While making contributions to the literature of corporate cash holding decisions,
this study cannot avoid limitations. Firstly, this study is limited to the period from 2011
to 2018 and to companies listed for at least nine years on the Ho Chi Minh City Stock
Exchange. Further studies may expand the scope of research over a longer period of
time and include businesses listed on other stock exchanges in Vietnam. This would
help reveal long-term trends and provide more evidence on Vietnamese firms' decisions
to hold cash. Secondly, this study shows that Vietnamese firms have reduced their cash
holdings in recent years. However, the reasons for this downward trend are still
unaccounted for. A number of studies in other countries have shown that changes in the
macro background may cause changes in corporate cash holdings over time (Almeida et
al., 2004). Therefore, follow-up studies should focus on discovering the causes of this
phenomenon to complete the picture of Vietnamese firms’ decisions to hold cash.
Thirdly, this study finds that the board of directors of firms, especially firms with low
growth and few investment opportunities, may have failed to monitor and discipline the
executives to the best interest of the shareholders. In order to have a more specific view,
further studies may analyse the impact of corporate governance quality or ownership
structure on the cash holding decisions of Vietnamese firms. Lastly, and most
importantly, this study shows that Vietnamese firms hold relatively less cash than firms
in other countries. This may be a sign of low internal investment capacity. If so, the
long-term growth and competitiveness of Vietnamese firms may be negatively affected.
For more conclusive evidence, future studies need to investigate the impact of cash
holdings on investment spending (especially investments in R&D) of Vietnamese firms.
29 trang |
Chia sẻ: hachi492 | Ngày: 14/01/2022 | Lượt xem: 258 | Lượt tải: 0
Bạn đang xem trước 20 trang tài liệu Why do vietnamese firms hold cash?, để xem tài liệu hoàn chỉnh bạn click vào nút DOWNLOAD ở trên
scipline of the board of directors or other
outside parties. In addition, previous studies have shown that businesses that have
excess money due to the self-interested behaviour of executives often squander money
on inefficient projects. Accordingly, from the free cash flow perspective, only the
factors of firm size, cash dividend policy, and investment opportunities are relevant to
the firm's decision to hold cash.
The remaining two theories, the trade-off theory and the pecking order theory,
both assume that firms’ executives make decisions to maximize the interests of the
firms, not their self-interests. However, from the trade-off theory perspective, the firm’s
decision to hold cash is made to serve short-term transactions and for precautionary
purposes. Therefore, factors such as long-term investment opportunities are not
relevant. In contrast, the pecking order theory explains the firm's decision to hold cash
on the basis of long term investment. Therefore, short-term considerations such as liquid
asset substitutes, cash flow risks, or financial crisis risks are not relevant factors. In
addition, because the pecking order theory does not project an optimal level of cash
reserves as in the trade-off theory, it is not possible to make predictions about the
impact of other internal funds (in this case, cash dividends) on the decision to hold cash.
Nguyen Thanh Hong An and Hoang Mai Phuong
13
Companies can accumulate cash and other liquid asset substitutes at the same time to
serve investment purposes in the future. Table 1 summarises the hypotheses about the
factors that affect the amount of cash held by firms according to the three theories.
Table 1. Theories and related hypotheses
Firm characteristics Trade off theory Pecking order theory Free cash flow theory
Future investment opportunities Positive Negative
Cash flow Positive Positive Positive
Cash flow uncertainty Positive
Cash dividend policy Positive Negative
Firm size Negative Negative Positive
Leverage Positive
Long-term debts Negative
Profitability Positive Positive Positive
Relationship with the bank Negative Negative Negative
Liquid assets substitutes Negative
3. DATA DESCRIPTIONS
To conduct the empirical analysis, we collected data on a sample of companies
listed on the Ho Chi Minh City Stock Exchange (HOSE). Banking and financial
services companies are excluded from the sample. The reason is that these companies
have to comply with additional regulations on holding cash. Thus, their decisions to
hold cash are not entirely based on economic considerations.
Table 2. Sample structure by industry
Industry
Number of
companies
Number of
observations
Percentage
Cumulative
observations
Cumulative
percentage
Wholesale 16 128 8.04 128 8.04
Retail 09 72 4.52 200 12.56
Information Technology 03 24 1.51 224 14.07
Accommodation and Catering 03 24 1.51 248 15.58
Mining 06 48 3.02 296 18.59
Production 76 608 38.19 904 56.78
Agriculture 04 32 2.01 936 58.79
Utilities 13 104 6.53 1,040 65.33
Transportation and Warehousing 16 128 8.04 1,168 73.37
Construction and Real Estate 53 424 26.63 1,592 100.00
Total 199 1,592 100.00 1,592 100.00
Notes: Companies are classified by their first registered type of business. Industries are classified in
accordance with NAICS (North American Industry Classification System) 2007.
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
14
In addition, we also chose companies listed from 2010 or earlier and that have
financial information released continuously until the time of data collection to ensure
the collection of adequate and reliable data, which help ensure the reliability of the
statistical analysis results. As a result, 199 enterprises in 10 industries were selected for
the survey, corresponding to 1,592 observations from 2011 to 2018. The relatively large
number of observations and the structure of the sample covering businesses in many
industries help ensure the sample representativeness. Table 2 presents information on
sample structure, classified by major business activities.
Tables 3 and 4 describe in detail the definitions and basic descriptive statistics of
the variables used in this study. As in Opler et al. (1999) and Ferreira and Vilela (2004),
we define the cash holding ratio as the amount of cash and cash equivalents divided by
net assets, where net assets are book values of assets minus the amount of cash and cash
equivalents.
