Faculty of Economics and Business Major track Finance and Organization
Bachelor Thesis
Do firms with higher stock liquidity perform better?
Case of Vietnamese stock market
Name Student number Supervisor
Ngan Ha Ngo 11106522 Dr. Evgenia Zhivotova
Abstract: With the recent growth of the emerging markets, more investment capital is pouring into these countries, and Vietnam is a case that is achieving outstanding economic development. The research is taken under the situation of Vietnamese stock market and focuses on the relation between the firm’s performance level with the liquidity of the stock. Using the Tobin’s Q ratio as the proxy for performance measure, and the turnover rate and bid-ask spread for liquidity measure, it is proved that firms whose stocks are more liquid have better performance than the ones with less liquid stock, and this result is robust.
Statement of Originality
This document is written by Student Ngan Ha Ngo, who declares to take full responsibility for the contents of this document:
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
I. Introduction
The emerging markets used to be ignored by investors in the 1990s as they believe these markets have extremely high risk premium and also returns were only at the moderate level (Kargin, 2002). Kargin (2002), in his research, found that the investment portfolios that included the stocks in emerging markets generate superior returns, but in general investors still underweight these stocks in their
portfolios. However, the recent return generated from these, according to the IMF, are two to three times higher than the investment in the mature ones, for instance, the US’s S&P 500, the UK’s FTSE or Japan’s Nikkei 500. Therefore, the recent growth of many emerging countries has again shifted the investors’ attention to these markets.
China is an outstanding case of an economy that has experienced impressive growth in the 21st century; however, there are more economies besides the Chinese one that also deserves the world’s attention, and Vietnam, which is the focus of this paper, is one of those. With the country’s average real GDP growth rate of 6% in the last five years, more than $12 billion FDI disbursed in 2017, and the appearance in the top 3 markets that have the highest stock return in 2017, the country now started to receive the global recognition. In World Bank’s report about Vietnam in 2016, the country, together with China, was called “a leading case of a country that has achieved rapid growth, poverty reduction, and shared prosperity.”
There are plenty of research about investment and portfolio management in developed markets but not so many looked at other emerging markets than China. No matter investing in developed or emerging market, liquidity is always an important factor that investors will keep in mind as it correlates directly and negatively to the stock returns (Bekaert, Harvey, & Lundblad, 2007; Duchin, 2010; Fang, Noe, & Tice, 2009; Lesmond, 2005). Fang et al. (2009) has looked at the relationship between the firms’ performance and their stock liquidity in case of the USA markets, and found out that there is a positive relationship: firms who perform better also have their stock trading more liquid. However, no research has been taken in emerging markets like Vietnam, to confirm if this relationship still holds in these so-called “riskier and poorer liquidity” markets (Kargin, 2002).
The main point of this paper is to indicate whether there is such a relation between the firm’s performance level and stock liquidity. Section II is the introduction of some essential features of the Vietnamese market, following by Section III which will include the literature review of theories that will be used in this paper. Section IV will cover the methodology of how this research had been taken before the regression results are discussed in section V and VI. The last section will be the summary of the results and some suggestions for further future research.
II. Basic features of the Vietnamese economy and stock market
1) The economy: from a low-income country to a high-potential one
The country had been through two continuous and long fights back to the French Colonists and the USA, which then left the severe poverty for the government in 1975. By 1986, the government announced the “Doi Moi” (means “renew”), which had changed the whole economy and started the developing period. Since then, the yearly GDP has not stopped to increase (see Figure 1 in Appendix), and the living standard has got better, proving by the continuous rising of the consumption expenditure (World Bank, 2017).
In 2006, the real GDP growth rate reached 6.21% (see Figure 2 in Appendix), which was approximately double of that of the upper-middle-income-level country and about five times relative to the USA (World Bank, 2017). Continuously achieving high growth, Vietnam starts to receive more
attention from the world and was added to the Next Eleven (N-11) list. Picked by the Goldman Sachs Investment Banker and economist Jim O’Neil, these N-11 countries, along with the BRICS countries, are predicted to become the world’s largest economy in the 21st century. The amount of FDI disbursed also experience the same rising trend, which is in line with the growth of the economy (see Figure 3).
2) The stock market and a prosperous 2017
The first and biggest stock exchange was open in 2000, which is named as Ho Chi Minh Stock Exchange (HOSE). According to the State Securities Commission of Vietnam (SSC), together with the Hanoi Stock Exchange (HNX, opened in 2005) and the Unlisted Public Company Market (UpCoM, opened in 2009), the market, till 2017, includes 2,016 stocks from the public and private companies and investment funds. The market then experienced the bubble-pricing pattern in 2006 – 2007 period (see Figure 4) when VNIndex, representing the index point of HOSE, achieved the peak of 1170.67 in March 2007, which is still never be reached again until now. With the occurrence of the 2008 USA’s economic crisis, the bubble was crashed, and the VNIndex fell sharply to roughly 370 points just a year after having achieved the highest peak ever. Ten years have passed, the market has slowly recovered, built up firmly, and again started its growing period at the beginning of 2017 when much confident prediction is given for a successful year. In fact, 2017 was indeed a successful year for the Vietnamese market stock with the rise of 46.9% of the VNIndex from 664.87 to 984.24, making it one of the stock index that had the highest return during the year. In just a few days after the beginning of 2018, the Vietnamese investors have just celebrated VNIndex finally surpassing 1000 after ten years, resulting even more positive prediction from World Bank and the IMF for the market in 2018 compared to 2017.
