Macroeconomic Analysis

Econ 6022

Lecture 1

Fall, 2021

1 / 96
 Course Information
• Instructor: Heng Chen (Henry)
- Email: hengchen@hku.hk
- Office: 915, KKL
- Office hour: 17:30 - 18:30 Wednesday (Online) and by
appointment
• Teaching Assistant: Ellen Tai and Xiaoting Mai
- Office hour: By email appointment
- Tutorial: TBA
• Exams and assignments
- 2-3 Assignments: 20%
- Mid-term: 20% (Sept 17)
- Final: 60%
• Course Website:
- Moodle course title: ECON6022 Macroeconomic Analysis

ECON6022_1AB_2021

2 / 96
 Prerequisites and Textbooks

Prerequisites
• Be comfortable with basic calculus and optimization
• Intro Macro
Textbooks
• Macroeconomics: Modern Approach (Robert Barro)
Readings
• Extra book chapters/notes
• Articles from magazines, newspapers or blog entries

3 / 96
 Topics and Methodology

Topics
• Economic growth
• Business cycles
• Money and prices
• Macroeconomic policies
Methodology
• Inter-temporal approach
• Dynamic models
• Overview of modern macroeconomics

4 / 96
 Preliminary Schedule
• Introduction and Growth Facts
• Economic Growth: Solow Model
• Consumption and Saving
• Business Cycles I: Facts
• Business Cycles II: A Framework
• Mid-Term Exam (Sept 17)
• Business Cycles III: Extensions
• Money and Prices I: Demand and Inflation
• Money and prices II: Non-Neutrality
• Money Supply and non-conventional policy

5 / 96
 Applications

• Growth prospect in China;

• QE in the U.S.;

• Monetary policy in China and the Great fall of China;

• The economic consequences of 4 Trillion stimulus plan;

• Saving puzzles in China and low savings rate in the U.S.

6 / 96
 Growth Facts

• World growth facts

- World growth experience
- Key facts and questions
• Kaldor’s facts
- Industrial economies
- Key facts and properties

7 / 96
8 / 96
 Measure income growth rates:
• We now consider in more detail how to measure
differences in the rate of income growth among countries.
• An (annual) growth rate of a variable X is defined as the
change of X from the first to the second year divided by
the value of X in the first year:
Xt+1 − Xt
g=
Xt
• The equation for the annual growth rate can also be
rewritten as follows:
Xt+1 = Xt × (1 + g)
• For two years we then have:
Xt+2 = Xt+1 × (1 + g)
= Xt × (1 + g) × (1 + g)
= Xt × (1 + g)2
9 / 96
• More generally, for n years:

Xt+n = Xt × (1 + g)n

• Thus, we can calculate a (geometric) average growth rate

over n years in the following way, by solving the more
general equation for g:
 1
Xt+n n
g= −1
Xt

10 / 96
Growth rates and changes in GDP over long periods:
• Even seemingly small differences in annual growth rates,
when sustained over long periods can give rise to
substantial changes in GDP levels.
• For instance, in 1950

GDPCH GDPCH
= 1.34, = 1.45
GDPUK GDPNOR
• The average annual growth rates 1950-2004 have been,
respectively,

gCH = 1.9%, gUK = 2.2%, gNOR = 2.9%.

• As a result, in 2004

GDPCH GDPCH
= 1.09, = 0.84
GDPUK GDPNOR
11 / 96
 Growth before 1820:

• Before 1820 data is even more difficult to find, but we know

some facts from historical sources:
1. growth in GDP pc was very slow
2. year-to-year fluctuations dominated
3. income differences were very small by modern standards
4. significant changes in the countries’ rankings occurred (for
example the relative economic decline of China in the 19th
century).

12 / 96
GDP per Capita by Country Group, 1820-1998

13 / 96
 Growth since 1820:

• For earlier periods the data is sketchier (misses for some

countries and / or some time periods).

• However, two facts are salient:

1. The pace of growth worldwide from 1820-1998 has
accelerated;
2. The gap between poor and rich has widened.

14 / 96
Stylized fact 1: large differences in living standards

• The fifth of the world that lives in rich countries receives

62% of the world income.
• The GDP per capita of Switzerland in 2006 is 37400 USD
(PPP-adjusted).
• The GDP per capita of poorest thirteen countries is below
1100 USD.

15 / 96
Stylized fact 2: very different growth experiences

• Growth miracles: e.g. Asian tigers.

