FACTORS AFFECTING THE LEVEL OF CASH HOLDING OF COMPANIES LISTED IN VIETNAM STOCK MARKET In Partial Fulfillment of the Requirements of the Degree of MASTER OF BUSINESS ADMINISTRATION In FINANCE By Mr. Nguyen Nhat Anh ID: MBA02002 International University - Vietnam National University HCMC Sep 2014 I INTERNATIONAL UNIVERSITY SCHOOL OF BUSINESS SOCIALIST REPUBLIC OF VIETNAM Independence - Freedom - Happiness ASSURANCE QUALIFIED THESIS Student’s Name: Nguyen Nhat Anh Student ID: MBA02002 Title of Thesis: Factors affecting the level of cash holding of companies listed in Vietnam stock market Advisor: PhD. Duong Nhu Hung I assure that the content of this thesis has been qualified all requirements for a research paper and able to participate in the final thesis defense. Approved by (Signed) PhD. Duong Nhu Hung II FACTORS AFFECTING THE LEVEL OF CASH HOLDING OF COMPANIES LISTED IN VIETNAM STOCK MARKET In Partial Fulfillment of the Requirements of the Degree of MASTER OF BUSINESS ADMINISTRATION In Finance By Mr. Nguyen Nhat Anh ID: MBA02002 International University - Vietnam National University HCMC Sep 2014 Under the guidance and approval of the committee, and approved by all its members, this thesis has been accepted in partial fulfillment of the requirements for the degree. Approved: ---------------------------------------------Chairperson ---------------------------------------------Committee member ---------------------------------------------Committee member --------------------------------------------Committee member --------------------------------------------Committee member --------------------------------------------Committee member III Acknowledge First and foremost, I would like to give sincere thanks to my advisor of this thesis, PhD. Duong Nhu Hung, Lecturer of International University, Vietnam National University, Ho Chi Minh City (VNU-HCM) for the valuable guidance and advice in such a long way for successful completion of my thesis within the time frame. He inspired me greatly to work in this thesis with his initial orientation. His willingness, faith and patience in my abilities always boost my confidence and motivate me contributed tremendously to this thesis. I highly appreciate when each time he correct each minor error. Without instructions from PhD. Duong Nhu Hung, I could neither stick right way to the topic nor find the direction to complete the research. Secondly, my gratitude goes to all lecturers of the School of Business of International University, Vietnam National University, Ho Chi Minh City (VNU-HCM), who has made me familiar, understand and passion on the concepts and knowledge of all courses under MBA program. Especially, Mr. Lai Tran Thanh Son, Ms. Pham Thi Anh Tho and Mr. Nguyen Hoang Phu – please receive my thanks for your kind supports when I get stuff with papers. Finally, I am forever indebted to my parents for their understanding, endless patience and encouragement when it was most required. I am also grateful to Mr. Ho Huu Tien and Mr. Le Minh Giac - my classmate in MBA program. My classmates have followed up and supported so much on quality of thesis. Last but not latest, I want to say thanks to my close friends, Ms. Pham Mai Tram, Mr. Nguyen Chi Thanh, Ms. Nguyen Phuong Trang, Mr. Nguyen Tran Phuong, and Ms. La Y Yen for their spirit support. I cannot fulfil this thesis without encourage, and sometimes criticisms to helps me stand up after failures. Without helps of the particular that mentioned above, I would face many difficulties while doing this thesis. IV Plagiarism Statements I would like to declare that, apart from the acknowledged references, this thesis either does not use language, ideas, or other original material from anyone; or has not been previously submitted to any other educational and research programs or institutions. I fully understand that any writings in this thesis contradicted to the above statement will automatically lead to the rejection from the MBA program at the International University - Vietnam National University – Ho Chi Minh City (VNU-HCM). V Copyright Statement This copy of the thesis has been supplied on condition that anyone who consults it is understood to recognize that its copyright rests with its author and that no quotation from the thesis and no information derived from it may be published without the author’s prior consent. © Nguyen Nhat Anh/ MBA02002/ 2010–2014 VI Table of content Acknowledge ........................................................................................................................... IV Plagiarism Statements ............................................................................................................... V Copyright Statement ................................................................................................................. VI Table of content....................................................................................................................... VII List of tables .............................................................................................................................. X List of Figures .......................................................................................................................... XI List of Abbreviations ............................................................................................................... XII Abstract ................................................................................................................................. XIII Chapter 1. Introduction ....................................................................................................... 1 1.1. Problem Statement .................................................................................................... 1 1.2. Research Objectives .................................................................................................. 3 1.3. Research Questions ................................................................................................... 4 1.4. Research Scope ......................................................................................................... 4 1.6. Theoretical Framework ............................................................................................. 5 1.7. Contribution and Implication of Thesis to Business and Society................................ 5 1.8. Structure of the Thesis .............................................................................................. 6 Chapter 2. 2.1. Literature Review .............................................................................................. 7 Definitions, Time Value and Reasons for Cash Holding ............................................ 7 2.1.1. Cash Definitions............................................................................................. 7 2.1.2. Time Value of Cash........................................................................................ 7 2.1.3. Reasons for Cash Holding .............................................................................. 8 2.2. Information Asymmetry Theory and Related Theories .............................................. 9 2.3. Factors Affecting Cash Holding from Related Theories .......................................... 12 2.3.1. Size of Business ........................................................................................... 13 2.3.2. Financial Leverage ....................................................................................... 14 2.3.3. Liquid Assets Substitutes ............................................................................. 16 2.3.4. Cash Flow .................................................................................................... 16 VII 2.3.5. Cash Flow Volatility .................................................................................... 17 2.3.6. Liability to Bank .......................................................................................... 18 2.3.7. Dividend Payout ........................................................................................... 20 Chapter 3. Research Methodology .................................................................................... 24 3.1. Data Sources ........................................................................................................... 24 3.2. Data Statistics ......................................................................................................... 26 3.2.1. Descriptive Statistics .................................................................................... 26 3.2.2. Univariate Analysis ...................................................................................... 26 3.2.3. Regression Analysis ..................................................................................... 27 3.2.3.1. Fama-MacBeth Regression Model ............................................................. 28 3.2.3.2. Cross-Sectional Regression Model with Mean Value of Variables ............. 29 3.2.3.3. Pooled Ordinary Least Squares Regression Model with Year Dummy ....... 30 3.2.3.4. Pooled Ordinary Least Squares Regression Model with Year Dummy and Industry Dummy ....................................................................................................... 31 3.2.3.5. Fixed Effects Regression Model ................................................................. 33 3.2.4. 3.3. Other Supported Tests for Regression Models .............................................. 34 Determined Dependent and Independent Variables ................................................. 35 3.3.1. Dependent Variable – Cash .......................................................................... 35 3.3.2. Independent Variables .................................................................................. 36 3.3.2.1. Size of Business (SIZE) ............................................................................. 36 3.3.2.2. Bank Debt (BANKDEBT) ......................................................................... 36 3.3.2.3. Cash Flow (CFLOW) ................................................................................. 37 3.3.2.4. Cash Flow Violability (CFVOLAT) ........................................................... 37 3.3.2.5. Liquid Assets (LIQ) ................................................................................... 37 3.3.2.6. Dividend (DIVIDEND) .............................................................................. 37 Chapter 4. 4.1. Data Analysis and Results ................................................................................ 39 Descriptive Statistics............................................................................................... 39 VIII 4.2. Univariate Analysis................................................................................................. 42 4.3. Regression Results .................................................................................................. 44 4.3.1. Fama–MacBeth Regression Model ............................................................... 44 4.3.2. Ordinary Least Squares Regression with 5 Years Average ............................ 48 4.3.3. Pooled OLS Regression with Year Dummies ............................................... 50 4.3.4. Pooled OLS Regression with Year Dummies and Industry Dummies ........... 53 4.3.5. Fixed Effects Regression Model ................................................................... 56 4.4. Result from Supporting Tests for Regression Models .............................................. 57 4.4.1. 4.5. Hausman Test .............................................................................................. 57 Summary of Results ................................................................................................ 59 Chapter 5. Research Conclusion........................................................................................ 63 5.1. Conclusion and Implication .................................................................................... 63 5.2. Limitation of the Study ........................................................................................... 65 5.3. Future Research ...................................................................................................... 66 References ................................................................................................................................ 68 Appendix.................................................................................................................................. 72 Appendix 1: 98 Listed Companies Chosen ............................................................................ 72 Appendix 2: Secondary Data for Excel and Stata Input ......................................................... 79 IX List of tables Table 1: Summary of Previous Studies ..................................................................................... 20 Table 2: List of Industry Dummies ........................................................................................... 31 Table 3: Summary of Formulas for Variables ........................................................................... 