Departing from the prior studies on the relationship between dividend policy and stock price volatility, we construct the hypothesis about the impact of dividend policy on stock price volatility of the Vietnam baking sector in financial integration. We considered the enactment of Basel II of SBV in 2016 as the exogenous variable. We also include the institutional foreign ownership in the model because the limitation of foreign ownership in
Vietnam banking sector (30%) reduces the ability to raise external equity from oversea investors, and put more burden on banks in increasing nominal share capital (Fitch Ratings, 2019).
2.1. The impact of dividend policy on stock price volatility
Stock price volatility reflects the movement in stock price during a course of time, and is used to measure the unforeseen changes in stock prices. From investors’ viewpoint, stock volatility is an indicator of the exposed risk level. Stock price volatility is the systematic risk faced by stockholders (Guo, 2002). It depends on the uncertainty of firm’s future cash flow and discounted rate. Dividend policy, which strongly correlates to future cash flow and cost of capital, is expected to make a profound impact on stock price volatility.
The study of Baskin (1989) is one of the most influential investigation on the impact of dividend policy on share price volatility. This study proposed four models which connect dividends to share price risk, including the duration effect, the rate of return effect, the arbitrage realization effect and the information effect. The duration effect predicts that the price of high dividend yield stocks is less volatile than the low ones, since high dividend yield implies more near-term cash flow, leading to the less sensitivity to the fluctuation in discount rate and a lower price volatility. The rate of return effect predicts the negative correlation between dividend yield and price volatility based on the pecking order theory of capital structure (Myers & Majluf, 1984). Myer & Majluf (1984) stated that managers follow a hierarchy when selecting sources of finance. Managers choose equity financing as a last resort due to the highest cost of information asymmetry for new share issuance. Dividend yield and payout ratio can be considered as proxies for future growth opportunities. Firms with high yield and high payout ratio convey to the market that they are lacking of promising investment, which could cause higher volatility. Regarding arbitrage realization effect, because of the inefficiency of the financial market, the higher the dividend yield, the greater such arbitrage profit. However, Adrian (2004) argued that this proposal depends on uncertainty about dividend policy. The last effect refers to the signaling function of dividend.
Managers can use dividend yield to adjust share price volatility. The dividend content theory also emphasized that dividend payment transmits the confidence of insiders about future cash flows. However, other indicators such as leverage, size, sales growth can provide a signal to the market, and the information effect of dividend policy might be negligible in total
In case of Vietnam banking sector, we proposed dividend yield to have a negative impact on stock price volatility due to three reasons. First, the Vietnamese stock market is frontier and no significant evidence for supporting three forms of EMH could be found (Xuan Anh et al. 2016). Manh et al. (2017) found the abnormal return around the announcement of dividend and reported the inefficiency of the Vietnamese stock market. Second, the banking sector in Vietnam is in the rapidly growing stage, which requires an enormous amount of internal equity capital. The dispersion in the Vietnam baking sector is rather wide (Fitch Ratings, 2019). Third, dividend policy is a reliable indicator of financial health in Vietnam, where the dividend policy is less stable and weaker investor protection (Nguyen & Bui, 2019).
H1 – There is the significantly negative relationship between dividend yield and stock price volatility in the Vietnam banking sector.
2.2. Different impact of cash dividend and stock dividend on stock price volality
In this study, we classify the impact of cash dividend and stock dividend on stock price volatility. Prior studies paid acute attention to cash dividend since cash disbursement is the way to distribute free cash flow to shareholders. There is a cash outflow from assets in place regarding to cash dividend while they consider stock dividend as accounting distribution selection. Stock dividends do not make direct influence on cash flow, which would lead to no impact on stock prices (Wang et al. 2013). However, abnormal returns around stock dividend announcement have been reported in previous researchers. In the following with the Retained Earnings Hypothesis, stock dividends can make effects on the share price since managers can send a signal about their confidence in the firm’s ability to make future cash distribution (Rankine & Stice, 1997). The study of Bechmann & Raaballe (2007) conducted on the Copenhagen Stock Exchange (CSE) found that firms, who pay stock dividends, financed its growth by debts and retained earnings, yet able to afford cash dividend permanently. Stock dividend also acts as a substitute for cash distribution (Lakonishok & Lev, 1987; Baker &
Phillips, 1993). Baker & Phillips (1993) found that 40.8% of the managers use stock dividends to temporary substitute for cash dividend. That means stock dividend can influence the changes in market share price. However, how stock dividend effects on the share price volatility is questionable. The Retained Earnings Hypothesis (REH) stated that the stock distribution enables the market to identify firms whose managers are optimistic about their firm’s future earnings. This signal could be controversy as the managers have to incur the costs as lacking ability to pay cash dividend. Crawford et al. (2007) emphasized the importance of research context when testing the significance of REH in stock dividend.