• Growth disasters: e.g. Argentina
- In 1900, the GDP pc of Argentina was comparable to that of
France and Canada. Argentina was well ahead of Sweden,
Norway, Italy and Japan. It was more than three times as
rich as Korea.
- In 2006, the GDP pc of Argentina was 1/2 of France, less
than 2/3 of Korea, and lower than Botswana.
• Part of the developing world is falling behind: the average
African household consumed 25% less in 1998 than a
quarter of century earlier.
• Even industrial countries have grown at very different rates
over long span of time (Table 6.1)

16 / 96
Table 6.1 Economic Growth in Eight Major Countries,
1870-2006

17 / 96
 Two basic questions about growth

• What’s the relationship between the long-run standard of

living and the saving rate, population growth rate, and rate
of technical progress?
- Motivated by stylized fact 1.
- A question about the difference in income levels!
• How does economic growth change over time? Will it
speed up, slow down, or stabilize?
- Motivated by fact 2.
- A question about income growth rates.

18 / 96
 Kaldor’s Facts: Industrialized Countries

• Both output per worker (y ) and capital per worker k grow

over time.
K
• They also grow at similar rates, so does not change
Y
much over time.
• The return to capital, i.e., the interest that firms have to pay
if they rent capital, is almost constant over time. r = const
• The labor share and capital share are almost constant,
w ·L r ·K
= const, = const
Y Y
Additionally:
• The convergence of per capita GDP of different countries
and regions over time

19 / 96
 Recap - Stylized facts:

• Recent explosive growth: In the last two centuries the

world has seen an explosion of economic growth
unprecedented in history.
• Unequal growth: The growth of income has been uneven
though.
• Growing income gap between rich and poor: This unequal
growth has led to a vast widening income gap between rich
and poor countries.
• Kaldor’s Facts about industrial economies for the last 100
years.

20 / 96
 Overview of Growth Theories

• Explaining all the stylized facts is beyond the scope of the

course.
• Three production factors: K , H, and A.
• We focus on capital accumulation:
- Exogenous growth model (lecture 2)
• Solow model: Accumulation of physical capital, K .
- Endogenous growth models:
• Lucas model: Accumulation human capital, H.
• Romer model: Technological progress, A.
• We outline other possible sources.

21 / 96
 Plan

• Physical capital
• Production function
• Solow model
- Dynamics: model mechanisms
- Reality checks: taking the model to data
- Extensions
• Growth accounting

22 / 96
 Physical Capital

• We define as capital the tools needed for production, i.e.,

the physical objects that extend our ability or do work for
us.
• Workers using more capital can produce more output.
• Differences in capital per worker can be responsible for
differences in productivity and standards of living.
• Evidence: High correlation between capital pw and GDP
pw.

23 / 96
24 / 96
 Five features of capital:
1. It is productive: it raises the amount of output that a
worker can produce.
2. It is produced: Capital has itself been produced through
the process of investment (private or public). Investment
corresponds to saving, which means that the resources
used to produce capital could have been consumed
instead. They are invested today to yield a higher
consumption tomorrow.
3. It is rival in its use: only a limited number of people can
use a given piece of capital at one time (different from
ideas which can be used by an infinite number of people at
the same time).
4. It yields a return: Since it makes a worker more productive,
the worker or its firm will be willing to pay to use it.
5. It wears out: Using capital causes it to wear down a little,
which is called depreciation.
25 / 96
 The Production Function

• Production function: describes how quantity of inputs are

turned into quantity of production.
• For simplicity we first consider only two inputs: capital, K
and labor, N.
• Let Y denote the total output, then we have:

Y = AF (K , N)

• A measures total factor productivity (TFP)

• We introduce some assumptions on the production
function.

26 / 96
 Assumption 1: constant returns to scale.
• If we multiply the quantities of each input by some factor,
the quantity of output will increase by the same factor:
F (zK , zN) = zF (K , N)
• Can be transformed into a relationship between output per
worker and capital per worker:
   
1 1 K N K
Y = AF (K , N) = AF , = AF ,1
N N N N N
K Y
...or defining k = N and y = N,
rewrite
 
1 K
Y = AF ,1
N N
as
y = AF (k , 1) = Af (k )
• Output per worker is a function only of capital per worker.
27 / 96
Assumption 2: diminishing marginal product

• MP is the extra output produced when we increase the

usage of an input (for example capital) by a small amount.
• Mathematically, it is a partial derivative

∂F (K , N) ∂F (K /N, 1) df (k )
MPK = = =
∂K ∂K /N dk
• Diminishing MP of capital: if we keep adding units of
capital, then the quantity of new output added by each new
unit of input is smaller than that added by previous units of
input.