38 Table 4: Results from Descriptive Statistics .............................................................................. 39 Table 5: Results from Univariate Analysis ................................................................................ 42 Table 6: Correlations Matrix of Variables from Unvariate Analysis .......................................... 43 Table 7: Results from Fama – MacBeth Regression Model ....................................................... 44 Table 8: VIF Test for Fama-MacBeth Model ............................................................................ 46 Table 9: Results from Ordinary Least Squares Regression with 5 Years Average ..................... 48 Table 10: VIF Test for Cross-Sectional Regression Model........................................................ 49 Table 11: Results from Pooled OLS Regression with Year Dummies ....................................... 50 Table 12: VIF Test for Pooled OLS Model with Year Dummies ............................................... 51 Table 13: Results from Pooled OLS Regression with Year Dummies and Industry Dummies ... 53 Table 14:VIF Test for Pooled OLS with Year and Industry Dummies ....................................... 54 Table 15: Results from Fixed Regression Model ....................................................................... 56 Table 16: Results from Hausman Test ...................................................................................... 57 Table 17: Result Summary of Regression Models ..................................................................... 59 Table 18: Thesis’s Result Summary .......................................................................................... 60 X List of Figures Figure 1: Theoretical Framework to Be Applied in This Research .............................................. 5 Figure 2: Selected Regression Models ...................................................................................... 28 Figure 3: Cash Ratio Over Five Years ...................................................................................... 40 Figure 4: % Standard Deviation of Cash Flow over Total Assets .............................................. 41 Figure 5: Graphical Correlation Matrix ..................................................................................... 43 Figure 6: Percentage of Cash Holding in Years Firms Paid Dividend ........................................ 47 Figure 7: Correlations between Cash, Income and Bank Debt ................................................... 47 XI List of Abbreviations BANKDEBT: Bank Debt BVND: Billion Viet Nam Dong (Currency) CFLOW: Cash Flow CFVOLAT: Cash Flow Volatility DF: Degrees of Freedom EBITDA: Earning Before Interest, Tax, Depreciation and Amortization EMH: Efficient Markets Hypothesis EMU: Euro Countries FV: Future Value of Cash HNX: Ha Noi Stock Exchange HOSE: Ho Chi Minh Stock Exchange LEV: Leverage LIQ: Liquidity OLS: Ordinary Least Square Regression Pooled OLS: Pooled Ordinary Least Square Regression Model PV: Present Value of Cash SIZE: Size of Firm VIF: Variance Inflation Factor XII Abstract Understanding factors affecting the level of cash holding is and has been a challenging task in finance to reach business objectives. That is the reason why cash management is as much an integral part of business cycle in any period. Statistical and econometrical models are broadly used in analysis of specific factors affecting. This paper investigates the factors affecting the level of cash holding with the combined application of descriptive statistics, univariate analysis and regression methods such as model Fama–MacBeth, regression using cross-sectional data of average value, pooled OLS regression with year dummies, pooled OLS regression with year dummies and industry dummies. This paper also outlines the practical steps which need to be undertaken to use above methods on companies listed in Vietnam stock market. A framework for each methodology is drawn up. The emphasis is on analysis trend of factors affecting the level of cash holding when being put in consideration and explanation with related theories. The main goal of this study is to present the result of each method and factors affecting the level of cash holding in order to bring general insights of Vietnam stock market to help individual investors make decisions. Keywords: Vietnam stock market, descriptive, univariate, regression, level of cash holding. XIII This page is intentionally left blank. XIV Chapter 1. Introduction The aim of this chapter is to provide the problem background and then to understand the purpose behind the research and questions, to which this thesis seeks and answer. It also states goals that the author will seek in respect to both theoretical and practical contribution. 1.1. Problem Statement The phrase “cash is king” is an expression commonly referred to finance, although its meaning might not be as simple as this expression suggests. Until now, there are two opposing views when looking at the level of cash that a company should hold. For the first view, cash is normally regarded as "just an asset that a firm need to help it to function” (Atrill & McLaney, 2004, p.124). In order to support for this statement, Berk and DeMarzo (2011) gave several reasons for a company to hold cash. Those researches believed that holding cash was one of the essential elements needed for a firm to grow and prosper. Cash is also considering as the life-blood of any firm and if so without it, the survival is very unlikely. Every firm needs cash daily to pay its bills and to pay off its liabilities on time, so that it can survive. The firm can make use of cash to invest; for strategic purposes such as acquisitions; for a marketing tool; to solve short-term obligations, to pay daily transactions, for repaying its loans, paying its employees, even to satisfy bank requirements and so on. Contrary to above arguments, the optimal level of cash holdings in a company should be zero, Maness and Zietlow (2004) supported. This relates to the old personal saying of Plautus is that people have to spend cash to earn cash. Instead of holding on cash, a company could has either invested in a new project to earn a potential return or distributed it among shareholders to let them invest elsewhere, also this resulting in a potential return. This was not just a theoretical perspective as around the new millennium increasingly many companies were striving for decreased cash balance; financial managers globally advocated the benefits of small cash 1 holdings, Mintz (2000) stated. Holding excess cash might lead to a loss of an opportunity that could have increased the company’s value. With holding high level of cash, it will decrease earning yield of business indirectly, or even directly, because it does not maximize profit or potential opportunity from daily transactions or long-term investments. If the management capability is weak in cash monitoring, it may lead to liquidity crisis, even to bankruptcy. If so, what factors will decide this optimal level of cash holding by making costs and benefits on balance? Company can hold cash for many reasons, short-term or long-term, finance or nonfinance, strategic or operational objectives. No matter what purpose, cash flow shows us how the company has performed in managing inflows and outflows of cash and provides a sharper picture of the company's ability to pay bills and creditors, and to finance the growth. Deep understanding on factors affecting the level of cash holding first, and then implementing effective cash management solutions are the most popular approach to improve business performance by now. For this reason, it is very important for every firm to monitor its cash flow in order to adequate plan expenditures. The author’s personal opinion is that cash management is always a hot spot for managers. With context of high inflation and economic downturn in Vietnam currently, high liquidity of cash will enhance the confidence of investors. Masan group (1,970 bVND); Techcombank (3,142 bVND); Dam Phu My (3,748 bVND); PetroGas service (3,621 bVND); Vinamilk (3,101.4 bVND); FPT (1,900 bVND); Bao Viet insurance (2,700 bVND); Hoa Phat group (1,064 bVND) and Hoang Anh Gia Lai group (2,335 bVND) were highlighted as bright spots for cash management and creating advantage with outstanding profit in their market segments. Looking into the financial reports at the end of 2011s, many companies cannot have good financial results like above companies. The factors, which affect cash holding, are investigated by many authors in the developed countries. Kim et al. (1998) and Opler et al. (1999) did use data of American companies. 2 Pinkowitz and Williamson (2001) did use data of American, German and Japan. Ferreira and Vilela (2014) were with Euro companies (EMU); Ozkan and Ozkan (2004) were with the UK companies; Custodio et al. (2005) and Bates et al. (2009) were with the US companies; and Meggison and Wei (2012) were with Chinese companies. These researches had been giving important resources to other researchers, investors and managers. As summary of financial reports of companies listed in Vietnam stock market from 2009 to 2012, the ratio of cash and cash equivalence over the total asset is decreased from 2009 (9.9%). This ration is lowest in 2011, with 8.9%. However, at the end of 2012 (9.2%), this ratio grow up, reach to the ratio of 2010 (9.2%). Beside some basic statistics on non-business magazines, there are few researches on cash management. The empirical studies are actually lack for demands of researchers and investors. In the context of Vietnam, rapid changes in business environment, but there are few academic researches on evaluating the factors that affect level of cash holding in firm. With this research, the author research and supplements a close analysis to cash management of companies listed in Vietnam stock market based on historical data. By combining and simulating approaches of previous empirical researches, the author also expects to provide the answer for management issue of what factors create strong influences on their cash levels. 1.2. Research Objectives This research will concentrate on clarifying factors which affect decision of how much cash was hold by companies listed in Vietnam stock market. The objectives are:  To identify factors that affect the level of cash holding and then measure on their impact to cash management of companies listed in Vietnam stock market.  To propose recommendations for the management. 3 1.3. Research Questions This research will provide a practice of 98 companies listed on Vietnam stock market resulted from related theories. It will focus on two questions:  What firm factors affect the level of cash holding of companies listed on Vietnam stock market?  What are the implications for cash management on companies listed on Vietnam stock market? 1.4. Research Scope Due to time limitation and data availability, this research focuses on companies listed on two stock exchanges in Vietnam: Ho Chi Minh and Hanoi Stock Exchange. Most of them are supervised by State Securities Commission of Vietnam. The data set is mainly extracted from annual consolidated financial reports, which we have checked for errors. Due to the size and early stage of the stock market, the sample could only cover in five years and for ninety-eight companies. The methodology part will clearly show other criteria for selecting and filtering data set 4 1.6. Theoretical Framework The author defines framework for this research as below: Research Objectives Literature Review Research Design Collect Data from HSX and HNX Data Analysis Decision and Implication Conclusion and Recommendation Figure 1: Theoretical Framework to Be Applied in This Research 1.7. Contribution and Implication of Thesis to Business and Society This research provides a review of Vietnam stock market based on examining the level of cash holding and cash management. It suggests an approach allowing local individual investors to make informed decisions in stock market. A suitable model for analyzing cash holding is really necessary for any local individual investor to make informed decisions in the local stock market. 5 1.8. Structure of the Thesis The research will be in such a way as below. Chapter 1, introduction, will includes the research problem and the importance of research. It also mentions to research’s scopes, objectives, questions, structure, and contribution. Chapter 2, literature review, will deal with the origin, terminology, and reasons why companies want to hold cash. The next part will point out factors, which affect the level of cash holding predicted by related theories and research’s results of previous authors, both in foreign countries and Vietnam. Chapter 3, methodology, describes approaches for statistics. It describes the methods by which the author used to collect primary and secondary data; as well as how to describes and check variables. Regression methods will be introduced to show how to analyze dataset; depending on analyzing the advantage and disadvantage of previous authors and current practice on Vietnam stock market. Chapter 4 discusses results at length and detail. Chapter 5 will be a summary of results, conclusions, implications and limitation also. The author will focus to answer the main questions mentioned in the beginning of this research. The author will also initiate several suggestions to make this research more complete and thorough. 