1140672

advertisement
Empirical Test of Asymmetric Wealth Effect of China Stock Market
De-cai Zhou, Hai-dong Xie, Zheng-feng Du
School of Economics & Management, Nanchang University, Nanchang, China. P. R.
(decaizhou@163.com)
Abstract - Basing on monthly data from January 2003 to
December 2011, Use Multivariate dynamic Markov regime
switching model to build the China stock market business
cycle index, then use cointegration and error correction
model to test its consumptions wealth effect. The test results
show that China's stock market not only has a significant
short-term wealth effects, but also has significant long-term
wealth effect. However the coefficient is negative, which is
actually a crowding out effect. Using of the autoregressive
dynamic distributed lag model, do further empirical analysis
and find the asymmetric characteristics of stock market
wealth effect in China. The main reason is maybe the
substitution effect of investment to consumption larger than
those of income effect in China, which may be related to the
immature China stock market, investors and instable
revenue expectation.
Keywords – China stock mark, asymmetric wealth effect,
business cycle index, empirical test
I.
LITERATURE REVIEW
Ando, Albert and Modigliani (1963) have first
proved that stock market has significant wealth effect in
the analytical framework of permanent income hypothesis
and life-cycle approach [1]. Since then, the overseas and
domestic research on stock market wealth effect has
focused on using different models and data to test
empirically the size of wealth effect.
The foreign research of stock market wealth effect
can be roughly divided into two overlapping stages: The
first stage is testing empirically wealth effect of house
-hold’ assets, such as stocks, bonds and real estate, such
as Hall (1978), Campbell (2000), Parker (1999), Boone,
Girouard and Wanner (2001) [2] [3] [4] [5]. The second stage
is empirically analyzing the asymmetry of the household’
assets wealth effect, which has sprung up in recent years.
Patterson (1993) and Shea (1995) discussed the
asymmetrical influence of wealth shock on consumption [6]
[7]
. Kuo and Chung (2002) connected the asymmetric
sensitivity of consumption with economic cycle [8]. Cook
(2002) and Stevans (2004) do empirical analysis to verify
the asymmetric effects on consumption [9] [10]. Poterba
(2000) thought that the stock market wealth effect is
potentially asymmetrical [11]. Zndi (1999) tested and
found that the stock market wealth increasing by $1 may
make consumption increase by about 4 cents, but $1
shrink may lead to a consumption reduction of 7 cents [12].
However, Nicholas and Stephen’ conclusion is completely
opposite to Zndi’s (1999) [13].
____________________
Subsidized by China’s national social science Fund (NO.10CJL025)
Empirical test of conclusions on the China stock
market wealth effect can be mainly divided into two
categories: The first is that the China stock market wealth
effect does not exist in the consumption sense, such as
Xin-ping Xia, Yi-xia Wang and Ming-gui Yu (2003),
Shao-xiang Tang, Yu-cheng Cai and Liang-qiu Xie (2008),
Ren-he Liu, Ying-na Huang and Ai-ming Zheng (2008) [14]
[15] [16]
.The second is that China stock market wealth effect
is existence, though less obvious, and relatively weak,
such as Zhen-ming Li (2001), Li Gao and Wei-dong Fan
(2001), Xue-feng Li and Hui Xu (2003), Zha-yan Luo
(2004), Hong Chen (2007), Chi-cheng Luo and Jian-jiang
Liu (2008) [17] [18[19] [20] [21] [22].
The researches results above are the basis for our
paper. However, the study of the China stock market
wealth effect is only concerned about the existence or the
size of the wealth effect, ignoring the inherent
characteristics of the wealth effect, such as asymmetry.
II. THE WEALTH EFFECT OF STOCK MARKET
ON HOUSEHOLD CONSUMPTION: EMPIRICAL
TEST BASED ON COINTEGRATION AND ERROR
CORRECTION MODEL
We study the wealth effect of stock market on
household consumption mainly using the variables of the
stock market index, household income and consumption.