28 / 96
29 / 96
 Assumption 3: Competitive economy.
Factors of production (labor, capital) are paid their marginal
products.
• The value of the MP of a worker is the maximum the firm is
willing to pay each worker.
• In a competitive environment, firms cannot pay wages
below the value of the MP of labor.
• Similarly, the MPK equals to the equilibrium “rental rate” of
capital, that is the cost of renting one unit of capital for one
unit of time.
∂Y
w = MPN =
∂N
∂Y
R = MPK =
∂K

30 / 96
 Cobb-Douglas Production Function
• The most common functional form is the Cobb-Douglas:
F (K , N) = K α N 1−α
• α (which is assumed to lie between 0 and 1) denotes the
elasticity of output per worker to capital per worker.
• Note that
ln(Y ) = ln A + α ln(K ) + (1 − α) ln(N)
and
ln(Y /N) = ln(Y ) − ln(N) ≡ ln (y )
ln(K /N) = ln(K ) − ln(N) ≡ ln (k )
thus
ln (y ) = ln (A) + α ln (k )
• Therefore
∂ ln y ∂y k
α= =
∂ ln k ∂k y
31 / 96
• α denotes then how much percent y increases when k is
increased by 1%.
• α can be estimated from the data (under the maintained
hypothesis of a Cobb-Douglas functional form).
• Left to you as an exercise (good exam question): show that
the Cobb-Douglas production function above indeed
features constant returns to scale.
• Express it in per-worker terms:
   α  1−α
Y F (K , N) K N K N
y= =A = AF , =A = Ak α
N N N N N N

32 / 96
• Under Cobb-Douglas:

∂Y
MPK = = αAK α−1 N 1−α
∂K
• Note that the MPK is indeed diminishing:

∂MPK ∂2Y
= = (α − 1)αAK α−2 N 1−α < 0
∂K ∂K 2

33 / 96
 Income shares of capital and labor:

• The total amount paid out to capital will equal the rental
rate per unit of capital times the total amount of capital:
R × K.
• The capital’s share of income is the fraction of total income
that is paid out as rent of capital:

R×K αAK α N 1−α

Capital’s share of income ≡ = =α
Y AK α N 1−α
• A similar calculation shows that the labor share of income
is equal to 1 − α.

w ×N (1 − α) AK α N 1−α
Labor’s share of income ≡ = = 1−α
Y AK α N 1−α

34 / 96
Cobb-Douglas property: Constant income shares.

• This is an important result: even though the quantities of

capital and labor in the economy may vary, changes in the
rental rate of capital and wage rate will be such that the
shares of national income paid out to each factor of
production will be unaffected.
• The result holds true under the assumptions of (i)
Cobb-Douglas production function and (ii) competitive
markets.
• Under these assumptions, we can estimate α by looking at
capital’s share of national income.

35 / 96
36 / 96
Empirical facts about the capital share of income:

• α is one of the crucial pieces of data that economists

examine in studying economic growth.
• The average is 0.35.
• Nevertheless, economists standardly use a value of 1
3 to
approximate for the capital share of income.

37 / 96
 The Solow model:

• Does capital drive long-run growth?

• How does the Solow model help us understand differences
in world income levels?
• How does the Solow model help us understand different
growth experiences?

38 / 96
 Solow model
• Assumptions that help to simplify our discussion.
- Economy is closed and G = 0.

Yt = Ct + It . (6.4)

- Saving rate is constant St = s · Yt .

- Saving equates investment: St = It .
- Work force, Nt , grows at a constant rate n.
• The change in the capital stock is the difference between
the amount of (gross) investment, I, and the amount of
depreciation, D:

∆Kt = Kt+1 − Kt = It − Dt = St − d · Kt

• Key equation: capital accumulation

∆Kt = s · Yt − d · Kt
39 / 96
 Solow model

• Further assumption on production function: Cobb-Douglas

• The capital accumulation equation becomes:

∆Kt = s · A · Ktα · Nt1−α − d · Kt

• But we care more about the capital stock per worker and
income per worker. Income per worker is a measure of
standards of living.
• We want to rewrite everything in per-worker terms: for
example, yt = Yt /Nt ; ct = Ct /Nt ; kt = Kt /Nt
• kt is also called the capital-labor ratio
• This transformation needs some math manipulation.