6 Chapter 2. Literature Review This chapter uses existing concepts, models, formulations and frameworks on the basis of previous authors and substantially develops own views and insights. It includes two main parts. First one explains definitions and reasons why companies want to hold cash. The next part explains factors predicted, which affect level of cash holding, by using financial theories, as well as presents academic results of previous authors. 2.1. Definitions, Time Value and Reasons for Cash Holding 2.1.1. Cash Definitions Everyone is actually familiar with the term cash daily, and more generally, the Oxford online dictionary defines cash as “coins and notes”. According to financial theories, they definite cash that include cash, short-term securities, and cash equivalences (Opler et al., 1999; Ferreira & Vilela, 2004; Bates, Bates et al. 2009). Many investors accept as true that cash equivalences convert into cash in short period. 2.1.2. Time Value of Cash Time value of cash is one of the most basic principles in finance. “One dollar today is worth more than a dollar tomorrow” is perhaps one of the most famous quotes in the finance. It means that if a person should invest one dollar today it will earn interest and will therefore be worth more the day after. Berk and DeMarzo (2011) postulated that the interest rate would be dependent on the investment and has to adjust by subtracting the amount of inflation on the currency. Both investors and companies use the concept of time value of cash in order to calculate the present as well as the future value (FV) of their operations. For example, they calculate the present value (PV) of the cost and benefits for a project in order to determine if the project is profitable or not. Berk and DeMarzo (2011) gave another example to calculate if their 7 current cash level is worth enough to cover their future expenditures. The formula shows the calculation for PV, which is the present value calculated by the future value FV, the interest i, and the period n as shown in the formulas. Thus, Maness and Zietlow (2004) postulated that this leads to the argument that the optimal level of cash that a company should hold is zero, since it is then considered as an idle resource for the company that do not provide any or little additional value. 2.1.3. Reasons for Cash Holding In summary, the four following reasons that companies want to hold cash. The first reason is for business operating transactions. Businesses do need cash to use for operating activities business. The demands for cash among different types of firm are also different. For example, the retail business has available cash in safe box to control operations. In addition, these businesses need cash to replenish inventory deficits or to pay salaries. Conversely, Damodaran (2008) confirmed that trades computer software could operate with much smaller cash balances. The second reason is for reservation motivation. According to Custodio et al. (2005), a business held cash to finance the activities and investments when no other financial resources available or too expensive. The firm holds cash to cover unexpected costs and potential liabilities cannot be determined. For example, Damodaran (2008) confirmed that is cyclical business to accumulate cash during the economic boom and use that cash in case of recession to cover deficit from operations. If the firm does not have available liquid assets, the cash flow shortage can stop investing in the profitable projects. Thus, Ozkan and Ozkan (2004) advised that keeping cash is to avoid financial cost and depletion. Managers consider this motivation to hold cash as reservation motivation. The third reason is the agency cost. In public businesses, it always exist problem between owners and managers. The board will decide whether to return into stock or to retain in the business. In many businesses, the managers, with available programs and activities, create funds to help them pursue these programs. Therefore, when the 8 management pays more attention to gain more control over the optimal allocation of resources, they will accumulation of cash. Damodaran (2008) gave his opinion that it is not for shareholders but for the purpose of opening wide funds. The management is favorite cash holdings, because they are afraid of risks. According to Opler et al. (1999), the management holds cash to get flexibility in pursuing their own goals. The final one is for investing in the future. If the capital market is efficient, businesses can raise capital to invest in new and potential profitable projects or other investments that has no transaction costs. However, in reality, businesses must always face with the limitations of borrowing and the cost to access capital markets. These restrictions may stem from inside, or can also come from outside the firm, but mostly, from outside the firm. According to Damodaran (2008), In order to face these challenges, firm must reserve cash available to cover investment needs in the future. If not, they run into the risk of reject the project and investment value. The author mainly uses the first two reasons in this study. Two main theories used to explain factors are the trade-off theory and the pecking order theory. 2.2. Information Asymmetry Theory and Related Theories In the 1960s, both Fama and Samuelson developed independently the efficient markets hypothesis (EMH), that market prices fully reflect all available information. In practice, behavioral economists attributed the imperfections because of cognitive biases such as representative bias or information bias. Information asymmetry theory is mainly used in this study. And other related theories are also referred, including agency theory, trade-off theory, pecking order theory, and cash flow theory. These theories play a central role in corporate finance and directly associate with level of cash holding in firms. Asymmetric information occurs when one group of participants has better or more timely information than other groups. According to Myers and Majluf (1984), asymmetric information makes raising capital more difficult and answer the reason why existence of asymmetric information makes external financing costly. The outside parties want to ensure that they 9 purchase securities not priced too high, so they will need a reasonable discount. Because parties outside the enterprise hold less information than the management, they can evaluate too low stock valuations based on the information provided by the management. Smith and Stulz (1985) delivered evidence that is consistent with asymmetric information problems of external financing. For example, investors discount the value of firms when they attempt to sell risky securities. As view of Opler et al., (1999) asymmetric information led to situation that outside capital becomes more expensive. This model predicts increased costs when raising capital securities, which sold, are sensitive and the problem of information asymmetry information becomes more important. Growth opportunities are intangible in nature and their value fall precipitously in financial distress and bankruptcy. This would in turn imply that firms with greater growth opportunities have greater incentives to avoid financial distress and bankruptcy and hence hold larger cash and marketable securities. Agency theory is using in many fields and concerns itself with the problem of ensuring that someone acting on the behalf of another, an agent representing a principal, serves the interest of the principal. Agency problem arises where the two parties have different interests and asymmetric information, such that the principal cannot directly ensure that the agent is always acting in the principal’s best interests. As opinion of Eisenhardt (1989), the issue at hand is that the agent might have other goals and interests than the principal and would act to achieve these at the expense of the principal. The theory assumes that in certain situations, if possible, the agent would act to serve his or her objectives and the solution in these situations comes from aligning the interests of the principal and the agent. Harris and Raviv (1978) showed that these relationships are numerous and cross-links, as such lawyer - clients, employer - employee and shareholders-managers. Agency problems that might arise in a shareholder-manager situation, concerns among other things the optimal level of cash holdings. Jensen (1986) worried that managers have motives for having higher levels of corporate cash holdings than what is optimal for shareholders. 10 Trade-off theory implies that businesses will set up an objective of cash level by balancing between marginal cost and marginal benefit of holding cash. The author summarizes three main benefits from holding cash as below. Firstly, holding cash is to reduce the ability of financial exhaustion, because the amount of cash held will act as a safety provision for unexpected losses or limitations from raising capital from outside sources. Secondly, cash holding allows businesses to pursue optimal investment policy, even whenever businesses encounter financial constraints. As enterprises face difficulties in raising external capital, businesses may be forced to abandon the investment project has a positive net present value. Thirdly, holding cash reduces the cost of raising capital from outside or avoid forced liquidation of existing assets of the enterprise. The role of cash is considering as a buffer between the source and use of firm resources. Ferreira and Vilela (2004) proved that traditional marginal cost of holding cash is the opportunity cost of capital because it reduces the rate of return calculated on current assets. Pecking order theory implies that the cost of financing increases with asymmetric information. Managers know more about their companies’ prospects, risks and value than outside investors. Pecking order theory of Myers and Majluf (1984) said that companies finance for investments in the order as follows: the first is retained earnings, then safe debt, followed by risky debt, and finally equity. The basic purpose of this order is to reduce funding costs due to problems of asymmetric information and other financial expenses. When having enough sufficient operating cash flow, it will fund new projects to seek potential returns. Otherwise, businesses will pay the debt and holding cash. As research of Ferreira and Vilela (2004), when retained earnings are not sufficient to fund existing projects, they will use the amount of reservation and, if necessary, will now borrowings. According to free cash flow theory, Jensen (1986) did define that free cash flow is cash flow exceeds the amount necessary to fund all projects with positive net present value, discounted at the cost of capital involved. Free cash flow theory is a subordination of agency cost 11 theory. Interests and motivations of shareholders and management have conflicting issues related to the optimal size of the firm and the payment of dividends to shareholders. The conflict is particularly severe in the enterprise with great free cash flow. Jensen (1986) believed that it is usually the enterprise free cash flow greater than the opportunity profitable investment. Free cash flow theory is also basing on the theory of asymmetric information and the theory of agency costs. According to Jensen (1986), the payment to shareholders will reduce the resources under management's control, thus reducing their strength. Therefore, the management tends to increase the amount of cash available under their control to achieve power and autonomy in investment decisions. When cash is available to serve the purpose of investment, management does not need to raise capital from outside and not provide the external parties information about the investment projects. It is called the independence in management power. Therefore, management can make the investment projects negatively impact the welfare of shareholders. In reality, the shareholder wishes to receive more cash back when having free cash flow. 2.3. Factors Affecting Cash Holding from Related Theories Based on related theories, the author will present clearly how firms reacted with; and changes on level of cash holding when having changes of independent factors. Related factors, which presented in previous academic researches, will be summarized along with above tradeoff, pecking order and/or cash flow theories, if any. Size of business, financial leverage, bank debt, cash flow, cash flow volatility, liquidity of working capital, and finally dividend payout yearly are exact main seven factors that previous authors reviewed. Those factors have close relationship with marginal cost and benefit; with the investment and reservation purpose, and with asymmetric information and agency cost. People can see the details in below review and result summary at the end of this chapter. 