The stock index is expressed by building the business
cycle index of China stock market. Different from
Standard practice that select only a single Shanghai Stock
Exchange Composite Index (SHSECI), we have chosen
three stock indexes: the SHSECI, Shenzhen Stock
Exchange Composite Index (SZSECI) and Hong Kong's
Hang Seng China Enterprises Index (HKCEI) as the
observed variables to build a business cycle index of
China stock market. We have introduced SZSECI in order
to contain the impact of the growing SME board of China
in recent years on the consumption; taking into the low
degree of internationalization and marketization of China
stock market, we have also put in HKCEI in order to
conclude the impact of international stock markets on
China's consumption. We select monthly urban household
per capita disposable income and consumption on behalf
of household income and consumption. There is because
urban household are the main investors of China stock
market, and rural household are less involved in it.
A. Construct business cycle index of China stock
market based on the Multivariate dynamic Markov
regime switching model
we make LS lagged for 6 orders.
8
6
4
2
0
2003M01
2003M06
2003M11
2004M04
2004M09
2005M02
2005M07
2005M12
2006M05
2006M10
2007M03
2007M08
2008M01
2008M06
2008M11
2009M04
2009M09
2010M02
2010M07
2010M12
2011M05
2011M10
Due to limited space, we don’t describe the
Multivariate dynamic Markov regime switching model
and the scholars who are interested in it can read related
works of Kim and Nelson(1998) [23]. Select monthly
arithmetic average of the daily closing price of three stock
indexes which are SHSECI, SZSECI and HKCEI from
January 1999 to December 2011 as the observed variables
to build Business cycle index of China stock market(S).
The reference index is CITIC A shares Composite Stock
Index (Z) which is transformed into the index which the
mean is 100. The results show in Figure 1, data from the
SINA net and CITIC index net. It can be seen from Figure
1 that the business cycle index of China stock market is
consistent with the reference index CITIC Composite A
Shares Stock Index.
LC
LY
LS
Fig. 2. Times series of LC, LY, and LS
2) Tests on data’s stability. Use unit root test to
examine the stabilities of the series. Reach the outcome at
the tableⅠ. The tableⅠshows that three variables LC, LY
and LS have unit roots under ADF test and PP test and
they are unstable time series. However their first
difference of DLC, DLY and DLS all reject the null
hypothesis on the 1% significance level, namely, they are
stable time series.
1999M02
1999M10
2000M06
2001M02
2001M10
2002M06
2003M02
2003M10
2004M06
2005M02
2005M10
2006M06
2007M02
2007M10
2008M06
2009M02
2009M10
2010M06
2011M02
2011M10
300
250
200
150
100
50
0
S
Z
Fig. 1. Business cycle index(S) and the reference (Z)
B. Empirical test to wealth effects of China stock
market on consumption based on cointegration and error
correction model
1) The data’s selection and disposal. We selected
urban per capita consumption as dependent variable and
business cycle index of China stock market and urban per
capita disposable income as independent variables. They
are all monthly data, which originate from CEInet and
range from January 2003 to December 2011. Amongst,
the data of the consumptions and the income after 2007
are calculated from the Quarterly data with per capita
social retail goods as weights. In order to eliminate the
effect of inflation, all data are adjusted to the actual value
by setting CPI’s base period at January 2003. Considering
household consumption and their disposable income turn
out significant seasonal properties, we adjusted them by
the method of X12.