Table 3. Definition of research variables
Variable Code Formula
The ratio of cash holdings
over net assets
CASH
Cash and cash equivalents
Total assets - Cash and cash equivalents
Market to book value MTB
Total assets - Equity + Market value of equity
Total assets
The ratio of cash flows over
net assets
CFRATIO
Net operating cash flows
Total assets - Cash and cash equivalents
Cash flows uncertainty CFUNCERTAINTY |CFRATIOit - CFRATIOi |̅̅ ̅̅ ̅̅ ̅̅ ̅̅ ̅̅ ̅̅
Cash dividend DIV
Equal 1 if the company pay cash dividend and
0 otherwise
Firm size LSIZE Log(Total assets)
The ratio of total debts over
net assets
LEVERAGE
Total debts
Total assets - Cash and cash equivalents
The ratio of long term debts
over net assets
MATURITY
Long term debts
Total assets - Cash and cash equivalents
Profitability ROA
Net profit after tax
Total asset
The ratio of bank
borrowings over Total assets
BORROWRATIO
Bank borrowings
Total assets
The ratio of net working
capital over net assets
NWC
Net working capital - Cash and cash equivalents
Total assets - Cash and cash equivalents
Nguyen Thanh Hong An and Hoang Mai Phuong
15
Table 4. Descriptive statistics
Variable N Mean Median Std. dev. Min Max
CASH 1,592 0.133731 0.059342 0.297952 0.000188 6.405646
MTB 1,592 1.288252 0.875000 3.601632 0.100000 121.340000
CFRATIO 1,592 0.071180 0.049837 0.164695 -0.749780 2.188948
CFUNCERTAINTY 1,592 0.081639 0.056605 0.097923 8.15E-06 1.199334
DIV 1,592 0.781407 1.000000 0.413422 0.000000 1.000000
LSIZE 1,592 9.163826 9.114875 0.536265 8.106911 11.459350
LEVERAGE 1,592 0.543082 0.555187 0.235398 0.030426 2.384434
MATURITY 1,592 0.138233 0.063652 0.199298 0.000000 4.490146
ROA 1,592 0.060375 0.047797 0.083371 -0.852590 0.783700
BORROWRATIO 1,592 0.253434 0.241600 0.189432 0.000000 0.975100
NWC 1,592 0.139682 0.115669 0.238880 -2.049630 0.925818
Notes: The statistics are calculated using 1592 observations from 199 companies listed on HOSE from
2011 to 2018; CASH is the ratio of cash and cash equivalents over net assets; MTB is the ratio of market
to book value; CFUNCERTAINTY is the absolute difference between cash flow ratios for two adjacent
periods; LSIZE is the logarithm of total assets; LEVERAGE is the ratio of total debts over net assets;
MATURITY is the ratio of long term debts over net assets; ROA is the ratio of net profit over total assets;
BORROWRATIO is the ratio of bank borrowings over total assets; and NWC is the ratio of net working
capital over net assets.
To estimate the impact of cash dividend policy, we define a dummy variable
with a value of 1 if the company pays cash dividends in the observed year and 0
otherwise. To measure the performance of the company, we use the ratio of net profit
after tax divided by the total value of the company's assets.
Because the book value does not reflect the potential investment opportunities,
investors often have to collect market information on companies’ investment
opportunities and incorporate this information into stock prices. Therefore, we use the
market value-to-book ratio to measure the company's potential investment
opportunities. The larger the ratio, the more investment opportunities the company has
and, accordingly, the higher the likelihood of the company's growth.
To measure the value of highly liquid assets that can be sold when a business
needs cash, we calculate the ratio of net working capital to net assets. Leverage is
calculated by the ratio of total debt to net assets. The logarithm of a company's book
value is used as a measure of its size.
To measure the firm's ability to generate cash, we calculate the ratio of net cash
flow from operating activities to the net assets of the corresponding year. Based on this
variable, the cash flow variability of each firm is measured by taking the absolute value
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
16
of the difference between the net cash flow from business activities on net assets and the
average of this figure for the entire survey period.
Long-term liabilities of the company are calculated by subtracting short-term
liabilities divided by net assets. In addition, we calculate the bank loan ratio by dividing
the business’s bank debt by its assets.
4. THE TRENDS OF CASH HOLDINGS OF VIETNAMESE FIRMS
Referring to Table 5, the average ratio of cash to net assets of the companies in
the research period is about 13.4%. This is lower than the average cash holding ratio of
countries in the European Economic and Monetary Union (EMU) (14.8%) according to
a study by Ferreira and Vilela (2004), or the corresponding US number (17.0%)
according to research by Opler et al. (1999). The median value of 5.9% indicates that
the typical cash holding ratio of firms in the sample is quite low and that high cash
holdings occur only in a few companies. Compared to some Asian countries/territories,
the median cash holding ratio of Vietnamese companies is higher than that of India
(3.4%), Thailand (3.8%), and the Philippines (4.9%); but lower than Malaysia (6.3%),
South Korea (8.9%), Indonesia (10.3%), Taiwan-R.O.C (11.6%), Hong Kong-P.R.C
(13.1%), and Japan (15.5%) (Dittmar, Mahrt, & Servaes, 2003).
Table 5. Cash holding ratio of Vietnamese firms in the period of 2011-2018
Year Mean Median Max Min Std. dev. Obs.
2011 0.133107 0.053759 1.092636 0.000987 0.185867 199
2012 0.145937 0.059293 5.238572 0.001605 0.398022 199
2013 0.150987 0.068639 3.513244 0.000467 0.292225 199
2014 0.165921 0.070360 6.405646 0.000188 0.488042 199
2015 0.152056 0.067763 2.280095 0.000576 0.289222 199
2016 0.118841 0.055353 2.171203 0.000769 0.210982 199
2017 0.112335 0.052354 2.634409 0.000564 0.223202 199
2018 0.090662 0.054060 0.899218 0.000788 0.120149 199
Total 0.133731 0.059342 6.405646 0.000188 0.297952 1,592
In addition, Table 5 also shows that the trend of holding cash increased
gradually after 2011 (13.3%) and peaked in 2014 (16.6%) before continuously declining
to 9.1% in 2018, with more than 50.0% of businesses having cash holdings less than
5.5%. Based on the descriptive statistics, we can infer that the cash holding ratio of
Vietnamese companies is generally low and has tended to decrease in recent years. The
increase in cash holding probably happens only in a few companies. If we look at the
median over time, the common cash holding ratio of companies in the sample only
fluctuates from 5.2% to 7.0% during the study period.
