Ảnh hưởng của tính thanh khoản đến tỷ suất sinh lời cổ phiếu của các công ty niên yết trên thị trường chứng khoán Việt Nam: Khóa luận đại học chuyên ngành tài chính - ngân hàng / Nguyễn Thị Bích Tiền ; Dương Thị Thùy An người hướng dẫn khoa học

NGÂN HÀNG NHÀ NƯỚC VIỆT NAM BỘ GIÁO DỤC VÀ ĐÀO TẠO

TRƯỜNG ĐẠI HỌC NGÂN HÀNG TP. HỒ CHÍ MINH

---------------------------------------------

NGUYỄN THỊ BÍCH TIỀN

ẢNH HƯỞNG CỦA TÍNH THANH KHOẢN ĐẾN TỶ SUẤT

SINH LỜI CỔ PHIẾU CỦA CÁC CÔNG TY NIÊM YẾT TRÊN

THỊ TRƯỜNG CHỨNG KHOÁN VIỆT NAM

MÃ SỐ: 52340201

KHÓA LUẬN TỐT NGHIỆP

CHUYÊN NGÀNH: TÀI CHÍNH – NGÂN HÀNG

NGƯỜI HƯỚNG DẪN KHOA HỌC

TS. DƯƠNG THỊ THÙY AN

TP. HỒ CHÍ MINH, NĂM 2020

i

ABSTRACT

Vietnam's stock market has undergone a development process for more than 20 years

and and has proven its important role in raising capital for the economy. However,

the Vietnam stock market is still new and contains many hidden risks that investors

must face, including liquidity risk when they transfer ownership of their securities.

Investors consider liquidity to be an important factor when making their investment

decisions. Many researchers investigate the return-illiquidity (liquidity) relation, but

evidence over the past two decades is generally inconsistent and mixed. There is a

large body of research that supports the view that the liquidity of securities affects

their expected returns. Since rational investors require a higher risk premium for

holding illiquid securities, cross-sectional riskadjusted returns are lower for liquid

stocks. This proposition has been empirically supported in various studies on mature

capital markets. While the majority of studies suggest that liquidity has a negative

correlation with stock returns, experiments on emerging markets such as Asia and

Africa have shown mixed results.

Therefore, I perform this study to investigate the effects of stock liquidity on stock

returns in Viet Nam stock market. I address the question of whether liquidity is an

important variable to capture the shared time-series variation in stock returns by

investigating whether the effect of liquidity on stock return remains after controlling

for the well-known stock return factors using monthly Vietnam data. The data set

contains 40 companies that was collectted on the Hose stock Exchange and HNX

stock Exchange from October 2014 to September 2019. I collect both stock price and

number of outstanding shares as well as book equity for each firm. The yiel ten –

years government bond is used to calculate the risk – free rate.

Using the Fama – French three – factor model including beta, size, book-to-market

ratios, along with liquidity factor as a five – factor model. The liquidity factor was

conducted based on turnover ratio that are calculated by the average of the monthly

of shares traded scale by the the average number of shares outstanding over three

ii

months before October. The market excess return is calculated based on the yiel ten

– years government bond.

Following the measurement of Lam and Tam (2011), I construct 9 portfolios for first

year using monthly VietNam stock data. I form two sets of portfolios based on (1)

size and liquidity, (2) book-to-market ratio (BM) and liquidity. At the end of

September 2014, I rank the stock data monthly by market capitalization and divide

the sample into three equal-size portfolios. Independently, I calculate the annual

respective liquidity proxy for each stock in the sample and assign each stock to three

liquidity portfolios. The 9 size-liquidity portfolio was formed by an intersection

between the size groups and liquidity groups. Then, I repeat the portfolio-formation

procedure using the book-to-market ratio and liquidity to form the 9 book-to-market

(BM) - liquidity portfolios. After forming the two sets of portfolios, I compute the

portfolios’equally-weighted monthly returns. The excess portfolio return is the

portfolio return minus the risk-free rate from. I apply the fortfolios for the remaining

years. The book-to-market equity (BV/P) is computed as the ratio between a firm’s

book equity and its market price. Size is calculated by the number of outstanding

shares multiplied by the share's market price at the end of September 2014.

To the liquidity factor (LIQ) given by turnover ratio, at the end of September 2014,

firms are sorted by size (market capitalization) and included in two portfolios: Small

(S) and Big (B). The same stocks are independently sorted into three portfolios

according to their turnover ratio: L1 (most illiquid), L2 (average liquidity), and L3

(most liquid). Six portfolios (S/L1, S/L2, S/L3, B/L1, B/L2, B/L3) are then formed at

the intersection of size and turnover ratio. The equally-weighted monthly returns on

the six portfolios are calculated each month over 5 years following portfolio

formation. LIQ is the simple average of the returns on the low-liquidity portfolios

minus the returns on the high-liquidity portfolios.

I follow Fama – French (1993) method to construct size and book-to-market factors.

At the end of September 2014, firms are classified into 2 portfolios: Big (B) and Small

iii

(S) by their size at the end of September 2014. There are stocks in the sample are

independently sorted into three portfolios of book-to-market: Low (BL), Medium

(BM), and High (BH) based on their book-to-market. Six portfolios (S/L, S/M, S/H,

B/ L, BM, and B/H) are then formed at the intersection of size and book-to-market

and in a way of having approximately equal numbers of stocks. The value-weighted

monthly returns on the six portfolios are calculated each month over 5 years following

portfolio formation.