40 / 96
 Transformation

• Divide both sides with Nt

∆Kt 1 Kt
= · s · A · Ktα · Nt1−α − d ·
Nt Nt Nt
• Using the def. of capital-labor ratio,

∆Kt
= s · A · ktα − d · kt
Nt
Kt
• Note that ∆kt = ∆( N ∆Kt
t
) 6= Nt , instead, we know,

∆kt ∆Kt ∆Nt

= −
kt Kt Nt

41 / 96
 Transformation
• Rearrange this equation, we have

∆Kt ∆Nt
∆kt = · kt − · kt
Kt Nt
• Using the def. of population growth rate and capital-labor
ratio,
∆Kt Kt
∆kt = · − n · kt
Kt Nt
• Now we find out that ∆Kt
Nt ,

∆Kt
= ∆kt + n · kt
Nt
• Therefore,

∆kt + n · kt = s · A · ktα − d · kt
42 / 96
 Solow model

• Finally, we arrive at

∆kt = s · A · ktα − (n + d) · kt

• Basically, the equation summarizes the Solow model.

• Let’s understand the equation:
- Actual investment and break-even investment
- break-even investment: the amount of investment that must
be done just to keep capital labor ratio at its existing level.
- To keep kt from falling:
- d · kt : maintain depreciate capital so that aggregate capital
stock K is constant, but the number of workers is growing ...
- n · kt : capital stock must grow to keep kt steady.

43 / 96
 Dynamics

• Here comes the dynamics of capital-labor ratio:

- If s · A · ktα > (n + d) · kt ⇒ ∆kt > 0 ⇒ kt+1 > kt
- If s · A · ktα < (n + d) · kt ⇒ ∆kt < 0 ⇒ kt+1 < kt
- If s · A · ktα = (n + d) · kt ⇒ ∆kt = 0 ⇒ kt+1 = kt

44 / 96
Figure 6.5 Determining the capital-labor ratio in the steady
state

45 / 96
 Solve for the steady state
• k ∗ is the steady state capital, such that ∆k ∗ = 0.
• Set ∆k ∗ = 0,

s · A · (k ∗ )α = (d + n) · k ∗

1
∗ s·A 1−α
k =( )
d +n
• Similarly, the steady state income

1 α
∗ ∗ α s
y = A · (k ) = A 1 − α · ( )1 − α
d +n
• High saving rate, s, leads to high k ∗ and y ∗ .

46 / 96
 Converging to the steady state

• If k begins at some level other than k ∗ , it will move toward

k∗
- For k below k ∗ , saving > the amount of investment needed
to keep k constant, so k rises
- For k above k ∗ , saving < the amount of investment needed
to keep k constant, so k falls
• Example: suppose kt < k ∗ ,
- how does the economy move towards the steady state?
- how do capital-labor ratio, output per worker evolve over
time?
- how about the growth rates of those variables?
- and how about aggregate variables?

47 / 96
Converging to the steady state

48 / 96
 Consumption at the Steady State (Tutorial)
• Consumption at the steady state is the difference between
output and investment at the steady state.
• Note that we are talking about consumption at the
steady state
• Trade-off:
- The higher k ∗ , the higher output, f (k ∗ )
- The higher k ∗ , the higher investment needed to keep k ∗
from falling
- c ∗ = (1 − s) · A · f (k ∗ )
- k ∗ is affected by saving rate, s, given A, n and d.
• Increasing k ∗ will increase c ∗ up to a point
- This is kG in the figure, the Golden Rule capital-labor ratio
- For k ∗ beyond this point, c ∗ will decline
- But we assume henceforth that k ∗ is less than kG , so c ∗
always rises as k ∗ rises.
- Why is it a reasonable assumption?
49 / 96
The relationship of consumption per worker to the capital-labor
ratio in the steady state

50 / 96
 Data I: Explain income differences

About differences in “level”

• Now we would like to explain income differences using the
Solow model.
• Consider the case where the only differences among
countries are in their saving rates, s.
• Thus we assume that countries have the same rates of
depreciation, d, population growth rate, n, and the same
productivity, A.
• Furthermore we assume (for now) that the countries are all
at their steady-state levels of income per worker.