Nguyen Thanh Hong An and Hoang Mai Phuong
17
Table 6 shows the difference in the amount of cash held among firms of
different industries. The sector with the highest average cash holdings is
Accommodation and Catering Services (86.0%), followed by Retail (25.5%) and
Transportation and Warehousing (18.1%). The two industries with the lowest average
cash holdings are Construction and Real Estate (7.4%) and Wholesale (10.0%).
However, the averages are often strongly influenced by outlier observations. This makes
the average trend measurement not accurately reflect the general trend, especially when
the number of observations is small. This problem happens with a number of industries,
such as Accommodation and Catering Services and Information Technology (with only
three companies), Agricultural Production (with four companies), or Mining (with six
companies). For a more comprehensive view, we refer to the median cash holding.
According to this statistic, the industries with the highest proportion of cash holdings
are Mining (12.6%) and Information Technology (12.5%). Meanwhile, Construction
and Real Estate, Accommodation and Catering Services, and the Wholesale industry
have the lowest proportion of cash holdings, with 3.5%, 3.7%, and 4.6%, respectively.
Table 6. Cash holding ratio by industry
Industry Mean Median Max Min Std. dev.
Wholesale 0.099665 0.045674 0.551947 0.002829 0.122727
Retail 0.255419 0.093824 2.634409 0.008216 0.498662
Information Technology 0.145489 0.125234 0.579960 0.029647 0.118543
Accommodation and Catering 0.859166 0.037202 6.405646 0.003146 1.745637
Mining 0.143934 0.120256 0.530283 0.000188 0.125662
Production 0.126660 0.060063 1.136987 0.000446 0.171812
Agriculture 0.104090 0.088507 0.516208 0.002673 0.099536
Utilities 0.153989 0.068390 1.311618 0.000576 0.241806
Transportation and Warehousing 0.181198 0.099092 1.645856 0.004071 0.224938
Construction and Real Estate 0.073545 0.035171 0.917798 0.000904 0.098082
Total 0.133731 0.059342 6.405646 0.000188 0.297952
Notes: Companies are classified by their first registered area of business;
Industries are classified in accordance with NAICS 2007.
When classifying the companies into four quartile groups by size, the results
show that companies in the first quantile (i.e., companies in the smallest 25.0% group)
have the highest cash holding ratio, with an average holding ratio of 22.7% (median of
8.5%). Larger firms (in the second, third, and fourth quantiles) have lower average cash
holding ratios, at 9.6% (median of 4.5%), 10.8% (median of 5.9%), and 10.3% (median
of 5.3%). In addition, the decline in the proportion of cash held during 2016-2018
occurred in companies in all four quantiles but the most serious was in the group of
small businesses (see Figure 1).
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
18
.05
.10
.15
.20
.25
.30
.35
2011 2012 2013 2014 2015 2016 2017 2018
Quantile 1 Quantile 2
Quantile 3 Quantile 4
Figure 1. Cash holding ratio by firm size
Referring to Table 7, firm performance seems to have a great impact on cash
holding policy. Specifically, companies with positive profits often accumulate more
than three times as much cash as loss-making companies (14.0% versus 4.0%).
Similarly, enterprises with positive cash flows also accumulate twice as much cash as
businesses with negative cash flows (16.0% compared to 8.0%).
Table 7. Cash holding ratio by cash dividend policy, profitability,
liquid asset substitutes, and cash flow
Year
Cash dividend Profit Liquid assets substitutes Cash flow
Yes No Positive Negative Positive Negative Positive Negative
2011 0.14 0.06 0.14 0.02 0.11 0.19 0.17 0.07
2012 0.17 0.05 0.16 0.03 0.12 0.23 0.17 0.09
2013 0.17 0.07 0.16 0.05 0.15 0.16 0.17 0.12
2014 0.20 0.05 0.17 0.07 0.16 0.19 0.21 0.07
2015 0.19 0.05 0.16 0.05 0.15 0.17 0.18 0.09
2016 0.14 0.06 0.12 0.04 0.10 0.17 0.13 0.08
2017 0.13 0.05 0.11 0.07 0.10 0.14 0.13 0.08
2018 0.10 0.04 0.09 0.05 0.09 0.10 0.10 0.08
Mean 0.16 0.05 0.14 0.04 0.12 0.17 0.16 0.08
Nguyen Thanh Hong An and Hoang Mai Phuong
19
In addition, payment needs can also cause companies to hoard more cash.
Businesses with negative liquid asset substitutes (measured by net working capital)
often have to reserve more cash than businesses with positive liquid asset substitutes
(17.0% versus 12.0%). Businesses that have pledged to pay cash dividends to their
shareholders also have cash holding ratios three times higher than businesses that do not
pay cash dividends (16.0% versus 5.0%).
As shown in Table 8, we find that firms with better access to external capital
often hold less cash. Specifically, firms with a high ratio of long term debt, bank debt,
and leverage (both short-term and long-term), which fall into the third and fourth size
quantiles, have cash holding ratios of 10.0%, 9.0%, and 12.0%, respectively, compared
to 17.0%, 18.0%, and 15.0% for the group with little external borrowing (which fall into
the first and second quartiles). Conversely, firms in the group with a lot of future
investment opportunities or high cash flow risks often have a higher level of cash
holding than businesses with few future investment opportunities or low cash flow risks
(18.0% and 17.0% compared to 8.0% and 10.0%).