In the latest announcement in July 2017 from the Morgan Stanly Capital Investment, Vietnam is still placed in the list of frontier markets by MSCI for qualitative reasons related to openness to foreign investors, despite meeting all the fundamental requirements regards of market capitalization, liquidity, PE ratio, etc. Nevertheless, the number of foreign investors granted the transaction codes in the Vietnamese market in December 2017, according to the Vietnam Securities Depositories (VSD), is 447 including both individuals and organizations, which is the highest ever for a monthly period. Currently, there are more than 23000 foreign individuals and organizations that participate in the Vietnamese stock market; in 2017, the purchasing and sale values from these foreign investors are approximately $767 million and $717 million respectively.
III. Theoretical Framework
1) Performance measure: Tobin’s Q
First introduced by an economist James Tobin, the Tobin’s Q is a favorite market-to-book ratio that has been used in several research to measure firm’s performance and growth opportunities (Duchin, 2010; Fang et al., 2009; Graham, Lemmon, & Wolf, 2002; Kaplan & Zingales, 1997; Kogan &
Papanikolaou, 2014; Semmler & Mateane, 2012; Tobin & Brainard, 1976; Yermack, 1996). It measures “the valuation of capital installed in the firm relative to its replacement cost” (Kogan & Papanikolaou, 2014).
The formula is simple and easily constructed: the numerator is just the market value of the total assets of the firm (AMV), and the denominator is the asset book value of the firm, which covered not only physical assets but also other items on balance sheet (ABV). In the later researches (Aslan, Easley, Hvidkjaer, & O’Hara, 2011; Kaplan & Zingales, 1997; Kogan & Papanikolaou, 2014; Lee & Masulis, 2009; Masulis, Wang, & Xie, 2007), for the ease in computation, the authors calculated the assets market value of the firm by taking the book assets value plus the market value of equity (EMV) (i.e. the market cap, which is number of outstanding shares times the share price) minus the book value of equity (EBV).
𝑄 =𝐴𝑀𝑉 𝐴𝐵𝑉 =
𝐴𝐵𝑉 + 𝐸𝑀𝑉 − 𝐸𝐵𝑉 𝐴𝐵𝑉
On the stock market, the stocks are exchanged daily at different price levels due to the different stock valuation of different investors; this indirectly affected the market value of the firm, and result in the volatility of the Q (Aslan et al., 2011). In equilibrium, Q should be equal to 1 (Tobin & Brainard, 1976); however, in reality, the stock market is also not entirely efficient and in some cases do not reflect the right value of the stocks, so the value of Q mostly different than 1. Previous research (Duchin, 2010; Graham et al., 2002; Lee & Masulis, 2009; Masulis et al., 2007; Semmler & Mateane, 2012) proved that the firms whose Q is higher than 1, which are referred as “higher performance” firm, attract more
investment, while the firms who have their Q less than 1 are referred as “lower performance”, discourage investment and have lower “replacement cost”.
The Tobin’s Q is the primary focus of this paper, however, to test the robustness of the result, another performance measure that is used later in this paper is the return on asset (ROA). The detailed discussion about this measure is provided in Section VI.
2) Liquidity measure: Bid-ask Spread & Turnover rate
There are different measurements of liquidity, but the two above-mentioned measure has been used in previous research on liquidity: the bid-ask spread (Fang et al., 2009; Goyenko, Holden, & Trzcinka, 2009; Lesmond, 2005; Levi & Zhang, 2015) and the turnover ratio (Datar, Naik, & Radcliffe, 1998; Fang et al., 2009; Lesmond, 2005; Liu, 2006).
The use of turnover rate as a proxy to measure stock liquidity is first proposed by Datar, Naik, & Radcliffe (1998). Turnover ratio is calculated by taking the number of shares traded in a certain period (for instance: daily, monthly, yearly, etc.) divided by the total number of shares outstanding (Datar et al., 1998). Since the degree of liquidity is correlated with trading frequency (Amihud & Mendelson, 1986), the turnover ratio can be used to measure liquidity, as liquidity sometimes cannot be observed directly (Datar et al., 1998). Another advantage when using turnover ratio is the ease of data collection and calculation; this allows investors to review the liquidity of a large number of stocks in an extended period (Datar et al., 1998).
Another measurement that is used in this paper is the bid-ask spread. Instead of looking at liquidity, Amihud & Mendelson (1986) focus on the illiquidity of the stock as the “cost of immediate execution.” Investors always face a problem when trading, as they may need to buy or sell the stock immediately but then still want to wait for a favorable price; when buyers only want to buy at the low price but sellers only agree to sell at high price, illiquidity is created and no orders are executed.
Therefore, buyers will need to pay higher than their favorable price, which is called buying premium, in order to immediately own the target stocks, and sellers will ask for a lower price, which is referred as selling concession, to get the stocks immediately sold; the bid-ask spread is then measured by the sum of the buying premium and the selling concession (Amihud & Mendelson, 1986). In other words, the bid-ask spread of stock is merely the difference between the highest bid and lowest ask for that stock.