51 / 96
• Consider two countries, which we denote by i and j.

 α/(1−α)
si
yiss = A1/(1−α)
n+d
 α/(1−α)
sj
yjss = A1/(1−α)
n+d
• Dividing the first equation by the second yields the ratio of
income per worker in Country i to income per worker in
Country j.
 α/(1−α)
yiss si
=
yjss sj

52 / 96
• For example, suppose that Country i has an investment
rate of 20% and Country j has an investment rate of 5%.

• Furthermore we assume α = 13 , so that 1−α

α
= 12 . For the
ratio of incomes per worker we then have:
1/2
yiss

0.20
= = 41/2 = 2
yjss 0.05
• A factor of four saving rate difference generates only an
income difference of factor 2.

53 / 96
• It smells bad! Example:

ss 1
yU.S.

sU.S. 2
ss =
ypoor spoor
ss
yU.S.
• ss = 32 and sU.S. ≈ 25% and implies ...
ypoor
• spoor ≈ 25%/322 ≈ 0.02%, if we assume the Solow model
could explain away all the income differences.

54 / 96
55 / 96
 Theory under-predicts actual differences:

• The figure takes the United States as the benchmark

(US=1).
• Horizontal axis: theory-predicted GDP per worker relative
to the US, given the data on investment rates.
• Vertical axis: actual data GDP per worker relative to the
US.
• High positive correlation.
• However, if the theory were "exactly right", all points would
be aligned on the 45-degree line.
• Predictions are highly imprecise. The theory underpredicts
the actual differences (almost all points are below the
45-degree line).

56 / 96
 Why are the predictions imperfect?

• Possible reasons:
1. Other important factors are left out (e.g., human capital,
technology).
2. Countries might not be in their steady state as assumed by
the model. Reasons:
2.1 shocks (e.g., wars),
2.2 changes in investment behavior over time.
• Capital accumulation accounts for only a small
fraction of cross-country income gaps.

57 / 96
Data II: Explain differences among growth rates

About differences in “growth”

Transition dynamics:
• In the Solow model, the farther away the economy is from
its steady state the faster it grows (shrinks).
• The growth rate of k , increases with the gap between the
current and steady state capital.
∆k k
• = s · A( ∗ )α−1 (k ∗ )α−1 − (n + d) = (n + d)[( kk∗ )α−1 − 1].
k k

58 / 96
 Evidence: OECD

Countries that were poorer in 1960 grew faster after 1960.

59 / 96
 Evidence: A broader set of countries

Countries that were poorer in 1960 did not grow faster!

60 / 96
 Convergence

• Absolute convergence:
• “Poor countries grow faster?”
• It seems not true.
• Actually, it is NOT what the Solow model says.
• evidence: no absolute convergence.

• Conditional convergence:
• grow faster, when it is further away from its own steady
state.
• different steady states, hold fixed country characteristics,
e.g. saving rate, technology level etc.
• evidence: some conditional convergence.

61 / 96
The Solow model and Kaldor’s Facts (Tutorial)

• Data: d = 7%, n = 1%, s = 25%, r = 4%

• Capital share (or Labor share) is constant,

MPK · K 1
=α=
Y 3
• Capital-output ratio,
k∗ s
∗
= ≈3
y n+d
• Return to capital,
α
R = r + d = MPK = ≈ 11%
k∗
( ∗)
y
• Long run growth

62 / 96
 Exogenous TFP Growth (Tutorial)

• In the Solow model, we see that capital accumulation

cannot drive long run growth.
• We can assume that the productivity, A, grows
exogenously:
∆A
=g
A
• The key equation becomes,
g
∆k̄ = s · k̄ α − (d + n + ) · k̄
1−α
−1
K
where k̄ = 1 = k · A 1−α
A 1−α N
• In this case, there is long run growth in steady state.

63 / 96
 Exogenous TFP Growth (Tutorial)

• Solow model cannot explain the long run growth.

• It is driven by TFP growth, which is not explained in this
model.
• How do we explain how the technology makes progress?
Endogenous growth theories.
• Now we want to understand how important the growth in
productivity is.
• Growth accounting.

64 / 96
 Sources of Growth

• Solow model cannot explain the long run growth.

• It is driven by TFP growth, which is not explained in this
model.
• What drives the growth in TFP?
• We want to quantify the importance of the growth in
productivity, human capital and others ...

65 / 96
 Growth Accounting

• It is just an accounting exercise.

