MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIET NAM Ho Chi Minh City, September 2021 BANKING UNIVERSITY OF HO CHI MINH CITY NGUYỄN HUỲNH TUYẾT TRINH THE RELATION BETWEEN STOCK LIQUIDITY, FOREIGN[.]
INTRODUCTION
Reason to research
In corporate financial management, maximizing firm value is one of the targets that most of the financial managers tend to do Therefore, firm valuation is really important to the firm owner, who knows well the actual value and thence determine the development strategies for their business Besides, firm valuation can bring many benefits depends on each purpose and subject: Government, investor or corporate governance It also is the basis in acquisition, merger, consolidation, equitization, making decisions on corporate management, effective financial decisions, decisions on investment or investment cooperation.
Over 20 years of establishment and development, stock market of Vietnam has made positive contributions to the development of the entire economy The stock market helps businesses access cheap and available capital and thence the process of managing and maximizing firm value is developed A business share price can fluctuate regularly depends on its actual market valuation and many factors as price volatility, risk, unexpected events in the future The convesion from cash by highly liquid assets will help investors feel more secure because this feature helps investors
2 recover cash and thereby recover capital quickly at low cost Selling the illiquid securities is very difficult, even impossible unless selling at a very low price. Therefore, analyzing the liquidity of shares in the past of the enterprise is a necessary task in corporate management and is the foundation for investment analysis and firm valuation.
The contribution of the stock market to every corner of the economy, bring positive influence, the factor can not be failed to mention on that overall succession is the foreign investment flows Foreign investors not only buy listed stocks but also shifted aggressively to new listed shares, Initial Public Offering (IPO) or Mergers and Acquisitions (M&A) Sometimes in the past, trading with the foreign investors might affect the behavior of many domestic investors in some stocks and even the whole market The Vietnam stock market is expected to continue to attract investors from abroad, enhence competitiveness The role and influence of foreign portfolio investment in financial market and the stock market are under researched in many papers, especially in developing and developed stock market countries Foreign portfolio investment plays a very important role in providing liquidity for market, improving management ability, competition, the lucidity for market and lowering the cost of capital for enterprises.
According to the evidence from recession period, when the economy goes down, the value of collateral for the main asset will decrease, leading to the decrease in stock prices and firm value due to the disbelief of the investors and eventual bankruptcy is also possible So bussinesses need to pay attention to stock liquidity and the consideration of the liquidity in the firm valuation process is one the first steps that the appraiser must do Starting from these aspects above, the author chose the topic
“The relation between stock liquidity, foreign investor trading and firm value”.
Research objective
The objective of the topic is to analyze the impact of firm value, stock trading volume of foreign investors ạd stock liquidity on portfolio’s return in Vietnam market and their relation.
Research questions
To fulfill the goal of reasearch, this thesis’s mission is to answer these following questions:
- How is the relation between the stock trading volume of foreign investors, firm value and stock liquidity in portfolio investment?
- Besides the 3 factors in Fama-French models, does the trading volume of foreign investors factor, firm value factor and stock liquidity factor affect the portfolio’s return?
Research subject and range
The author uses the common stock data of non-financial companies, which is listed on Ho Chi Minh City Stock Exchange (HOSE).
The reason for choosing the period between 2015-2019 of the research is due to theCovid pandemic happened in 2020, which can lead to the less accuracy of statistic in stock market.
Methodology
The main research method used in this topic is to use the Fama – French model, combined with liquidity factor, company value factor and foreign investors's trading volume factor using data of non-financial company stocks listed on the Ho Chi Minh Stock Exchange.
Moreover, the author uses Ordinary Least Squares (OLS) regression, to estimate the model The author also uses meta-analysis method to review the related definition and theoretical basis Besides, to descibe the secondary data, the author uses descriptive statistics method and to serve the calculation and data extraction in this research, the author uses Excel 2019 and Stata software.
Research outline
The thesis outline has five chapters:
Chapter 1 (Introduction) presents the reasons of the author to research on this
4 topic Besides the research objective, research questions, subjects and scope, methodology and structure of this research to give the reader an overall picture of the whole thesis Chapter 2 (Literature review) presents the definition and theoretical basis related to stock liquidity, firm value and stock volume of foreign investors in the stock market This chapter provides some previous empirical research to clarify the urgency of the topic and a basis for choosing the research model and analyzing research results simultaneously.
Chapter 3 (Methodology of research) presents in detail the content of the research model, detailed description of collected data and research method process.
Chapter 4 (Empirical results) presents the results and the author discuss the empirical results of the model, the relationship among three main factors mentioned in the topic.
Chapter 5 (Conclusion and recommendations) depends on the results presented in
Chapter 4, presents the conclusion and suggests some petitions for the topic In addition, the thesis also gives the limitations and outstanding problems, from which proposing the next research directions.
LITERATURE REVIEW
Definition
Stock liquidity describes the degree to which a stock can be quickly bought or sold in the market The large volume of trading is a sign of high liquidity of the stock. (Amihud, 2002)
Liquidity refers to an asset is sold immediately after that asset is purchased without reducing the price and transaction costs It means whenever an investor considers investing in an asset, the investor will look carefully at the possibility of that asset to be resold again and whether the asset will incur transaction cost when resale or not. (Nguyen Anh Phong, 2012)
Liquidity is related to the valuation of an asset, representing the extent of an asset can be converted into money depends on the supply and demand of the market (Yeh et al., 2015)
There are many definitions of liquidity In general, the liquidity is the conversion into money of an asset depending on the supply and demand of that asset without incurring any transaction costs The faster the property can be converted into money, the higher liquidity and backwards.
A firm is an organization with its own name, assets, transaction office which was registered and established under the provisions of law for business purposes (National Asssembly, 2020) 1 Thus, a firm is an economic organization that gathers the resources of financial, physical and human to perform activities as investment, production and consumption of products to maximize the owner’s benefits or for social development goals.
The going concern firm value is the operating value of that business with the assumption that the business will continue to operate after the valuation date. (Ministry of Finance, 2017) 2
1 According to Clause 10, Article 4, Enterprise Law 2020
2 Vietnam Valuation Standard No 12 – Enterprise valuation promulgated together with Circular No
The value of a fixed-term operating firm is the value of an operating firm with the assumption that the life cycle of that firm is finite, because the firm is forced to dissolve after a specified time in the future (Ministry of Finance, 2017) 3
Therefore, firm value is understood as the total value of all assets (including tangible and intangible assets) that the firm owns, estimated at market value at the time of firm valuation Firm value can also be the total of the present value of future income that the firm can generate during its production and business process.
In empirical researches, foreign investments are generally measured as the proportion of foreign assets to total assets in domestic firm Foreign direct investments include long-term physical investments made by a company in a foreign country, such as opening plans or purchasing buildings Foreign investments in the form of purchasing sercurities and holding financial assets (stocks, bonds, mutual funds, exchange traded funds), which typically are known as foreign porfolio investments (FPI), this type of foreign investments is theoretically different, it does not participate in corporate management Therefore, foreign portfolio investments should be separately examined FPI is primarily a short-term investment and is an important source of international capital for many developing countries, but it also has more risk than the other sources, especially when there is a currency crisis, the capital flow is extremely sensitive to the general market and news.
The role of foreign investors has been an important topic In the perspectives of emerging markets, foreign investors are normally considered as sophisticated investors with well-equipped monitoring techniques A huge volume of papers analyzes the impact of trading behavior of different types of investors on stock prices Moreover, price and volume relationships are reported in previous studies to be instrumental in the assessment of their combined impact on stock market volatility – research of Chakraborty & Kakani (2016) In many aspects, the benefits
122/2017/TT-BTC dated November 15, 2017 of the Ministry of Finance.
3 Vietnam Valuation Standard No 12 – Enterprise valuation promulgated together with Circular No
122/2017/TT-BTC dated November 15, 2017 of the Ministry of Finance. from FPI are seen as: FPI makes the domestic capital market more liquid and lead to risk diversification, reduce the dependence of economy on the banking system; FPI flows improve quality of the market and product diversification with the appearance of high management experience and skillful foreign investors; Furthermore, FPI enables businesses to increase their operating capital Attracting foreign investors can carry out the industrialization of the country with limited capital mobilization.
