Research on the Influencing Effect between CHVA and CPI in China Based on VAR Models

doi:10.4236/ajibm.2013.34044

Paper Menu >>

Journal Menu >>

American Journal of Industrial and Business Management, 2013, 3, 378-381

http://dx.doi.org/10.4236/ajibm.2013.34044 Published Online August 2013 (http://www.scirp.org/journal/ajibm)

Research on the Influencing Effect between CHVA and

CPI in China Based on VAR Models*

Jinge Zhou1, Juan Chen2, Xiuli Yu1,3, Yifan Li4, Qifeng Lin4

1Guangdong University of Technology School of Management, Guangzhou, China; 2Guangdong University of Technology School of

Economics and Trade, Guangzhou, China; 3Psychological Education and Research Department, Guangdong University of Technol-

ogy, Guangzhou, China; 4School of Applied Mathematics, Guangdong University of Technology, Guangzhou, China.

Email: [email protected]

Received April 25th, 2013; revised May 25th, 2013; accepted June 25th, 2013

which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT

The cointegration test, granger causality test, VAR model, impulse response function and other econometric methods

are used in this paper to analyze the influencing effect between commercial housing vacancy rate and CPI and its delay

impact. The results show that there is a long-term equilibrium relationship between commercial housing vacancy rate

and CPI in China. There are at least one cointegration relationship between CHVR and CPI. The past values of the CPI

appear to contain information which is useful for forecasting changes in the CHVR. CPI has a significant effect on

CHVR and CPI rising drives CHVR.

Keywords: CHVR; CPI; VAR; Cointegration

1. Introduction

The real estate industry is a leading industry of the entire

China economy and influences the quality of life of the

residents. The commercial housing vacancy rate fluctua-

tion entails real estate with commodity house prices. In

recent years, the price of China’s real estate has im-

proved extremely, and CPI also increased sharply. Hence,

some scholars believe that the commercial housing can’t

avoid the risk of inflation.

In the past few years, both domestic and overseas re-

searchers made an empirical analysis of the relationship

between the real estate and CPI. Raymond’s thesis

showed that there is no causal relationship between land

supply and housing prices and their estimations, which

based on the annual data of Hong Kong’s public land

sales, found that the government acts to maximise land

revenue [1]. Jack H. Rubens, Michael T. Bond and James

R. Webb believed that assets have the ability to protect

investors from the effects of inflation are generally la-

belled inflation hedges; the real estate has been regarded

as one of the best inflation hedges of past years [2]. Wil-

lian C. Wheaton holds the view that the vacancy rate,

fixed in the short run, determines the expected length of

sale and search, which plays a central role in the reserva-

tion prices of buyer and seller [3]. SHEN Yue and LIU

Hong-Yu researched the relationship between the real

estate development investment and GDP in China [4].

ZHOU Zhi-Chun, LI Zheng and MAO Jie researched

how real estate industry corresponds to economic vicis-

situdes, which is an empirical analysis based on Chinese

data [5]. WANG Yao-Wu, Jin Haiyuan considered the

real estate supply and land supply is the most important

factor. They also thought the second important factor is

the interest rate [6].

To conclude, both domestic and overseas reseachers

made an empirical analysis of the relationship between

real estate and CPI. However, related researching did not

underline the commercial housing vacancy rate, which is

the cornerstone of China real estate market. According to

the fact above, this paper made some research from the

perspective of effect size and effect time lag between

commercial housing vacancy rate and CPI on China real

estate market. The research is of positive value and po-

litical reference under the back ground of China’s current

economic system and real estate industrial integration.

2. Data Processing

All data series are annually begins in 1995 and ends in

2010, it shows by Table 1. Data on China’s CHVR

(which is commercial housing vacancy rate) and consumer

*This is an extended version of the paper at the 2011 International

Conference on Networks and Information.

Research on the Influencing Effect between CHVA and CPI in China Based on VAR Models 379

Table 1. 1995-2010, China’s CHVR and CPI’s annual data.

Years CHVR

The chain

growth of

CHVR

The chain

growth of CPI

1995 0.1209 0.0239 0.5450

1996 0.1397 0.0188 0.5170

1997 0.1654 0.0257 −0.0500

1998 0.1802 0.0148 −0.1330

1999 0.1960 0.0158 −0.1580

2000 0.1670 −0.0290 0.0250

2001 0.1540 −0.0130 −0.0330

2002 0.1400 −0.0140 −0.0330

2003 0.1400 0.0000 0.2670

2004 0.1179 −0.0221 0.1920

2005 0.1211 0.0032 0.1330

2006 0.1010 −0.0201 0.2330

2007 0.0820 −0.0190 0.5330

2008 0.0950 0.0130 0.1080

2009 0.1083 0.0133 0.1420

2010 0.1130 0.0047 0.3750

prices (CPI) are all from the China National Bureau of

Statistics (CNBS).

