3.1 Sample
In this section, we examine daily stock prices of Taiwan and the United State from 7
November 1995 to 26 March 2009. In the case of the Taiwan stock market, we use the Taiwan
Weighted Stock Index and Over-the-Counter index. For the United States we use S&P500.
Because of different closing hours, we delete the data which only has partial market
information. There are 3178 daily returns included finally.
We retrieve the raw data of the Taiwan Weighted Stock Index and Over-the-Counter
index for the entire period from TEJ, and S&P500 from Yahoo Finance
(www.yahoo.com/finance).
In this paper, we use close-to-close returns of US to investigate whether the former
returns of US effects the returns of Taiwan. We discuss three market segments of Taiwan
stock which comprise close-to-open, open-to-close and close-to-close returns respectively.
We wanted to discuss the influences of different periods, especially between the Asian
Financial Crisis in 1997and recent Financial Tsunami recently, so we considered four periods
as follows: The first period is before the Asian Financial Crisis, because Over-the-Counter
index of TEJ starts from November 1995, we choose the data from 7 November1995 to 30
June 1997. According to Cheung, Cheung and Ng (2007) and Chang, M. C. (2008), the
second period is during the crisis from 2 July 1997 to 30 June 2000. We defined Financial
Tsunami starts from the bankruptcy of New Century Financial Corporation in March 2007 by
some research (Ashcraft and Schuermann, 2008) and the international financial news of
Yahoo. The third period is between the crisis and Financial Tsunami from 3 July 2000 to 27
February 2007. The last period is during the tsunami from 1 March 2007 to 26 March 2009.
We use the following picture to understand the trading periods between US and Taiwan
easily. The “Ot” represents the opening stock price of t day, and the “Ct” represents the closing
stock price of t day. Because the trading dates of US are later than Taiwan’s, we can say that t
day of Taiwan stock market is influenced by t-1 day of US stock market. For example, the
return of US’s t day (the period from Ct-1 to Ct)affects three periods of Taiwan stock market:
(1).The period from the closing price of t day to the opening price of t+1 day (from Ct to Ot+1.)
(2). The period from the opening price of t+1 day to the closing price of t+1 day (from Ot+1 to
Ct+1.) (3).The period from the closing price of t day to the closing price of t+1 day (from Ct to
Ct+1,) and so on.
Figure 1: The trading periods between US and Taiwan stock markets
3.2 Descriptive Statistics
In Table 1, we show the descriptive statistics of the returns of Taiwan and US stock
markets from 1995/11/7 to 2009/3/26. Panel A shows the descriptive statistics for the
close-to-open returns of Taiwan stock markets and close-to-close returns of US stock markets.
The daily close-to-open returns are calculated by 100×log (ΡtOpen/ΡtClose−1 ).
＜Panel A is inserted about here＞
Panel B shows the descriptive statistics for the open-to-close returns of Taiwan stock
markets and close-to-close returns of US stock markets. The daily open-to-close returns are
calculated by 100×log (ΡtClose/ΡtOpen).
＜Panel B is inserted about here＞
Panel C shows the descriptive statistics for the close-to-close returns of Taiwan stock
markets and close-to-close returns of US stock markets. The daily close-to-close returns are
calculated by 100×log (ΡtClose/ΡtClose−1 ). And all the descriptive statistics of the returns of the
series are given in Table1.
＜Panel C is inserted about here＞
In Table 1, we can know the difference between close-to-open, open-to-close and
close-to-close returns of TAIEX and OTC over the sample period from 1995/11/7 to
2009/3/26. The means of close-to-open returns of TAIEX and OTC are positive (Panel 1),
while the means of open-to-close of TAIEX and OTC are negative (Panel 2). The means of
close-to-close returns of TAIEX is positive, however, OTC’s is negative (Panel 3).
For volatility, it is shown that the standard deviation of TAIEX and OTC are the smallest
in close-to-open returns. And the standard deviation of TAIEX and OTC are the highest in
close-to-close returns.
Close-to-open and close-to-close returns of TAIEX and OTC are negatively skewed, so
we know they shift to the left. And open-to-close returns of TAIEX and OTC are positively
skewed, so we know they shift to the right. Moreover, all moments of the returns have excess
kurtosis.
The means of close-to-close returns of NASDAQ, Dow Jones and S&P 500 are positive
and have excess kurtosis. Dow Jones and S&P 500 are negatively skewed and shift to the left.
However, NASDAQ is positive and shifts to the right.
In Panel 1, it is indicated that the volatility of NASDAQ return is the highest and the
close-to-open return of TAIEX is the smallest. In Panel 2, the volatility of NASDAQ return is
the highest and the Dow Jones is the smallest. And it is shown that the volatility of OTC
return is the highest and the Dow Jones is the smallest in Panel 3.
All the variables we use are returns, so we will not have any spurious regression. We can
use quantile regression directly.