In 1952, economist Harry Markowitz published an article titled "Portfolio Selection" in the journal Finance, Volume 7 No 1 He was the founder of modern investment theory Later, many asset pricing models based on his theory were researched, tested and developed Theory suggests that investors can minimize market risk at an expected rate of return by building a diversified portfolio.
The author's Markowitz portfolio theory is based on the following assumptions:
- Firstly, each investment choice is viewed as a probability distribution of the expected rate of return;
- Secondly, every investor always wants to maximize the expected utility and utility curve;
- Thirdly, the portfolio's risk is based on the variance of the rate of return;
- Fourthly, investors' decisions are based on expected return and risk. Therefore,
- he utility curve is a function of the expected return and the variance of the rate of return;
- Finally, investors will choose a higher rate of return if the same level of risk. Conversely, investors will choose lower risk if the same level of return.
A portfolio is considered efficient when it offers the lowest risk for a given expected return or the highest return for a given level of risk
In the financial field, the term return is understood as the percentage between the income generated by interest or the difference between the buying and selling price divided by the initial investment amount When investors hold shares in a listed company, they promise to earn profits or earnings per share (EPS) based on the company’s operating performance Thus, when a company is profitable, it means that its stock is rising in price, making it more attractive in the eyes of investors.However, the calculation of stock returns still has limitations such as in the case of businesses that can choose to buy back their shares to increase the EPS index in order to increase the share price and manipulate investors into thinking that the business is making good profits, Besides, in case the enterprise issues convertible securities such as convertible preferred shares or stock options, the real EPS will always be lower than the earnings per share In addition to the company's internal factors that are the main basis for stock valuation, there are also some external factors, such as disruptive information and investor behavior, which can also affect the rise and fall of the company's stock price.
Theoretical basis
The concept of liquidity in the stock market is very abstract and has many different definitions.
According to Black (1971), the market has liquidity appears in stock has these conditions:
- The market always has a bid and ask price for investors who want to buy or sell a small number of securities imediately.
- The difference between the bid and ask price (The spread) is always small.
- An investor who is selling or buying a large number of securities, in case of special information may expect to do so in a long period of time at a price that is not different from the average price compared with current market price.
- An investor can buy or sell a large volume of securities immediately but accompanied with a complement or discount depending on the volume of securities The higher the volume of securities, the greater the complement or discount.
A liquid market is a continuous market, it means most of the securities can be bought and sold instantly An efficiant market means a small number of securities can always be bought and sold at a price that close to the market price at that time It also means a certain number of securities can be bought and sold over a long period of time at an average price which is close to the current market price (Black, 1971). Pastor and Stambaugh (2003) argues that stock liquidity is a very broad and ambiguous concept, but in general it reflects the ability to trade large volumes of shares quickly at low fees and without price movement.
Harris found that a stock is considered to be liquid in the market when the stock has low spreads, large trading volume, short trading time, and the deviation of arbitrage from competition is rapidly corrected These components do not exist independently, but interact and influence each other, determining the liquidity of stocks Nguyen Phuong Loan (2018)
Brennan et al (1997) measures liquidity using volume data and shows that dollar volume traded in stocks has a negative effect on stock returns and it adds to the negative effect of size (market capitalization of shares) There are many different studies interested in trading volume, the relationship between trading volume and stock returns.
Lesmond, Ogden & Trzcinka (1999) introduced a measure of illiquidity as the ratio of days of unchanged price during the year The model presented in the paper is based on the reason investors are prevented from trading when transaction costs are high and effective transaction costs can be inferred from the frequency of non- trading days.
From the definitions of liquidity of the above researchers, the article summarizes the concept of stock liquidity, which is the ability to quickly and easily convert into cash on the stock market and the share price is relatively stable over time and has a high ability to recover capital as originally invested The liquidity of shares in the stock market is characterized by a large number of transactions.
2.2.2 The relationship between stock liquidity and firm value
In financial economics field, the relation between liquidity and firm performance has received considerable attention in many perspectives The effect of liquidity on performance as well as the dependence of liquidity on firm performance are both considered by the researchers The causative theories motivate many distinct mechanisms through which liquidity affects firm performance Most of the researches focus on the effect of liquidity on operating performance through agency- based theories The important theories in this relation include Maug (1998) – models a large relationship investor’s monitoring decision The investor monitors and trades with an aim to earn the profit from the price appreciation caused by his monitoring activities Maug concludes that liquid stock markets, far from being a hindrance to corporate control, it tends to support effective corporate governance Both agency- based and feedback-based causality focus on the impact of liquidity on firm performance.
However, liquidity can also affect firm value by changing the discount rate If investors value liquidity, illiquid stocks should trade at falling prices This implies a positive correlation between stock liquidity and market-based performance measures such as Tobin's Q Baker & Stein (2004) suggesting that liquidity may be related to valuation as an indicator sentiment In their model, highly liquid stocks are overvalued Since they trade at the premium, they have a lower expected future return In contrast to the agency-based causative theories, Subrahmanyam and Titman (2001) and Khanna and Sonti (2004) show liquidity can positively affect firm performance even when agency conflicts do not exist In this case, the liquidity stimulates the entry of informed investors who make prices more informative to shareholder Khanna and Sonti (2004) showed that informed traders is known as a factor that effect their trades on managerial behaviour into their trading strategy, more aggressively and thus making prices more informative This effect improves the operating performance and relaxes the financial constraints Both effects increase firm performance.
Furthermore, decision to stay or go of shareholders affects the firm cash flows This is particularly valuable when the relationship between stakeholders and the firm is fragile or there is high cash flow uncertainty in existing projects These theories imply that the effect of liquidity is positive to the sensitivity of firm operations to the information content of stock prices.
Fang et al (2009) find a positive effect of stock liquidity on firm value, which can be attributed to more informative stock prices and better management incentives. However, there is no evidence that stock liquidity influences the value of the company through external control and monitoring channels The following points may explain the lack of consensus on the effects of stock liquidity on firm value in the empirical literatures on corporate finance: The first is an endogeneity problem, particularly the unobservable characteristics of the company that affect the stock liquidity and firm value; The second is the heterogeneity in the relationship between stock liquidity and firm value across different corporates or industries; and the third is identify the specific mechanism underlying the impact of stock liquidity on firm value.
In conclusion, causative theories are either operating-performance-based, showed that liquidity affects operating performance, or pricing-based, the performance proceeds from an illiquidity premium or mispricing Operating performance theories can be devided into agency or feedback theories Moreover, the relation between liquidity and performance might not be based on a causal effect from liquidity First,liquidity may simply be correlated with other variables that affect firm value.Second, a strong case can be made for liquidity being the dependent variable in the liquidity/performance relation rather than the independent variable The supporting dependent liquidity is that high performance firms will have high book ratos on market and that attracts institutional investors Such trades increase market depth and increase stock liquidity Therefore, high performance firm generates liquidity by create institutional investor demand Under this theory, the relationship between liquidity and performance should be promoted because high performance firm may be attractive to institutional investors.
2.2.3 The relationship between foreign investments and stock liquidity
Tesar and Werner (1995) and Vagias and Van Dijk (2010) find that an increase in foreign holdings, estimated from accumulated capital inflows, improves local stock market liquidity, and Wei (2010) also reached the same conclusion using data held by foreign institutional investors.
Goldstein and Razin (2006) showed in their theoretical model that there is a trade-off between foreign direct investment (FDI) and FPI and between management efficiency and liquidity Both compromises are based on information mismatch as the reason Foreign direct investors have ownership and management positions in domestic companies and therefore have access to the company’s private information, enabling them to supervise management activities effectively However, their access to inside information will incur liquidity costs related to the impact of their trading prices In contrast, Foreign Portfolio Investors acquire ownership without controlling domestic companies and expand the company’s shareholder base, thereby increasing liquidity through trading activities.
Lilian Ng et al (2015) studied the effect of foreign investors on stock liquidity around the world, their study provided evidence that foreign investors influence the liquidity of stocks through trading and information channels However, the presence of foreign portfolio ownership has increased the stock liquidity of domestic companies through increased trading activities Foreign Portfolio Investors reduce the degree of information asymmetry, thereby increasing the liquidity of stocks. Foreign investors have a larger influence on stock liquidity than their domestic peers, with their liquidity effects varying with firm exposures to global markets Stock liquidity decreases with foreign direct ownership but increases with foreign portfolio ownership.