It is widely known that liquidity is an essential feature in asset pricing: assets with higher transaction costs or low degree of liquidity are often traded at a lower price than the more liquid one. In case of emerging market like Vietnam, liquidity appears to be an even more crucial price-determining factor than the developed market, and models that included liquidity variables outperformed those that included only market risk variables (Bekaert, Harvey, & Lundblad, 2007). Another research by Lesmond (2005) on liquidity in emerging markets showed that emerging markets, usually under countries with
weak political and legal institutions, have higher liquidity premium than the developed one with the stable political system.
IV. Methodology
1) Data and variable construction
Fundamental data about firms are collected from the database of Techcombank Vietnam, one of the largest banking company in Vietnam. Trading data (price and volume) are taken from the CafeF, the so-called “Vietnamese Bloomberg,” which is a well-known public and high-quality source for investors and researchers to use.
The initial sample contains all 342 stocks that are traded on the HOSE. All firms which have stocks traded on the HOSE are considered, except firms that are having “warning sign” from the SSC due to frauds in their financial statement or recent negative profits. Firms whose stocks got involved in some price-manipulation scandals that happened in Vietnam recently or have missing values are also excluded. After cleaning the data, small firms with the amount of market cap less than 50 billion VND (about 2.2 million USD) are crossed out of the list. Consequently, the sample is left with a cross-sectional data with 109 observations represent different 109 stocks traded on the HOSE (see Table 1 for the definition of all variables).
i. The primary variables: 𝑸, 𝑻𝑶𝑽𝑹 and 𝑺𝑷𝑹𝑫
The dependent variable 𝑄, which is the Tobin’s Q ratio, has been explained in previous section and can be calculated by the above formula:
𝑄 =𝐴𝑀𝑉 𝐴𝐵𝑉 =
𝐴𝐵𝑉 + 𝐸𝑀𝑉 − 𝐸𝐵𝑉 𝐴𝐵𝑉
In accounting term, asset is the sum of liability and equity, and since the market value and book value of liability (LMV and LBV) are considered to be equal (Lee & Masulis, 2009; Masulis et al., 2007), the formula can be written as below and will be used in this paper to calculate the value of variable 𝑄. Like most of the reference related to Tobin’s Q in this paper, the value of 𝑄 is calculated at the beginning of the fiscal year, which is the beginning of 2017 in this case.
𝑄 =𝐴𝑀𝑉 𝐴𝐵𝑉 = 𝐴𝐵𝑉 + 𝐸𝑀𝑉 − 𝐸𝐵𝑉 𝐴𝐵𝑉 = 𝐿𝐵𝑉 + 𝐸𝐵𝑉 + 𝐸𝑀𝑉 − 𝐸𝐵𝑉 𝐿𝐵𝑉 + 𝐸𝐵𝑉 = 𝐿𝐵𝑉 + 𝐸𝑀𝑉 𝐿𝐵𝑉 + 𝐸𝐵𝑉 Data of price and trading volume in 2016 and 2017 are used to calculate the two liquidity measures; if stocks are recently traded then the minimum required period is eight month. The value of variable 𝑇𝑂𝑉𝑅 is constructed as follows: the total stock trading volume of each month is averaged and then divided by the total number of shares outstanding of that stock, which is the same method that was used in the research of Fang et al. (2009)
𝑇𝑂𝑉𝑅𝑖=
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝑡𝑟𝑎𝑑𝑖𝑛𝑔 𝑣𝑜𝑙𝑢𝑚𝑒 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠ℎ𝑎𝑟𝑒𝑠 𝑜𝑢𝑡𝑠𝑡𝑎𝑛𝑑𝑖𝑛𝑔
Variable 𝑆𝑃𝑅𝐷 represents the bid-ask spread value in the trading activities of each stock. Following Bekaert et al. (2007), differences in daily trading bid and ask price in 2016 – 2017 are
calculated and then averaged by 501 trading days throughout the period. If stocks are recently traded, the minimum number of data point is 200 trading days.
𝑆𝑃𝑅𝐷𝑖 =
∑𝑛𝑖=1𝑆𝑝𝑟𝑒𝑎𝑑 𝑜𝑓 𝑑𝑎𝑦 𝑖
𝑛 ; 𝑆𝑝𝑟𝑒𝑎𝑑 = 𝐻𝑖𝑔ℎ𝑒𝑠𝑡 𝑎𝑠𝑘 − 𝑙𝑜𝑤𝑒𝑠𝑡 𝑏𝑖𝑑 ii. Other explanatory variables: 𝑰𝑫𝑰𝑶𝑹𝑰𝑺𝑲, 𝑳𝑶𝑮_𝑨𝑮𝑬 and 𝑽𝑵𝟑𝟎.
The 𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾 represents the firm-specific risk, calculated by the standard deviation of OLS residuals taken from regression of the daily stock return to market risk premium (CAPM). The firm-specific risk could possibly emerge from firm’s activities and CEO’s decision; additionally, higher volatility implies higher expected return on stocks, which could also have an impact on the firm’s market cap and the Q value. This had been proved in the research of Spiegel & Wang (2005): the authors confirmed that liquidity and the idiosyncratic risk of firms are negatively correlated, and while both variables affect the stock returns, the idiosyncratic risk has a much stronger effect compared to the degree of stock liquidity, and often clear out the effect of liquidity on return. Therefore, following Spiegel & Wang (2005) and Fang et al. (2009), the 𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾 is included to control for the possible underlying factors that drive the relationship between firm’s performance and liquidity.