transforming data into logarithm form does not
change the original cointegration relationships and the
linearize the trend, eliminate heteroscedasticity among
time series, we take natural logarithm to them, marking
household consump -tion, disposable income and stock
market business cycle index as LC, LY and LS. The final
outcome of the disposal is displayed in Fig 2, which
indicates that the lag 4 to 8 of business cycle index of
China stock market take effects on household
consumption and the effects are asymmetrical. Therefore,
TABLE I
Unit root test
ADF test
PP test
variables
T-statistics
Prob
stability T-statistics
LC
-0.8184
0.8094
stable
DLC
-9.4011
0.0000
stable
LY
-0.1603
0.9388 unstable
DLY
-9.6525
0.0000
LS
-1.3695
0.5946 unstable
-1.3103
0.6228 unstable
DLS
-5.8209
0.0000
-5.8402
0.0000
stable
stable
-0.6897
Prob
0.8440 unstable
-123.7329 0.0001
-1.6781
stability
stable
0.4394 unstable
-125.8168 0.0001
stable
stable
TABLE II
Johansen cointegration test
Trace test
Max-eigenvalue test
Null hypothesis Eigenvalue
statistics Prob
Statistic
Prob
112.3502
0
None *
0.664045
121.9182
At least 1
0.065637
9.567978 0.6821 6.992752
0.6704
At least 2
0.024692
2.575226 0.6626 2.575226
0.6626
0
3) Cointegration test. The variable unit root test
results show that the LC, LY and LS are all one order
integrations. In order to avoid spurious regression, we
examine whether there are cointegration relationships
among three variables. The results of Johansen
cointegration relationship test based on multiple VAR
model are shown in table Ⅱ. Trace test results show that,
on the 5% significance level, we reject the hypothesis that
there is no cointegration equation and accept the
assumption that there exists 1 cointegration equation. This
indicates that there is one cointegration relationship
among the three variables. The result of Max-eigenvalue
test also shows that, on the 5% level, we reject the
assumption that there exists no cointegration equation, the
three variables have 1 cointegration equation. It can be
seen that this two kinds of test results show that LC, LY,
LS exist cointegration relations, namely that three there is
a long-term equilibrium relationship between them.
4) Granger causality test. Cointegration test results
show that there is a long-term stable equilibrium
relationship among LC, LY and LS ranging from January
2003 to December 2011. The following granger causality
tests further prove their short-term relationships; the
details are shown in Table Ⅲ. The table Ⅲ shows the
stock business cycle index is the granger reason of
household consumption and disposable income, and there
are no granger causalities between household
consumption and disposable income.
TABLE III
Granger Causality Tests
Null hypothesis
optimal lag
F-stat
P-value
DLY doesn’t Granger Cause DLC
3
0.52529
0.6659
DLC doesn’t Granger Cause DLY
3
0.70339
0.5523
DLS doesn’t Granger Cause DLC
3
3.72185
0.014
DLC doesn’t Granger Cause DLS
3
0.37106
0.7741
DLS doesn’t Granger Cause DLY
3
4.15309
0.0082
DLY doesn’t Granger Cause DLS
3
0.47803
0.6983
5) Estimation of cointegration and error correction
equations. Cointegration test and granger causality test
results show that there are equilibrium relationships
among LC, LY and LS both in the short-term and in the
long-term; therefore we use cointegration and error
correction model to study the stock market’s wealth effect
to the household consumption. According to lag order
tests, coefficients that lag 1 order are basically not
significant. This paper choose lags order 2 ~ 4 in error
correction model, and then use Eviews7.0 to estimate the
relationship among LC, LY LS, the results are as follows:
Cointegration equation:
𝐿𝐶𝑡
= 0.6979 + 0.8656𝐿𝑌𝑡 − 0.0165𝐿𝑆𝑡 +𝐸𝐶𝑀 𝑡
(14.2944∗ ) (85.5859∗ ) (−2.3638∗ )
(9)
Error correction equation:
𝐷𝐿𝐶𝑡 = −0.5009𝐸𝐶𝑀𝑡−1 − 0.0106𝐷𝐿𝐶𝑡−2 − 0.2165𝐷𝐿𝐶𝑡−3
(−0.0056)
(−2.5399∗ )
(−0.9692)
− 0.1348𝐷𝐿𝐶𝑡−4 +0.2018𝐷𝐿𝑌𝑡−2 + 0.3792𝐷𝐿𝑌𝑡−3 + 0.2463𝐷𝐿𝑌𝑡−4
(−0.8011)
(1.2464)
(1.8976∗∗ )
(1.9296∗∗ )
+ 0.2206𝐷𝐿𝑆𝑡−2 − 0.3631𝐷𝐿𝑆𝑡−3 + 0.1579𝐷𝐿𝑆𝑡−4
(2.3784∗ )
(−3.5757∗ )
(1.7071∗∗ )
(10)
Amongst the numbers in brackets are t-statistics,
those with * and ** represent significant on 5% and 10%
levels respectively.