Overview of previous empirical researches
Amihud and Mendelson (1986) developed a pioneering study examining the role of illiquidity/liquidity in asset pricing, using the bid-ask spread represent for illiquidity. They found a positive correlation between expected return and illiquidity Then Eleswarapu and Reinganum (1993) re-examined Amihud and Mendelson's study with a more up-to-date time period and showed that the positive relationship between illiquidity and returns is only limited in January.
Peterson and Fialkowski (1994) and Brennan and Subrahmanyam (1996) raised the concern that the bid-ask spread is a representation for liquidity This has led to the use of other measures of liquidity such as trading volume (Brennan et al., 1997), turnover ratio (Datar et al., 1998; Chan and Faff, 2005), the standard deviation of the return rate, of the trading volume, and the regression coefficient of the volatility of the income ratio and of the trading volume (Chordia et al., 2001), a measure of liquidity of Pastor and Stambaugh (2003) (Ho and Hung, 2009), measure of illiquidity of Amihud (2002) (Lagoarde-Segot, 2009) and Liu's liquidity ratio (2006). Generally, these studies support the findings of Amihud and Mendelson (1986). Nguyen et al (2007), studied whether the use of the 3moment CAPM capital asset pricing model can explain liquidity risk They also compared the four-factor model based on Fama - French and Pastor Stambaugh factor with the model based on stock characteristics The results indicate that there is a liquidity premium, and these models do not capture the liquidity premium On the other hand, it does not indicate that the properties of securities represent for liquidity Using the rate of return as a proxy for liquidity, the author also found that securities with a low rate of return (less liquidity) require a higher rate of return than other securities with higher rate of return (more liquidity) The strength of this relationship is not affected by the existence of other variables The results of this paper confirm the results of the separate effects of Amihud and Mendelson on liquidity and are not influenced by other explanatory factors of stock returns The sensitivity of stock returns to fluctuations in market liquidity does not depend on specific liquidity effects.
Model of Amihud (2002) used daily data of stock returns and dollar trading volume to measure the illiquidity of sercurities However, the data used in Amihud’s model
(2002) is just available in developed markets and not in emerging markets as Taiwan To overcome the problem of short time series in data, the model of Bekaert et al (2007) used the observed percentage of daily average returns over the month to measure the liquidity for 19 emerging stock markets.
The model of Pastor and Stambaugh (2003) used the price reversal coeficient to measure liquidity, due to the relationship of temporary price and trading volume in line The measure assumes that low liquidity is caused by the reversal in rate of returns of the higher trading volume The measurement method is proposed by model of Roll (1984), price arises from reality is return price through a transaction between bid and ask prices Therefore, the volatility change in price covariance provides a simple method to measure the liquidity Then, the model of Bao et al
(2011) provides a method to measure liquidity, summarized from investment bonds using method of Roll (1984) and examines the low liquidity effects of price.
Mona Al-Mwalla (2017) tested whether book-to-market ratio, size and trend could explain the stock return in emerging market The author tested the Fama-French three factors model and the Fama-French four factors model (which is Fama-French three factors model and the trend factor) at the Amman - ASE stock exchange in Jordan Data were collected according to the monthly return of each security for the period from June 1999 to June 2010 The number of companies varied from 114 stock samples in 1999 to 205 stock samples in 2010 The author divided the portfolio by size factor (small – S; medium – M; large – B) and BE/ ME ratio (high – H; medium– M; low – L), forming 9 portfolios: S/L, S/M, S/H, M/L, M/M, M/H, B/L, B/M, B/
H The results show that the Fama-French three-factor model better explains stock return volatility than the Fama-French four-factor model.
In the US, in the study "Testing the CAPM model and the Fama French three-factor model" in 2004, author Nima Billou compared and tested the effectiveness of the two models The effectiveness of the model will be compared based on a and the mean of the absolute value of a In a 1996 study, two authors, Fama and French, proved that the Fama French three-factor model is better than the CAPM due to its smaller a This study uses 25 stock portfolios divided by size and value that Fama and French used to retest to see if the model is still effective over a longer period of time Research data is taken from Ken French's website, which gathers all stocks from 3 major stock exchanges in the US, NYSE, AMEX and NASDAQ With the research period from 7/1963 to 12/2003, a of CAPM = 0.3, a of Fama French = 0.13, besides with 95% confidence, R2 of CAPM model is 72% and R2 of model 3 Fama French factors are 89%, proving that the Fama French model is still more effective than the CAPM model After Nima Billou expanded the sample from 7/1926 to 12/2003, a of CAPM = 0.23, a of Fama French = 0.19, R2 of CAPM was 77% and R2 of Fama French was 88% The results show that the two factors of size and value are very influential in the US stock market, so the Fama French model still proves to be more effective than the CAPM model in explaining stock returns.
In India, the CAPM and Fama French Models were studied by two authors, Gregory Connor and Sanjay Sehgal, under the title "Testing of Fama and French Models in India" This study shows that the Fama French three-factor model is only suitable in two of the three results compared to the US stock market, there are: Firstly, the market, size and value factors exist commonly to explain stock returns Secondly, there is a linear relationship between stocks and the above factors in explaining the dispersion of average returns Whereas market, size, and value factors do not universally affect earnings growth rates, and therefore do not affect stock returns,this is in contrast to the US stock market This paper takes data from the month-end returns of 364 stocks from June 1989 to March 1999 The average R2 in the Fama
French model is 84.22%, while the R2 of the CAPM model is only 75% This study shows that running the linear regression of these two models can explain and predict the returns of securities and portfolio of securities in stock market in India With this level of significance, investors can consider and apply these two models to make securities trading more effective.
But in contrast, in Korea, Kyong Shik Eom and Jong-Ho Park ran a linear regression model with data from 868 stocks during the period from July 1981 to December
2007 with the research topic "Equal evidence of the three-factor model in Korea”. This group of authors shows that during the period (1984-1994) the Fama French three- factor model is not suitable to explain the returns of securities, only the CAPM model is suitable In addition, the study shows that the Fama French model is only suitable for forecasting returns in the Korean stock market for a short period of time but not for a long period of time (26.5 years) This result indicates that the R2 of the three- factor model is 2.52% with a confidence interval of 99% This low level of significance shows that market variables, size, and value do not explain and predict the returns of a security and its portfolio In addition, the authors also extended the model and found that the main factors affecting the model to explain the rate of return are liquidity, information disclosure and spread of credit.
The research paper (2010) by Master of Business Administration Chun-Wei Huang titled "Applying the three-factor model CAPM and Fama French to the Taiwan stock market" Chun-Wei Huang shows that the CAPM model can be applied to the Taiwan stock market because market risk factors have a strong influence on stock returns The Fama French model has only two variables of market risk and size that are statistically significant, while the value variable is less significant in explaining the rate of return Therefore, Fama French's three-factor model cannot be fully applied in this country's stock market The results of the regression model are as follows, the R2 of the CAPM model is 55.8% (99% confidence), the Fama French model is 69.9%(95% confidence) This study only uses 90 stocks while there are about 700 in the market, so it is not representative of the entire market Not only that, according to Chun-Wei Huang, the market risk factor is affected by the impact of politics and diplomacy The size factor is influenced by market risk The value factor is influenced by the company's financial statements.
Lam and Tam (2011) study “Liquidity and Asset Valuation: Evidence on the Hong Kong Stock Market” with a dataset of 769 companies listed on the Hong Kong Stock Exchange in the period 7/1981 – 6/ 2004 Using the research model of Fama – French including 3 factors and adding the liquidity factor represented by the ratio of trading volume divided by the volume of shares outstanding After controlling for the determinants of stock returns, the author discovers a negative correlation between liquidity and stock illiquidity risk premium and asserts that liquidity is an important factor for valuing stock returns on Hong Kong stock market.
2.3.2 Previous empirical researches in Vietnam
Thus, through studying and applying CAPM and Fama-French 3 factors models in developed countries and in emerging market countries, we see that in order for the CAPM and Fama French models to be successfully applied in Vietnam, it is necessary to have a sample time The study is long enough and the number of securities listed on the exchange in the sample is relatively large Particularly for the Vietnamese market, some typical studies use multi-factor models to determine the rate of return on stock are also done For example, research by Vuong Duc Hoang Quan & Ho Thi Hue (2008) on a 3-core Fama-French model element The study points out the impact of market factors, stock returns It is also affected by business characteristics such as company size, BE/ME ratio To speak in other words, the factors of the 3-factor Fama-French model all affect the rate stock returns in Vietnam.