12 2.3.1. Size of Business Based on trade-off theory, Faulkender (2002) argued for that if a firm was larger, the demand for cashing holding is lower. A similar result also presented by Bover and Watson (2005) that larger firms were likely to have lower demand of cash holding. Actually, as to Kim et al. (1998), large firms were less likely to face borrowing constraints than small firms, because of scale economies resulting from a fixed cost component of security issuance costs. Furthermore, Ferreira and Vilela (2004) protested that small firms with high business risks and strong growth opportunities tended to hold more cash because it will be more expensive for small firms’ to raise funds. The transaction fees, which usually accompany with raised funds, were fixed, and thus, the marginal cost were higher for small firms. Besides that, Ferreira and Vilela (2004) advocated that larger firms were more likely to be diversified to be in probability of financial distress so larger firms hold less cash. These arguments proposed a negative relation between the size of a firm and the demand of cash holding. In the research of Pinkowitz and Williamson (2001), the determinants of cash holding for the United States, Japan, and Germany were investigated. A regression for all the three countries showed that there was a negative relation between firm size and cash holding. Nevertheless, when they tested the three countries separately, evidence shows that Japan and the US had a negative relation whereas Germany had a positive relation between firm size and cash holding. Bates et al. (2009) investigated why cash holding for US industrial firms doubled from the year 1980s to 2000s, and related factors. Initially, they found evidence that there is a negative relation between firm size and cash holding in the 1980s and 1990s and they are consistent with models of a transaction demand for cash. However, in the year 2000s they concluded that there is a positive relation between firm size and cash holding due to agency problems. One more study, Ozkan and Ozkan (2004) found a positive but insignificant relation between firm size and cash holding. This positive coefficient suggested that there might be other factors affecting the way in which size of firms applies influence on their cash holding decisions. A typical case is that it may be that large firms are 13 more successful in generating cash flows and profit so that they can go back to accumulate more cash. In addition, large firms have greater growth opportunities and smaller liquid assets besides cash. In this case, they may decide to hold more cash, Ozkan & Ozkan (2004) argued. However, these arguments were not strongly supported by evidences. According to pecking order theory, Ferreira and Vilela (2004) and Opler et al. (1999) found no evidence to support for a positive relation between firm size and cash holding. They presumed that large firms had more successful, and have more cash after controlling for investment. But it is weak confirmation. Pointing to cash flow theory, Ferreira and Vilela (2004) debated that larger firms are more likely to have more stakeholders, which will increase the management’s self-determination of making decisions. In addition, larger firms are not likely to be the target of a takeover, which requires more resources. Thus, it was expected that managers of large firms have powers that are more discretionary over the investment and financial policies of the firm, which leads to a greater amount of cash holding. So there is a positive relation between firm size and cash holding. However, at the end Ferreira and Vilela (2004) and Opler et al. (1999) did not still find any evidence to support this positive relation. They are their subjective consider for reference only. In Vietnam, Trinh (2012) did use financial data in order to look into the percentage of holding cash over total assets. The author explained that the market is volatile much, so big firms did tend to hold more cash. 2.3.2. Financial Leverage According to trade-off theory, Ferreira & Vilela (2004) recognized that leverage ratio performs as a proxy for the ability of the firms to issue debt. In this way, the firm has a higher ability to increase debt so that firm will hold less cash. On the other hand, a firm with less capability to increase debt holds more cash. Ozkan and Ozkan (2004) observed and saw that 14 firms with high leverage have lower cash holdings in order to lower the cost of cash holding. This means that leverage has a negative relation on cash holding. In common, it is believed that leverage increases the probability of bankruptcy due to the stress that amortization plans placed on the firm treasury management. To reduce the chance of experiencing financial distress, firms with higher leverage are expected to hold more cash. There was the same result as Ozkan and Ozkan (2004), however Thomas et al. (2007) mentioned the defensive purposes for cash holdings when firms could settle up all of their debt duties at the end of the sample period. Thomas et al. (2007) saw that the considerable average ratio of cash over assets for the U.S. industrial firms increased rapidly, by 129%, from 1980 to 2004. Because of this increase in the average cash ratio, firms could settle up all of their debt duties with their cash holdings at the end of the sample period, so that the average firms had no leverage when leverage is measured by net debt. This change in cash ratios and net debt was the result of a secular trend, but is more pronounced for firms that do not pay dividends. The average cash ratio increased over the sample period because firms change: their cash flow became riskier, they hold fewer inventories and accounts receivable, and were increasingly R&D intensive. The defensive purposes for cash holdings appeared to explain the increase in the average cash ratio. However, Ferreira & Vilela (2004) found no evidence for this positive relation between leverage and cash holding. These arguments finally suggested that leverage may have an unknown (positive and negative) relation between leverage and cash holding. Relying on pecking order theory, Ferreira & Vilela (2004) perceived that debt grows when investment exceeds retained earnings and falls down when investment is less than retained earnings. Cash holding fall when investment exceeds retained earnings and grow when investment is less than retained earnings. That was a reason why Ferreira & Vilela (2004) suggested that there was a negative relation between leverage and cash holdings. According to Opler et al. (1999), a firm’s debt reacted to changes in the internal funds of the firm. When the 15 firm’s internal funds increased, its leverage decreased. For more details, most of the time firms gained internal funds instead of issuing equity because it was expensive due to adverse selection. Because with internal funds the firm often spent more cash than receiving cash, the firm decreases cash holdings and raises debt. In short, this relationship between cash holdings, debt, and internal resources suggested that there is a negative relation between leverage and cash holdings. Relating to cash flow theory, Ferreira and Vilela (2004) noticed that firms with low leverage were less subject to monitoring, allowing for superior managerial discretion. This meant that firms with less leverage hold more cash which is a negative relation between leverage and cash holding. 2.3.3. Liquid Assets Substitutes In balance sheet, firms can have liquid assets substitutes. These assets can be converted into cash easily in short time and with low costs. These include accounts receivable and inventories and this is in fact the net working capital minus cash. They are substitutes for cash and therefore trade-off theory predicted a negative relationship between liquid assets and cash holding, Bigelli and Sanchez (2010); Ferreira and Vilela (2004); Ozkan and Ozkan (2004) examined. Furthermore, Opler et al. (1999) found evidence that large firms hold liquid assets substitutes so that they could be able to keep investing when cash flow is too low and when outside resources are costly. Those evidences came from analysis on balance sheet of firms researched. It has a relation with pecking order and cash flow theories. 2.3.4. Cash Flow Ferreira and Vilela (2004) believed that areal reason for firm to hold cash is cash flows. Cash flow was recognized after corporate tax profit plus depreciation. If there is large cash flow means that there is enough liquidity so the firm holds less cash. However, they had found no 16 evidence to support this relation. It is clear to say that short-term and long-term abilities are balance with large cash flow. According to the trade-off model, cash flow provides a ready source of liquidity. Kim et al. (1998) alleged that cash flow could be seen as cash substitutes. Consequently, they expected that there is a negative relation between cash flow and cash holdings. When cash flow is high, means that the operating activities are going well and positive cash flow can buffer for any loss from financing or investing activities. The firm can invest more in order to grow, so that the firm has to hold more cash, Ferreira & Vilela (2004) confirmed one more time. Likewise, when there is high cash flow are expected to hold more cash because the firm prefer internal finance more than external finance, Ozkan & Ozkan (2004) analyzed. It helps to reduce cost much. Due to pecking order theory, these arguments strongly pointed out that there were a positive relation between cash flow and cash holding. In Vietnam, Phuong (2013) did research on special factors that affect cash level of 125 listed firms in Viet Nam from 2009 to 2012. As author’s result, seven factors that affects to level of cash in those firms are size of business, leverage, bank liability, cash flow, cash flow violability, and dividend. The pecking order theory is concluded as the most suitable theory to explain for changes on corporate cash management. 2.3.5. Cash Flow Volatility According to trade-off theory, cash holding can be very important for a company when it is suffering because of lower cash flows or worse firm circumstances. Bigelli and Sanchez (2010); and Ozkan and Ozkan (2004) dealt with that literature therefore expected a positive relation between volatility of cash flows and cash holding. The greater the volatility of cash flow, the greater the possibility that the firm will be short of liquid assets. It is easy to see that if the firm has to pass up some valuable growth opportunities, it will be costly to be short of cash. Therefore, firms with high cash flow volatility hold more cash in order to avoid the expected 17 costs of liquidity constraints, Ozkan & Ozkan (2004) concluded. Another supporting is that, Ferreira & Vilela (2004) also found evidence that there is a positive relation between cash flow volatility and cash holding in the EMU countries. Firms with more cash flow volatility faced a higher probability of experiencing cash shortages due to unexpected bad cash flows, they supported. In their paper, Opler et al. (1999) also had found evidence that US firms with more cash flow volatility hold larger amount of cash. Two remaining theories do not much relate to cash flow volatility. Megginson and Wei (2012) studied the relation between state ownership and cash holdings in China’s share-issue privatized firms from 1993 to 2007. In their research, level of cash holding had negative relation with size of business, financial leverage, and net non-fixed assets, but had positive relation with cash flow fluctuation. In Vietnam, Khanh (2013) did support the opinion that many firms in samples held much cash, because of economic crisis and fluctuation. Firms held cash to avoid risks, not because of finding good opportunities. They balance their operations and focus on their strength, not diversify investment. The result showed that cash flow, cash flow volatility, level of leverage, and debts were factors affecting to cash level. 