Table 8. Cash ratio by future investment opportunities, long term debts, bank
borrowings, cash flow uncertainty, and leverage
Year
MTB Long term debts Bank borrowings
Cash flow
uncertainty
Leverage
Q3-4 Q1-2 Q3-4 Q1-2 Q3-4 Q1-2 Q3-4 Q1-2 Q3-4 Q1-2
2011 0.19 0.11 0.08 0.18 0.08 0.19 0.16 0.09 0.11 0.16
2012 0.25 0.09 0.07 0.22 0.06 0.24 0.21 0.08 0.10 0.21
2013 0.22 0.09 0.09 0.21 0.10 0.21 0.17 0.13 0.13 0.17
2014 0.23 0.08 0.08 0.24 0.10 0.23 0.24 0.09 0.14 0.19
2015 0.20 0.09 0.13 0.18 0.12 0.19 0.17 0.14 0.16 0.14
2016 0.16 0.07 0.11 0.12 0.09 0.15 0.15 0.09 0.13 0.11
2017 0.14 0.07 0.11 0.12 0.09 0.14 0.12 0.11 0.11 0.11
2018 0.12 0.06 0.08 0.10 0.06 0.12 0.10 0.08 0.07 0.11
Mean 0.18 0.08 0.10 0.17 0.09 0.18 0.17 0.10 0.12 0.15
5. WHY DO VIETNAMESE FIRMS HOLD CASH?
To answer this question, we perform three analysis steps. First, the correlation
coefficient for the research variables is calculated to preliminarily evaluate the
relationship between the ratio of cash holdings and certain characteristics of firms. Next,
we conduct a regression analysis to measure and verify the relationship between the
firm-specific variables and the firm's cash holding ratio, thereby identifying potential
explanations for the cash holdings of Vietnamese firms. Finally, we conduct a robust
test to reinforce the reliability of the conclusions drawn from the regression analysis.
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
20
5.1. Correlation analysis
The correlation coefficients and corresponding statistical significance levels are
shown in Table 9. Specifically, the correlation coefficients between the cash holding
ratio (CASH) and cash flow (CFRARIO) and the cash holding ratio (CASH) and firm
performance (ROA) are positive and statistically significant, showing that firms with
favourable conditions for accumulating cash often hoard more cash. In contrast, the
negative and statistically significant correlation coefficients between the cash holding
ratio (CASH) and net working capital (NWC), the cash holding ratio (CASH) and the
ratio of bank loans (BORROWRATIO), the cash holding ratio (CASH) and the ratio of
long term borrowings (MATURITY), and the cash holding ratio (CASH) and the size of
the firm (LSIZE) show that firms with abundant cash replacement resources or those in
a good position to access external capital often have lower amounts of cash. In addition,
the positive correlation coefficients between cash ratio (CASH) and the change of cash
flow (CFUNCERTAINTY) and between cash ratio (CASH) and the leverage ratio
(LEVERAGE) indicate that companies with higher levels of financial risk often hold
more cash. Finally, Table 9 also shows that firms paying cash dividends often hold a
higher amount of cash than firms that do not pay dividends, and the firm's investment
prospects do not seem to be related to the amount of cash held.
Table 9. Correlation coefficients between research variables
CASH MTB ROA CFRATIO
CFUN-
CERTAINTY
DIV
CASH 1.000000
MTB 0.018134 1.000000
ROA 0.309230*** 0.064337** 1.000000
CFRATIO 0.536636*** 0.021469 0.349558*** 1.000000
CFUNCERTAINTY 0.371144*** -0.006400 0.045956* 0.347767*** 1.000000
DIV 0.143828*** -0.050720** 0.299470*** 0.122890*** -0.022010 1.000000
LSIZE -0.139780*** 0.094052*** -0.058580** -0.129810*** -0.180540*** 0.065471***
LEVERAGE 0.093301*** 0.014309 -0.364080*** -0.140770*** 0.045740* -0.034040
MATURITY -0.063880** -0.031240 -0.152400*** -0.023890 -0.134850*** 0.018585
BORROWRATIO -0.212970*** -0.000890 -0.387700*** -0.248590*** -0.032160 -0.104450***
NWC -0.205230*** 0.014549 0.210194*** -0.071380*** -0.020350 -0.021040
LSIZE LEVERAGE MATURITY BORROWRATIO NWC
LSIZE 1.000000
LEVERAGE 0.229393*** 1.000000
MATURITY 0.244207*** 0.294988*** 1.000000
BORROWRATIO 0.272129*** 0.568452*** 0.255953*** 1.000000
NWC -0.067950*** -0.552140*** -0.081990*** -0.375470*** 1.000000
Notes: *, **, and *** correspond to the 10%, 5%, and 1% levels of significance, respectively.
Nguyen Thanh Hong An and Hoang Mai Phuong
21
The results of the above correlation analysis seem to show that Vietnamese
companies hold money primarily for transactional and precautionary purposes rather than
for long term investment or for the benefit of the executives. They increase cash holdings
when business conditions are favourable, but also consider cash substitutes, such as net
working capital or bank loans. This view is underpinned by the very low average cash
holding ratio of most companies and the decreasing trend of cash holdings in recent years.
While the correlation analysis reveals much interesting information, the result is
not conclusive. The reason is because correlation analysis only describes the
relationship between two variables without taking into account the interaction between
them and other variables. Furthermore, correlation analysis does not take into account
the causality of the relationships. Ignoring these two characteristics may cause the
interpretation of the statistical analysis results to be misleading. In order to have a more
comprehensive and accurate view of the impact of factors on the company's cash
holding ratio, regression analysis is performed in the next section.
5.2. Regression analysis
To analyze the impact of the research variables on the ratio of cash holding, we
perform a regression analysis as follows. First, the usual regression model with the
pooled data (denoted as POLS (Pooled Ordinary Least Squares)) is estimated:
CASHit = β1 + β2MTBit + β3CFRATIOit + β4CFUNCERTAINTYit + β5DIVit
+ β
6
LSIZEit + β7LEVERAGEit + β8MATURITYit + β9ROAit (1)
+ β
10
BORROWRATIOit + β11NWCit + εit
However, the ordinary least squares regression model with pooled data can
produce inconsistent results as it ignores the impact of unobserved factors at the firm
level. The estimation results of Model (1) are presented in Column A of Table 10.