Tran Thi Hai Ly (2010) tests the Fama - French model, to see how the model works in the Vietnamese stock market The number of companies in the sample varied from
25 companies on December 31, 2004 to 136 companies on December 31, 2007 The author builds into 4 investment portfolios: S/L, S/H, B/L, B/H The portfolios are a combination of the size factor (small – S; large – B) and the BE/ME ratio (high – H; low – L) Regression results show that the CAPM model has an adjusted R2 from 0.83 to 0.90 But the Fama-French 3-factor model with adjusted R2 is all higher than 0.92 The author also replaces the scale factor with the state ownership factor to test in the Vietnamese market The results show that the three-factor model adjusted for the Vietnamese market better explains the changes in the returns of the portfolios, the adjusted R2 is very high from 97.7% or more.
Nguyen Thu Hang and Nguyen Manh Hiep (2012) tested the Fama-French model in Vietnam's stock market, adding a clearer and more complete explanation of the Fama- French model to the volatility of stock returns on the Vietnamese market The article used weekly returns of 68 stocks in 2007 to 235 stocks in 2012, listed on HOSE The author tested the model in two periods: the first was from July 2007 to March 26, 2008 and the second was from August 18, 2008 to June 2012 The portfolios were divided based on the size factor (small – S; medium – M; large – B) and the BE/ME ratio (high – H; medium – M; low – L), forming 9 portfolios: SL,
SM, SH, ML, MM, MH, BL, BM, BH Regression results show that, in the first period when the market was growing, the Fama-French model explains well the volatility of stock returns in the Vietnamese market, with the adjusted R-squared from 0.55 to 0.99, while the CAPM model was only from 0.34 to 0.98 For the second period, when the market was in recession, the Fama-French model was still not completely explainable as in the first period, but the adjusted R-squared was from 0.66 to 0.96, still higher than the CAPM model which was only from 0.47 to 0.94.
METHODOLOGY OF RESEARCH
Research process
Step 1: From the research objective, the author identified issue of the research and determined the appropriate research method;
Step 2: The author reviewed and summarized the previous related empirical researches and theories;
Step 3: Based on scientific theories and related empirical researches, the author identified the research model Then, the author chose an appropriate research sample, collected data and identified the variables in the research model;
Step 4: From the collected data, the author analysed data according to the specific quantitative methods to select the appropriate result;
Step 5: From the research results, the author made conclusions and recommendations on the topic.
Research model
After researching in previous studies, the author found that currently to measure stock returns, there are two famous models, the CAPM model and the 3 factors Fama
- French model The CAPM has been applied by some authors in the previous researches, but the results show that there are still many shortcomings and do not explain the influence of factors on stock returns Therefore, it can be seen that theFama – French model is an effective solution to the current time when it is added with two more company specific factors, namely the size of the business and the ratio of book value to market value Besides, the liquidity of stocks has also been mentioned in many previous studies on many different markets Firm value and foreign trading volume are also factors that many investors are interested in when investing today.
To analyse the relation of firm value, foreign trading and stock liquidity, the research is based on the Fama French 3 factors model to build the porfolio and calculate the SMB and HML factors, completed with foreign portfolio investment factor and firm value factor The research model is presented as below:
R tt -R f t = a t + p t (R mt -R ft ) + P is (SMB) t + MfiML) t + fl if (FV) t +
Pi.ẶQV) + p tt (Liq) c + ■„ (3.1) where:
R it : is the return rate of the portfolio i; a i : is the constant intercept;
(R mt — R ft ): is the excess market return;
SMB t (Small minus Big): is the size risk factor;
(HML) t (High minus Low): is the value risk factor;
(PV): is the foreign trading factor;
(ỌV)t: is the firm value factor;
(Liq): is the stock liquidity factor; £ it : is the error term of porfolio i at time t;
0 i , P is , 0 ih , P if , P iq , P ii : are the sensibility coefficients.
Data of research
The data used by the author in the research is collected from the data of non- financial companies listed on the Ho Chi Minh Stock Exchange from 2015 to 2019
The selected companies must operate in a stable and sustainable manner The reason why the author excludes companies with interrupted transactions and negative book values is because it affects the liquidity of stocks, and because of the particularity of these organizations, we exclude banks and financial institutions Finally, the data used in this study are data from 63 companies from January 2015 to December 2019. The author forms investment portfolios according to the way of portfolio division in the study of the 3-factor model of Fama & French (1993) The details are as follows: According to market capitalization (MC) at the time of portfolio establishment, companies are divided into 2 groups If a company's market capitalization is less than or equal to 50% of the overall market median, then that company is classified as small (S) and vice versa, big companies (B) How to calculate the market capitalization value in the author's research paper is as follows:
MC = p t x N t (3.2) Where: p t : is the share price at time t;
N t : is the number of outstading shares at time t.
The book-to-market (BE/ME) ratio of each stock is determined using the ratio of the book value of equity to firm size Stocks are arranged by ascending BE/ME: 1/3 of the stocks with the smallest BE/ME are placed in the low BE/ME group (L), the next 1/3 of stocks are in the medium BE/ME group (M) and 1/3 of the stocks with the highest BE/ME ratio are placed in the high BE/ME group (H).
Thus, in each quarter (quarter t), the companies are classified according to two above factors and divided into portfolios (6 categories) as follows:
SH Portfolio: Stocks with small size firm and high BE/ME ratio;
SM Portfolio: Stocks with small size firm and medium BE/ME ratio;
SL Portfolio: Stocks with small size firm and low BE/ME ratio;
BH portfolio: Stocks with big size firm and high BE/ME ratio;
BM Portfolio: Stocks with big size firm and medium BE/ME ratio;
BL portfolio: Stocks with big size firm and low BE/ME ratio.
Table 3.1: BE/ME ratio and Market Capitalization combination portfolios
BE/ME ratio Low (L) Medium (M) High (H) Market Small (S) SL porfolio SM porfolio SH porfolio
Capitalization Big (B) BL porfolio BM porfolio BH porfolio Source: Data collected by the author
To the porfolio that combined between the firm value and foreign porfolio investments, the author uses the same way with portfolio division in the study of the 3-factor model of Fama & French (1993) After having the result of Q and F, the two variables are calculated by the formula (3.5) and formula (3.6), according to normalized foreign trading volume (F) at the time of portfolio establishment, companies are divided into 2 groups If foreign trading volume is 50% higher than the market median will be classified as high (H) and vice versa, low (L) The same with firm value (Q), stocks are arranged by ascending firm value: 1/3 of the stocks with the smallest firm value are placed in the small value group (S), the next 1/3 of stocks are in the medium firm value group (M) and 1/3 of the stocks with the biggest firm value are placed in the big value group (B) Finally, in each quarter (quarter t), the companies are classified according to two above factors and divided into portfolios (6 categories) as follows:
LS Portfolio: Stocks with low foreign trading volume and small firm value;
LM Portfolio: Stocks with low foreign trading volume and medium firm value;
LB Portfolio: Stocks with low foreign trading volume and big firm value;
HS portfolio: Stocks with high foreign trading volume and small firm value;
HM Portfolio: Stocks with high foreign trading volume and medium firm value;
HB portfolio: Stocks with high foreign trading volume and big firm value.
Table 3.2 Firm value and Normalized foreign trading volume combination portfolios Firm value
Low (L) LS porfolio LM porfolio foreign LB porfolio trading
High (H) HS porfolio HM porfolio volume HB porfolio
Source: Data collected by the author
Measurement of variables
Stock return R it : is the rate of return of a stock for a quarter, calculated as the average of the stock's daily returns of that quarter The formula for calculating daily rate of return is shown below:
R it : is the daily rate of return of the stock at day t;
P t : is the close price of the stock at day t;
Pt-1: is the close price of the stock at day t-1.
Market return R mt : is measured based on the VN-Index collected on the Ho Chi Minh Stock Exchange
R mt : is the market return at quarter t;
V N Index t : is the VN-Index at quarter t;
VNIndex t-1 : is the VN-Index at quarter t-1.
Risk-free rate of return R f t : equal to the interest rate of 5-year Government bonds converted at quarterly interest rate from January 2015 to December 2019.