𝐿𝑂𝐺_𝐴𝐺𝐸 is simply the natural logarithm of the firm’s current age until 2018. The use of this variable is similar to the one included in the research of Ishii et al. (2003), with the explanation that managers of firms which have been operated for longer period are more experienced than the newer one, and the asset value of these “older” firms are possibly larger than the small one as they have accrued it through out years. Therefore, the variable 𝐿𝑂𝐺_𝐴𝐺𝐸 is added to control this possible effect on the relationship of firm’s performance and stock liquidity.
Finally, dummy variable 𝑉𝑁30 is added into the model, which is equal to 1 if the stock is named on the VN30, and is 0 otherwise. The VN30 is the list of top 30 most potential stocks traded on the HOSE, which is revised and announced every half year; usually these stocks belong to firms with large asset value, large market cap and have some certain upcoming advantages that possibly result in a high growth. These stocks, therefore, often received more attention from investors and their trading volume are also higher than other stocks that are not included. Consequently, being included in the VN30 possibly increases the expectation of firm’s performance in the upcoming period, which can drive up the stock price and results in the higher Q value, so the variable 𝑉𝑁30 is added to control that effect. The latest VN30 list, which is announced on July 2017, is used in this paper.
iii. Descriptive statistics and correlation
The descriptive statistics of the cross-sectional sample are shown in Table 3. All variables have reasonable means; however, 𝑄 and 𝑆𝑃𝑅𝐷 have an obviously wide range, while the range for other variables are much smaller. The smallest value of 𝑄 is 0.068, while the largest value reaches 8.199; for 𝑆𝑃𝑅𝐷 the smallest and largest value are 0.122 and 6.22 respectively. Both two variables have high kurtosis (14.011 for 𝑄, 10.535 for 𝑆𝑃𝑅𝐷), while 𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾 and 𝐿𝑂𝐺_𝐴𝐺𝐸 have their kurtosis close to zero; 𝑇𝑂𝑉𝑅 has kurtosis of 2.948, nearest to the kurtosis of 3 for normal distribution. 𝑄, 𝑆𝑃𝑅𝐷 and 𝑇𝑂𝑉𝑅 have the skewness value larger than 1, especially 𝑄 (skewness 3.372) and 𝑆𝑃𝑅𝐷 (skewness 2.789), implying that the distributions of these variables are highly skewed to the right.
The correlation matrix is shown in Table 2, which indicates the first insights into the relationship between 𝑄 and the other variables. All variables have positive correlation with 𝑄, except 𝑆𝑃𝑅𝐷 with negative correlation. These correlations are reasonable and compliment with previous findings that have been mentioned in this paper, as higher-performance firms should have their stocks more liquid (so positively correlated with 𝑇𝑂𝑉𝑅 and negatively correlated with 𝑆𝑃𝑅𝐷). The correlation matrix also suggests that firms with higher Q should have higher idiosyncratic risk (positively correlated with
𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾) and have been under operation for a longer period (positively correlated with 𝐿𝑂𝐺_𝐴𝐺𝐸. The positive correlation of dummy variable 𝑉𝑁30 and 𝑄 implies that the VN30 firms perform at a higher level than the other ones.
2) Model and Hypothesis
With the variables mentioned above, the model is now constructed:
𝑸𝒊 = 𝝏𝟎+ 𝝏𝟏𝑻𝑶𝑽𝑹𝒊+ 𝝏𝟐𝑺𝑷𝑹𝑫𝒊+ 𝝏𝟑𝑰𝑫𝑰𝑶𝑹𝑰𝑺𝑲𝒊+ 𝝏𝟒𝑳𝑶𝑮_𝑨𝑮𝑬𝒊+ 𝝏𝟓𝑽𝑵𝟑𝟎𝒊+ 𝜺𝒊
(𝜀𝑖: error term; 𝜕0, 𝜕1, 𝜕2, 𝜕3, 𝜕4, 𝜕5: coefficients)
Since turnover rate measures the liquidity of the stock, while bid-ask spread measure illiquidity of stock, the coefficients of these two variables estimated in next sections should have opposite sign. Based on previous relevant research, the following hypotheses are suggested:
H0: The liquidity of stock does not impact the firm’s performance (that is 𝜕1 = 0; 𝜕2 = 0)
H1: Firms with liquid stock have higher performance level (that is 𝜕1> 0; 𝜕2< 0)
Possibly explanation to form this hypothesis is that: firms with higher performance level, that often have lower required return for stock, have a higher value of 𝑄; and as liquidity is priced, investors must pay a liquidity premium if they want to own these more liquid stocks (Holmström & Tirole, 2001). Another possible reason is that liquid stocks also attract more investment, and raise investors’ expectation about firms’ performance; informed investors about the performance of the firms will also trade more aggressively which then affect the liquidity of stock. (Holmström & Tirole, 2001). Aslan et al. (2011) also found out in their research that firms with less risk of information asymmetry have higher turnover rate and higher Tobin’s Q ratio.
H2: Firms with less liquid stock have higher performance level (that is 𝜕1 < 0; 𝜕2 > 0)
The formation of this hypothesis is based on the work of Goldstein & Guembel (2008) when they explained that inefficient investment (not choosing the right stocks or the non-optimal allocation of capital in each stock) could happen. In reality, not all investors are rational, and the rational ones can still invest inefficiently knowing a possible situation that they could be exploited by speculators. The
inefficiency could result in the case that the firms with less liquid stocks can still be the high-performing type, while the firms with much more liquid stocks are the ones with lower performance level (Goldstein & Guembel, 2008).