2.3.6. Liability to Bank Based on trade-off theory, Ferreira and Vilela (2004) provided evidence that EMU firms had a closer relationship with banks. In EMU countries, banks owned a significant proportion of firm’s stock; they provided more. In addition, according to Krivogorsky, Grudnitski, and Dick (2009), firms in Continental Europe relied more on bank debt than bonds for their external funds. It seem that firms that rely on bank loans as major source of financing are less likely to experience agency and asymmetric information problems associated with other kinds of debt. Ferreira & Vilela (2004) argued that banks were in a better position to evaluate the firm’s credit quality and to monitor and control the firm’s financial policies. Pinkowitz and Williamson 18 (2001) used industrial firms from the Germany, U.S, and Japan to explain effects of bank power on cash holdings. The first notice is that Japanese firms hold a greater percentage of their assets in cash, which is slightly greater than the U.S. The author recorded that Japanese cash balance were affected by the monopoly power of banks. During periods of powerful banks, firms’ high cash holdings are steady with banks extracting charges. The explanation also suggested for this finding is that banks encourage firms to maintain large cash balances in order to extract charges from firms or to reduce their monitoring costs. When banks weakened, because of war, Japanese cash levels became more like U.S. firms. This author went to conclusion that strong Japanese banks persuade firms to hold large cash balances. This is contrary to widely held beliefs about the Japanese governance system. According to Ozkan and Ozkan (2004), banks could minimize the information costs and could get access to information not otherwise publicly available. Banks could monitor borrower’s private information more effectively than other lenders. When a bank provide a loan or renew a loan to a firm means that there is positive information about that firm. So when a firm has bank debt means that it decreases the probability of experiencing financial distress. In this case, firms with bank debt hold less cash. It seems that it is a secure and very popular relationship. According to pecking order theory, Ferreira and Vilela (2004) noted that bank debt was negative related to cash holding of a firm for precautionary reasons. It is expected that firms that rely on bank loans as major source of financing are less likely to experience agency and asymmetric information problems associated with other kinds of debt. This is because banks are in a better position to evaluate the firm’s credit quality and to monitor and control the firm’s financial policies. Finally, Ferreira and Vilela (2004) gave a clear opinion that because agency and asymmetric information problems were a source of significant indirect financing costs, which may limit the access to capital markets, one would expect that firms with a greater proportion of bank debt had less cash holdings for precautionary motives. 19 Same evaluation as leverage factor, relying on cash flow theory, Ferreira and Vilela (2004) concluded that firms with a good relation with banks were more subject to monitoring, which could decrease superior managerial discretion. This means that firms with bank debt hold less cash. In Vietnam, according to Trinh (2012), firms which had more debts and long-term debts, did hold less cash. 2.3.7. Dividend Payout In relation with trade-off theory, according to Ferreira and Vilela (2004); Ozkan and Ozkan (2004); and Pinkowitz and Williamson (2001), there is a negative relation between dividend payment and cash holdings. Their viewpoints are that firm pays dividends could afford to hold less cash when they are more capable of raising funds when needed by cutting dividend. The same opinion, a firm that does not pay dividends had to use the capital markets to raise funds, Opler et al. (1999) and Ferreira and Vilela (2004) agreed. Therefore, an important job of top management level is to balance profit and price of common shares. Current volume and price of common share are the positive input factor dividend paid-out, however profit is the key factor to decide how much firm gave out. Hereby is the summary table of above related studies, which provided ideas, concepts, and theories for this study. Table 1: Summary of Previous Studies Author Kim et al. Research Source of Research Result Year Data 1998 U.S Level of cash holding has positive relation with investment opportunity and cash flow; however it has negative relation with size of business, financial leverage, cash conversion cycle and financial crisis 20 possibility. Opler et al. 1999 America Group businesses of high growth opportunity, low risks, or small size hold more cash than other businesses. Simultaneously, firm which easily got entrance to capital market holds less cash than others did. Pinkowitz 2001 U.S, Japanese firms hold more cash than others do in and German German and the U.S. The author noted that cash Williamsion and Japan balance of Japanese firms did effected by bank’s monopoly power. Ferreira and 2004 EMU Vilela Research’s result showed the positive relation between the level of cash holding and investment opportunity; and has negative relation with total assets that not include cash, leverage and size of business. Ozkan and 2004 UK Ozkan The author did find evidences of inverse correlation between ownership structure and level of cash holding. Level of cash has positive relation with cash flow from operations and growth opportunity, but it has negative relation with non-fixed assets and liabilities to bank. Custodio, et 2005 al. U.S The author found that level of cash holding increased during economic crisis. Firms hold more cash or kept cash balance stable while risk-free interest in short-term went down. 21 Bates et al. 2009 U.S The author found some main reasons for high level of cash holding, such as decrease in inventory (decrease in non-fixed assets); high fluctuation of cash flow; increase in investment opportunity and R&D expenditures. Meggison 2012 China and Wei Level of cash holding had negative relation with size of business, financial leverage, and net nonfixed assets, but had positive relation with firm profit, growth opportunity, expenditures and cash flow fluctuation. Trinh 2012 Vietnam Vietnamese companies did not concentrate on opportunity cost. The author explained that in the testing period, the market is volatile, so big firms did tend to hold more cash. Firms, which had more debts and long-term debts, did hold less cash. Khanh 2013 Vietnam Cash flow, cash flow volatility, level of leverage, credit limit, outsider finance, debts, business cycle, economic risks and investment opportunity were factors affecting to cash level. Firms held cash to avoid risks, not because of finding good opportunities. Phuong 2013 Vietnam Seven factors that affects to level of cash in firms are size of business, leverage, bank liability, cash flow, cash flow violability, opportunity and dividend. 22 The pecking order theory is concluded as the most suitable theory to explain for changes on corporate cash management. 23 Chapter 3. Research Methodology This chapter will explain the theoretical and logical thinking of the author regarding the approach and structure of the thesis 3.1. Data Sources In this research, the author has used secondary sources of data. The author has collected data of the selected firms and design own database structure. The data is collected from official website of Ho Chi Minh (HOSE) and Ha Noi (HNX) stock exchange and Cafef website (http:// http://cafef.vn/). Those data are considered by the author to be reliable due to the requirement of publicly companies listed to be independently audited. Financial statements of selected companies are collected. Companies listed in both stock exchanges are selected as followings criteria. Firstly, Companies must be listed in one of two stock exchanges at least five years back from 2014. Five years are selected because of time limit and long enough to do thesis. Secondly, companies which operates in field of finance, bank, insurance, stock market, real estate and energy. Thirdly, construction or energy companies, which operate on trading real estate or hotel, are also eliminated. The financial condition of those firms has under special control of government and has different kind of operations. 98 listed companies were finally selected. The main industries of companies listed are clearly defined in a stated document of government. It is called VSIC 2007. Revenue is the united reason to decide what company is in what industry. Twenty one industries are showed in Pooled OLS regression analysis below for united presentation purpose only. The sample size is reviewed again to ensure that they are enough for thesis analysis. Since multi-regression is mainly used in this thesis, practical theory of Tabachnick and Fidell, (1996) guided that minimum sample size could be calculated as formulas 50 + 8*m (in which m: 24 number of independent variables. In this thesis, the author will tend to use 33 variables, so minimum sample size is 314. In reality, total sample size of thesis is 490, as formulas 98*5 (in which 5: five years). Therefore, current sample size of thesis can be satisfied. Data set is gathered from financial reports, and then grouped as tabular table. The purpose is mainly using for below statistics, such as descriptive statistics, unvariate analysis, and four multi-regression. As following empirical models of previous researches, the author uses six variables, including size of firm, leverage, bank debt, cash flow, liquidity, and dividend. The author combines Stata 12 software to generate complex statistics such as percentile, quartiles, Rsquare, P-value, F test, and regression data. The author also uses Excel 2007 as supporting tool for Stata 12 to compute and draw out tendency of cash holding over the years. Software has specific strength and then when the author combines, they will save much time for this thesis. Other than the created database, a number of other secondary sources have been used. There are both advantages and disadvantages with having secondary sources according to Bryman & Bell (2011). Firstly, it has saved both cost and time in the data collection process. Secondly, the data collected has higher quality than the author would have been able to produce by myself and finally it allows the author to conduct their study over a number of years in the past. The disadvantage is that the author might lack familiarity with the data since they has not collected the data by themselves, might lack understanding on the complexity of the data and do not has any control over the data (Bryman & Bell, 2011, p.320-321). However, the data collected was of the nature that the author had previously encountered academically. Given the advantages and disadvantages, the author has chosen to use the following secondary sources. The historical database of stock closing prices of the listed firms that has been investigated is a widely used database when it comes to business research and has a high quality of data. Stock prices at the end of each companies and each year are collected from Cafef website (http:// http://cafef.vn/). Since the author is a student at Ho Chi Minh International University, the author has authorized to buy the access to the university library database and 25 related academic link for relevant literature. Keywords that has been used in searches has ranged from, cash holdings, cash to corporate finance. In addition to articles, a wide variety of textbooks has been used in this thesis. These books are mostly related to business research, cash, cash holding, statistics and corporate finance. 3.2. Data Statistics This part says clearly about how statistics and models will be used for analyzing dataset to develop results. They include three main methods. They are descriptive statistics, univariate analysis, and fours models of regression. Reliability and validity of variables and the united techniques of the data treatment will be concerned carefully. 3.2.1. Descriptive Statistics Descriptive statistics will be completed basically by computing quantities from research data. The author uses percentile and quartiles to measure the tendency of data set and observe the balance of entire distribution of data points. The author will essentially use the mean and median. A graphical technique, area chart is used to extend and visualize data set and then help the author to see the initial tendency of cash holding. The author combines Stata 12 software to generate percentile and quartiles and Excel 2007 to compute and draw out tendency of cash holding over the years. 3.2.2. Univariate Analysis In statistics, the Pearson correlation is a measure of the linear correlation (the dependence) between two variables X and Y. It was developed by Karl Pearson from a related idea introduced by Francis Galton in the 1880s. Checking univariate will be done by comparisons of independent variables on the quartile points, through the observation on percentages of cash to the total assets. Built quartile for each year in period, will help to explain why their positions are maybe overlap. The construction and comparison across quartiles are to 26 check whether there is a significant difference statistically between the company which has large cash balance (the fourth quartile) and small balance (first quartile). T-statistic is to test hypotheses about statistically significant differences between fourth quartiles and first quartiles, as Opler et al. (1999) and Ferreira and Vilela (2004) wrote down. To calculate the quartiles and t-statistics as well as p_value, the author uses Stata 12 software. According to Pearson (1895), correlation coefficients between variables, go from -1 to 1 and closer to 1 means strong correlation. A negative value indicates an inverse relationship. In Stata 12, author can use the command “pwcorr Y X1 X2... Xn, star (0.05) sign” or pairwise correlation tool available. The confidence level is 95%. Graphical presentation of those correlation is shown by command “graph matrix Y X, half maxis(ylabel(none) xlabel(none))”. 