Firm value (Ọ)t : for firm value, the author uses Tobin's Q variable calculated according to the following formula:
Market value of equity xBook value of liabilities
Book value of total assets
When Market value of equity is calculated by the multiplication of close price and number of outstanding shares at that quarter.
Foreign trading: Transactions of foreign investors have been considered in many studies, and net foreign capital inflows are calculated as total purchases minus total foreign sales, then net foreign inflows are normalized by divided by the current market capitalization as in the studies of Bekaert et al (2002), Dahlquist and Robertsson (2004), Griffin et al (2004) Richards (2005) and ĩlkỹ and İkizlerli
(2012) This normalization allows for market-to-market comparison and is also useful in showing the magnitude of the need for comparison between net foreign capital inflows and total stock supply.
Stock liquidity: According to Brennan et al (1997) measures liquidity using volume data and shows that stock trading volume has a negative impact on stock returns. This study is consistent with the author's expectation, so the author divides the stock group based on liquidity using the volume of shares traded on the Ho Chi Minh Stock Market The author divides them as follows: 1/3 of the stocks have low trading volume will be put into the low liquidity group (LL), the next 1/3 of stocks will be
F = ^ buy MC ^ sell (3.6) put into the middle liquidity group (LM) and 1/3 of the stocks with the largest trading volume will be put into the high liquidity group (LH).
Size risk factor (SMB) and value risk factor (HML) measure method
On the basis of the 3-factor model of Fama - French, in this study, the SMB factor is calculated as the average return of the small-sized portfolio (portfolio S) minus the average return of the large-sized portfolio (portfolio B).
SMB — SL + SM + SH BL + BM + BH
Similarly, the HML factor is calculated as the average return of the high BE/ME portfolio (portfolio H) minus the average return of the low BE/ME portfolio (portfolio L).
Firm value factor (QV) and foreign trading factor (FV) measure method
On the basis of the 3-factor model of Fama – French, in this study, FV factor and
QV factor is calculated base on SMB and HML factor The FV factor is calculated as the average return of the Low portfolio (portfolio L) minus the average return of the High portfolio (portfolio H)
LB + LM + LS HB + HM + HS _
Similarly, the QV factor is calculated as the average return of the Big portfolio (portfolio B) minus the average return of the Small portfolio (portfolio S).
Stock liquidity factor (Liq) measure method
The Liq factor is calculated as the average return of the low liquid portfolio (LL) minus the average return of the high liquid portfolio (LH).
HML _ SH + BH SL + BL
Description of research data
The research sample includes 63 companies, including sectors such as Real Estate and Construction, Materials, Agriculture, Consumer Goods, Services during the period 2015-2019 listed on Ho Chi Minh City Stock Exchange (HOSE) The author extracts experimental research data for each variable through the Table 3.3.
Table 3.3: Source of experimental research data
Book value of total assets
Book value of liabilities Quarterly data is extracted
Close price from Thomson Reuters
Quarterly data is extracted from website Fiintrade
Link: fiintrade.vn Quarterly extracted from website Cafef
HML Book value of equity
Link: Cafef.vn Quarterly data is extracted
Rm VN-Index from Thomson Reuters Quarterly data is extracted from website Fiintrade
Rf Interest rate of 5-year Government
Link: fiintrade.vn Quarterly data is extracted from website Investing bonds
Source: Data collected by the author
Testing method
In the first step, the author uses descriptive statistical methods to preliminarily analyze the properties of the research sample such as: maximum value, minimum value, mean value and standard deviation, combined with using Correlation matrix to test the correlation between the independent variables in the model.
The author establishes 12 OLS regression equations based on the research model of Fama – French 3 factors combined with stock liquidity, firm value factor and foreign trading volume factor The author uses the Ordinary least squares regression method
- the most widely used method to estimate the parameters in the regression equation. The simplest approach is to assume that the regression coefficients (intercept and slope) are constant between observations and do not change over time.
In the study, the author uses the following tests:
If the time series do not show any trend or seasonal effects, they are stationary The summary statistics calculated from the time series remain consistent over time, such as the mean or variance of observations When the time series is stationary, modeling is easier Augmented Dickey – Fuller unit-root test is applied in this research to test the stationary of the models It is a general statistical test used to test whether a particular time series is stationary, one of the most commonly used statistical tests when analyzing the stationarity of a series.
Multicollinearity occurs when the independent variables are correlated, causing the sign of the regression coefficient to be reversed or the independent variables to lose statistical significance To test this phenomenon, the author uses the variance magnification factor VIF Multicollinearity occurs when one of the VIFs found is
The Breusch - Pagan test is a test for heteroscedasticity of regression errors.Heteroskedasticity means "different distribution” Homoscedasticity in regression is an important assumption, if the assumption exists, regression analysis cannot be used With hypothesis H0: there is a phenomenon of Heteroskedasticity If the resulting P- value is greater than 5%, the null hypothesis H0 is rejected.
Autocorrelation is intended to measure the relationship between a variable's present value and any past values due to the inertia of the time series With hypothesis H0:there is a phenomenon of Autocorrelation If the resulting P-value is greater than 5%,the null hypothesis H0 is rejected The author uses the first-difference transformation method of the portfolio's excess return variable into the regression model to eliminate this phenomenon in some portfolio regressions.
In chapter 3, the author presented the research process to summarize the steps to carry out the topic The author also presented the data used in the research, the measurement of each variable, the source of data and the specific research methods besides the testing method in model regression to solve the objectives of the topic.The contents presented in chapter 3 will be the basis for the author to present research results and conclusions in the next two chapters.
EMPIRICAL RESULTS
Descriptive statistics
Table 4.1: Descriptive statistics of independent variables
Variable Mean Standard deviation Minimum Maximum
Source: data collected by the author and calculated on Stata software
Table 4.1 summarizes the basic parameters of research data, specifically the independent variables, including mean, standard deviation, minimum and maximum values The average results of HML, Liq and QV are positive, indicating that investors focus on companies with high BE/ME ratios, high stock liquidity and high firm value Otherwise, the investors pay less attention on firm size and foreign net trading volume The average excess return value of the market is -4.797232 showing that the stock market situation in the research period January 2015 – December 2019 has a downward trend.
Correlation matrix
Table 4.2: Correlation matrix of independent variables
Rm-Rf SMB HML QV FV Liq
Source: data collected by the author and calculated on Stata software
Correlation analysis provides an overview of the relationship between the research variables It can be seen from the matrix of correlation coefficients that the correlation coefficients are all less than 0.6, so the possibility of multicollinearity is relatively small when the estimation method is used in the research model This is the same with the result in Table 4.3, where the author finds that there is no multicollinearity problem in the model that will affect the regression results Most of the model variables are positively correlated, except variable F is negatively correlated with variable Rm-Rf and SMB, this is understandable in the case of a down market, foreign investors are more careful in investing in stocks Besides, the results show that liquidity has a negative correlation with the market's return The negative correlation between stock liquidity (Liq) and Rm-Rf value is also presented in the study of Lam and Tam (2011) Moreover, the Lig factor also has negative correlation with the size and value factor.
Regression analysis result
Before performing model regression, the author performs multicollinearity test and stationarity test of each model with portfolios as dependent variable To test multicollinearity phenomenon, the author uses the variance inflation factor VIF The results after using the test are presented as Table 4.3.
Table 4.3: VIF coefficients of the variables in the research model
Variable VIF SQRT VIF Tolerance R-
Source: data collected by the author and calculated on Stata software
In this table, the author shows the association between measurement variables After measuring the VIF, the result shows that the highest VIF is 1.90 and the average VIF is 1.50 This means that multicollinearity problems have no effect on the regression results.
Table 4.4 Stationary test results for 12 porfolios and independent variables Porfolio T - Statistic Porfolio T - Statistic Variable T - Statistic
Source: data collected by the author and calculated on Stata software
The results of the portfolio test show that the absolute values of the t-statistics are larger than the absolute values of the rejected values at the 1%, 5% and 10% significance levels Therefore, the author asserts that the value chain of the portfolios in the model is stationary Similarly for the independent variables Ri-Rf, SMB, HML, QV, FV and Liq also results in a stationary series.
Table 4.5 The results of the test of Heteroskedasticity for 12 porfolios
Portfolio Prob>Chi2 Portfolio Prob>Chi2
Source: data collected by the author and calculated on Stata software
The author performs Heteroskedasticity test for each portfolio group, the results show that the values of Prob > chi2 are all greater than 5%, which shows that there is no Heteroskedasticity in the model.