The results indicate that: (1) as it is shown in
equation (9), household consumption (LC) and disposable
income (LY), stock market index (LS) have a long-term
equilibrium relationship. The marginal propensity of
consume (MPC) of current period disposable income (LY)
is 0.8656, the marginal propensity of consume (MPC) of
stock index (LS) is -0.0165, and t-statistics are significant,
indicating that the stock market have dramatic long-term
negative wealth effect to household consumption, namely
the crowding out effect, which is opposite to developed
countries’ basically positive wealth effect. In the long run,
stock market’s excessive boom may take up great deals of
consumption funds, posing a crowding out effect on
consumption. This may be due to the fact that only about
20 years have passed since the establishment of China
stock market. So the market is not mature, has too much
speculation and volatility, it is hard for consumption to
form stable long-term earnings expectation.
(2) It can be seen from the error correction equation
(10) that there exist short-term dynamic relations among
changes in household consumption (DLC) and in
household disposable income (DLY), in the stock market
(DLS). First of all, the coefficients of 3 ~ 4 order lagged
disposable income are significant at 7% level, indicating
that China household consumes according to permanent
income; Second, all the coefficients of changes in stock
market index (DLS) are significant at 5% and 10% levels
respectively, indicating that the stock market has
significant short-term wealth effect to household
consumption, but the effect is not stable. This may have
something to do with the China stock market’s drastic
volatility and fluctuation.
6) Analysis of generalized impulse response function.
Generalized impulse response function is used to measure
the effect of random perturbation terms from a standard
deviation shocks on endogenous variable’s current and
future values. From the figure 3, it is known that the
household consumption has a strong positive reaction to
one of its own standard deviation in period 1, about
0.0395, then it increase gradually, and stabilize at about
0.10 after the period 23; it has a strong positive reaction
to one of household disposable income standard deviation
in period 1, about 0.0351, then it increase gradually, and
stabilize at 0.098 after the period 23; it also has a strong
positive reaction to one of household disposable income
standard deviation in period 1, about 0.0015, then it f
fluctuate gradually, and stabilize at 0.008 after the period
12. This shows that there have short and long terms
wealth effects of stock index.
0.12
0.1
0.08
0.06
0.04
0.02
0
-0.02 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
LC
LY
LS
Fig. 3. Response of LC to Genalized one S.D innovations
IV. ASYMMETRICAL WEALTH EFFECTS OF
STOCK MARKET: BASED ON AUTOREGRESSIVE
DISTRIBUTED LAG MODEL
In order to study stock market’s asymmetrical wealth
effect on household consumption, referring to Nicholas
Apergis and Stephen M. Miller(2004), this paper adopt
the error correction model as follows[13]:
𝐽
𝐷𝐿𝐶𝑡 = 𝐸𝐶𝑀𝑡−1 + ∑𝐼𝑖=1 𝛼1𝑖 ∗ 𝐷𝐿𝐶𝑡−𝑖 + ∑𝑗=0 𝛼2𝑗 ∗
𝐷𝐿𝑌 + ∑𝐾 (𝛼 ∗ 𝐷𝐿𝑆 + + 𝛼 ∗ 𝐷𝐿𝑆 − ) +
𝑡−𝑗
𝑘=0
3𝑘
𝑡−𝑘
4𝑘
𝑡−𝑘
∑𝑀
(11)
𝑚=0 𝜃𝑚 𝜇𝑡−𝑚
Amongst 𝐷𝐿𝑆 + and 𝐷𝐿𝑆 − represent positive
impact and negative impact of changes in stock market
respectively. In order to eliminate multi col-linearity and
self-correlation caused by multi-orders lags, this paper
estimate the above model by using auto regression
dynamic distribution model. Amongst error correction
term is the one estimated from cointegration lag 0-2.
