
j(t+1)
χYt+1
LWtQt+1+ π(1− gt+2
γα
1−α
)]
When σ > 0⇒ Positive Relationship betweenYt+1
Qt+1and gt+1
Market Clearing: ct = Yt ⇒ct+1
ct=
Yt+1
Yt= gt+1π
−α1−α
Monetary Policy: 1 + it = max{(1 + i)Yt
Qt, 1}
Jang-Ting Guo Discussion on “Stagnation Traps” 11 / 12
(yu, gu)
(1, gf )
AD
GG
gro
wth
g
output gap y
Objective Findings Comments
At Unemployment Steady State
(1) Baseline π < 1⇒ Deflation
Extension with Precautionary Savings, but UnemployedHouseholds Cannot Borrow or Trade Firms’ Shares
(2) Zero Nominal Interest Rate iu = 0
Negative Nominal Interest Rates Observed in Europe: ECB’sDeposit Rate of −0.2%, and Swiss National Bank’s DepositRate of −0.75%
⇒ 1 + it = max{
(1 + ı) yφt , i}
, where i¯< 1
Jang-Ting Guo Discussion on “Stagnation Traps” 12 / 12
Objective Findings Comments
At Unemployment Steady State
(1) Baseline π < 1⇒ Deflation
Extension with Precautionary Savings, but UnemployedHouseholds Cannot Borrow or Trade Firms’ Shares
(2) Zero Nominal Interest Rate iu = 0
Negative Nominal Interest Rates Observed in Europe: ECB’sDeposit Rate of −0.2%, and Swiss National Bank’s DepositRate of −0.75%
⇒ 1 + it = max{
(1 + ı) yφt , i}
, where i¯< 1
Jang-Ting Guo Discussion on “Stagnation Traps” 12 / 12
(1, gf )
AD
GG
gro
wth
g
output gap y