Table 4.6 The results of the test of Autocorrelation for 12 porfolios
Portfoli o Prob>Chi2 Portfolio Prob>Chi2
Source: data collected by the author and calculated on Stata software
Table 4.6 presents the results of the Autocorrelation test, showing that the values of Prob > chi2 are greater than 5% except SM, LM and HS portfolio, which shows that there is Autocorrelation phenomenon in these portfolios Therefore, the author uses the first-difference transformation method of the portfolio's excess return variable into the regression model to eliminate this phenomenon The Prob>chi2 after using this method of SM portfolio is 0.7987, LM portfolio is 0.2781 and HS portfolio is 0.2733, greater than 5% significance level.
After the regression, the author performed the VIF multicollinearity test for the independent variables in the model, the results showed that the independent variables in the model did not have multicollinearity phenomenon To consider the phenomenon of Heteroskedasticity of the model, the author performed BreuschPagan Test found that there is no Heteroskedasticity in the model Autocorrelation test results showed that there are 3 portfolios have Autocorrelation in the model.After using the first-difference transformation method, the phenomenon is no longer exist in the model The regression results in Table 4.7 and Table 4.8 are the results after overcoming the above violations The author has a 5-factor regression (SMB,HML, QV, FV, Liq) for a group of 12 portfolios classified by company size -BE/ME and firm value - trading volume of foreign investors Regression results of 6 portfolios all give Adjusted R-squared results ranging from 0.2530 to 0.7866.
Table 4.7 Regression results of the model for 6 portfolios of BE/ME ratio and firm size
Port Rm-Rf SMB HML QV FV Liq
Source: data collected by the author and calculated on Stata software
We can see from the Table 4.7, the SL portfolio – small size firm and low BE/ME ratio portfolio model is significant below 10% significance level With the regression results of the portfolio, the market risk premium, the value factor and the size factor have the impact on the return of the stocks included in the portfolio at a significance level of less than 5% In which, the size factor has a positive impact on the return of the portfolio with a high coefficient (.792540), while the value factor and the market factor have a negative impact on the stock return with coefficients are (-.925894) and (-.000711), this result is similar to the study of Nguyen Thanh Phong (2012), where the market variable is negatively correlated and the regression value is very small for the return of the portfolio The company value factor and the foreign investment factor, the liquidity factor in this model have no impact on the return of stocks in the portfolio.
In general, the portfolios in the article have a negative correlation between the market risk premium and the returns of stocks In which, the risk premium of 2 portfolios SM and BM (small size firm with medium BE/ME ratio and big size firm with medium BE/ME ratio) is significant at 10% level, while BH portfolio (big size firm with high BE/ME ratio) is significant at 5% level All regression coefficients of the SMB variable tend to decrease as the firm size increases, which is similar to the results presented in the study of Lam and Tam (2011) However, the P-value of this variable is only significant at 5% in SL portfolio and 10% in BL portfolio.
In the portfolio model with a combination of BE/ME ratio and firm size in this research, it seems that the firm value factor and the foreign investment factor have no effect on the profitability of stocks in the portfolio For stock liquidity factor, the results show that liquidity has a negative impact on stock return in portfolio SM (small size firm with medium BE/ME ratio) at 5% significance level In this model,
SM portfolio after regression has significance level of 0.0004 (smaller than 1% of significance level) and has Adjusted R-squared up to 0.7866 In the BL portfolio (big size firm with low BE/ME ratio), stock liquidity also has a negative correlation with stock returns at the 10% significance level The negative correlation coefficient of the liquidity factor appears in all 6 portfolios This result is consistent with the author's initial expectation, that stock trading volume has a negative impact on stock return.
Table 4.8 Regression results of the model for 6 portfolios of foreign trading volume and firm value
Port Rm-Rf SMB HML QV FV Liq
Source: data collected by the author and calculated on Stata software
The results are similar to those shown in Table 4.7, market risk premium is negatively correlated with the return on stocks in the 6 portfolios combining by firm value factor and foreign trading factor which shown on Table 4.8, the significance levels are below 1% and 5% It can be seen that investors are only interested in firm value more than the volume of foreign transactions This is shown by the positive correlation coefficient of the variable QV to the LB and HB portfolios with p-values that are significant below 5% and 10% and the correlation coefficients are 5745966 and 4669322, respectively When considering the coefficients of the FV regression results, the author also found that there is a positive correlation between foreign investor trading volume and stock return in these 6 portfolios, these coefficients tend to increase as firm value increases with 2 portfolios (HS and HM) has a significance level of less than 1% and 5%.
Regression results for 6 portfolios all show a fairly high Adjusted R-squared value from 0.3975 to 0.6303 In this research, when adding 3 variables QV and FV, Liq to the 3-factor Fama French model, there are no statistically significant results in 3 variables SMB, HML and Liq for the return of stocks in 6 portfolios combined between firm value and foreign trading volume The HML and Liq factors in these portfolios don't tend to be obvious Although the results of Liq factor is the same with the results in the regression of 6 portfolios built on firm size and BE/ME ratio, showing there are negative coeficients between Liq and 6 portfolios.
Generally, in all 12 portfolios, when building portfolios by combining firm value and foreign trading or combining BE/ME ratio and firm size, most of the regression models give P-value significance results are less than 5% and Adjusted R-squared values are quite high Especially in portfolios that combined by firm value and foreign trading volume, there are 2 portfolios with significance below 1% and 3 portfolios with significance below 5% This shows that the Fama French model adds firm value, stock liquidity and foreign trading factors have a good explanation for the volatility of return in investment portfolios.
In 12 portfolios, most of the regression coefficients of the risk premium factor are negatively correlated and have significance below 5% The enterprise value variable has a positive effect on stock return, especially significant in stocks with a high amount of foreign investment and tends to increase when foreign trading increases.Besides, the volume of foreign investment has a positive effect on stock returns in11/12 stocks, especially significant in stocks with small and medium firm value, tends to increase when firm value increases and tends to decrease when BE/ME ratio increases Regression results also show that stock liquidity has a negative impact on stock return in all 12 portfolios, in the portfolio of companies of the same size, the portfolio of stocks with medium BE/ME ratio is affected by the larger liquidity factor compared to the stock portfolio with low BE/ME ratio and the liquidity factor has no effect on the stock portfolio with high BE/ME ratio The study also shows that market risk factors have a strong influence on stock returns, no matter what criteria portfolios are built on.
In chapter 04, the author presents the results of empirical research on the 3 factors:stock liquidity, firm value, transactions of foreign investors which are added in Fama– French 3 factors model in chapter 4 Firstly, the author performs describe the independent variables in the research models including the meaning, standard deviation, maximum value, minimum value and correlation statistics between independent variables by correlation coefficient matrix Then, the author implements the OLS linear regression method to regress the model The author uses AugmentedDickey – Fuller unit-root test to check the stationary of variables in the model Then,the author checks the model's defects including multicollinearity test by the Variance inflation factor (VIF) measure and check the phenomenon of Heteroskedasticity andAutocorrelation of the model For some models suffering from Autocorrelation, the author uses the first-difference transformation method to overcome this phenomenon Finally, the author discusses the results of the chosen estimation method.
CONCLUSION AND RECOMMENDATIONS
Summary of main research results and conclusions
The study relies on the 3-factor Fama - French model to add to the 3-factor model
QV, FV and Liq, in which variables QV and FV are built similar to 2 factors SMB and HML, after building factors the author get 12 portfolios The sign statistical results of the three-factor regression coefficient added to the model are summarized in Table 5.1.
Table 5.1 Sign statistical results of the QV, FV, Liq regression coefficients of 12 portfolios
BH BL BM SH SL SM
LS LM LB HS HM HB
Source: data collected by the author and calculated on Stata software
Based on Table 5.1 and previous analysis, it shows that liquidity has a negative correlation with the stock return on the Vietnamese stock market Specifically,liquidity is negatively correlated with stock returns in all portfolios in the model.This correlation is especially clear and significant for those portfolios built on two factors BE/ME ratio and firm size, companies with medium and low BE/ME ratio.Stock liquidity does not show a significant correlation for portfolios formed from the combination of firm value and foreign trading volume Even though the correlation coefficients in these 6 portfolios also show negative results as BE/ME ratio and firm size portfolios and the author's initial expectation, the Pvalue and Adjusted R- squared values also give the model results of high significance Therefore, it can be concluded that stock liquidity is not affected by the value of the company and the transactions of foreign investors.