Theoretically, two potential sources of constraints on a firm’s ability to pay cash dividends are contractual restrictions in the debt covenant and statutory restriction imposed by the firm’s state of incorporation. These assumptions seem to be much relevant to the current situation of the Vietnam banking sector due to the pressure of increasing internal equity capital to meet the CAR in Basel II, and the requirement of debt restructuring. Bessler & Nohel (2000) expressed the difference in the content of dividend policy made by banks and non-bank organizations. Legal regulation may affect the propensity of paying dividends of banks, for example, the capital adequacy requirements implemented by banks after the credit-crunch (Asharf et al., 2016). Therefore, banks with high dividend ratios signal the market, and the market reacts significantly to the dividend payment. The impact of cash dividend payment on stock volatility differs from that of stock dividend. We made the hypothesis about the impact of cash dividend and stock dividend as follows:
H2 – There is the significant impact of dividend payout ratio on price volatility of listed commercial banks in Vietnam.
H3 – The impact of cash payout ratio on price volatility of listed commercial banks in Vietnam differs significantly from that of stock payout ratio.
2.3. The impact of external factor and institutional foreign ownership on stock price volatility
We also investigated the impact of Basel II enactment and institutional foreign ownership on stock price volatility of listed banks in Vietnam. We tested the moderating effect of Basel II enactment and institutional foreign ownership with dividend policy on stock price volatility. Through including dummy variable Basel II in the model, we want to examine the impact of the regulatory change on the price movement of Vietnam's bank stock price.
World Bank (Global Financial Development Report 2019/2020) stated that significant weaknesses in the regulatory and supervisory system were the main reason for major reform effort in the banking industry after the global financial crisis of 2007-2009. The lack of capital regulation led to the situation that banks did not have equity capital enough to weather the crisis. Minimum Capital requirements play an important role since it allows banks to sustain unexpected losses, offset the incentive of moral hazard, increase the likelihood of survival during periods of financial turmoil. Capital requirement also an important tool for monitoring banks. When properly implemented, capital requirements incentive banks to improve their risk management. Empirical evidence reported that, in countries where supervision and regulation are costlier, the role of capital in systemic stability is stronger. Basel I was the first international initiative to define and regulate capital. In 2004, a revised capital framework, Basel II, replaced Basel I, is built on three pillars: (i) minimum capital requirements; (ii) supervisory oversight on behalf of national regulators; (iii) stronger market discipline as information disclosure on capital, risk exposure, and risk assessment processes. In this context, approaching the three pillars of Basel II is highly recommended for Vietnam's banking sector to be prudence and competitive in the global banking system (Vo, 2018a, Vo
& Nguyen 2018b). Empirical studies reported the positive impact of bank capital adequacy on profitability in Vietnam (Batten & Vo, 2019; Dang, 2019; Do & Vu, 2019; Nguyen, 2020).
Hence, adopting Basel II could signal positively to the market about risk management and sustainable performance of commercial banks, contributing to the market stability. However, Basel II seems to be far reaching for small private commercial banks; and big banks continue the race to Basel III after Basel II. This occurence would make bank managers cut cash dividend or pay dividend in stocks to retain internal sources of finance. Therefore, we hypothesized that the enactment of Basel in Vietnam's banking sector contributes significantly to the stock price volatility; and the impact of payout ratio on the stock volatility before and after Basel II adoption is significantly different.
H4- The enactment of Basel II makes a significant impact on the stock price volatility in Vietnam banking sector.
H5 – The impact of dividend policy on stock price volatility of Vietnam banking sector is significantly different under Basel II enactment.
Regarding to the impact of Institutional Foreign Ownership, Ciner & Karagozoglu (2007) stated that the participation of foreign will reduce the asymmetric information in emerging markets; hence, it is expected that higher foreign ownership contributes to the stock market stabilization. Rhee & Wang (2009) showed that the transaction of foreign investors causes some issue for stock market, such as high volatility, greater asymmetric information, and inactive trading. This problem becomes severe in emerging markets since the trading of foreign investors can be a pattern for domestic ones. Empirical evidences found negative relationship between foreign ownership and stock volatility, such as Li et al. (2011), Chen et al. (2013), Chiang & Chan (2019), Vo (2015), Phan & Tran (2019), Truc et al. (2020). In particular, Vo (2016) suggested that institutional investor ownership stabilizes the stock return volatility, and this effect is greater in firms paying dividends. Even though this study does not classify the effect of foreign investors from domestic ones, it affirms that block ownership is an important proxy for stock volatility in small emerging market with highly dominated by institutions.