V. Empirical result and discussion
Based on the regression results presented in Table 4, it is evident that some patterns consistently appear in all regression models. The coefficients of 𝑇𝑂𝑉𝑅 and 𝑆𝑃𝑅𝐷 have opposite signs in all five regression results, which is in line with theories mentioned in previous sections. The coefficient estimators for 𝑆𝑃𝑅𝐷 are negative highly significant in all regression, proving that there is a strong negative relationship between the bid-ask spread in stock trading price and firm’s Q ratio: firms with higher Q have smaller stock bid-ask spread. This result again confirmed the previous findings about the relation between stock liquidity and firm’s performance level: higher performance firms with higher Q have their stock more liquid compared to the lower performance firms, as investors need to pay a premium for owning or selling the stock immediately (Holmström & Tirole, 2001); informed investors about the high performance level of firms will also trade aggressively and thus increase the stock
liquidity, and uninformed investors also got stimulated by the stock with good liquidity level and then increase the liquidity even more (Holmström & Tirole, 2001).
The estimators for the coefficient of 𝑇𝑂𝑉𝑅 also appear to be positive in all regression, which also suggest that higher-Q firms have more liquid stocks; however, this relationship is not significant. This insignificant relationship is possibly due to the systematic risk (global economy, political features, etc.) or delay in stock prices reaction to new information (Jegadeesh, N.; Titman, 1993). Another explanation is that the action of stocks buying and selling is not only dependent on the fundamental value of stocks, but also on the price of that stock, and the price of stocks also have their demand elasticity (Chan &
Lakonishok, 1993). In other words, a firm with higher performance level and strong fundamental values of stock often traded at higher price than other stocks on the market, and even if the stock is trading at the reasonable price, it still possibly faces illiquidity problem as the stock price is too expensive to encourage investors to buy it.
All three other explanatory variables 𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾, 𝐿𝑂𝐺_𝐴𝐺𝐸 and 𝑉𝑁30 have positive coefficient estimators, which imply a positive relationship with the firm’s Q value. However, only the estimator of 𝑉𝑁30 is significant, while those of the other two is insignificant. The positive but insignificant of the coefficient estimator of 𝐼𝐷𝐼𝑂𝑅𝐼𝑆𝐾 showing that the firm-specific features do have a relation to firm’s performance level, however this effect is not as significant as the stock liquidity. The estimation of 𝐿𝑂𝐺_𝐴𝐺𝐸 coefficient, which also positive and insignificant, points out an interesting view: the “older” and more experienced firms not necessarily have higher Q value and higher performance level than the “younger” ones. Finally, the positive and significant estimators of 𝑉𝑁30 implies that firms in the list of VN30 perform at a higher level and have higher degree of stock liquidity than those are not included; this is reasonable as otherwise they are already not included in the list. These VN30 stocks are already received more attention from investors as they are judged carefully by the SSC and considered to have high growth potential in the upcoming period, so it is understandable that these firms perform higher than other normal firms and their stocks are also more liquid.
In conclusion, the regression results suggest the rejection of H0 and H2 and acceptation of H1: firms with more liquid stocks have higher performance level than the ones with less liquid stocks. However, the effect of bid-ask spread appears to be more significant than the turnover ratio. The firm’s idiosyncratic risk, age and the presence in VN30 also have a positive relation to the performance level, but the inclusion in VN30 implies the significant signal of firm’s high-performance level, while the other two variables do not give significant effect.
VI. Robustness check: test for ROA
Beside the main test of Tobin’s Q that has been shown above, another test which uses a different performance measure as the dependent variable of the model is performed to test the robustness of the relationship between firm’s performance and stock liquidity. In this section, the firm’s performance is now measured by return on asset, measured by variable 𝑅𝑂𝐴 instead of the variable 𝑄 like the above-mentioned regression model. The use of return on asset as a proxy for performance level is similar to many researches from the older ones (Core, Holthausen, & Larcker, 1999; Mandelker, 1974) to the more recent ones (Bertrand & Betschinger, 2012; Jenter & Kanaan, 2015; Ma, Whidbee, & Zhang, 2011).