When considering the foreign trading volume factor, the FV variable is mostly positively correlated with the return of stocks in the portfolio In more detail, there are 2 portfolios with significant positive correlation (the HS portfolio and the HM portfolio) This occurs clearly with explanatory significance in portfolios with high foreign investment volume and small and medium firm value.
In the firm value variable, when considering the correlation coefficient of the QV variable with the stock's return, the author finds that firm value has a positive effect on stocks in the list of stocks with big firm value, regardless of whether the factor of foreign investment is high or low It can be seen that stock investors in Vietnam are interested in firm value, there are companies with great value that often receive more attention and investment from investors, while in companies with medium and small value, investors will pay attention in buying and selling of foreign investors.
The Adjusted R-squared and P-value regression results of 12 portfolios which is showed in the Table 5.2 presents a clearly comparison of significance level and descriptive power of each portfolio in the model.
Table 5.2: The Adjusted R-squared and P-value regression results of 12 portfolios P-value Adjusted R-squared
Source: data collected by the author and calculated on Stata software
Recommendations
Based on the results of the research paper with data collected on the Ho Chi Minh City Stock Exchange (HOSE), the author gives some suggestions for the investors and State management agencies in Vietnam.
The results of the study show that the volatility of stock returns is not the same for the size of each company There is a difference in the volatility of returns of large and small companies Therefore, when choosing to invest in stocks of large companies, the rate of return will be lower than investing in stocks of small companies, because the risks between the two investments are different Based on this, investors get a more suitable return expectation.
Investors can determine the liquidity of stocks based on the trading volume of stocks This factor is very important for investors Due to the liquidity of stocks, it is safer for investors to hold stocks Because stocks with high liquidity can be traded at any time and are easily converted into cash This helps to ensure that investors’s capital can always flow easily.
Different ways of dividing the portfolio will lead to different research results (in terms of statistical significance and magnitude of the impact as well as estimates in line with expectations (positive/negative)) The research paper is the basis for investors to refer to in analyzing, choosing in investment, deciding and adjusting the portfolio based on market and company analysis However, the market risk premium factor - the only factor used in the traditional CAPM model - is always statistically significant and carries the expected sign, regardless of the way the portfolio is divided Therefore, when determining stock returns in Vietnam, the traditional CAPM model still seems to be appropriate as a starting point Multifactor models can then be used to provide additional evidence for any correction that investors expect This adjustment completely depends on the specific situation of the market at the time investment decisions are made.
The stock market in Vietnam is increasingly developing and this is also a channel to raise capital for the Government, businesses and the economy, as well as a capital mobilization channel to attract domestic and international investors Therefore, policies related to the stock market must be improved The asymmetric information on the Vietnamese stock market has caused stock prices to fluctuate erratically,making it difficult to apply financial valuation models, this requires a reliable data system for the entire market The government needs to make more solutions to stabilize the macro-economy, avoid major shocks that affect businesses and investors confidence, and ensure policies to encourage and create faith for foreign investors when participating in the Vietnamese stock market.
Research limitations and future research directions
Research data is almost the most important factor that determines the reliability of the model However, due to limited time, the research only collected quarterly data from the beginning of 2015 to the end of 2019 The author used 63 non-financial companies listed on the Ho Chi Minh City Stock Exchange Therefore, studying the impact of these factors only partially reflects the profitability of the stock portfolio,and there are restrictions on the objective assessment of the entire Vietnamese market Therefore, further research may consider and involve companies from theHanoi Stock Exchange to improve the objectivity of the impact assessment.
In chapter 5, the author provides important observations from the research From there, conclusions are drawn to reflect the overview of investor investment methods on the Vietnamese stock market In addition, the research is the basis for investors to refer to for their own purposes Finally, the limitations of the study have also been pointed out and the author has also proposed further research directions to improve the study on the profitability of stocks, as well as the consideration of adding additional measures factors in the model.
Amihud Meldenson (1986) ‘Asset pricing and the bid-ask spread’, – 17.
Amihud, Y (2002), ‘Illiquidity and stock returns: Cross-section and time-series Effects’, Journal of Financial Markets, 5(1), 31–56.
A Subrahmanyam, S Titman, (2001), ‘Feedback from stock prices to cash flows’,
Baker, M., & Stein, J C (2004) ‘Market liquidity as a sentiment indicator’ Journal of Financial Markets, 7(3), 271–299.
Bao, J., Pan, J., & Wang, J (2011), ‘The Illiquidity of Corporate Bonds’, Journal of
Bekaert, G., Harvey, C.R., Lumsdaine, R.L., (2002), ‘The dynamics of emerging market equity flows’, Journal of International Money and Finance 21, 295-350.
Bekaert, G., Harvey, C R., & Lundblad, C (2007), ‘Liquidity and expected returns: Lessons from emerging markets’, Review of Financial Studies, 20(6), 1783–1831. Black (1971), ‘Toward A Fully Automated Stock Exchange’, 28(11), 2157–2160.
Brennan M.J., Subrahmanyam, A (1996), ‘Market microstructure and asset pricing on the compensation for illiquidity in stock returns’, Journal of Financial
Brennan, M.J., Cao, H., (1997), ‘International portfolio investment flows’, Journal of Finance 52, 1851–1880.
Chan H., Faff, R., (2005), ‘Asset pricing and illiquidity premium’, Financial Review,
Chordia T., Subrahmanyam, A., Anshuman, V.R., (2001), ‘Trading activity and expected stock returns’, Journal of Financial Economics, - 59.
Dahlquist M., Robertsson, G., (2004), ‘A note on foreigners trading and price effects across firms’, Journal of Banking and Finance 28, 615-632.
Daltar V.T., Naik, N.Y., Radcliffe, R., (1998), ‘Liquidity and stock returns: an alternative test’, Journal of Financial Markets, - 1.
Eleswarapu V.R., Reinganum, M (1993) ‘The seasonal behavior of liquidity premium in asset pricing’, Journal of Financial Economics, - 34.
Fama and French (1992), ‘The cross-section of expected stock returns’, Journal of
Fama, E., & French, K (1993), ‘Common Risk Factors in the Returns on Stocks and Bonds’, Journal of Financial Economics, 33, 3-56.
Fang, V., Noe, T., Tice, S., (2009), ‘Stock market liquidity and firm value’, Journal of Financial Economics, 94 (1), 150-169.
Goldstein, I and Razin, A (2006), ‘An information-based trade-off between foreign direct investment and foreign portfolio investment’, Journal of International
Griffin, J.M., Nardari, F., Stulz, R.M., (2004), ‘Are daily cross-border equity flows pushed or pulled?’, Review of Economics and Statistics 86(3), 641-657.
Gujarati, D.N (2003) Basic Econometrics 4th Edition, McGraw-Hill, New York.
Ho C., Hung, C., (2009), ‘Investor sentiment as conditioning information in asset pricing’, Journal of Banking and Finance, - 33.
Khanna, N., & Sonti, R (2004), ‘Value creating stock manipulation: Feedback effect of stock prices on firm value’, Journal of Financial Markets, 7(3), 237–270.
L Pastor, R.F Stambaugh, (2003), ‘Liquidity risk and expected stock returns’,
Lagoarde-Segot T., (2009), ‘Financial reforms and time-varying microstructures in emerging equity markets’, Journal of Banking and Finance, 2009 – 33.
Lam, Keith and Tam (2011), ‘Liquidity and asset pricing Evidence from the Hong Kong stock market’, Journal of Banking and Finance, 35, 2217-2230.
Lilian Ng et al (2015), ‘Foreign Investor Heterogeneity and Stock Liquidity around the World’, Review of Finance, (2016) 1867-1910.
Liu W A, (2006), ‘Liquidity-augmented capital asset pricing mode’, Journal of
Maug, E (1998) ‘Large shareholders as monitors: Is there a trade-off between liquidity and control?’, Journal of Finance, 53(1), 65–98.
Mona Al-Mwalla, (2017), ‘Can Book-to-Market, Size and Momentum be Extra Risk Factors that Explain the Stocks Rate of Return? Evidence from Emerging Market’,
Journal of Finance, Accounting, Volume 3 Number 2, 42-68.
