Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation...
Transcript of Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation...
![Page 1: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/1.jpg)
Supplementary Notes on Chapter 5 of D. Romer’sAdvanced Macroeconomics Textbook (4th Edition)
Changsheng Xu & Ming Yi
School of Economics, Huazhong University of Science and Technology
This version: April 14, 2020
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 1 / 19
![Page 2: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/2.jpg)
Section 5.3 Interpreting ρA and ρG
Interpreting ρA and ρG
(5.9) At = ρAAt−1 + εA,t, −1 < ρA < 1
(5.11) Gt = ρGGt−1 + εG,t, −1 < ρG < 1
Merits of Stationary processes? Effects of a sudden shock keepdiminishing. Extreme case: ρ = 0.E(At) = E(Gt) = 0.Q: Considering the Endogenous Growth ingredients, ρA ⪋ 0?
Q: Considering the Government spending rules, ρG ⪋ 0?
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 2 / 19
![Page 3: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/3.jpg)
Section 5.4 Interpreting Equation (5.24)
Interpreting Equation (5.24)
(5.23)1
ct= e−ρEt
[1
ct+1(1 + rt+1)
](5.24)
1
ct= e−ρ
[Et
(1
ct+1
)Et(1 + rt+1) + Cov
(1
ct+1, 1 + rt+1
)]Euler Equation with uncertainty.Cov(X,Y ) = E[(X − E(X))(Y − E(Y ))]. Interpreted as how thepatterns of the realization bundles of ct+1 and rt+1 (they are bothrandom variables) would look like.If Cov < 0, households know that it is more likely (compared tothe baseline case where Cov = 0) that if a higher next-periodinterest rate occurs, a lower next-period marginal utility (onconsumption) also occurs. As a result, given other conditionsfixed, i.e, given Et(
1ct+1
) and Et(1 + rt+1), the households want tosave less (higher ct).What if Cov > 0? Economic intuitions.Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 3 / 19
![Page 4: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/4.jpg)
Section 5.5 Interpreting Equation (5.31)
Interpreting Equation (5.31)
(5.23)1
ct= e−ρEt
[1
ct+1(1 + rt+1)
](5.31) ln st − ln(1− st) = −ρ+ n+ lnα+ lnEt
(1
1− st+1
)
If households choose s = lnα+ n− ρ in each period, then there isno uncertainty in st+1 and s = st = st+1 solves (5.31).Economic intuition? In this specific and oversimplified model(with δ = 0 and Gt ≡ 0), the technological shock, say a positiveone, leads to both a decrease in 1
ct+1(given other things fixed) and
an increase in rt+1 (why?). Magically, in this model, the twoeffects offset each other:
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 4 / 19
![Page 5: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/5.jpg)
Section 5.5 Interpreting Equation (5.31)
Interpreting Equation (5.31) (Continued)
1 + rt+1
ct+1=
α(At+1Lt+1
Kt+1
)1−α
(1− st+1)Yt+1 · 1Nt+1
=αYt+1K
−1t+1
(1− st+1)Yt+1.
The two Yt+1’s offset each other. According to (5.27), there is nouncertainty in Kt+1 at time t. As a result, if st+1 is fixed at s,there is no need for the households to adjust ct according to thecontingent technological shock At+1.Any other equilibrium paths? There may be, but isn’t anequilibrium path with st ≡ s beautiful and intuitive?
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 5 / 19
![Page 6: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/6.jpg)
Section 5.5 Interpreting Equation (5.31)
Interpreting Equation (5.37)
it is thus straightforward to understand equation (5.37), givenct
1−ℓt= wt
b (5.26), ct = Yt1−sNt
, wt = Yt1−αℓtNt
: First, ℓt does not needto adjust according to At+1. Second, shock At does not change ℓt,a current-period positive technological shock tends to increaseincome and to decrease marginal utility on consumption, these twoeffects offset each other. Consequently, we haveℓt ≡ ℓ = 1−α
1−α+b(1−s) .
Again, we get fixed ℓ and s only in this oversimplified model!
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 6 / 19
![Page 7: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/7.jpg)
Section 5.5 Interpreting Fluctuations in Yt
Interpreting Fluctuations in Yt
(5.39) lnYt = α ln s+ α lnY trendt−1 + (1− α)(A+ gt+ ln ℓ+ N + nt)
+αYt−1 + (1− α)At
= lnY trendt + Yt ,
where
lnY trend0 = α lnK0 + (1− α)(A+ ln ℓ+ N) (1)
Y0 = (1− α)A0 = (1− α)εA,0 (2)(5.42) Yt = (α+ ρA)Yt−1 − αρAYt−2 + (1− α)εA,t
Numerical examples: α = 23 , ρA = 1
3 , A = 1, g = 0.03, b = 12 , n =
0.01, ρ = 0.1, N = 10,K0 = 10, V ar(εA,t) = 0.05.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 7 / 19
![Page 8: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/8.jpg)
Section 5.5 Interpreting Fluctuations in Yt
Interpreting Fluctuations in Yt (Continued)
0 20 40 60 80 100
6.6
6.8
7.0
7.2
7.4
7.6
7.8
Quarter
log
valu
e
lnY(t)lnYtrend(t)
0 20 40 60 80 100
−10
−5
05
Quarter
Percentage
Y~(t)
Figure 1: Numerical example with α = 23
, ρA = 13
, A = 1, g = 0.03, b = 12
, n = 0.01,ρ = 0.1, N = 10, K0 = 10, V ar(εA,t) = 0.05.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 8 / 19
![Page 9: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/9.jpg)
Section 5.5 The oversimplified model, Pros and Cons
The oversimplified model, Pros and Cons
ProsAnalytically solvable.Delivers the idea of Real Fluctuations under parsimonioussettings.