The construction of variable 𝑅𝑂𝐴 is simple: taking the net income divided by the total asset, which is the same as asset book value that we use to calculate variable 𝑄 as above. Data is again taken from the database of Techcombank Vietnam, and the data in the beginning of fiscal year 2017 will be used. Therefore, we have the following model that is used for robustness check test:
𝑹𝑶𝑨𝒊 = 𝜽𝟎+ 𝜽𝟏𝑻𝑶𝑽𝑹𝒊+ 𝜽𝟐𝑺𝑷𝑹𝑫𝒊+ 𝜽𝟑𝑰𝑫𝑰𝑶𝑹𝑰𝑺𝑲𝒊+ 𝜽𝟒𝑳𝑶𝑮_𝑨𝑮𝑬𝒊+ 𝜽𝟓𝑽𝑵𝟑𝟎𝒊+ 𝝎𝒊
(𝜔𝑖: error term; 𝜃0, 𝜃1, 𝜃2, 𝜃3, 𝜃4, 𝜃5: coefficients)
Like the previous section, three hypotheses are used again:
H0: The liquidity of stock does not imply the firm’s performance (that is 𝜃1= 0; 𝜃2= 0)
H1: Firms with liquid stock have higher performance level (that is 𝜃1> 0; 𝜃2< 0)
H2: Firms with less liquid stock have higher performance level (that is 𝜃1< 0; 𝜃2> 0)
The regression model using ROA gives out similar results (see Table 5) to the model using Tobin’s Q (see Table 4). The signs of 𝑇𝑂𝑉𝑅 (positive) and 𝑆𝑃𝑅𝐷 (negative) are opposite, which is in line with theories and previous regression model in the paper. 𝑆𝑃𝑅𝐷 remains to be the variable which has a strong effect on the performance measure 𝑅𝑂𝐴, while 𝑇𝑂𝑉𝑅 still has some positive impact but not significant. This suggests that firms with higher turnover rate and less bid-ask spread having higher return on asset; in other words, firms with more liquid stocks performs better as they have better ability to increase their asset value. Same as the previous model with Tobin’s Q, the result implies the rejection of H0 and H2, and possibly accepting H1 considering the insignificant of 𝑇𝑂𝑉𝑅’s coefficient estimators. The differences in the results of two models is that: the effect of 𝐿𝑂𝐺_𝐴𝐺𝐸 now becomes more significant, which indicates “the older the better” pattern, while the explanator power of 𝑉𝑁30 now turns
insignificant, which implies that firms included in VN30 are not necessarily having higher profitability ratio than the others; however, these variables are not the main scope of this paper and do not affect the robustness of the results.
To conclude, the regression of both the main model which include the paper’s focus on Tobin’s Q and the other model using ROA instead show the similar results: the same sign of 𝑇𝑂𝑉𝑅 (positive) and 𝑆𝑃𝑅𝐷 (negative) and the same significant level for these variables; the only difference is the significance level of 𝐿𝑂𝐺_𝐴𝐺𝐸 and 𝑉𝑁30. The consistency in the results of two models suggests that: firms whose stocks have higher degree of liquidity performs better than the ones whose stocks are less liquid, and the result is robust.
VII. Conclusion
The paper performs research in the case of Vietnam, which is an emerging market that is achieving much attention from the foreign investors. The economy is in its high-growth period, and the stock market just has the auspicious year of 2017 with the rise of 46.9% of the VNIndex (SSC, 2018). More positive reports have been published by prestigious sectors like World Bank and IMF, which raise investors’ expectation about an even more successful on the stock market of 2018. The projection of VNIndex reaching 1300 at the end of the year, given by the Saigon Securities Inc. (SSI), a popular financial services sector in Vietnam, is possible as now the VNIndex is only 100 points less than its highest-peak ever achieved in the bubble period in 2007, and 2018 has just begun. Therefore, the
Vietnamese stock market has recently experienced an extreme increase in the number of foreign investors that opened the stock trading account in both the HOSE and HNX. With the effort of bringing Vietnam into the list of MSCI’s next emerging market on 2020, the government have announced new policies that are more open and beneficial to foreign investors and increase the amount of information and data in English to create a better trading experience for them (SSC, 2017).
Liquidity is always an essential factor of stocks that investors will have to look at: they want to execute their buying or selling order immediately but also want to wait for a favorable price (Amihud &
Mendelson, 1986). The investors may either have to pay a buying premium when executing their buying orders or experience a selling concession when they want to sell their stocks immediate, which will apparently affect the total return (Amihud & Mendelson, 1986; Holmström & Tirole, 2001). However, liquidity means more than just the ease of trading at the favorable price, which one of them, the firm’s performance, is the primary focus of in this paper. Using Tobin’s Q as the proxy for the performance level of firms and two popular liquidity measures that are turnover ratio and bid-ask spread, the result suggests that firms which have more liquid stocks perform at a higher level than the ones with less liquid stock, and this relationship is robust. The bid-ask spread turns out to have a much stronger explanatory power on the firm’s performance than the turnover ratio does, but both variables are consistent with the theories. Some other results are the older firms not necessarily perform better than the younger one, and the power of being included in VN30 give a significant signal of a higher performance level of firms.
The limitation of this paper is that the chosen period to research is not so long and only stocks traded on HOSE are added pooled in, due to many missing values that are not included in the original database. There are possibly a lot more variables that affect the firm’s performance but are not added to the model due to the scope of this research. Future research taken under the situation of Vietnamese markets should take a more in-depth look to each different industry sectors, add more explanatory variables to these models, and a more extended period of research with extension to HNX and UPCOM stocks should be performed.
Bibliography
Amihud, Y., & Mendelson, H. (1986). Asset pricing and the bid-ask spread. Journal of Financial
Economics, 17(2), 223–249.
https://doi.org/10.1016/0304-405X(86)90065-6 Aslan, H., Easley, D., Hvidkjaer, S., & O’Hara, M.
(2011). The characteristics of informed trading: Implications for asset pricing. Journal of
Empirical Finance, 18(5), 782–801.
https://doi.org/10.1016/j.jempfin.2011.08.001 Bekaert, G., Harvey, C. R., & Lundblad, C. (2007).
Liquidity and expected returns: Lessons from emerging markets. Review of Financial Studies,
20(6), 1783–1831.
https://doi.org/10.1093/rfs/hhm030 Bertrand, O., & Betschinger, M. A. (2012).