3.2.3. Regression Analysis Opler et al. (1999), Ferreira and Vilela (2004) and Pinkowitz and Williamson (2001) had used regression models in their empirical studies, includes: models of Fama-MacBeth, crosssectional regression using average values, pooled OLS regression with year dummies and finally pooled OLS regression with both year and industry dummies. We can see selected regression models in figure 2 below. 27 Figure 2: Selected Regression Models In the study, the author will use the same four regression models to find the correlation between the level of cash hold and specific factors of the listed firms in Vietnam stock market. In each regression model, the null and alternative hypotheses will be set up to test about the linear correlation coefficient between level of cash holding and factors. The null hypotheses is Ho: β = 0 (the linear correlation coefficient is zero) and alternative hypotheses is H1: β ≠ 0 (the linear correlation coefficient is different from zero). These tests are two-tailed. P-value is also used to support final conclusion making. The author will assume that variables cash and factors are normally distributed. Since each regression has different pros and cons, so that combination of four regressions will help to explain the relation between variables better. 3.2.3.1. Fama-MacBeth Regression Model In a research in 1973, Fama-MacBeth model is the financial model used in the study, "Risk, profitability and balance: the empirical test" which was performed by two authors, Fama and MacBeth. On the other studies of factors affect the level of cash holdings, these authors had also used this approach. Under this method, the cross-sectional regressions will be run for each year. This approach eliminates the problem of residual correlation sequence that be often found in time series data regression over time. The Fama-MacBeth model was effective when viewed every year as an independent cross-sectional data, as Opler et al. (1999) confirmed. The author will conduct regression over two steps. First step, the author runs cross-sectional data for each year separately. Study period are five years (from 2009 to 2013), so model will be performed fifth times. Where the regression model’s variance has changes, the author will use the Robust option in Stata 12 to fix. Second step, the author will compare the t-statistic as well as p_value well as correlation coefficient of 5 result to reach conclusion. Cross-sectional regression for each year is determined as follows: 28 CASHi= β0+ β1SIZEi+ β2LEVi+ β3BANKDEPTi + β4CFLOWi + β5CFVOLATi+ β6LIQi+ β7DIVIDENDi+ ui Where i = 1, 2, …., N. And where: CASHi is the amount of cash or cash equivalence hold by firm i. SIZEi is the natural logarithm cash over total assets of firm i. LEVi is ratio of the total liabilities over total assets of firm i. BANKDEPTi is the ratio of bank liabilities over total liabilities of firm i. CFLOWi is the ratio of cash flow over total assets of firm i. CFVOLATi is the ratio of average standard deviation of cash flow (from 2009 to 2013) over average total assets (from 2009 to 2013) of firm i. LIQi is the ratio of working capital exclude cash & equivalent over total assets of firm i. DIVIDENDi is the dummy variable of firm i with 1 value if dividend paid and 0 value if not paid. Ui is the variance of firm i. 3.2.3.2. Cross-Sectional Regression Model with Mean Value of Variables In this regression model, each variable of each firm is average value for period 2009-2013. OLS regression is run in Stata 12. Cross-sectional regression for each year is determined as follows: CASHi= β0+ β1SIZEi+ β2LEVi+ β3BANKDEPTi + β4CFLOWi + β5CFVOLATi+ β6LIQi+ β7DIVIDENDi+ ui Where i = 1, 2, …., N. And where: CASHi is the amount of cash or cash equivalence hold by firm i. SIZEi is the natural logarithm cash over total assets of firm i. LEVi is ratio of the total liabilities over total assets of firm i. BANKDEPTi is the ratio of bank liabilities over total liabilities of firm i. CFLOWi is the ratio of cash flow over total assets of firm i. CFVOLATi is the ratio of average standard deviation of cash flow (from 2009 to 2013) over average total assets (from 2009 to 2013) of firm i. LIQi is the ratio of working capital exclude cash & equivalent over total assets of 29 firm i. DIVIDENDi is the dummy variable of firm i with 1 value if dividend paid and 0 value if not paid. Ui is the variance of firm i. 3.2.3.3. Pooled Ordinary Least Squares Regression Model with Year Dummy Pooled OLS is a variant of OLS regression. Dummy variables are added to OLS regression to see the result over the years. This action also makes samples increased. Ferreira and Vilela (2004) used in their research with the same purpose. The author uses closing stock prices at the end of each year for year dummies. Pooled OLS regression for each year is determined as follows: CASHit= β0+ β1SIZEit+ β2LEVit+ β3BANKDEPTit + β4CFLOWit + β5CFVOLATit+ β6LIQit+ β7DIVIDENDit+ β82009it + β92010it + β102011it + β112012it+ β122013it + uit Where i = 1, 2, …., N; t= 1, 2,…., Ti And where: CASHi is the amount of cash or cash equivalence hold by firm I in year t. SIZEi is the natural logarithm cash over total assets of firm i in year t. LEVi is ratio of the total liabilities over total assets of firm i in year t. BANKDEPTi is the ratio of bank liabilities over total liabilities of firm i in year t. CFLOWi is the ratio of cash flow over total assets of firm i in year t. CFVOLATi is the ratio of average standard deviation of cash flow (from 2009 to 2013) over average total assets (from 2009 to 2013) of firm i in year t. LIQi is the ratio of working capital exclude cash & equivalent over total assets of firm i in year t. DIVIDENDi is the dummy variable of firm i with 1 value if dividend paid and 0 value if not paid. Uit is the variance of firm i in year t. β82009it is the dummy variable of year 2009, with 1 is year 2009 and 0 is not year 2009. β92010it is the dummy variable of year 2010, with 1 is year 2010 and 0 is not year 2010. β102011it is the dummy variable of year 2011, with 1 is year 2011 and 0 is not year 2011. β112012it is the dummy variable of year 2012, with 1 is year 2012 and 0 is not year 2012. β122013it is the dummy variable of year 2013, with 1 is year 2013 and 0 is not year 2013. 30 3.2.3.4. Pooled Ordinary Least Squares Regression Model with Year Dummy and Industry Dummy This method is similar to Pooled Ordinary Least Squares Regression (Pooled OLS) with year dummy. Industry dummy variables are added to OLS regression to see the change of result if have due to the change of industry. Beside above dummies, 21 industry dummies are in the list below. Total variables will be 33. Table 2: List of Industry Dummies Ord. Industry 1 Wholesaler Trading 2 Retailer Trading 3 Textiles - Leather 4 Chemicals - Pharmaceutical 5 Service activities related to transportation 6 Other mining and quarrying 7 Exploitation, processing, purchase ore 8 Warehouse 9 Processing Metal products and non-metallic mineral 10 Machinery - Vehicles 11 Furniture and related products 12 Paper Products and Printing 13 Processing Other products (medical equipment, toys, jewelry, ...) 14 Wood products 15 Products from plastic and rubber 16 Electrical equipment - Electronics - Telecommunications 31 17 Food - Beverage - Tobacco 18 Cultivation 19 Waterway Transport 20 Telecommunications 21 Construction Pooled OLS regression for each year is determined as follows: CASHit= β0+ β1SIZEit+ β2LEVit+ β3BANKDEPTit + β4CFLOWit + β5CFVOLATit+ β6LIQit+ β7DIVIDENDit+ β82009it + β92010it + β102011it + β112012it+ β122013it + β13Industry1it + β14Industry2it + β15Industry3it + β16Industry4it + β17Industry5it + β18Industry6it + β19Industry7it + β20Industry8it + β21Industry9it + β22Industry10it + β23Industry11it + β24Industry12it + β25Industry13it + β26Industry14it + β27Industry15it + β28Industry16it + β29Industry17it + β30Industry18it + β31Industry19it + β32Industry20it + β33Industry21it + uit Where i = 1, 2, …., N; t= 1, 2,…., Ti And where: CASHi is the amount of cash or cash equivalence hold by firm I in year t. SIZEi is the natural logarithm cash over total assets of firm i in year t. LEVi is ratio of the total liabilities over total assets of firm i in year t. BANKDEPTi is the ratio of bank liabilities over total liabilities of firm i in year t. CFLOWi is the ratio of cash flow over total assets of firm i in year t. CFVOLATi is the ratio of average standard deviation of cash flow (from 2009 to 2013) over average total assets (from 2009 to 2013) of firm i in year t. LIQi is the ratio of working capital exclude cash & equivalent over total assets of firm i in year t. DIVIDENDi is the dummy variable of firm i with 1 value if dividend paid and 0 value if not paid. Ui is the variance of firm i in year t. β92009it is the dummy variable of year 2009, with 1 is year 2009 and 0 is not year 2009. β102010it is the dummy variable of year 2010, with 1 is year 2010 and 0 is not year 2010. 32 β112011it is the dummy variable of year 2011, with 1 is year 2011 and 0 is not year 2011. β122012it is the dummy variable of year 2012, with 1 is year 2012 and 0 is not year 2012. β132013it is the dummy variable of year 2013, with 1 is year 2013 and 0 is not year 2013. β14Industry1it is the dummy variable of industry 1. β15Industry2it is the dummy variable of industry 2. β16Industry3it is the dummy variable of industry 3. β17Industry4it is the dummy variable of industry 4. β18Industry5it is the dummy variable of industry 5. Β19Industry6it is the dummy variable of industry 6. β20Industry7it is the dummy variable of industry 7. β21Industry8it is the dummy variable of industry 8. β22Industry9it is the dummy variable of industry 9. β23Industry10it is the dummy variable of industry 10. β24Industry11it is the dummy variable of industry 11. β25Industry12it is the dummy variable of industry 12. β26Industry13it is the dummy variable of industry 13. βΒ27Industry14it is the dummy variable of industry 14. β28Industry15it is the dummy variable of industry 15. β29Industry16it is the dummy variable of industry 16. β30Industry17it is the dummy variable of industry 17. β31Industry18it is the dummy variable of industry 18. β32Industry19it is the dummy variable of industry 19. β33Industry20it is the dummy variable of industry 20. β34Industry21it is the dummy variable of industry 21. 3.2.3.5. Fixed Effects Regression Model In panel data analysis, the term fixed effects is used to refer to an estimator for the coefficients in the regression model. If we assume fixed effects, we impose time independent effects for each entity that are possibly correlated with the regressors. The panel data of 98 firms is presented from 2009s to 2013s with independent variables to analyze the impact of variables that vary over time. First step is to convert time-series data into panel data by using the command xtset “xtset comp year” then use the stata command “xtreg” to run the fixed effects. Firms “COMP” represents the entities or panel (i) and “year” represents the time variable (t). Panel variable refers to “strongly balanced”. It means to the fact the all firms have data for all 33 five years. If, one firm does not have data for one year, then the data is unbalanced. Ideally, people want to have a balanced dataset but this is not always the case, however you can still run the fixed effects model. Fixed effects regression model is determined as follows: CASHit= β0+ β1SIZEit+ β2LEVit+ β3BANKDEPTit + β4CFLOWit + β5CFVOLATit+ β6LIQit+ β7DIVIDENDit+ αi + uit Where i = 1, 2, …., N; t= 1, 2,…., Ti And where: CASHi is the amount of cash or cash equivalence hold by firm i in year t. SIZEi is the natural logarithm cash over total assets of firm i in year t. LEVi is ratio of the total liabilities over total assets of firm i in year t. BANKDEPTi is the ratio of bank liabilities over total liabilities of firm i in year t. CFLOWi is the ratio of cash flow over total assets of firm i in year t. CFVOLATi is the ratio of average standard deviation of cash flow (from 2009 to 2013) over average total assets (from 2009 to 2013) of firm i in year t. LIQi is the ratio of working capital exclude cash & equivalent over total assets of firm i in year t. DIVIDENDi is the dummy variable of firm i with 1 value if dividend paid and 0 value if not paid. αi (i=1…n) is the unknown intercept for each entity. Ui is the variance of firm i in year t. 3.2.4. Other Supported Tests for Regression Models Each regression model has advantages and disadvantages, however its regression result cannot show possibilities of high correlation among variables; the heteroscedasticity in a linear model; or multicollinearity. These tests were designed to give the signal for the author, then in order to reduce negative effects on regression models above and to support for the conclusion. As Kumar (1975), multicollinearity (also collinearity) is a statistical phenomenon in which two or more independent variables in the multiple regression model are highly correlated. According to Farrar (1967), multicollinearity does not reduce the predictive power or reliability of the model as a whole, at least within the sample data themselves; it only affects calculations 34 regarding individual predictors. A high degree of multicollinearity can also prevent computer software packages from performing the matrix inversion required for computing the regression coefficients, or it may make the results of that inversion inaccurate. An important assumption for the multiple regression model is that independent variables are not perfectly multicolinear. One regressor should not be a linear function of another. When multicollinearity is present standand errors may be inflated. Stock and Watson (2007) said that Stata software will drop one of the variables to avoid a division by zero in the OLS procedure. The Stata command to check for multicollinearity is VIF (variance inflation factor), and is used right after running the regression. A vif > 10 or a 1/vif < 0.10 indicates multicollinearity. In statistics, the Breusch–Pagan test is used to test for heteroscedasticity in a linear regression model. It tests whether the estimated variance of the residuals from a regression are dependent on the values of the independent variables. The null hypothesis is that variances across entities is zero. This is, no significant difference across units, or no panel effect. The command in Stata is xttset0 type it right after running the random effects model. 