𝐷𝐿𝐶𝑡 =−0.6901𝐸𝐶𝑀𝑡−1 − 0.5068𝐷𝐿𝐶𝑡−1 − 0.1056𝐷𝐿𝐶𝑡−2
(−7.500∗ )
(−5.6788∗ )
(−1.2512)
+ 0.6561𝐷𝐿𝑌𝑡 + 0.3001𝐷𝐿𝑌𝑡−1 +0.0823𝐷𝐿𝑌𝑡−2
(4.4966∗ )
(1.3329)
(13.5913∗ )
+
+
+
− 0.1318𝐷𝐿𝑆𝑡 + 0.0136𝐷𝐿𝑆𝑡−1
+ 0.1523𝐷𝐿𝑆𝑡−2
∗
(−2.3400)
(0.8036)
(2.7276 )
−
−
+ 0.1448𝐷𝐿𝑆𝑡− − 0.00269𝐷𝐿𝑆𝑡−1
− 0.1852𝐷𝐿𝑆𝑡−2
∗
∗
(2.1228 )
(−1.2035)
(−2.6051 )
+ 𝜇𝑡 − 0.3450𝜇𝑡−2
(12)
(−2.9991∗ )
Where the bracketed figures are t-statistics, those
with * represent significant at 5% significance level.
The results of model indicate that: 1) the
development of stock market has both positive and
negative wealth effect on household consumption in
short-term, but negative in long-term. It is asymmetrical.
In the short-term, stock market’s positive impact on
household consumption is negative and significant at the
5% level, The effects of the positive impact of the stock
market index of the lag 1 and 2 on household
consumption is positive. But only the coefficient of lag 2
is significant at the 5% level. All positive effects are
0.0341.Spot stock market index’s negative impact on
household consumption is positive and significant at the 5%
level. The effects of the negative impact of the lag 1 and
2‘s stock market index on household consumption is
negative. But only the coefficient of lag 2 is significant at
the 5% level. All negative effects are -0.0673.The
negative effects are 0.9719 times higher than the positive
effects. These figures illustrate that the household
consumption is more sensitive about the stock market’s
bad news than good news. It is obviously asymmetrical.
All the positive and negative effects are -0.0332. It turns
out that in the long run wealth effect of stock market on
household consumption is negative which actually a kind
of crowding out effect. In other words, as the stock
market develop, the household do not use stock returns to
consume, but to increase investment in the stock market.
It is incompatible with the situation that the wealth effects
of stock markets are mostly positive in developed
countries. This may be due to several reasons: Firstly, the
development level of China's stock market is relatively
low. The stock market is easy to soaring prices, then
result that the stock market investment become high-risk
ones and its income is uncertain and cannot form a stable
income expectations. Secondly, it is poor maturity of the
China stock investors, who are chasing shadie seriously.
It makes the investment substitution effect is greater than
the income effect. Thirdly, China’s stock market
environment is not ideal and it has the phenomenon of the
short bull market and long bear market. The bull markets
is generally one year and bear market for 3-5 years, and it
also appeared the famous 28 phenomenon, that is, 20% of
investors make money, 80% of investors lose money. So
it naturally cannot be better to boost the consumption.
2) The coefficient of the error correction term is
-0.6901 and significant in the 5% level, indicating that
short - term has a large reverse correction to long-term.