Performance of domestic and cross-border acquisitions: Empirical evidence from Russian acquirers. Journal of Comparative Economics,
40(3), 413–437.
https://doi.org/10.1016/j.jce.2011.11.003 Chan, L. K. C., & Lakonishok, J. (1993). Institutional
trades and intraday stock price behavior.
Journal of Financial Economics, 33(2), 173–
199. https://doi.org/10.1016/0304-405X(93)90003-T
Core, J. E., Holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief executive officer compensation, and firm performance.
Journal of Financial Economics, 51(3), 371–
406. https://doi.org/10.1016/S0304-405X(98)00058-0
Datar, V. T., Naik, Y. N., & Radcliffe, R. (1998). Liquidity and stock returns: An alternative test.
Journal of Financial Markets, 1(2), 203–219.
https://doi.org/10.1016/S1386-4181(97)00004-9
Duchin, R. (2010). Cash Holdings and Corporate Diversification. The Journal of Finance, 65(3), 955–992. https://doi.org/10.1111/j.1540-6261.2010.01558.x
Fang, V. W., Noe, T. H., & Tice, S. (2009). Stock market liquidity and firm value. Journal of
Financial Economics, 94(1), 150–169.
https://doi.org/10.1016/j.jfineco.2008.08.007 Goldstein, I., & Guembel, A. (2008). Manipulation and the allocational role of prices. Review of
Economic Studies, 75(1), 133–164.
https://doi.org/10.1111/j.1467-937X.2007.00467.x
Goyenko, R. Y., Holden, C. W., & Trzcinka, C. A. (2009). Do liquidity measures measure liquidity? Journal of Financial Economics,
92(2), 153–181.
https://doi.org/10.1016/j.jfineco.2008.06.002 Graham, J. R., Lemmon, M. L., & Wolf, J. G. (2002).
Does Corporate Diversification Destroy Value?
The Journal of Finance, 57(2), 695–720.
https://doi.org/10.1111/1540-6261.00439 Holmström, B., & Tirole, J. (2001). LAPM: A
liquidity-based asset pricing model. Journal of
Finance, 56(5), 1837–1867.
https://doi.org/10.1111/0022-1082.00391 Ishii, J. L., Glaeser, E., Gyourko, J., Holthausen, R.,
Kaplan, S., Mullainathan, S., … Wulf, J. (2003). Corporate governance and equity prices. Quarterly Journal of Economics,
118(February), 107–155.
Jegadeesh, N.; Titman, S. (1993). Returns to Buying Winners and Selling Losers : Implications for Stock Market Efficiency. The Journal of
Finance, 48(1), 65–91.
Jenter, D., & Kanaan, F. (2015). CEO Turnover and Relative Performance Evaluation. Journal of
Finance, 70(5), 2155–2184.
https://doi.org/10.1111/jofi.12282
Kaplan, S. N., & Zingales, L. (1997). Do Investment-Cash Flow Sensitivities Provide Useful Measures of Financing Constraints? Journal of
Economics, 112(1), 169–215.
https://doi.org/10.1162/003355397555163 Kargin, V. (2002). Value investing in emerging
markets: risks and benefits. Emerging Markets
Review, 3(3), 233–244.
https://doi.org/10.1016/S1566-0141(02)00023-7
Kogan, L., & Papanikolaou, D. (2014). Growth opportunities, technology shocks, and asset prices. Journal of Finance, 69(2), 675–718. https://doi.org/10.1111/jofi.12136
Lee, G., & Masulis, R. W. (2009). Seasoned equity offerings: Quality of accounting information and expected flotation costs. Journal of
Financial Economics, 92(3), 443–469.
Lesmond, D. A. (2005). Liquidity of emerging markets. Journal of Financial Economics,
77(2), 411–452.
https://doi.org/10.1016/j.jfineco.2004.01.005 Levi, S., & Zhang, X. J. (2015). Asymmetric
decrease in liquidity trading before earnings announcements and the announcement return premium. Journal of Financial Economics,
118(2), 383–398.
https://doi.org/10.1016/j.jfineco.2015.08.003 Liu, W. (2006). A liquidity-augmented capital asset
pricing model. Journal of Financial
Economics, 82(3), 631–671.
https://doi.org/10.1016/j.jfineco.2005.10.001 Ma, Q., Whidbee, D. A., & Zhang, A. W. (2011).
Value, valuation, and the long-run performance of merged firms. Journal of Corporate
Finance, 17(1), 1–17.
https://doi.org/10.1016/j.jcorpfin.2010.07.001 Mandelker, G. (1974). Risk and return: The case of
merging firms. Journal of Financial
Economics, 1(4), 303–335.
https://doi.org/10.1016/0304-405X(74)90012-9
Masulis, R. W., Wang, C., & Xie, F. (2007). Corporate Governance and Acquirer Returns Corporate Governance and Acquirer Returns.