ConsUnrealistic predictions. Recall that s(t) ≡ s, so consumptions andinvestments fluctuate at the same rate as the output does. Table5.2...
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 9 / 19
![Page 10: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/10.jpg)
Equation (5.54), an example.
An Example
α = 23 , g = 0.02, n = 0.0025, δ = 0.025, ρA = 0.95, ρG = 0.95, r∗ =
0.015, ℓ∗ = 13 .
aLA = 0.35, aLK = −0.31, aCA = 0.38, aCK = 0.59, bKA = 0.08, bKK =0.95, aCG = −0.13, aLG = 0.15, bKG = −0.004.For China, set bKG = 0.4 ∗ 0.95 + 0.6 ∗ (−0.004). (Why?)
Yt = lnYt − lnY trendt
= lnYt − lnYt−1 + lnYt−1 − lnY ∗
= lnYt − lnYt−1 + lnY ∗t−1 + Yt−1 − lnY ∗
= ln YtYt−1
− ln Y ∗t
Y ∗t−1
+ Yt−1
≃ gY (t)− gfundamentalY (t) + Yt−1 (3)
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 10 / 19
![Page 11: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/11.jpg)
Equation (5.54), an example.
Figure 2
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 11 / 19
![Page 12: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/12.jpg)
Equation (5.54), an example.
0 10 20 30 40
68
1012
1416
Number of Quarter starting from 2005:Q2
Per
cent
age
Y~
(t)
Y~1
(t)Y~2
(t)Y~3
(t)
Figure 3: Period: 2nd quarter of 2005 - 1st quarter of 2015. Y is the real outcome, Y 1
neglects the 2008 financial crisis and 2009 bailout plans, while Y 2 considers the former andY 3 further considers the latter.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 12 / 19
![Page 13: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/13.jpg)
Equation (5.54), an example.
This is, of course, an informal and incorrect study because I am fittingdata by treating randomly generated paths of stochastic processes,rather than observed ones, as the underlying realized shocks, besidesthe problem that all parameters are not calibrated to China economy.The codes and figure, however, can serve as a simple illustratingexample.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 13 / 19
![Page 14: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/14.jpg)
Hodrick-Prescott Filter
Hodrick-Prescott Filter
Let {yt}Tt=1 be the logarithms of a time series variable, which is madedup of a trend component {y∗t }Tt=1 and a deviating component {yt}Tt=1.Given a positive value λ, there is a trend component that solves
min{y∗t }Tt=1
(T∑t=1
(yt − y∗)2 + λ
T−1∑t=2
[(y∗t+1 − y∗t )− (y∗t − y∗t−1)]2
)(4)
The first term penalizes tye deviations while the second termpenalizes variations in growth rate (a trend should not beunsmooth).It is often suggested that λ = 1600 for quarterly data,λ = 1600
44= 6.25 for annual data, and λ = 1600× 34 = 129600 for
monthly data.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 14 / 19
![Page 15: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/15.jpg)
Hodrick-Prescott Filter
Revisit the previous numerial example
0 20 40 60 80 100
6.6
6.8
7.0
7.2
7.4
7.6
7.8
Quarter
log
valu
e
lnY(t)lnYtrend(t)
0 20 40 60 80 100
6.6
6.8
7.0
7.2
7.4
7.6
7.8
Quarter
log
valu
e
lnY(t)HP−filtered lnYtrend(t)
Figure 4: Real trend on left, Hodrick-Prescott filtered trend on right.Numerical example with α = 2
3, ρA = 1
3, A = 1, g = 0.03, b = 1
2, n = 0.01, ρ = 0.1, N = 10,
K0 = 10, V ar(εA,t) = 0.05.Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 15 / 19
![Page 16: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/16.jpg)
Hodrick-Prescott Filter
Revisit the previous numerial example (continued)
0 20 40 60 80 100
−10
−5
05
Quarter
Per
cent
age
Y~(t)HP−filtered Y
~(t)
Figure 5: Real cycles versus Hodrick-Prescott filtered cycles.
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 16 / 19
![Page 17: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/17.jpg)
Replicating Figures 5.2, 5.3, and 5.4
Replicating Figures 5.2, 5.3, and 5.4Practice with the codes!
10 20 30 40
−0.2
0.0
0.2
0.4
0.6
0.8
1.0
Quarters
Percentage
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 401
A~(t)
K~(t)
L~(t)
Figure 6: Replicate Figure 5.2Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 17 / 19
![Page 18: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/18.jpg)
Replicating Figures 5.2, 5.3, and 5.4
Replicating Figures 5.2, 5.3, and 5.4
10 20 30 40
−0.2
0.0
0.2
0.4
0.6
0.8
1.0
Quarters
Percentage
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 401
Y~(t)
C~(t)
Figure 7: Replicate Figure 5.3
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 18 / 19
![Page 19: Supplementary Notes on Chapter 5 of D. Romer’s Advanced ... · Section 5.5 Interpreting Equation (5.31) Interpreting Equation (5.37) it is thus straightforward to understand equation](https://reader030.fdocuments.net/reader030/viewer/2022040306/5ebe5b6dac3d2376bf032371/html5/thumbnails/19.jpg)
Replicating Figures 5.2, 5.3, and 5.4
Replicating Figures 5.2, 5.3, and 5.4
10 20 30 40
−0.
20.
00.
20.
40.
60.
81.
0
Quarters
Per
cent
age
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 401
w~(t)annual r(t)−r*
Figure 8: Replicate Figure 5.4
Xu & Yi (HUST) Advanced Macroeconomics Supplementary Notes 19 / 19