(3) Whether spot or lag 1 and 2 , household disposable
income impact on household consumption is positive and
significant at the 5% level , but the extent of this impact is
gradually decreasing , indicating that the household is
based on permanent income consumption.
V.
CONCLUSION
Based on the latest monthly time series data, we used
the cointegration and error correction model and dynamic
distributed lag model to study China stock market’s
wealth effect. The cointegration and error correction
model indicate that the wealth effect on household
consumption is significant both on short-term and
long-term, but the value is mostly negative which is
incommensurate that the wealth effect is mostly positive
in developed countries. Further, taking advantage of the
studies using the dynamic distributed lag model, China
stock market shows both short-term and long-term wealth
effect on household consumption, and the overall effect is
negative. That confirms the conclusion of our
cointegration and error correction model once again. We
also find that the wealth effect of China stock market has
an asymmetric characteristic. The main conclusions are as
follows:
1) The conclusions of cointegration and error
correction model. The cointegration equation shows that
the stock market index (LS) has negative impact on
household consumption (LC) in the long term, and the
coefficients are significant at the 5% level. Reasons that
make the investments in China’s equity market unable to
play a role of blood transfusion to household
consumption, and impede the transition from China stock
market to consumptions are the companies in it abusing
the power of financing, as well as the market environment
of bull short and bear long and 28 phenomenon.
Disposable income (LY) for long-term impact on
household consumption (LC) is positive and is significant
at the 5% level, indicating that in the long term, China
household pay with permanent income in consumption.
From the error correction model, the changes of stock
market index (DLS) and disposable income (DLY) have a
significant effect on one of household Consumption
(DLC). But the impacts are both positive and negative.
Combined with the analysis equation decomposition and
impulse response, we find that the impact of changes of
stock market index (DLS) and household' disposable
income (DLY) to one of household consumption (DLC) is
increasing at the beginning and then tend to stability
gradually in the short term (about 1-2 year). Those mean
that the wealth effect of stock index (LS) and disposable
income (LY) to household consumption exists and keeps
forever.
2) The conclusions of dynamic distributed lag model.
Firstly, we have confirmed two conclusions of
cointegration and error correction model: one is that the
change of China stock market index (DLS) has a
significant wealth effect on one of household
consumption (DLC), and the effect are both positive and
negative, but is negative in general, that is -0.0176; the
other is that the changes of household disposable income
(DLY) has a significant effect on one of household
consumption (DLC) and is gradually decreased with
increasing order lag, which shows that China household
pay with permanent income in consumption. Secondly,
the change of China stock market index (DLS) has
significant and asymmetric wealth effect on the changes
household consumption (DLC) .The negative impact of
the change of stock market index (DLS) is 1.9720 times
of the positive one. So the difference is very large.
China stock market wealth effect is mainly manifested as
crowding out effect, which does harm to the long-term
development of China stock market. Therefore, we need
to improve the system of stock market, especially making
the flow from investment to consumption channels
smoothly, say, the establishment of bonus system.
REFERENCES
[1] Ando, Albert and Franco Modigliani, “The life cycle
hypothesis of saving: aggregate implication and tests
(Periodical style),” American Economic Review, 53, pp.
55–84, March 1963.
[2] Robert E. Hall, “Stochastic implications of the life
cycle-permanent income hypothesis: theory and evidence
(Periodical style),” The Journal of Political Economy,
Volume 86, pp. 971- 987, June 1978.
[3] Campbell, John Y, “Asset pricing at the millennium
(Periodical style),” NBER working paper, No. w7589,
March 2000.
[4] Parker, J. A, “Spendthrift in America: on two decades of
decline in the U.S. personal saving rate (Periodical style),”
University of Wisconsin, Working Paper, 1999.
[5] Boone L., N. Girouard and I. Wanner, “Financial market
liberalization, wealth and consumption (Periodical style),”
OECD Economics Department Working Paper, No. 308,
2001.