To compare and select a more effective model, we modify the structure of
Model (1) to include unobserved factors at the firm level:
CASHit = β1+ β2MTBit + β3CFRATIOit + β4CFUNCERTAINTYit + β5DIVit
+ β
6
LSIZEit + β7LEVERAGEit + β8MATURITYit + β9ROAit (2)
+ β
10
BORROWRATIOit + β11NWCit + μi + εit
Model (2) is estimated by two methods. First, Model (2) is estimated using the
GLS (Generalized Least Squares) method with the assumption that the unobserved
factors at the firm level are random with zero average and are independent of the
explanatory variables in the model. This model is denoted as REM (Random Effects
Model) and the estimation results are presented in Column B of Table 10. Model (2) is
re-estimated with the assumption that the unobserved firm-level factors are not random
and possibly correlated with the explanatory variables in the model. This model is
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
22
denoted as FEM (Fixed Effects Model) and the estimation results are presented in
Column C of Table 10
To choose between the three models (POLS, REM, and FEM), we perform the
Breusch-Pagan LM (Lagrange multiplier) test to choose between the POLS model and
the REM model and the Hausman test to choose between the REM model and the
FEM model.
One of the issues that can affect the test results is the distribution properties of
the error term of Model (2). Previous studies have also shown that because the
proportion of cash held is always positive, the regression results of Model (2) can
produce errors that have a distribution which is different from the normal distribution,
thereby affecting the validity of the tests (Bates et al., 2009). To remedy, we regress
Model (2) using the logarithm of CASH (denoted LCASH) as the dependent variable.
This approach is deemed to help the estimates achieve distribution properties that are
closer to the normal distribution. Therefore, we estimate Model (3) and the estimation
results are presented in Column D of Table 10:
LCASHit = β1+ β2MTBit + β3CFRATIOit + β4CFUNCERTAINTYit + β5DIVit
+ β
6
LSIZEit + β7LEVERAGEit + β8MATURITYit + β9ROAit (3)
+ β
10
BORROWRATIOit + β11NWCit + μi + εit
To account for the change in the amount of cash holdings over the years, dummy
variables that encapsulate the impact of the time factor are included in the structure of
the three models (not presented in the formulae). In addition, the estimated standard
errors of the models are adjusted for heterogeneity and serial correlation using the
method presented in Arellano (1987).
To test for the existence of a multi-collinearity problem, the variance inflation
factors (VIF)’s of the independent variables in the models are calculated. The results
(not presented) show that the average of the VIF’s is 1.57 and no VIF of any variable
exceeds the value of 3. Referring to Table 9, the correlation coefficients between the
independent variables are also very low. The statistical evidence mentioned above
shows that multi-collinearity is not a serious problem.
Based on the results of Breusch-Pagan LM test from Table 10, the hypothesis
that the POLS model is a suitable model is rejected. Between the other two models, the
Hausman test shows that the FEM model structure is, indeed, a more suitable one.
Therefore, we use the FEM model to perform the next analysis steps to figure out the
factors that affect the amount of cash held by the firms.
Based on the FEM model, we find that the regression coefficients of the
CFRATIO and ROA variables are both positive and statistically significant at 1%. This
shows that businesses tend to hold more cash when business conditions are favourable.
The regression coefficients of the BORROWRATIO, LSIZE, and NWC variables are all
negative and statistically significant at 5% or higher, suggesting that companies often
Nguyen Thanh Hong An and Hoang Mai Phuong
23
reduce the amount of cash held when other sources of cash substitution (excess working
capital or loans from banks) become more abundant or more accessible.
Table 10. Regression results
Variable POLS (A) REM (B) FEM (C) Log(Cash) (D)
C 0.157399 0.354104** 0.451913** 0.2088445*
MTB 0.000023 -0.000164 0.000266 -0.020206***
CFRATIO 0.631320*** 0.5335301*** 0.472683*** 0.732942***
CFUNCERTAINTY 0.634263*** 0.305838*** 0.190045 0.321758
DIV 0.022486 0.027349** 0.020857* 0.156602***
LSIZE -0.026103** -0.051097*** -0.067221** -0.170040
LEVERAGE 0.335456*** 0.448380*** 0.579273*** 2.175882***
MATURITY -0.026489 0.004876 -0.005535 -0.142808
ROA 0.728976*** 0.600218*** 0.477420*** 1.918083***
BORROWRATIO -0.367070*** -0.337246*** -0.320349*** -2.6611105***
NWC -0.202528*** -0.223244*** -0.221754** -0.340727*
Year dummies Yes Yes Yes Yes
N 1,592 1,592 1,592 1,592
R2 0.452096 0.365879 0.733057 0.745819
F 76.39791 53.422020 17.57516 18.77891
Prob(F) 0 0 0 0
Breusch-Pagan LM 1245.090000
Prob(LM Chi-sqr) 0
Hausman Chi-sqr 110.685317
Prob(Chi-sqr) 0
Notes: *, **, and *** correspond to the10%, 5%, and 1%levels of significance, respectively; The
statistics are calculated using 1592 observations from 199 companies listed on HOSE from 2011 to 2018;
CASH is the ratio of cash and cash equivalents over net assets; MTB is the ratio of market to book value;
CFUNCERTAINTY is the absolute difference between cash flow ratios of the two adjacent periods;
LSIZE is the logarithm of total assets; LEVERAGE is the ratio of total debts over net assets;
MATURITY is the ratio of long term debts over net assets; ROA is the ratio of net profit over total assets;
BORROWRATIO is the ratio of bank borrowings over total assets; and NWC is the ratio of net working
capital over net assets.
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
24
On the other hand, the regression coefficient of the LEVERAGE variable has a
positive sign, indicating that the more debt a firm has, or the higher its financial risk,
the more cash it holds. This result is inconsistent with the results of Bates et al. (2009)
for the US market or research results of Chen, Dou, Rhee, Truong, and Veeraraghavan
(2015) for a multinational dataset. Unlike the initial observations in the correlation
analysis, the CFUNCERTAINTY variable, another variable representing the volatility
of financial ability, does not seem to have a significant effect on the business decision
to hold cash.