Because of the data acquisition is more difficult. We

use the annual data. This could have an impact on the

accuracy of the article.

3. Empirical Analysis

3.1. Unit Root Test

The ADF test was the first test developed for testing the

null hypothesis of root and was the most commonly used

test in practice [7].

This value is just under less than 5% critical value in

Table 2. CPI and CHVR are stationary time series. So

we can undertake next inspection. Because the serials are

same-order single integral serial. Hence, we can further

test the long-term equilibrium relationship between all

variables.

3.2. VAR Model’s Cointegration

Two time series with stochastic trends can move together

so closely over the long run that they appear to have the

same trend component, that is, they appear to have a

common trend, which are said to be cointegrated [7]. In

this section, we introduce a test for whether cointegration

is present.

CHVR and CPI share a common stochastic trend, be-

cause their prod under 0.05, it shows by Table 3. The

spread or the difference between the two rates does not

exhibit a trend. They appear to be cointegrated. Accord-

ing to the cointegration test results we can estimate that

there are at least one cointegration relationship between

CHVR and CPI.

3.3. Granger Causality Test

One useful application of the F-statistic in time series

forecasting is to test whether the lags of one of the in-

cluded regressors has useful predictive content, above

and beyond the other regressors in the model. The claim

that a variable has no predictive content corresponds to

the null hypothesis that the coefficients on all lags of that

variable are zero. This is called the Granger causality

statistic, and the associated test is called Granger causal-

ity test [8].

We consider the relationship between the CHVR and

CPI. Based on the OLS estimates (Table 4), the F-statis-

tic testing the null hypothesis that the coefficients on all

lags of the CPI is 3.83 (p < 0.1): we can conclude (at the

0.1 significance level) that the CPI Granger-causes

changes in the CHVR. It do means that the past values of

the CPI appear to contain information that is useful for

forecasting changes in the CHVR, beyond that contained

in the past values of the CHVR.

4. VAR Model Estimation

Vector autoregression (VAR) is a set of k times series

regressions, in which the regressors are lagged values of

all k series. A VAR model extends the univariate autore-

gression to a list, or “vector”, of time series variables.

The equation is called a VAR model.

Table 2. ADF test results.

value 5% level

critical value Conclusion

CPI −2.20 Stable

CHVR −2.67 Stable

Table 3. Johansen cointegration test results.

Hypothesized

No. of CE (s)Eigenvalue

Trace

Statistic 0.05 Critical

Value Prob.

None 0.50 15.77 12.32 0.01

At most 1 0.35 6.01 4.13 0.02

Table 4. Test results of granger causality.

Null HypothesisF-StatisticAssociated

Prob. Conclusion

KONG does not

Granger Cause

CPI 0.45 0.65

Agree null

hypothesis

CPI does not

Granger Cause

KONG 3.83 0.06

Refuse null

hypothesis

Research on the Influencing Effect between CHVA and CPI in China Based on VAR Models

380

In the case of two time series variables, Yt and Xt, the

VAR (p) consists of the two equations:

10 years, the impulse responses of CPI to a one standard

deviation shock in the CHVR equation. The main results

of this contractionary shock on the other variables in the

system can be summarized as follows:

10 11

1111 1

ttp

tptp tpt

YX X



 



 



 

(1)

CPI has a great influence on the CHVR. When the CPI

rises, the vacancy rates increase accordingly. The re-

sponse of CPI to CHVR is significant negative response

appears firstly—as the CHVR level rises gradually to

reach a trough two years after the initial shock. The de-

cline becomes significant after three years. Finally,

CHVR decreases steadily after the seven years to reach

its lowest level in the three years, and returns to its

pre-shock level ten years after CPI impulse.

2021 1

22112

tptp tpt

YX X



 



 



 

(2)

where the β’s and γ’s are unknown coefficients and μ1t

and μ2t are error terms.

