國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
4
CSI will grant certain inclusion factor to the total A shares according to the percentage of free float shares in total A shares to insure index stability.
The component of CSI300 is adjusted regularly where the standard: (1) Meetings of CSI Index Advisory Committee are usually held in the end of May and November every year and constituents adjustment are implemented on the next trading day after the close of the second Friday in June and December each year; (2) Number of constituents adjusted at each periodical review will not exceed 10%. (3) CSI 300 adopts buffer zone rules for the sake of minimum turnover. New candidate stocks ranked top 240 will be given priority to add into the index and old constituents ranked top 360 will be given priority to remain in the index. (4) Reserve list is established at each periodical review of CSI 300, which is used to implement temporary adjustment during two adjacent periodical reviews.
During 2005, till 2014 (last announcement date taken in this study) 455 companies have been included (excluded) to (from) the index as shown in Figure 1. In 2013, CSIL has decided to change the regular adjustment of CSI300 from Jan and Jul to Jun and Dec. As shown in the figure 1, others (pattern in vertical line) are the stocks included or excluded on irregular day which are those top 10 market value of IPO(11) or because of exclusion of Merging(7) companies.
3.
Literature Review and Model ConstructionIn most stock markets, the effects of index composition changes on constituent stocks have been proved by several studies. There are significant and well-documented stock price increases when a stock is added to an index. In order to explain this phenomenon, several hypotheses have been advanced. This section presents a brief review of the relevant literature and attempts to develop the hypotheses for this study.
‧
Figure 1 Composition adjustment of the CSI300 index from 2005-2014 3.1 Downward Sloping Demand Curve Hypothesis
The first explanation is the Downward-Sloping Demand Hypothesis (DSCDH). When a stock is added to an index, there are additional demand from index-related users to hold the stock, resulting in short-term upward price pressure. According to this hypothesis, the demand curve is downward sloping not only in the short run, but also in the long run (Scholes, 1972; Shleifer, 1986; Lynch and Mendenhall, 1997). Several studies provide consistent empirical evidence for stocks in the S&P 500 index (Harris and Gurel, 1986;
Wurgler and Zhuravskaya, 2002), and the hypothesis is also supported by evidence from other US indices and markets, such as the FTSE 100 index (Mase, 2007), and the ISE100 and ISE-30 indices (Bildik and Gulay, 2008).
3.2 Price Pressure Hypothesis
The second explanation is the Price Pressure Hypothesis (PPH) which assumes that the long-term demand is perfectly elastic at full-information price. It holds if stock prices reverse to their ex-ante level after the index change, and recognizes that immediate information about non-information-motivated demand shifts may be costly and, consequently, the short-term demand curve may be less than perfectly elastic. Harris and
0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Frequency
Jan Jul Others Jun Dec
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
6
Gurel (1986), Woolridge and Ghosh (1986), Dhillon and Johnson (1991), Liu (2001), Madhavan (2003), Chen et al. (2004), Vespro (2006, and Yun and Kim (2010), among others, support the PPH.
3.3 Liquidity Effect Hypothesis
The third explanation is the Liquidity Effect Hypothesis (LEH), which predicts that liquidity will improve (deteriorate) after a stock is added to (deleted from) an index (Chen
et al., 2004). The amount of information on a stock increases upon its addition to an index due to greater attention from investors and greater coverage from analysts, the media, and other financial intermediaries. As a result, the information asymmetry declines and more liquidity become available. The concurrent decline in the liquidity premium causes a positive price movement. Furthermore, the presence of more investors trading the stock reduces the inventory cost component of liquidity, resulting in a further positive price adjustment (Chen et al., 2004). Various studies provide empirical support the LEH for the S&P 500 index (Hegde and McDermott, 2003; Becker-Blease and Paul, 2006), the Dow Jones index (Beneish and Gardner, 1995), and the TSE 300 index (Chung and Kryzanowski, 1998).
3.4 Investors Recognition Hypothesis
The third explanation is the investor recognition or ‘shadow cost’ hypothesis (Merton,
1987). Shadow cost states that when investors hold incompletely diversified portfolios in
segmented markets. The return required by less than fully diversified investors is higher
than that required in a full-information setting, with the difference between the two returns
representing the shadow cost. When a stock is added to an index, this increases the
awareness of investors, who will hold it to achieve diversification. The shadow cost of the
stock thus falls, resulting in an increase in the stock price (Chen et al., 2004). Elliott et al.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
7
(2006) reports that increased investor awareness explains the cross-section of abnormal announcement returns for stocks on the S&P 500 index.
3.5 Information Hypothesis
Another theory, which was investigated by some
index change researchers, is the Information Hypothesis (IH). According to this, the Standard & Poor’s Corporation uses at least some non-public information in selecting stocks for addition to the index. Therefore, the event may be seen as a positive signal to the market in support of the added stock. This potentially reduces the information asymmetry between the firm and the market, thereby resulting in permanent price impact. Jain (1987) and Dhillon and Johnson (1991) find evidence consistent with this hypothesis.While the above information hypothesis relies on improved firm-to-investor information asymmetry changes in stock prices are more closely related to inter-investor information asymmetry. A large body of extant empirical literature suggests that firm-to-investor-IA and inter-firm-to-investor-IA are linearly positively related. This view is intuitively supported by a commonly held view, which suggests that information asymmetry results from some traders being privy to information prior to it being publicly available (Hasbrouck;
1991). However, Hong and Ravi (2014) presents empirical support for a non-linear relationship between the two types of information asymmetry. They argue that, if a firm is completely transparent, all market participants would know everything about the firm, hence, the inter-investor-IA should be zero. If the firm is completely opaque, all participants are uninformed; hence, the inter-investor-IA should also be close to zero. Somewhere between the two extremes, the information asymmetry between investors should attain a maximum. Understanding the changes in investor information asymmetry around index changes should help us in developing an investor-centered explanation for the documented
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
8
effects on price formation.
3.6 Arbitrage Risk Hypothesis
The firth explanation is the Arbitrage Risk Hypothesis (ARH) (Wurgler and Zhuravskaya, 2002), which states that risk discourages arbitrage from flattening demand curves for stocks. Two recent papers suggest this view of arbitrage. Wurgler and Zhuravskaya (2002) points out that a specialized arbitrageur's demand for a stock is inversely related to the stock's arbitrage risk, which is essentially its idiosyncratic risk. In traditional portfolio theory, portfolio risk, and therefore a diversified investor's demand, is not affected by a stock's idiosyncratic risk. Shleifer and Vishny (1997) also suggests that idiosyncratic risk impedes arbitrage. The models of Wurgler and Zhuravskaya and Shleifer and Vishny imply that stocks with higher idiosyncratic risk should be less attractive to arbitrageurs. On observing equivalent earnings surprises, arbitrageurs may, therefore, take smaller positions in stocks with higher idiosyncratic risk. If other investors underreact to earnings announcements, then high idiosyncratic‐risk stocks will be more mispriced and therefore exhibit greater drifts.
There are several fundamental reasons to expect a price effect from the composition change in an index. In the next section, we provide empirical evidence for both added and deleted stocks.