Ordinary Least Squares (OLS) method is used to estimate model. I conducted Lam

and Tam (2011) regressions for each month. For each of the 59 months in the sample,

I regressed stock returns on the liquidity measures. First, I perform descriptive

statistics and correlations on the variables that will be employed in my regression

estimation. As expected, all the variables do not exhibit any specific correlation. This

shows that the variable liquidity is likely to be an independent variable to add in the

Fama – French three factor model. Then I performed hypothetical tests of the OLS

method. Using the Unit Root Test method to check the stationary of the return value

chain of 9 portfolios and the independent variables in the model including MP, SMB,

HML, LIQ also results in stationary series, so the data is consistent with the original

assumptions of OLS. Multicollinearity test (VIF) for independent variables in the

model, the results show that the independent variables in the model do not have multi￾collinearity phenomenon. Considering the phenomenon of variance change of the

model by the Unit Root Test method, there are 2 portfolios of variance changed and

I have overcome this phenomenon by the Standard errors or robust standard errors.

The serial correlation test results with 2 portfolios of violation, so I use the method

of including the lag variable of the portfolio's outstanding rate of return variable into

the regression model to eliminate the phenomenon.

Regression results for the Fama – French three-factor model (MP, SMB, and HML)

along with liquidity factor as follows. For the size- and liquidity-sorted results, most

of the MP, SMB, and HML factor coefficients are significant except MP factor.

iv

Besides, the regression coefficients of liquidity variables show that small and illiquid

firms tend to have positive intercepts, while big and liquid firms tend to have negative

intercepts. The coefficients of SMB tend to decrease as the firm size increases, which

is similar to result in the study of Lam and Tam (2011). If the portfolios are sorted by

book-to-makert and liquidity, we find similar results. Illiquid firms tend to have

positive intercepts while liquid firms tend to have negative intercepts, the coefficient

on HML increases as book-to-market ratio increases. The coefficient of SMB

generally increases as liquidity increases. In addition, the Fama – French three -

factors model combines the liquidity factor proves effective in measuring fluctuations

in stock returns on the stock market in Vietnam, as all of the intercepts are

insignificant in the regression model.

The research results support most of abroad studies in that liquidity plays an

important role in pricing returns. Specifically, the research shows that liquidity is

negatively correlated with stock returns, except low liquidity and small firms. This

trend is especially evident for the highly liquid portfolios, the more liquid a company

is and the larger its size, the lower its share of return. This is consistent with the theory

of risk premium, low-liquidity stocks require higher returns to cover liquidity risk.

The study also denied the results in the research of Nguyen Anh Phong (2012) and

Vo Xuan Vinh (2017) that liquidity is positively correlated with stock returns. This

proves that Vietnamese investors tend to value liquidity risks in their stock

investments and this feature is consistent with the general trend of development of

stock markets around the world. The study also supports Bui Thi Le’s finding (2018)

in the VietNam stock market. Based on the results of regression analysis using data

of 40 securities firms on 2 stock exchanges, I advise investors to consider the stock's

liquidity in their investment decisions to minimize the potential for losses.

However, research still has certain limitations. Stock data is limited and mostly

collected from the HOSE, thus it does not cover the characteristics of the whole stock

market. Research data is collected from October 2014 to September 2019 is the time

v

when the market is going down with great fluctuations in political and economic

situation. Therefore, the results may be affected and inconsistent with up market times

or markets unconditional.

vi

TÓM TẮT

Nghiên cứu này kiểm định tác động của tính thanh khoản đến tỷ suất sinh lời cổ phiếu

trên thị trường chứng khoán Việt Nam, sử dụng dữ liệu cổ phiếu của các công ty niêm

yết trên 2 sàn chứng khoán Hose và HNX trong giai đoạn 5 năm từ tháng 10/2014 –

tháng 10/2019. Nghiên cứu sử dụng mô hình 3 nhân tố Fama – French bổ sung thêm

nhân tố thanh khoản đại diện bởi chỉ số khối lượng giao dịch trên khối lượng cổ phiếu

lưu hành (tỷ lệ turnover), giống như trong nghiên cứu của Wang và Iorio (2005) hay

Amihud (2002). Phương pháp phân chia danh mục tương tự như trong nghiên cứu

của Lam và Tam (2012), các cổ phiếu được phân loại dựa thành 9 nhóm danh mục

dựa trên quy mô – thanh khoản và tỷ lệ sổ sách trên giá thị trường – thanh khoản. Áp

dụng phương pháp hồi quy bình phương nhỏ nhất OLS để kiểm định mô hình, kết

quả hồi quy ở từng nhóm danh mục cho thấy tính thanh khoản được định giá trên thị

trường chứng khoán Việt Nam. Cụ thể, thanh khoản có mối tương quan âm đối với

tỷ suất sinh lời của cổ phiếu ngoại trừ các danh mục có tính thanh khoản thấp và quy

mô nhỏ. Kết quả này ủng hộ phần lớn các nghiên cứu nước ngoài và cho thấy các nhà

đầu tư Việt Nam xem xét tính thanh khoản là một yếu tố rủi ro trong việc định giá cổ

phiếu. Đồng thời kết quả nghiên cứu giúp các nhà đầu tư có cơ sở để đánh giá và dự

đoán tình hình thị trường chứng khoán trong tương lai để có quyết định đầu tư sáng

suốt.