The VAR assumptions are the time series regression

assumptions of Key Concept (1), applied to each equa-

tion. The coefficients of a VAR are estimated by esti-

mating each equation by OLS [9]. The response of the CHVR to CPI is smaller in mag-

nitude than the response of CPI to CHVR. More impor-

tantly, the shock dies out very quickly, five years after

the initial impulse so that there is no innovation paradox.

The system is estimated with annually data from 1995

to 2010, under the baseline system. In selected period,

we are constrained it by the availability of data for the

China National Bureau of Statistics. As suggested by the

relevant lag selection criteria (Akaike Information Crite-

rion, Schwartz Bayesian Criterion) we use two lags. The

CPI allows for a contemporaneous response of the

CHVR (In the Figure 1 the red lines mean error bars, the

blue one mean line of impact).

5. Summaries

The results show that there is a long-term equilibrium

relationship between CHVR and CPI in China. There are

at least one cointegration relationship between CHVR

and CPI. The CPI Granger-causes changes in the CHVR.

It is certain that the past values of the CPI appear to con-

tain information that is useful for forecasting changes in

The orthogonalized residuals of the CPI equation are

identified as CHVR. Figure 1 reports, over a period of

-.01

.0 0

.0 1

.0 2

12345678910

Response of KZL to KZ L

-.01

.00

.01

.02

12345678910

Response of KZL to CPI

-.3

-.2

-.1

12345678910

Response of CPI to KZL

-.3

-.2

-.1

12345678910

Response of CPI to CPI

Response to Cholesky On e S.D. Innovati ons ± 2 S.E.

0.02

0.01

0.00

−0.01

0.02

0.01

0.00

−0.01

0.3

0.2

0.1

0.0

−0.1

−0.2

−0.3

0.3

0.2

0.1

0.0

−0.1

−0.2

−0.3

Figure 1. Impulse responses to a CHVR shock-system with CPI (1995-2010).

Research on the Influencing Effect between CHVA and CPI in China Based on VAR Models 381

the CHVR, beyond that contained in the past values of

the CHVR, not vice versa. CPI has a great influence on

the CHVR. When the CPI rises, the vacancy rates in-

crease accordingly. The response of CPI to CHVR is

significant response appears for the first—as the CHVR

level rises gradually to reach a trough two years after the

initial shock. The decline becomes significant after three

years. Finally, CHVR declines steadily after the seven

years to reach its lowest level in the three years, and re-

turns to its pre-shock level ten years after CPI impulse.

REFERENCES

[1] Y. C. Tse Raymond, “Housing Price, Land Supply and

Revenue from Land Sales,” Urban Studies, Vol. 35, No. 8,

1988, pp. 1377-l392．

[2] J. H. Rubens, M. T. Bond and J. R. Webb, “The Infla-

tion-Hedging Effectiveness of Real Estate,” Journal of

Real Estate Research, Vol. 2, 2009, pp. 45-55.

[3] W. C. Wheaton, “Vacancy, Search, and Prices in a Hous-

ing Market Matching Model,” The Journal of Political

Economy, Vol. 98, No. 6, 1990, pp. 1270-1280.

doi:10.1086/261734

[4] Y. Shen and H.-Y. Liu, “The Relationship between Real

Estate Development Investment and GDP in China,”

Journal of Tsinghua University (Science and Technology),

Vol. 44, 2004, pp. 1205-1208.

[5] Z.-C. Zhou, Z. Li and J. Mao, “Researched How Real

Estate Industry Corresponds to Economic Vicissitudes?

An Empirical Analysis based on Chinese Data,” Journal

of Nanting University, Vol. 47, 2010, pp. 47-57.

[6] Y.-W. Wang and H. Y. Jin, “An Empirical Analysis of

the Impart of Real Estate Macro-Control Policy,” China

Civil Engineering Journal, Vol. 41, 2008, pp. 105-114.

[7] J. H. Stock and M. W. Watson, “Introduction to Econo-

metric,” Shanghai University of Finance Economics Press,

Shanghai, 2004, pp. 448-463

[8] C. W. J. Granger, “Investigating Causal Relations by

Econometric Models and Cross-Spectralmethods,” Eco-

nometrica, Vol. 37, No. 3, 1969, pp. 424-438.

doi:10.2307/1912791

[9] H. H. Pesaran and Y. Shin, “Gneralized Impulse Re-

sponse Analysis in Linear Multivariate Models,” Eco-

nomics Letters, Vol. 58, No. 1, 1998, pp. 17-29.

doi:10.1016/S0165-1765(97)00214-0