[6] Patterson, K. D., “The impact of credit constraints, interest
rates and housing equity withdrawal on the Intertemporal
pattern of consumption-a diagrammatic analysis (Periodical
style),” Scottish Journal of Political Economy, Volume 40,
PP. 391-407, 1993.
[7] Shea, J. Myopia, “Liquidity constraints, and an aggregate
consumption: a simple test (Periodical style),” Journal of
Money, Credit, and Banking, Volume 27, PP. 798-805,
1995.
[8] Kuo, B. S. and Chung, C. T, “Is consumption liquidity
constrained? The asymmetric impact from business cycles
(Periodical style),” Academia Economic Papers, Volume
30, PP. 443-472, 2002.
[9] Cook, S, “Asymmetric mean reversion in consumption:
evidence from OECD Economies (Periodical style),”
Applied Econometrics and International Development,
Volume 2, PP. 27-34, 2007.
[10] Stevans, L. K, “Aggregate consumption spending, the stock
market and asymmetric error correction (Periodical style),”
Quantitative Finance, Volume 4, PP. 191-198, 2004.
[11] Poterba, J. M, “Stock market wealth and consumption
(Periodical style),” Journal of Economic Perspectives,
Volume 14, PP. 99-118, 2000.
[12] Mark R. Zandi, “Wealth worries (Periodical style),”
Regional Financial Review, Volume 8, PP.1-8, August
1999.
[13] Nicholas Apergis, Stephen M. Miller, “Consumption
asymmetry and the stock market: new evidence through a
threshold adjustment model (Periodical style),” University
of Macedonia, Working Papers, 2005.
[14] Xin-ping Xia, Yi-xia Wang, Ming-gui Yu, “Empirical
research on the wealth effect of stock market in Chinese
(Periodical style),” (in Chinese), Science & Technology
Progress and Policy, Volume 5, PP. 126-128, May 2003.
[15] Shao-xiang Tang, Yu-cheng Cai, Liang-qiu Jie, “Chinese
stock market wealth: an empirical study based on the
dynamic distributed lag model and the state space
model(Periodical style),” (in Chinese), The Journal of
Quantitative & Technical Economics, Volume 6, PP. 79-89,
June 2008.
[16] Ren Liu, Ying-na Huang, Ai-ming Zheng, “The empirical
test of stock market wealth effect in Chinese (Periodical
style),” (in Chinese), On Economic Problems, Volume 8,
PP. 98-101, August 2008.
[17] Zhen-ming Li, “Empirical analysis of stock market wealth
effect in Chinese (Periodical style),” (in Chinese),
Economic Science, Volume 3, PP. 58-61, March 2001.
[18] Ri Gao, Wei-dong Fan, “New challenges for Chinese stock
market and monetary policy (Periodical style),” (in
Chinese), Journal of Financial Research, Volume 12, PP.
29-42, December 2001.
[19] xue-feng Li, Hui Xu, “Study on the weak effect of stock
market wealth in Chinese (Periodical style),” (in Chinese),
Nankai economic studies, Volume 3, PP. 67-71, 2003.
[20] Zuo-yan Luo, “Empirical analysis of Chinese's stock
market wealth effect in recent years (Periodical style),” (in
Chinese), Contemporary Finance and Economics, Volume
7, PP. 10-13, July 2004.
[21] Chen Hong, Nong Tian, “Stock market wealth effect in
Chinese: theory and practice (Periodical style),” (in
Chinese), Journal of Guangdong University of Finance,
Volume 4, PP. 76-80, April 2007.
[22] Chi-cheng Luo, Jian-jiang Liu, “Empirical study on the
wealth effect of stock market (Periodical style),” (in
Chinese), Statistics and Decision, Volume 1, PP. 137-139,
January 2008.
[23] Chang-Jin Kim and Charles R. Nelson, “State-space
models
with
regime-switching:
classical
and
gibbs-sampling approaches with applications (Book style),”
Cambridge: MIT Press, 1998.
Download