3.3. Determined Dependent and Independent Variables 3.3.1. Dependent Variable – Cash In order to measure the ratio of cash, the authors used two common measures. The first measure is the most popular and used, ratio of cash, cash equivalents divided by the book value of total assets. This measure was first used by Kim at el. (1998), followed by the authors Ozkan and Ozkan (2004), and Bates et al. (2009). The second is a measure for the natural logarithm of the ratio of cash, cash equivalents and total assets excluding. It was the first used by Opler et al. (1999), followed by the authors Pinkowitz and Williamson (2001) and Custodio, et al. (2005). In this research, the author will use the ratio of cash and cash equivalents divided by the book value 35 of total assets, as to Kim et al.(1998), as the dependent variable to measure the correlation between the amount of cash holdings and the specific factors of the firm. 3.3.2. Independent Variables Compiled from the predictions of factors that affect the amount of cash holding from three financial theories - tradeoff theory, pecking order theory, the theory of free cash flow-and the empirical studies of the previous authors, Kim et al. (1998); Opler et al. (1999); Pinkowitz and Williamson (2001); Ferreira and Vilela (2004); Ozkan and Ozkan (2004); Custodio, et al. (2005); and Bates et al. (2009). The author draws the factors that affect the amount of listed corporate cash holdings are: firm size, financial leverage, bank debt, cash flow, cash flow volatility, liquid assets not including cash, the chances of dividend payments. These variables serve as independent variables to explain the variation amount of the firm holdings. 3.3.2.1. Size of Business (SIZE) Firm size is measured by taking the natural logarithm of the total assets of each firm in the last year of the study period and was used as constants for all years in the studied period (Opler et al. (1999); Pinkowitz and Williamson (2001); Ferreira and Vilela (2004); Ozkan and Ozkan (2004); and Custodio, et al. (2005)). When be converted to logarithm, the number and data become significant figures. Total property is very large, however the logarithm makes it smaller, more compact, reducing the variance, and making the normally distributed variables. In this research, firm sizes are calculated by taking the natural logarithm of the total assets over the firm in 2013 and used this data as constants for the year from 2009 to 2013. Total asset balance amount is collected on their balance sheet at the end of the financial year. 3.3.2.2. Bank Debt (BANKDEBT) As Ferreira and Vilela (2004), and Ozkan and Ozkan (2004), bank debt is calculated as the ratio of total bank debt divided by total debt. As the same way of calculation, the author uses 36 the ratio of bank debt as the way of previous authors. The balance of total debt, both short-term loan and long-term loans, will be collected on companies’ balance sheets at the end of the financial year. Debt is collected from payable notes on the financial statements. The author excludes firms which have not present or not present clear information of bank debt. 3.3.2.3. Cash Flow (CFLOW) Cash flow variables are represented by the ratio of cash flow divided by total assets, as the way that Ozkan and Ozkan (2004) and Bates et al. (2009) did. Cash flow is calculated by taking the profit before tax, interest expense and depreciation (EBITDA) then deducting taxes, interest expense and dividends. The author calculates cash flow ratio as the way previous authors had done. 3.3.2.4. Cash Flow Violability (CFVOLAT) Fluctuations in cash flow variables are calculated by dividing the standard deviation of cash flow by average total assets. It was way that two authors, Kim et al. (1998) and Ozkan and Ozkan (2004) used. The author calculates the ratio cash flow volatility by taking the standard deviation of cash flow in five years (2009 - 2013) divided by average total assets of five years (2009-2013). Total assets were collected from firms’ balance sheets at the end of the financial year. 3.3.2.5. Liquid Assets (LIQ) Liquid assets variable is represented by the ratio of net current assets minus cash and cash equivalents divided by total assets. It was way that two authors, Ozkan and Ozkan (2004 ) and Bates et al. (2009) did use. The author now uses the same way like above. Current assets, current liabilities, cash and cash equivalents, total assets were collected from their balance sheet at the end of the financial year. 3.3.2.6. Dividend (DIVIDEND) 37 Dividend dummy is composed of two sets of value, 0 and 1. If within financial year, firms have to pay dividends, the dividend variable takes the value 1 and the value 0 if not paid in dividends. It was the way of Opler et al. (1999), Pinkowitz and Williamson (2001), Ferreira and Vilela (2004), Ozkan and Ozkan (2004), Custodio, et al. (2005) and Bates et al. (2009). The author determines value of dividend dummies by seeing dividend paid in the cash flow statement, items dividends and interest paid to the owners. In case of having dividend payments, dividend dummy variable takes the value 1. Conversely, variable takes value 0. Table 3: Summary of Formulas for Variables Dependent Variable Calculation Formulas CASH-Cash ratio Cash and equivalent / total assets Independent Variables SIZE-Size of business Natural logarithm of total assets LEV-Financial leverage Total liabilities / total assets BANKDEBT-Bank Debts Bank liabilities / total liabilities CFLOW-Cash flow (EBITDA – interest – corporate tax – dividend) / total assets CFLOWAT-Cash violability LIQ-Liquidity exclude cash flow Average standard deviation of cash flow (2009-2013)/ average total assets (2009-2013) Assets, (Current assets – current liabilities – cash & equivalent) / total assets DIVIDEND-dummy dividend If dividend paid, variable value is 1. If not, it is 0. paid 38 Chapter 4. Data Analysis and Results The findings present the research result respectively research methodology is presented in chapter three, including descriptive statistics, univariate analysis, and regression methods. The first section presents the preliminary analysis of the data, the descriptive statistics, and tendency of dependent variable, univariate outliers, normality, and examination of relationship between variables. The second section analyzes the hypotheses and then presents analyses of the proposed models. Data were analyzed using Stata version 12 and Excel version 2007. 4.1. Descriptive Statistics Descriptive statistics are very important and meaningful for this thesis allows simpler interpretation because dataset is large and hard to visualize. In table 05, we can see easily that average rate of cash holding over total assets of 98 firms over five years is 11.9%, value 89 billion VND (bVND). It is quite a big balance at the end of year. Additionally, mean of leverage and bank debt ratio show that firms’ borrowing ratio is only 25.2%, however debt from banks is quite high, 83.9% if any. Table 4: Results from Descriptive Statistics variable mean p25 p50 p75 sd CASH SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND .1192199 26.56168 .252598 .8399878 .0854143 .0534755 .1365245 .8163265 .0339252 25.818 .069 .647 .031 .021 -.001 1 .0796032 26.527 .225 1 .071 .038 .1305 1 .1748642 27.3 .417 1 .123 .06 .263 1 .1160025 1.184751 .2108854 1.662004 .1111144 .0632884 .5123252 .3876134 39 Dividend paid-out ratio is 81.6%, even when firms bore losses. The ratio of loss is nearly 7%. Liquidity ratio is in safe range of normal firm, 13.6%. All descriptive statistics show that firms keep high level of cash at the end of year. Figure 3: Cash Ratio Over Five Years In the figure 3, tendency of cash holding decrease slightly from 2009, but it did increase from 2012. According to a statistics, Rodion (2012) observed that the Vietnam's economic growth has slowed down to between four and five percent. That is not enough to create jobs for the rapidly growing population and the rising number of people entering the job market. Hoang (2012) said that the Vietnam’s leading experts have voiced concern that Vietnam is entering into a fresh battle to stave off yet another economic downturn, citing the central bank’s latest ceiling deposit rate reduction. Hoang (2012) also cited statement of Cao Sy Kiem, chairman of the Association of Small and Medium Enterprises, said small and medium firms were still paying an annual lending rate of 18% to 19% even though the ceiling deposit rate has been lowered to 12%. The mean of cash holding ratio over 5 years is reasonable if we observe that economic condition last 5 years. Firms have had tried to keep cash available and stable, but also must limited cash out flow in operations due to economic downturn. 40 Figure 4: % Standard Deviation of Cash Flow over Total Assets Figure 04 describes the downward tendency of cash flow. This ratio is calculated by taking standard deviation of cash flow in 5 years divided total assets of each year. This tendency was positive with level of cash holding in figure 03. The ratio decreased rapidly from 2009-2011s, and slowly went down in 2012s and 2013s. A notice was that while level of cash holding went down, the total assets increase, average 11%. It seems that firms accumulated assets and expand business even downturn of economy in recent years. 41 4.2. Univariate Analysis Each quartile has 98 firms. Most of independent variables, which are belonging to first or fourth quartile, are significant statistics with confidence level 95%. Table 5: Results from Univariate Analysis Variable Obs CASH 490 SIZE 490 LEV 490 BANKDEBT 490 CFLOW 490 CFLOVAT 490 LIQ 490 DIVIDEND 490 Percentile 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 Centile .0004672 .0339191 .0796032 .1748773 .9225227 23.244 25.8065 26.527 27.30025 30.761 0 .068 .225 .419 .757 0 .645 1 1 34.804 -.621 .031 .071 .123 1.029 .002 .021 .038 .06 .531 -5.984 -.00225 .1305 .264 8.414 0 1 1 1 1 Binom. Interp. [95% Conf. Interval] .0004672 .0294214 .0722549 .1493352 .9225227 23.244 25.64807 26.40561 27.20596 30.761 0 .0384379 .1984122 .3779768 .757 0 .2008758 1 1 34.804 -.621 .025 .063 .1089884 1.029 .002 .0184379 .035 .057 .531 -5.984 -.0335621 .111 .232 8.414 0 1 1 1 1 .0004672* .039873 .0912877 .1926788 .9225227* 23.244* 25.92701 26.71359 27.41712 30.761* 0* .0880116 .2483918 .452 .757* 0* .8050347 1 1 34.804* -.621* .037 .079 .13 1.029* .002* .023 .0421959 .0665621 .531* -5.984* .023 .1491959 .281 8.414* 0* 1 1 1 1* 42 Comparing the average of variables over first and fourth quartile is much significant when cash ratio changes. So, the author needs to run regression to observe more in later. However, in this step, the author runs another descriptive statistics to see the relationship between variables. Pairwise correlations of variables were run with confidence level 95% for displaying coefficients in the table 06. Table 6: Correlations Matrix of Variables from Unvariate Analysis CASH CASH SIZE LEV BANKDEBT CFLOW CFLOVAT DIVIDEND SIZE LEV BANKDEBT 1.0000 -0.1075* 1.0000 -0.3280* 0.3786* 1.0000 -0.1299* 0.0710 0.0415 1.0000 0.3110* -0.0856 -0.2340* -0.0683 0.1469* -0.2188* -0.1312* -0.0775 0.1498* 0.0913* -0.1200* -0.0021 CFLOW CFLOVAT DIVIDEND 1.0000 0.3609* 1.0000 0.0930* -0.0405 1.0000 According to Tabacknick & Fidell (1996), when correlation coefficient is high, or reach to 0.8, the multicollinearity may happen, however the coefficients between independent variables are quite low, fluctuating from -0.328 to 0.3786. It helps to show that multicollinearity is low possibility to happen in the dataset. The highest correlation coefficient is 0.3786, which show the strong correlation between size of firms and leverage. By scanning the table 06, we can see that correlation between cash holding and other variables as followings:  Cash has positive relation with cash flow, cash flow volatility, and dividend.  