In case of the Vietnam banking sector, the banks in Vietnam are small compared with foreign banks (Batten & Vo, 2019). Institutional foreign investors, characterized by professional and strong capital, can bring benefits to the Vietnam banks such as dynamic corporate governance, more transparency, and higher equity capital. The stock price becomes less fluctuated since total risk is reduced. Thus, we hypothesized that institutional foreign ownership will positively contribute to the stock market stabilization.
H6 – The institutional foreign ownership positively contributes to the stock price stabilization.
3. Data and Methodology
Data for this study are collected from the consolidated financial statements, annual reports, and resolutions of General Meeting of Shareholders of 13 commercial banks listed on the stock exchange in Vietnam in a course of 10 years, from 2010 to 2020. We selected commercial banks listed until 2017 to ensure sufficient data for dividend policy and stock price volatility.
- Dependent variable: Price_Std:
To measure the stock price volatility of a particular bank, we constructed dependent variable Price_Std as the following equation:
𝑃𝑟𝑖𝑐𝑒_𝑆𝑡𝑑𝑖𝑡 = 𝜎(𝑀𝑜𝑛𝑡ℎ𝑙𝑦 𝑃𝑟𝑖𝑐𝑒𝑖)
Most of the previous study on the relationship between dividend policy and stock dividend adopted the method of Parkison (1980) to measure the stock price volatility. Some studies used the coefficient of covariation for stock price volatility. Iradoost et al. (2013) measured price volatility using short-term and long-term indicator. Long-term price volatility equals to the ratio between standard deviation of stock price and mean value of stock price in a particular year. Similarly, Anh & Nhi (2016) used this ratio as a proxy for the stock price
volatility in Vietnam. Husainey et al. (2011) showed extreme values could influence that standard deviation. However, using Highest Price and Lowest Price to measure the volatility could lead to loss of information and overestimation of volatility. Therefore, we adopted the above equation to measure price volatility in the Vietnam banking sector. We observed the daily stock price of a particular bank was rather stable in a month; thus, we used monthly stock price.
- Explanatory variables:
Cash_ratio: This variable measures the cash dividend payout ratio of a particular bank.
Stock_ratio: This variable measures the stock dividend payout ratio of a particular bank.
We got the information about Cash_ratio and Stock_ratio from the resolutions of General Meeting of Shareholders. The data were aggregated annually.
DY_gen: This variable measures the Dividend Yield of a particular bank in general.
The fomula for DY_gen is:
𝐷𝑌_𝑔𝑒𝑛𝑖𝑡 =𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑖𝑡 (𝑖𝑛 𝑐𝑎𝑠ℎ 𝑎𝑛𝑑 𝑖𝑛 𝑠𝑡𝑜𝑐𝑘𝑠) 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑀𝑎𝑟𝑘𝑒𝑡 𝑆ℎ𝑎𝑟𝑒 𝑝𝑟𝑖𝑐𝑒𝑖𝑡
Basel 2: This is a contextual variable, referring to the impact of the enactment of Basel II on the Vietnam banking sector. Basel 2 is a dummy variable, in which:
𝐵𝑎𝑠𝑒𝑙 2 = { 0 𝑓𝑜𝑟 𝑦𝑒𝑎𝑟 𝑝𝑟𝑖𝑜𝑟 2016 1 𝑓𝑜𝑟 𝑦𝑒𝑎𝑟 𝑓𝑟𝑜𝑚 2016 𝑡𝑜 2020
Inst_Foreign_Own: this variable measures the ownership of foreign institutions. The value is extracted from the annual reports of the listed banks.
- Control variables:
Generally, previous studies on the impact of dividend policy on stock price volatily control the SIZE, LEVERAGE, EARNINGS of a firm. In this study, due to the distinct features of the Vietnam banks, we used ln(EQUITY) for SIZE, LER for leverage, and EPS for EARNINGS. The expected impacts of control variables on the stock volatility are in line with findings of previous studies3.