Journal of Finance, 62(4), 1851–1889.
https://doi.org/10.1111/j.1540-6261.2007.01259.x
Semmler, W., & Mateane, L. (2012). Equity Market or Bond Market — Which Matters the Most for Investment ? Revisiting Tobin ’ s q Theory of Investment, 2012(November), 203–209. https://doi.org/10.4236/ti.2012.34029 Spiegel, M., & Wang, X. (2005). Cross-sectional
variation in stock returns: Liquidity and idiosyncratic risk. Working Paper, Yale
University, 1–49.
https://doi.org/citeulike-article-id:12440568
Tobin, J., & Brainard, W. C. (1976). Asset Markets and the Cost of Capital. Cowles Foundation
Discussion Papers. Retrieved from
http://ideas.repec.org/p/cwl/cwldpp/427.html Yermack, D. (1996). Higher market valuation for
firms with a small board of directors. Journal
Appendix
Appendix
0 40 80 120 160 200 1986 1991 1996 2001 2006 2011 2016Vietnam GDP and consumption (billion USD)
GDP Final Consumption Expenditure
6.70% 3.47% 1.62% 6.21% -3.00% 0.00% 3.00% 6.00% 9.00% 12.00% 15.00% 1985 1990 1995 2000 2005 2010 2015
GDP real growth rate (%)
China Upper middle income United States Vietnam
Figure 2. GDP and Final Consumption Expenditure in some countries (World Bank, 2017)
Figure 2. GDP real growth rate of some countries (World Bank, 2017)
0% 2% 4% 6% 8% 10% 12% 14% 0 2 4 6 8 10 12 14 1985 1990 1995 2000 2005 2010 2015
FDI disbursed in Vietnam
FDI disbursed (bill USD) %GDP
Figure 3. FDI disbursed as percentage of GDP (World Bank, 2017)
Beginning Period Recovery Period
Bubble market Recession 0 3 0 0 6 0 0 9 0 0 1 2 0 0 VN In d e x 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Year
Variables Description Q
Tobin’s Q ratio of firm, computed by taking the sum of liability book value and equity book value divided by the sum of liability book value and equity market value
TOVR
The average monthly turnover rate of firm’s stock, computed by taking the average of 24 monthly trading volume in 2016 – 2017, divided by a total number of shares outstanding. Minimum required for the number of trading months is eight months.
SPRD
The average daily bid-ask spread of firm’s stock, computed by taking the average value of all daily bid-ask spread of firm’s stock. Minimum required for the number of trading days is 200 days.
IDIORISK
Idiosyncratic risk of the firm, which is the variance of residual in the OLS regression result of the excess stock return on the VNIndex excess return. The risk-free rate is taken as the interest rate of 10-year Vietnamese government bond.
LOG_AGE The firm’s age, counting from the foundation year till 2018 VN30
Dummy variable which equals to 1 if the firm is in the VN30 list, 0 otherwise. VN30 is revised and published every half year, includes the list of top 30 most potential stocks in that period.
Table 1. Definition of the cross-sectional variables
Q TOVR SPRD IDIORISK LOG_AGE VN30
Q 1 TOVR 0.1073 1 SPRD -0.4497 -0.0888 1 IDIORISK 0.0361 0.3395 -0.0973 1 LOG_AGE 0.0267 -0.1691 0.0279 -0.3324 1 VN30 0.2343 -0.000 0.0764 -0.2935 0.1687 1
Count Mean Median Mode
Std
Error Std Dev Variance Kurtosis Skewness Min Max
Q 109 1.082 0.662 N/A 0.124 1.292 1.668 14.011 3.372 0.068 8.199 TOVR 109 0.100 0.084 N/A 0.007 0.074 0.005 2.948 1.437 0.007 0.415 SPRD 109 0.916 0.607 0.451 0.090 0.937 0.878 10.535 2.789 0.122 6.220 IDIORISK 109 0.022 0.022 0.032 0.001 0.006 0.000 -0.311 0.453 0.012 0.036 LOG_AGE 109 1.308 1.301 1.041 0.021 0.220 0.048 -0.498 0.015 0.845 1.785 VN30 109 0.275 0.000 0.000 0.043 0.449 0.201 -0.977 1.021 0.000 1.000
Regression 1 Regression 2 Regression 3 Regression 4 Regression 5 TOVR 1.8683 1.1814 1.1500 0.8284 (1.6741) (1.5126) (1.4489) (1.5627) SPRD -0.6197*** -0.6114*** -0.6400*** -0.6359*** (0.1190) (0.1197) (0.1150) (0.1160) IDIORISK 14.9055 (21.9241) LOG_AGE 0.1259 (0.5234) VN30 0.7766*** 0.8212*** (0.2392) (0.2543) 𝑹𝟐 1.15% 20.23% 20.68% 27.92% 28.24% Adj 𝑹𝟐 0.23% 19.48% 19.19% 25.86% 24.76% *, **, ***: significant at 1%, 5%, 10%
Table 4. Cross-sectional regression result. Dependent variable: Q
Regression 1 Regression 2 Regression 3 Regression 4 Regression 5
TOVR 0.1469 0.0329 0.0320 0.1014 (0.0241) (0.1762) (0.1764) (0.1872) SPRD -0.0704*** -0.0702*** -0.0711*** -0.0712*** (0.0138) (0.0139) (0.0140) (0.1839) IDIORISK -0.4175 (2.6266) LOG_AGE 0.1171** (0.0627) VN30 0.0236 0.0124 (0.0291) (0.0305) 𝑹𝟐 0.31% 19.53% 19.56% 20.06% 23.00% Adj 𝑹𝟐 0.62% 18.78% 18.04% 17.77% 19.26% *, **, ***: significant at 1%, 5%, 10%