In the same manner, the regression coefficients associated with the MTB
variable are not statistically significant, implying that potential future investment
opportunities do not matter in the decision-making process about the amount of cash
held. Finally, the regression coefficient associated with the DIV variable is only
statistically significant at 10%, suggesting that the dividend policy does not seem to
have an impact on the decision to hold cash. Reconciling the results of the regression
model with CASH as the dependent variable (Column C of Table 10) with the
regression model using LCASH as the dependent variable (Column D of Table 10),
we can see some noticeable differences. The MTB variable is now statistically
significant at 1% and negatively correlated with the dependent variable, meaning that
firms with lower prospects of future investment seem to accumulate more cash. This is
consistent with the predictions of the free cash flow theory (Ferreira & Vilela, 2004).
In contrast, the LSIZE variable is no longer statistically significant, indicating that
firm size has no effect on the decision to hold cash. This result is inconsistent with the
results from the US market, where large companies often hold less cash (Bates et al.,
2009), or evidence from research by Dittmar et al. (2003)for a multinational dataset.
In addition, the DIV variable (cash dividend policy), which is not statistically
significant in Model (3), is now statistically significant at 1% in Model (4), implying
that companies paying cash dividends tend to hold more cash. This result is
inconsistent with Almeida et al. (2004), Ferreira and Vilela (2004), and Opler et al.
(1999). In these studies, the researchers argue that firms paying dividends are often
considered to have less difficulty accessing external capital and, as a result, there is no
need to hoard much cash. Normally, dividend policy is usually kept stable for a long
period (Lintner, 1956). Firms are very reluctant to change dividend policy, especially
to reduce dividends, because this action may convey negative information about the
performance of the company (Miller & Modigliani, 1961). Therefore, for firms that
have a cash dividend payment policy, it is likely that they consider this as a spending
item that needs to be planned in advance. This statistical result seems to support the
trade-off theory of the company's decision to hold cash.
Taking the above results together, we argue that the decision to hold cash is a
calculated decision, based on the consideration of costs and benefits to the business.
This behaviour is in line with the prediction of the trade-off theory. However, there is
also evidence that companies hoard cash for the benefit of the executives. Specifically,
the results show that companies with less investment opportunities in the future often
accumulate more cash than those with more investment opportunities. The act of
Nguyen Thanh Hong An and Hoang Mai Phuong
25
hoarding cash in this situation seems to indicate that firms with less investment
opportunities hold cash to serve the interests of the executive rather than the interests of
the shareholders.
The simultaneous existence of evidence supporting the three theories on the
purposes of holding cash is not surprising. Theoretically, the three most common
theories about corporate motivations to hold cash have many overlapping predictions
(Opler et al., 1999). In fact, a business can still determine the amount of cash held to
serve a variety of purposes. To determine which cash holding motivation is stronger, we
carry out the robust test.
5.3. Robust tests
A prediction specific to the trade-off theory that differentiates it from the other
two theories is that it predicts an optimal level of cash holdings (Opler et al., 1999;
Orlova & Rao, 2018). When the amount of cash holdings is considered too high,
companies tend to decrease holdings in the next period. Likewise, when cash holdings
are lower than desired, companies tend to increase holdings in the next period. This
results in a negative correlation between the change in current cash holding ratio
(CASHt) and changes in the cash holding ratio of the previous period (CASHt-1)
(Opler et al., 1999). Thus, if the conclusion in the regression analysis section is correct,
i.e., that companies consider the benefits and costs when deciding on the ratio of cash
held, we expect to observe a significant negative regression coefficient between the two
variables, CASHt and CASHt-1, as specified by Model (4):
∆CASHit = β1 + β2∆CASHit-1+ μi + εit (4)
Technically, we estimate Model (4) using two methods. For the first method, we
perform regression Model (4) for each company. A new data set is formed with 199
observations, consisting of the regression coefficients between CASHt and CASHt-1
for each enterprise. Then, a univariate test is performed to decide whether the average
regression coefficient between CASHt and CASHt-1 of the companies is significantly
negative or not. This method is also used in Opler et al. (1999).
The results of this method are presented in Figure 2. We can see that the average
regression coefficient of 199 companies in the sample is -0.302576 (with standard
deviation of 0.535707) and the median is -0.298583. Performing a univariate test to
decide whether the average regression coefficient is actually negative or not, we have
the test statistic t = -7.967714 (prob(t) = 0.0000), indicating that the mean of the slope
coefficients is negative and statistically significant at the 1% level. Thus, we can
conclude that the average regression coefficient between CASHt and CASHt-1 of the
companies is negative and statistically significant. This shows that the business has
adjusted the holding rate of cash to the desired level, which is consistent with the
forecast of the trade-off theory.
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
26
0
10
20
30
40
50
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Figure 2. Histogram of the regression coefficients between CASHt and CASHt-1
Notes: The regression coefficient of each individual company is estimated using the following model:
∆CASHt = β1+ β2∆CASHt-1+ εt.
However, this method also has the limitation that there are only eight
observations (over eight years) for each company. Thus, the estimation of Model (4) at
the company level may yield inaccurate results.
To achieve more reliable results, we regress Model (4) with panel data.
Although the results are more reliable due to the larger number of observations, this
estimate is still biased (unrealistically small) and may be inconsistent due to the
endogeneity problem when the lag dependent variable is used as an explanatory variable
(Nickell, 1981). To remedy, we use the Dynamic Generalised Method of Moment
(DGMM) as presented in Arellano and Bond (1991) to estimate the regression
coefficients. Regression results using these two techniques are presented in Table 11.
Table 11. Regression results between CASHt and CASHt-1 using DGMM
Variable POLS FEM Dynamic GMM
C -0.010395** -0.010452*** -
CASHt-1 -0.341588*** -0.358129*** -0.305148***
Year dummies Yes Yes Yes
N 1,194 1,194 995
R2 0.155357 0.235195
F 38.387850 1.490881
Prob(F) 0 0
Hansen J 19.623750
Prob(J) 0.142457
AR(2) -0.874611
Prob(AR2) 0.381800
Notes: *, **, and *** correspond to the 10%, 5%, and 1%levels of significance, respectively.