Nguyen D., Mishra, S., Prakash, A., Ghosh, D., (2007) ‘Liquidity and asset pricing under the three moment CAPM paradigm’, Journal of Financial Research, - 30.
Nguyễn Anh Phong (2012), ‘Tác động của thanh khoản đến suất sinh lời các cổ phiếu niêm yết trên thị trường chứng khoán Việt Nam’, Tạp chí phát triển kinh tế, Số
Nguyễn Phương Loan (2018), ‘Kiểm định mô hình kết hợp ba nhân tố của Fama French và thanh khoản trên thị trường chứng khoán Việt Nam’.
Nguyễn Thu Hằng và Nguyễn Mạnh Hiệp, (2012), ‘Kiểm định mô hình Fama-French tại thị trường chứng khoán Việt Nam’, Tạp chí Công nghệ ngân hàng, 81, 49- 56.
Peterson M., Fialkowski, D (1994) ‘Posted versus effective spreads: good prices or bad quotes?’, Journal of Financial Economics, - 35.
Richards, A., (2005) ‘Big fish in small ponds: The trading behavior and price impact of foreign investors in Asian emerging equity markets’, Journal of Financial and
Roll, R (1984), ‘A Simple Implicit Measure of the Effective Bid-Ask Spread in an Efficient Market’, The journal of Finance, XXXIX(4), 1127–1139.
Tesar, L L and Werner, I M (1995) ‘Home bias and high turnover’, Journal of
Trần Thị Hải Lý (2010), ‘Mô hình 3 nhân tố của Fama và French hoạt động như thế nào trên thị trường chứng khoán Việt Nam’, tạp chí Phát triển kinh tế, 239, 50-57. ĩlkỹ, N., İkizlerli, D., (2012) ‘The interaction between foreigners’ trading and emerging stock returns: evidence from Turkey’ Emerging Markets Review 13, 381- 409.
Vagias, D and van Dijk, M A (2010) ‘International capital flows and liquidity’.
Vo, Xuan Vinh & Batten, Jonathan, (2011), ‘An Empirical Investigation of Liquidity and Stock Returns Relationship in Vietnam Stock Markets during Financial Crisis,’ MPRA Paper 29862, University Library of Munich, Germany.
Võ Hoàng Oanh, (2013), ‘Mối quan hệ giữa tính thanh khoản và tỷ suất sinh lợi của chứng khoán: Bằng chứng thực nghiệm từ thị trường Việt Nam’, Tạp chí khoa học
Trường Đại học Thủ Dầu Một.
Vương Đức Hoàng Quân & Hồ Thị Huệ (2008), ‘Mô hình Fama-French 3 nhân tố: Một nghiên cứu thực nghiệm đối với thị trường chứng khoán Việt Nam’, Tạp chí
Wei, C (2010) ‘Have foreign institutions improved liquidity worldwide?’ Working
Yeh, S K., Yang, T H., & Chen, R.-R (2015), ‘The liquidity impact on firm values:The evidence of Taiwan’s banking industry’, Journal of Banking and Finance, 82,191–202.
APPENDIX APPENDIX 1: LIST OF COMPANIES IN THE RESEARCH
Stock symbol Company name Field
ASM Sao Mai Group Corp
BMP Binh Minh Plastics JSC
CTI Cuong Thuan Idico Development
DAG Dong A Plastic Group JSC
DXG Dat Xanh Group JSC
HAR An Duong Thao Dien Real estate
HDC Ba Ria Vung Tau House Development Real estate and
Tan Tao Investment and Industry Corp
KBC Kinhbac City Development Holding
KSB Binh Duong Mineral and Construction
NLG Nam Long Investment Corp
NNC Nui Nho Stone JSC
NTL Tu Liem Urban Development JSC
Phat Dat Real Estate Development Corp
PHR Phuoc Hoa Rubber JSC
PTL Petro Capital and Infrastructure Investment JSC
Petroleum Equipment Assembly and Metal Structure JSC
SJS Song Da Urban & Industrial Zone Investment and
TDC Binh Duong Trade and Development JSC
TDH Thu Duc Housing Development Corp
Vietnam Electricity Construction Joint Stock Corp
VPH Van Phat Hung Corp
VRC Vrc Real Estate and Investment JSC
PNJ Phu Nhuan Jewelry JSC
PTB Phu Tai JSC Service
VNG Thanh Thanh Cong Tourist JSC
DQC Dien Quang Lamp JSC
GIL BinhThanh Import Export Production and Trade JSC
TCM Thanh Cong Textile Garment Investment Trading JSC
VNM Vietnam Dairy Products JSC
BMC Binh Dinh Minerals JSC
CSV South Basic Chemicals JSC Materials
DHC DongHai of Bentre JSC
HPG Hoa Phat Group JSC
LCM Laocai Mineral Exploitation and
TLH Tien Len Steel Corporation JSC
APC An Phu Irradiation JSC
DCM PetroVietnam Ca Mau Fertilizer JSC
FMC Sao Ta Foods JSC Agriculture
TSC Techno-Agricultural Supplying JSC
Descriptive statistics sum 1IQ RmRf SMB HML Q F
Variable Cbs Me an std Dev Min Max
pwcorr LIQ RmRf SMB HML Q F
collin LIQ RnRf SMB HML Q F (obs )
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Variables: fitted values of SL chi2{l) = 2.40
Breusch-Pagan / Cook-Weisberg test for heteroskedasticitỵ
Variables: fitted values of ML
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Variables: fitted values of BL
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Variables: fitted values of SH
Breusch-Pagan / Cook-Weisberg test for heteroskedasticitỵ Ho:
Constant variance Variables: fitted values of MH
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Constant variance Variables: íitted values of BH
Breusch-Pagan / Cook-Weisberg test for heteroskedasticitỵ Ho:
Constant variance Variables: fitted values of HB chi2(l) = 0.15
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Constant variance Variables: íitted valu.es of LB
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Variables: íitted values of MB
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Variables: fitted values of HS
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Variables: fitted valu.es of LS
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Variables: fitted values of MS
LM test for autocorrelation lags
Breusch-Godfrey LM test for autocorrelation lags(p) chi2 df Prob > chi2
LM test for autocorrelation lags(p) chi2 df Prob > chi2
Breusch-Godfrey LM test for autocorrelation lags(p) chi2 df Prob > chi2
LM test for autocorrelation lags(p) Chi2 df Prob > chi2
LM test for autocorrelation lags(p) Chi2 df Prob > chi2
Breusch-Godfrey LM test for autocorrelation lags(p) chi2 df Prob > chi2
Breusch-Godírey LM test for autocorrelation lags(p) chi2 df Prob > chi2
LM test for autocorrelation lags(p) chi2 df Prob > chi2
LM test for autocorrelation lags(p) chi2 df Prob > chi2
Breusch-Godfreỵ LM test for autocorrelation lags(p) chi2 df Prob > chi2
Breusch-Godírey LM test for autocorrelation lags(p) chi2 df Prob > Chi2
reg SL LIQ RmRí SMB HML Q F
Source ss df MS Number of obs = 20 p (6/ 13} 2.73
L Coef std Err t p>| t| [95% Conf Interval
reg ML d_ML LIQ RmRf SMB HML 2 F
Source ss f d MS Number of obs
ML Coef std Err t p>|t| [95% Conf Interval] d_ML 48164
reg BL LIQ RmRf SMB HML 2 F
Source ss df MS Number of obs = 20
(1 missing value generated) reg SH d_SH LIQ R®f SMB HML Q F
Source ss d f MS Number of obs = 19
H Coef std Err t p>1t1 [95% Conf Interval
reg MH LIQ RmRf SMB HML Q F
Source ss d f MS Number of obs = 20
H Coef std Err t p>|t| [95% Conf Interval
HB portfolio reg BH LIQ RmRf SMB HML Q F
Source ss f d MS Number of obs = 20
H Coef std Err t p>|t| [95% Conf Interval
reg HB LIQ RmRf SMB HML Q F
Source ss f d MS Number of obs = 20
reg LB LIQ RmRf SMB HML Q F
Source ss d f MS Number of obs = 20
reg MB LIQ RmRf SMB HML Q F
Source ss d f MS Number of obs = 20
B M Coef std Err t p> t [95% Conf Interval
reg HS LIQ ĩ tolRf SMB HML
Source ss f d MS Number of obs = 20