• start with the production function, we have
∆Y ∆A ∆K ∆N
= +α· + (1 − α) ·
Y A K N
• The output growth is equal to a weighted average of capital
and labor growth plus the growth rate of A, which refers to
total factor productivity.
∆K ∆N
• It is easier to find out and
K N
∆A
• By construction, is the residual term, “the Solow
A
residual”

66 / 96
Hong Kong’s takeoff

67 / 96
 Understanding China’s Growth

(Zhu, 2012)

68 / 96
 Understanding China’s Growth

• Was China’s growth largely driven by investment?

• Was China’s growth largely driven by trade?

• Where did the large amount of foreign reserve originate?

• The long shadow of the 4-trillion stimulus plan.

69 / 96
Investment-driven growth in China?

70 / 96
 Understanding China’s Growth (Zhu, 2012)
Y = AK α (hL)(1−α)
Rearranging it leads to:

Y L K (α/(1−α)) (1/(1−α))
= ( ) hA
Pop Pop Y

Further,

71 / 96
72 / 96
 Trade-driven growth in China?

• Most productivity gains come from a reduction in internal

trade and migration costs, rather than a reduction in
external trade costs, contrary to the general narrative of
China’s export-oriented development story.
• Trade between its provinces has increased more than
trade between China and the rest of the world, and the flow
of workers across regions within China represents the
largest migration in human history.

73 / 96
 Trade-driven growth in China?

• In the early 2000s, China had substantial policy-induced

migration costs and internal trade costs. Since then, the
Chinese government has undertaken policy reforms and
infrastructure investments that reduced both migration and
trade costs and, at the same time, the Chinese economy
has experienced significant growth in aggregate
productivity.
• The internal trade and migration cost reductions, and the
associated increases in trade and migration within China,
account for 28 percent of China’s aggregate labor
productivity growth between 2000 and 2005. The reduction
in international trade costs, on the other hand, accounted
for only 8 percent of the growth.

74 / 96
Spatial distribution of income

75 / 96
Spatial distribution of income

76 / 96
 Immigration and interregional trade

• There was also a nation-wide administrative reform in 2003

that greatly streamlined the process for getting a
temporary residence permit in other provinces.
• Since 2000, trade barriers between regions have fallen
sharply.

77 / 96
Migration

78 / 96
Inter-regional trade

79 / 96
Inter-regional trade

80 / 96
 Quantify trade and immigration barriers

• The trade cost in 2002 was high, but it dropped sharply

between 2002 and 2007.
• On average, internal trade costs have fallen by 10-15
percent, non-agricultural international trade costs have
fallen by nearly 10 percent, and agricultural trade costs
have fallen by nearly 25 percent.
• For immigration costs, the average cost of moving from
rural to urban areas in a province in 2000 is equivalent to
reducing actual income to one-third; the cost of action
between provinces is an order of magnitude higher.
• However, between 2000 and 2005, immigration costs fell
by a significant 18 percent, while interprovincial mobility fell
by nearly 40 percent.

81 / 96
Quantify trade and immigration barriers

82 / 96
Quantify trade barriers

83 / 96
Potential benefits of continuing reform

84 / 96
 Growing like China (1992-2007)

• High economic growth

• High return to capital
• Employment reallocation between private and public
sectors
• Large amount of foreign reserves

85 / 96
Return to capital

86 / 96
Private employment

87 / 96
Foreign reserves

88 / 96
 Why is it a puzzle?

• Current account surplus: saving and investment;

• High investment rate and return to capital;

• Higher return and capital flow.

89 / 96
Distortions in capital market

90 / 96
Distortions in capital market

91 / 96
 Understanding the puzzle

• High productivity private firms cannot borrow from banks

and rely on its own savings to expand;
• Low productivity SOE can borrow to finance their
investment;
• High growth is a result of resource reallocation from SOE
to PE;
• When the public sector shrinks, its demand for fund from
banks shrinks.
• It results in a gap between saving and investment, which
has to manifest itself in current account surplus.

92 / 96
 The 4-trillion stimulus plan (2008-2019)

• Worsened the distortion toward heavy industries.

• TFP growth slows down while misallocation of resources

deepens.
• Short run stimulus (4%) v.s. long run consequences.

• The 4 trillion stimulus plan and asset markets.

93 / 96
The stimulus plan

94 / 96
The stimulus plan

95 / 96
96 / 96