Chapter 4 Empirical Results
4.3 Results of Vector Error Correction Model
4.3 Results of Vector Error Correction Model
According to the cointegration results in sector 4.2, we can conclude the long-run equilibrium relationship exists in T-REITs, stock price index, commercial rent and commercial price. To explore the short-run changes under cointegration condition, it would take advantage of VECM analysis. The most important thing of VECM is to understand the long-term equilibrium value adjusting speed between variables. This sector will illustrate the VECM results between T-REITs and stock price index, T-REITs and commercial rent as well as T-REITs and commercial price.
In Table 4-4, under the cointegration between T-REITs and stock price index when T-REITs is the dependent variable, the parameter at lag one of T-REITs and at lag one and three of stock price are significant, implying the current T-REITs are affected by lag one T-REITs and lag one and three stock price. The result also shows T-REIT and stock price are positive correlated, which supports the past conclusion that REITs and stocks are positive. We can also learn from the portfolio theory that when the investors benefit from the stock markets, they will transfer part of the funds to invest in other financial products in order to achieve the purpose of risk diversification. Also based on the theory of wealth effects, the real estate market will be the first choice when investors want to invest in other markets, and REITs can act as a proxy for real estate market. Therefore, the positive correlation also confirms the wealth effect existed between T-REITs and stock price.
Under the cointegration between T-REITs and commercial rent when T-REITs is the dependent variable, the parameter at lag one to three of T-REITs and at lag three of commercial rent are significant, suggesting the current T-REITs are affected by lag one to three T-REITs and lag three commercial rent. Because the commercial rents are on a quarterly basis, in this study three periods is about a quarter, implying the last quarter commercial rent changes impact the current T-REITs. This result shows commercial rent changes are positive to T-REITs, which confirms the positive correlation between REITs and rent in previous studies.
Under the cointegration between T-REITs and commercial price when T-REITs is the dependent variable, the parameter at lag one and three of T-REITs and at lag one and two of commercial price are significant, suggesting the current T-REITs are
‧
affected by lag one and three T-REITs and lag one and two commercial price. In addition, the impact of commercial price to T-REITs is negative, indicating that when the real estate market is at prosperity, the investors will choose to invest in the real property rather than the substitute financial products. Therefore, investors will invest in commercial real estate products directly when rises, and the investment in REITs is relatively reduce, resulting in T-REITs price decrease. When commercial price decreases, the investors will look forward to the development if T-REITs, and choose to invest in T-REITs rather than the real products, making T-REITs price rises.
In the adjusting speed, if the test results of the factor (CE1) is significant, indicating an error correction term is not 0 and a deviation from the long-run equilibrium value exists. At this point, the series will return to long-run equilibrium value will be a specific rate of adjustment. When the coefficient is not significant, the deviations will not be adjusted immediately, but over time, it will ultimately be adjusted towards the long-term equilibrium value. In this study, the adjusting speed for T-REITs and stock price index, T-REITs and commercial rent as well as T-REITs and commercial price are 33%, 54%, 42%, respectively. The adjustment rate between T-REITs and commercial rent is the highest, the reason might be that the rent value directly affect the total revenue of the investment and thereby affect the REITs price.
So the T-REITs price could quickly adjust to the equilibrium value when it moves away from the value.
The results suggest that the current T-REITs are not only affected by itself, but also the past stock price, commercial rent and commercial price. Surprisingly, the lag three stock price, commercial rent and commercial price are still impact the current T-REITs. We believe the possible reasons are (1) The inconsistency of the data basis.
Because the result can only reflect the information of the past quarter when using the quarter rent data to do the test. Therefore, the result shows T-REITs is affected by lag three rent. (2) Subject to the impact of agency problems, making the information reflected by T-REITs prices is limited. The current T-REITs managers are not the impartial and objective third party but the originators, so it is easy to operate by the impact of the managers and generate the agency problems. This will limit the ability of T-REITs reflecting the market information, and thereby affect T-REITs price volatility. Therefore, the T-REITs price fluctuations arising from the impact of other markets will be delayed. (3) The T-REITs market is still inefficient so that the
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
46
T-REITs price changes speed to the economic changes is slow. Thus, the T-REITs price changes response to economic variables is at limit speed.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
47
Table 4-4 Results of VECM
Dependent variable △REITs
Stock Price Index Results of commercial properties (rent) Results of commercial properties (price)
Variables Coefficient t-value Variables Coefficient t-value Variables Coefficient t-value CE1 -0.3270 *** -2.4773 CE1 -0.5441 *** -4.2243 CE1 -0.4184 ** -2.1321
△REITs(-1) 0.6943 *** 3.8371 △REITs(-1) 0.8488 *** 6.2951 △REITs(-1) 0.8754 *** 4.3633
△REITs(-2) -0.0785 -0.3384 △REITs(-2) 0.3408 ** 1.7592 △REITs(-2) -0.0829 -0.3224
△REITs(-3) 0.3171 1.2929 △REITs(-3) 0.6423 *** 3.1898 △REITs(-3) 0.7051 *** 2.5625
△REITs(-4) 0.0658 0.2514 △REITs(-4) 0.1527 0.6255 △REITs(-4) 0.0650 0.2076
△STOCK(-1) 0.0027 ** 1.8089 △CRENT(-1) -0.0867 -1.2406 △CPRICE(-1) -0.2343 * -1.6260
△STOCK(-2) -0.0013 -0.9398 △CRENT(-2) -0.0327 -0.4317 △CPRICE(-2) -0.2356 ** -2.1107
△STOCK(-3) 0.0034 *** 2.4297 △CRENT(-3) 0.2739 *** 3.4305 △CPRICE(-3) -0.0996 -1.0312
△STOCK(-4) 0.0007 0.4823 △CRENT(-4) -0.0936 -1.1154 △CPRICE(-4) -0.0279 -0.3637
C -0.0232 -0.0498 C -0.4254 -0.7314 C -0.1604 -0.3258
R-squared 0.4982 R-squared 0.7140 R-squared 0.4392
Adj. R-squared 0.3525 Adj. R-squared 0.6309 Adj. R-squared 0.2764
Note 1: △represents series at 1st differential order
Note 2: *, **, and ***represent significant level of 10% , 5%, and 1% respectively.