The baseline model to examine the impact of dividend policy on the stock price volatility of listed banks in Vietnam:
𝑷𝒓𝒊𝒄𝒆_𝑺𝒕𝒅𝒊𝒕 = 𝜷𝟎+ 𝜷𝟏∗ 𝑪𝒂𝒔𝒉_𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟐∗ 𝑫𝒊𝒗_𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟑∗ 𝑫𝒀_𝒈𝒆𝒏𝒊𝒕+ 𝜺𝒕 (1) In the second model, we include Basel 2, Inst_Foreign_Own, and the Interaction term of Basel 2 and Cash_ratio. The formula of the model 2 is:
3 Based on the findings of previous studies, we hypothesized that banks with stronger equity have lower stock volatility. Leverage is an indicator of risk, so the higher the leverage, the higher price volatility. Alternatively, Ler moves positively with stock price fluctuation. EPS reflects the net income for common shareholders.
Therefore, the higher EPS, the less risky the stock. So we expected that EPS showed negative relationship with stock price volatility.
𝑷𝒓𝒊𝒄𝒆_𝑺𝒕𝒅𝒊𝒕 = 𝜷𝟎+ 𝜷𝟏∗ 𝑪𝒂𝒔𝒉𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟐∗ 𝑫𝒊𝒗𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟑∗ 𝑫𝒀_𝒈𝒆𝒏𝒊𝒕+ 𝜷𝟒∗ 𝑩𝒂𝒔𝒆𝒍 𝟐 + 𝜷𝟓∗ 𝑩𝒂𝒔𝒆𝒍𝟐 ∗ 𝑪𝒂𝒔𝒉_𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟔∗ 𝑰𝒏𝒔_𝒇𝒐𝒓𝒆𝒊𝒈𝒏_𝒐𝒘𝒏𝒊𝒕+ 𝜺𝒊𝒕 (2)
In the third model, we control for EQUITY, LEVERAGE, and EPS. The full equation is as follow:
𝑷𝒓𝒊𝒄𝒆_𝑺𝒕𝒅𝒊𝒕 = 𝜷𝟎+ 𝜷𝟏∗ 𝑪𝒂𝒔𝒉𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟐∗ 𝑫𝒊𝒗𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟑∗ 𝑫𝒀𝒈𝒆𝒏𝒊𝒕+ 𝜷𝟒∗ 𝑩𝒂𝒔𝒆𝒍 𝟐 + 𝜷𝟓∗ 𝑩𝒂𝒔𝒆𝒍𝟐 ∗ 𝑪𝒂𝒔𝒉_𝒓𝒂𝒕𝒊𝒐𝒊𝒕+ 𝜷𝟔∗ 𝑰𝒏𝒔𝒕_𝒇𝒐𝒓𝒆𝒊𝒈𝒏_𝒐𝒘𝒏𝒊𝒕+ 𝜷𝟕∗ 𝐥 𝐧(𝑬𝑸𝑼𝑰𝑻𝒀)𝒊𝒕+ 𝜷𝟖∗ 𝑳𝑬𝑹𝒊𝒕+ 𝜷𝟗∗ 𝑬𝑷𝑺𝒊𝒕+ 𝜺𝒊𝒕 (3)
Descrptive summary of variables in the above models is presented in table 1:
Table 1. Descriptive summary of variables in the empirical models
Variable Mean Std.Dev. Min Max
Dependent variable
Price_Std 1.84311 1.92040 0.13963 10.85482
Explanatory variables
Cash_ratio 0.048673 0.062723 0 0.32
Stock_ratio 0.0403686 0.084583 0 0.41923
DY_gen 0.0719833 0.0854067 0 0.423654
Inst_foreign_own 0.1843069 0.1348717 0 0.7148
Control variables
Ler 0.9198424 0.0360756 0.6564691 1.02200
ln(Equity) 9.360883 0.7683835 7.506592 10.60212
EPS 1,898.874 1,575.777 -4,577.611 9,112.764
(Source: Authors’ calculation) Table 1 reported descriptive statistics for variables in the empirical models. The standard deviation of stock price in the Vietnam banking industry is 1,843.11 VND on average. The maximum price movement is 10,854.82 VND and the minimum one is merely 139.63 VND. The average ratio of cash disbursement is 4.8673%. Some banks do not pay the cash dividend while other banks pay high cash dividend. The maximum value of cash dividend ratio is 32%. The average stock dividend ratio is 4.03686% yet the maximum ratio reaches 41.923%. The general dividend yield (DY) of bank stocks is 7.19833%; however, some banks have a top level of DY, leading the maximum DY is at 42.3654%. The ownership of foreign investors takes 18.43% on average, yet the dispersion is wide. The maximum institutional foreign ownership is 71.48%, which belongs to a private commercial bank.
Instead, some State-owned banks showed no foreign ownership or a small proportion because
of the regulation of Vietnam Government on the state ownership domination in the banking- finance-insurance industry.