Nguyen Thanh Hong An and Hoang Mai Phuong
27
According to the results in Table 11, we find that the regression coefficient
corresponding to the variable CASHt-1 is negative and statistically significant at 1% by
all three estimation methods. As expected, the method of estimating POLS and FEM
ignores endogenous factors in Model (4), so the impact of CASHt-1 is estimated lower
than reality (-0.340 and -0.350 compared with -0.305). The regression coefficient of the
CASHt-1 variable has a negative sign, indicating that the enterprise adjusts the amount
of cash holdings in the current period downward when the amount of money in the
previous period is too high. Conversely, companies would increase their cash holdings
in the present period if the amount of cash holdings in the previous period is too low.
The absolute value of 0.30 for the regression coefficient also indicates that the change in
the cash ratio of the following year is about 30% of the previous year. This also means
that the rate of adjustment of the holding rate of the business is about 30% per year,
which is higher than the corresponding value of 24.2% for the US market as estimated
by Opler et al. (1999).
The regression results with the FEM model and the robust tests support the
trade-off theory, implying that listed companies in Vietnam hold cash mainly for
transactional and precautionary purposes. This conclusion is also consistent with the
recent research results of Chen et al. (2015) for a dataset of businesses from 72
countries. Moreover, the cash holdings of Vietnamese firms are generally low and have
been on a downward trend in recent years. This may be a bad signal for the
competitiveness and growth of Vietnamese firms in the future. The reason is that,
according to the pecking order theory, cash is also used as an additional capital source for
corporate investments, especially investments in R&D. Investments in R&D are generally
hard to finance by external sources of capital. The fact that Vietnamese enterprises hold
too little cash is an indication of low internal investment capacity. This can lead to a lack
of investment in research and development. To the extent that this is true, the growth and
competitiveness of businesses in the future may be negatively affected.
Last, but not least, the regression results also show that businesses with less
investment and growth opportunities in the future (proxied by low MTB values) tend to
increase cash holdings. According to free cash flow theory, this is a negative sign. In
particular, the executives of these companies may have decided to hoard cash to serve
their own interests rather than the shareholders’ and the board of directors has not
fulfilled their monitoring and disciplining responsibilities.
6. CONCLUSIONS
Based on statistical analysis, this study has shown that the amount of cash held
by Vietnamese firms in the period of 2011-2018 is quite low (an average of about
13.4% of net assets), with more than 50.0% of enterprises holding less than 5.9% of net
assets. This value is low compared to businesses in western countries, such as 14.8% of
businesses in EMU countries (Ferreira & Vilela, 2004) or 17.0% of US businesses in
the period before 1999 (Opler et al., 1999) or more than 20% on average between 2000
and 2006 (Bates et al., 2009). Another noteworthy point is that the average amount of
cash held by Vietnamese companies has been on a downward trend in the last three
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
28
years, with the average amount of cash held decreasing from a peak of 16.6% of net
assets in 2014 to below 9.1% of net assets in 2018 (with more than 50.0% of businesses
holding less than 5.4% of net assets). The hoarding of large amounts of cash only occurs
in a few specific businesses and is not a common feature of listed companies in
Vietnam.
Results of regression analysis between the ratio of cash held and a number of
business characteristics show that it is likely that Vietnamese firms make decisions on
the amount of cash held based on cost and benefit considerations and the main
motivation of holding cash is probably to serve transactions and prevent short-term
business volatility rather than to accumulate long-term investment or to serve the self-
interest of the executives. Analysis of cash adjustments shows that firms often increase
the amount of cash held in the next period when the amount of cash in the previous
period is lower than desired and vice versa. This is an additional evidence supporting
the trade-off theory of corporate cash holdings in the context of Vietnam.
The research results have two implications for researchers and investors. Firstly,
the analyses show that for certain businesses, namely firms with low growth potential
and limited investment opportunities in the future, managers may have made decisions
regarding cash holdings to serve their own self-interest rather than the shareholders’.
This also implies that the board of directors of these firms may have failed to fulfil their
over-sight and disciplinary functions regarding the behaviour of executives toward the
best interest of the shareholders. Secondly, the fact that Vietnamese enterprises have a
relatively low cash holding rate that mainly serves transactional and precautionary
purposes could be a negative sign. It shows that the internal investment capacity of
Vietnamese firms is low. This may limit their ability to invest in Research and
Development (R&D), which, in turn, negatively affects the long-term growth and
competitiveness of Vietnamese firms.
While making contributions to the literature of corporate cash holding decisions,
this study cannot avoid limitations. Firstly, this study is limited to the period from 2011
to 2018 and to companies listed for at least nine years on the Ho Chi Minh City Stock
Exchange. Further studies may expand the scope of research over a longer period of
time and include businesses listed on other stock exchanges in Vietnam. This would
help reveal long-term trends and provide more evidence on Vietnamese firms' decisions
to hold cash. Secondly, this study shows that Vietnamese firms have reduced their cash
holdings in recent years. However, the reasons for this downward trend are still
unaccounted for. A number of studies in other countries have shown that changes in the
macro background may cause changes in corporate cash holdings over time (Almeida et
al., 2004). Therefore, follow-up studies should focus on discovering the causes of this
phenomenon to complete the picture of Vietnamese firms’ decisions to hold cash.
Thirdly, this study finds that the board of directors of firms, especially firms with low
growth and few investment opportunities, may have failed to monitor and discipline the
executives to the best interest of the shareholders. In order to have a more specific view,
further studies may analyse the impact of corporate governance quality or ownership
structure on the cash holding decisions of Vietnamese firms. Lastly, and most
Nguyen Thanh Hong An and Hoang Mai Phuong
29
importantly, this study shows that Vietnamese firms hold relatively less cash than firms
in other countries. This may be a sign of low internal investment capacity. If so, the
long-term growth and competitiveness of Vietnamese firms may be negatively affected.