4. Results and Discussion
The Pearson correlation matrix of the variables in the models is presented in Table 2.
Price_Std negatively correlated with DY_gen (-0.1708), similar to the study of Baskin (1989) which was -0.643; Husainey et al. (2011) at -0.2853; Phan & Tran (2019) at -0.04; Truc et al.
(2020) at -0.17. The correlation coefficient between Stock_ratio and DY_gen is 0.5024, significant at 1%, which raises the question of multicolinearity in the regression model.
According to Drury (2008), when the correlation coefficient between two explanatory variables is from 0.7, multicollinearity exists. Therefore, no need to worry about the potential violation of the regression model assumptions. However, we postestimate the VIF to affirm the empirical results.
Table 2. Correlation analysis
Price_Std Cash_ratio Stock_ratio DY_gen Ler lnEquity Inst_
foreign_
own EPS
Basel2
Price_Std 1
Cash_ratio 0.0639 1
Stock_ratio 0.1039 -0.2348*** 1
DY_gen -0.1708* 0.5317*** 0.5024*** 1
Ler 0.0875 -0.2542*** -0.018 -0.2638*** 1
lnEquity 0.5377** 0.1301 0.0429 -0.1485 0.094 1
Inst_foreign_own 0.0426 0.0656 0.2508** 0.1999 -0.2565*** -0.0514 1
EPS 0.5367*** 0.1187 0.2536*** 0.0917 -0.0462 0.3423*** 0.1705 1
Basel2 0.4093*** -0.4319*** 0.1181 -0.3790*** 0.1476 0.3826*** 0.1177 0.2713*** 1
(Source: Authors’ calculation) Notes: (*) significant at 10%; (**) significant at 5%; (***) significant at 1%
According to Wooldrige (2010), Pooled OLS regression is relevant for a different sample in each period of time while fixed effects or random effects are employed in case of the same sample4. Therefore, we also employed FEM and REM to estimate the baseline
4 The Pooled OLS regression for the baseline model has no problem of multicollinearity (Mean VIF = 2.81), yet the problem of heteroskedasticity (Chi2(1) = 19.49, p-value = 0.0000). We generated the new dependent variable Price_Std_1 = log (Price_Std) and estimated the baseline model. The new model does not violate the assumption of multivariate OLS regression. On average, the dividend policy of banks listed on the stock exchange of Vietnam can explained 30.76% the price volatility of the banks. The impacts of exploratory variables on the dependent variable are in line with the results of both FEM and REM. The result of the baseline regression model is follow:
Variable Coef. Std.Err. t p-value
Cash_ratio 15.05012 6.637876 2.27 0.043
model (The model 1). The Hausman test for model selection indicated that REM is more reasonable than FEM (Chi2(3) = 2.17, p-value = 0.5373). However, we discovered the problem of heteroskedasticity (Modified Wald test for groupwise heteroskedasticity in POOL regression model reported the Chi2 (13) = 194.67 and p-value = 0.0000). We employed the robust regression to deal with the problem of variance difference. The Wooldridge test reported that there is no first-order autocorrelation in the model: F(1,12) = 0.086, p = 0.774.
The result of the baseline model was reported in the following table:
Table 3. The regression result of the model (1) Dependent variable: Price_Std
Normal estimation Robust estimation
Variable Coef. Robust Coef.
Cash_ratio 19.1163 19.1164
(4.10)** (2.78)**
Stock_ratio 15.7542 15.7542
(4.95)*** (3.65)**
DY_gen -18.9369 -18.9369
(-4.86)*** (-2.52)*
_cons 1.6527 1.6527
(5.88)*** (5.73)***
Model summary Wald chi2(3)=26.90 Wald chi2(3)=23.02 p-value = 0.0000 p-value =0.0000
Notes: (*) significant at 10%; (**) significant at 5%; (***) significant at 1%; t-statistics is in parentheses.
Table 4. The regression result of the model (1) Dependent variable: Price_Std_1 = log (Price_Std)
Variable Normal estimation Robust estimation POOL estimation
Coef. Robust Coef. OLS Coef.
Cash_ratio 9.0184 9.0184 11.3037
(4.02)*** (3.03)*** (5.79)***
Stock_ratio 7.4783 7.4783 8.5095
Stock_ratio 13.96532 4.654685 3.00 0.011
DY_gen -15.90595 7.693464 -2.07 0.061
_cons 1.668485 0.1686953 9.89 0.000
F (3,100) = 14.81; Prob > F =0.0000; R-squared = 0.3076; Adjusted R-squared =0.2868