For more conclusive evidence, future studies need to investigate the impact of cash
holdings on investment spending (especially investments in R&D) of Vietnamese firms.
ACKNOWLEDGEMENT
This research is funded by a Dalat University research grant.
REFERENCES
Acharya, V. V., Almeida, H., & Campello, M. (2007). Is cash negative debt? A hedging
perspective on corporate financial policies. Journal of Financial Intermediation,
16(4), 515-554.
Almeida, H., Campello, M., & Weisbach, M. S. (2004). The cash flow sensitivity of
cash. Journal of Finance, 59(4), 1777-1803.
Arellano, M. (1987). Computing robust standard errors for within-groups estimators.
Oxford Bulletin of Economics and Statistics, 49(4), 431-434.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte
Carlo evidence and an application to employment equations. The Review of
Economic Studies, 58(2), 277-297.
Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009). Why do U.S. firms hold so much
more cash than they used to? Journal of Finance, 64(5), 1985-2021.
Baumol, W. J. (1952). The transactions demand for cash: An inventory theoretic
approach. The Quarterly Journal of Economics, 66(4), 545-556.
Chen, Y., Dou, P. Y., Rhee, S. G., Truong, C., & Veeraraghavan, M. (2015). National
culture and corporate cash holding around the world. Journal of Banking and
Finance, 50(1), 1-18.
Dittmar, A., Mahrt, S. J., & Servaes, H. (2003). International corporate governance and
corporate cash holdings. Journal of Financial and Quantitative Analysis, 38(1),
111-133.
Faulkender, M. (2002). Cash holdings among small businesses. Retrieved from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=305179.
Fazzari, S. M., Hubbard, G. R., & Petersen, B. C. (1988). Financial constraints and
corporate investment. Brooking Papers on Economic Activities, 1988(1), 141-206.
Ferreira, M. A., & Vilela, A. S. (2004). Why do firms hold cash? Evidence from EMU
countries. European Financial Management, 10(2), 295-319.
Harford, J. (1999). Corporate cash reserves and acquisitions. Journal of Finance, 54(6),
1969-1997.
DALAT UNIVERSITY JOURNAL OF SCIENCE [ECONOMICS AND MANAGEMENT]
30
Harford, J., Mansi, S. A., & Maxwell, W. F. (2008). Corporate governance and firm
cash holdings in the US. Journal of Financial Economics, 87(3), 535-555.
Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance and takovers.
The American Economic Review, 76(2), 323-329.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency
costs and ownership structure. Journal of Financial Economics, 3(4), 305-360.
Keynes, J. M. (1936). The general theory of employment, interest, and money. London,
UK: Macmillan and Co. Limited Publishing.
Kinh, K. (2018). "Đau đầu" vì nhiều tiền: Chân dung những doanh nghiệp đang nắm
giữ lượng tiền mặt lên tới cả chục nghìn tỷ. Retrieved from
dau-vi-nhieu-tien-chan-dung-nhung-doanh-nghiep-dang-nam-giu-luong-tien-
mat-len-toi-ca-chuc-nghin-ty-20180220093509968.chn.
Lintner, J. (1956). Distribution of incomes of corporations among dividends, retained
earnings, and taxes. The American Economic Review, 46(2), 97-113.
Malmendier, U., & Tate, G. (2005). CEO overconfidence and corporate investment.
Journal of Finance, 60(6), 2661-2700.
Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation of
shares. The Journal of Business, 34(4), 411-433.
Miller, M. H., & Orr, D. (1966). A model of the demand for money by firms. The
Quarterly Journal of Economics, 80(3), 413-435.
Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporate finance, and the
theory of investment. The American Economic Review, 48(3), 261-297.
Myers, S., & Majluf, N. (1984). Corporate financing and investment decisions when
firms have information that investors do not have. Journal of Financial
Economics, 13(2), 187-221.
Nguyen, T. H. A., & Nguyen, V. T. (2018a). Free cash flow and corporate profitability
in emerging economies: Empirical evidence from Vietnam. Economics Bulletin,
38(1), 211-220.
Nguyen, T. H. A., & Nguyen, V. T. (2018b). Working capital management and
corporate profitability: Empirical evidence from Vietnam. Foundations of
Management, 10, 195-206.
Nguyễn, T. U. U., & Từ, T. K. T. (2015). Ảnh hưởng của việc nắm giữ tiền mặt vượt
trội đến các quyết định tài chính của các doanh nghiệp Việt Nam. Tạp chí Phát
triển và Hội nhập, 25(35), 36-45.
Nguyễn, T. U. U., & Từ, T. K. T. (2017). Ảnh hưởng của dòng tiền đến độ nhạy cảm
tiền mặt nắm giữ trong điều kiện hạn chế tài chính của các công ty Việt Nam.
Tạp chí Phát triển Kinh tế, 28(11), 26-53.
Nickell, S. (1981). Biases in dynamic models with fixed effects. Econometrica, 49(6),
1417-1426.
Nguyen Thanh Hong An and Hoang Mai Phuong
31
Okzan, A., & Okzan, N. (2004). Corporate cash holdings: An empirical investigation of
UK companies. Journal of Banking and Finance, 28(9), 2103-2134.
Opler, T., Pinkowitz, L., Stulz, R. M., & Williamson, R. (1999). The determinants and
implications of corporate cash holdings. Journal of Financial Economics, 52(1),
3-46.
Orlova, S. V., & Rao, R. P. (2018). Cash holdings speed of adjustment. International
Review of Economics and Finance, 54(C), 1-14.
Phạm, H., & Đinh, P. Q. T. (2018). Ảnh hưởng hạn chế tài chính đến quyết định nắm
giữ tiền mặt của các doanh nghiệp niêm yết tại Việt Nam. Journal of Science Ho
Chi Minh City Open University, 60(3), 133-143.
Các file đính kèm theo tài liệu này:
- why_do_vietnamese_firms_hold_cash.pdf