Cash has negative relation with size of firm, leverage, and bank debt. Below graph matrix produces a graphical representation of the correlation matrix by presenting a series of scatter plots for all variables as below to support for table 06: Figure 5: Graphical Correlation Matrix 43 CASH SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND 4.3. Regression Results 4.3.1. Fama–MacBeth Regression Model In this regression model, the confidence level is 95%. Table 07 is the result on below. Table 7: Results from Fama – MacBeth Regression Model 44 Number of obs 490 F( 8, 481) 13.49 Prob > F 0 R-squared 0.2003 Root MSE 0.10459 CASH Coef. SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND Y2009/10/11/12/13 _cons Number of obs 490 F( 8, 481) 13.47 Prob > F 0 R-squared 0.2009 Root MSE 0.10456 y2009 Robust Std. Err. t 0.001507 -0.1537 -0.0069357 0.2627144 0.0210049 -0.0281746 0.0326652 -0.0062815 0.1147423 0.0059038 0.0449128 0.0025485 0.0946552 0.1069747 0.0141362 0.0115613 0.0136272 0.1429969 0.03 -3.42 -2.72 2.78 0.2 -1.99 2.83 -0.46 0.8 Coef. Number of obs 490 F( 8, 481) 13.46 Prob > F 0 R-squared 0.1999 Root MSE 0.10462 y2010 Robust Std. Err. t 0.0002374 -0.1544514 -0.0069563 0.2611944 0.0194858 -0.0281493 0.0335371 -0.0091633 0.1127174 0.0057812 0.0444135 0.0025469 0.0923798 0.1074733 0.0141238 0.0116016 0.0109753 0.1394758 0.04 -3.48 -2.73 2.83 0.18 -1.99 2.89 -0.83 0.81 Coef. Number of obs 490 F( 8, 481) 14.32 Prob > F 0 R-squared 0.1999 Root MSE 0.10462 y2011 Robust Std. Err. t 0.0003384 -0.1544419 -0.0069034 0.2597771 0.0208069 -0.0280813 0.0327497 0.0008517 0.1086686 0.0057577 0.0440732 0.0025646 0.0926527 0.1068932 0.014188 0.0114906 0.0132549 0.137859 0.06 -3.5 -2.69 2.8 0.19 -1.98 2.85 0.06 0.79 Coef. Number of obs F( 8, 481) Prob > F R-squared Root MSE y2012 Robust Std. Err. t 0.0003853 -0.1546363 -0.0069145 0.2584282 0.0212768 -0.0279999 0.0327599 -0.0025653 0.1082372 0.0057987 0.044661 0.0025377 0.0930318 0.1066027 0.0142913 0.0115444 0.0119788 0.1391186 0.07 -3.46 -2.72 2.78 0.2 -1.96 2.84 -0.21 0.78 Coef. 490 13.67 0 0.2033 0.1044 y2013 Robust Std. Err. t -0.0000627 -0.1530942 -0.0072215 0.2601507 0.022222 -0.0277504 0.0345745 0.0172194 0.1143344 0.0057622 0.0445218 0.0024858 0.092137 0.1074552 0.0142288 0.0115988 0.0117352 0.1383883 -0.01 -3.44 -2.91 2.82 0.21 -1.95 2.98 1.47 0.83 R-squared is from 0.1999 to 0.2009. The low value of R-squared states that only ~20% of the total variation in level of cash were explained by variation of above factors. The low value of R-squared indicated that there might be many other important variables that contribute to the determination of level of cash holding. The result of five regressions show the similar result and correlation between variables. The null and alternative hypotheses are: Ho: β = 0 (the linear correlation coefficient is zero) and H1: β ≠ 0 (the linear correlation coefficient is different from zero). As mentioned in research methodology chapter, the author assumes that variables are normally distributed. From the alternative hypothesis, people know that this test is two-tailed. Hence, area in each tail of the t distribution equal 0.025 and degrees of freedom df are 4,408. From the t distribution table of Merrington (1941), the critical values of t are -1.960 and 1.960. In the figure run from Stata 12, the value of leverage, bank debt, cash flow, liquidity and dividend are in the rejection region. So that, the linear correlation coefficient between cash holding and leverage, bank debt, cash flow, liquidity and dividend are different from zero. While keeping other factors unchanged, level of cash holding was changed much due to change of two factors, leverage (negative) and cash flow (positive). In addition, when comparing p-value of 45 each factor with the significance level 0.05, p-value of firm’s size (~0.940) and cash flow volatility (~0.802) are greater then 0.05. As the result, the author rejects the null hypothesis. So, level of cash holding has no significant correlation with size of firms and cash flow volatility. The author uses the Stata command VIF to check for multicollinearity. Table 8: VIF Test for Fama-MacBeth Model . vif Variable VIF 1/VIF LEV CFLOW SIZE CFLOVAT LIQ DIVIDEND BANKDEBT 1.28 1.26 1.25 1.24 1.12 1.06 1.01 0.780779 0.793287 0.798763 0.809633 0.890777 0.945986 0.988317 Mean VIF 1.17 The mean vif (1.17) and each vif of variables are quite low, far to multicollinearity (10). The negative correlation between level of cash holding and liquidity (coefficient = negative 0.028) is consistent with the trade-off theory. On this theory, firm holds more liquidity, excluding cash and equivalence; balance of cash will go down. Level of cash holding has negative correlation with both leverage and bank debt. This result is supported by result of Ferreira and Vilela (2004). It means that borrow cash was not kept on hand, but was utilized for business operation. One of the above utilization is that firms did not only pay dividend, but also reserve cash for other purposes. 46 Figure 6: Percentage of Cash Holding in Years Firms Paid Dividend The figure 05 indicates that even firms paid dividend, the amount of cash reserved at the end of years is double amount of dividend paid out. The positive correlation of cash holding and dividend here is not strong for opinion of using excess cash to pay dividend. However, this is consistent with results of Opler et al (1999) and Custodio et al. (2005). The author continues to group dataset and use Excel 2010 to draw graph and show the correlation between cash holding and bank debt when firms’ income changes. It will help to clarify the correlation between cash and both leverage and bank debt. Figure 7: Correlations between Cash, Income and Bank Debt 47 Comparing to period of positive income, when firm was in negative income, the cash holding decreased much, from 12.5% to 4.4%, and also the bank debt increase from 74.3% to 83.5%. This is consistent with pecking order theory, when retain for earning decreases and cash holding cannot buffer for business activities, firms react to change and increase bank debt. However, this is oppose to indication of Kim et al (1998) that the lower profit, the higher cash holding is. As to Opler et al. (1999), because internal funds the firm often spent more cash than receiving cash, the firm decreases cash holding and raises debt. Positive relation between cash holding and leverage is consistent with cash flow theory. As opinion of Ferreira and Vilela (2004), firms with low leverage were less subject to monitoring, allowing for superior managerial discretion. This relation between cash holdings, debt, and internal resources suggested that there is a negative correlation between leverage and cash holdings. In summary, the correlation is consistent with all three theories. However, with low level of R-square (0.1999), the model would produce poor predictions, so the author should generate other regression models to measure how well the regression consistent with the data. 4.3.2. Ordinary Least Squares Regression with 5 Years Average R-squared of regression result is 0.4182 is quite high when compares with previous Fama-Macbeth regression model, 0.1999. It means that this model is more consistent with dataset, compared to Fama-Macbeth regression model. With this R-square, 41.82% changes of cash mean can be explained by the variation of independent variables. By another way, 59.18% changes of cash mean cannot be explained. The reason is because of unpredictable factors that are not concerned in this study. Table 9: Results from Ordinary Least Squares Regression with 5 Years Average 48 Equation CASH_MEAN Obs Parms RMSE "R-sq" F P 98 8 .0673208 0.4182 9.241256 0.0000 CASH_MEAN Coef. SIZE_MEAN LEV_MEAN BANKDEBT_MEAN CFLOW_MEAN CFVOLAT_MEAN LIQ_MEAN DIVIDEND_MEAN _cons .0014847 -.2104057 -.0162702 .3317434 -.1428694 -.0780228 .0287456 .1130525 Std. Err. .0067465 .042707 .0070669 .1083331 .1312201 .0273962 .0257753 .1733525 t 0.22 -4.93 -2.30 3.06 -1.09 -2.85 1.12 0.65 P>|t| 0.826 0.000 0.024 0.003 0.279 0.005 0.268 0.516 [95% Conf. Interval] -.0119184 -.2952507 -.0303098 .1165207 -.403561 -.1324503 -.0224615 -.2313426 .0148877 -.1255607 -.0022307 .5469661 .1178223 -.0235954 .0799527 .4574475 The null and alternative hypotheses are: Ho: β = 0 (the linear correlation coefficient is zero) and H1: β ≠ 0 (the linear correlation coefficient is different from zero). The confidence level is 95%. Area in each tail of the t distribution equal 0.025 and degrees of freedome df are 782. From the t distribution table of Merrington (1941), the critical values of t are -1.960 and 1.960. In the figure run from Stata 12, the value of leverage, bank debt, cash flow, and liquidity are in the rejection region. So that, the linear correlation coefficient between cash holding and leverage, bank debt, cash flow, and liquidity are different from zero. And the linear correlation coefficient between cash holding and firm’s size; cash flow volatility; and dividend are zero. With mean value of variables, this regression model shows that leverage, bank debt, cash flow, liquidity affect level of cash holding. In this model, the linear correlation coefficient between cash holding and dividend is zero. This result is not consistent with empirical studies of Opler et al. (1999), Custodio, et al. (2005), and Phuong (2013). It is the difference between this model and Fama-Macbeth model. The author uses the Stata command VIF to check for multicollinearity. Table 10: VIF Test for Cross-Sectional Regression Model 49 . vif Variable VIF 1/VIF LEV_MEAN CFLOW_MEAN SIZE_MEAN CFVOLAT_MEAN LIQ_MEAN DIVIDEND_M~N BANKDEBT_M~N 1.47 1.38 1.33 1.30 1.27 1.18 1.04 0.678273 0.722950 0.753895 0.772199 0.787195 0.847007 0.960249 Mean VIF 1.28 The mean vif (1.28) and each vif of variables are quite low, far to multicollinearity (10). 4.3.3. Pooled OLS Regression with Year Dummies In this Pooled OLS regression model, the confidence level is 95%. R-squared of regression result is 0.2039 is not much higher, when comparing with Fama-Macbeth regression model, 0.1999. It indicates that 05 year dummies added did not affect to R-squared much. Table 11: Results from Pooled OLS Regression with Year Dummies 50 note: Y2013 omitted because of collinearity Equation Obs Parms RMSE "R-sq" F P CASH 490 12 .1046882 0.2039 11.12814 0.0000 CASH Coef. SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND Y2009 Y2010 Y2011 Y2012 Y2013 _cons -.00018 -.1530947 -.0072192 .2629966 .0213117 -.0279361 .0346445 -.0188896 -.021254 -.0132005 -.0156736 0 .1344688 Std. Err. .0044955 .0254807 .0028737 .0484582 .083174 .0098009 .0126547 .0151387 .0151145 .0150821 .0150357 (omitted) .1178747 t P>|t| [95% Conf. Interval] -0.04 -6.01 -2.51 5.43 0.26 -2.85 2.74 -1.25 -1.41 -0.88 -1.04 0.968 0.000 0.012 0.000 0.798 0.005 0.006 0.213 0.160 0.382 0.298 -.0090134 -.2031628 -.0128658 .1677792 -.1421202 -.0471944 .0097789 -.0486362 -.050953 -.0428358 -.0452177 .0086535 -.1030267 -.0015726 .358214 .1847436 -.0086779 .0595102 .0108569 .008445 .0164348 .0138706 1.14 0.255 -.0971479 .3660855 The null and alternative hypotheses are: Ho: β = 0 (the linear correlation coefficient is zero) and H1: β ≠ 0 (the linear correlation coefficient is different from zero). Area in each tail of the t distribution equal 0.025 and degrees of freedome df are 5,878. From the t distribution table of Merrington (1941), the critical values of t are -1.960 and 1.960. In the figure run from Stata 12, the value of leverage, bank debt, cash flow, liquidity, and dividend are in the rejection region. So that, the linear correlation coefficient between cash holding and leverage, bank debt, cash flow, liquidity and dividend are different from zero. This result is consistent with result from FamaMacbeth model and previous authors Kim et al. (1998), Opler et al (1999), Pinkowitz and Williamson (2001), Ferreira and Vilela (2004), Ozkan and Ozkan (2004), Custodio et al. (2005), Bates et al. (2009), Meggison and Wei (2012), and Phuong (2013). Additionally, the result is also consistent with pecking order theory and cash flow theory. And the linear correlation coefficient between cash holding and firm’s size; cash flow volatility and closing stock price of year 2009-2012s are zero. The author uses the Stata command VIF to check for multicollinearity. Table 12: VIF Test for Pooled OLS Model with Year Dummies 51 . vif Variable VIF 1/VIF Y2012 Y2013 Y2011 Y2010 CFLOW LEV SIZE CFLOVAT LIQ DIVIDEND BANKDEBT 1.65 1.64 1.63 1.62 1.29 1.29 1.27 1.24 1.12 1.07 1.02 0.605246 0.609964 0.613993 0.618904 0.773058 0.776195 0.790081 0.808840 0.888915 0.931509 0.982531 Mean VIF 1.35 The mean vif (1.35) and each vif of variables are quite low, far to multicollinearity (10). However, in regression model, dummy variable, year 2013, is omitted because of multicollinearity. According to Tabacknick & Fidell (1996), there was some evidence of multicollinearity that could weaken regression analyses. This is a reason for the author to consider whether we can remove closing price of year 2013 from the regression analysis or not, due to the high correlation with other variables. Therefore, the author will increase the number of variables and run one more Pooled OLS regression with both year dummies and industry dummies to have another regression relation when industry dummies are pooled. It will help to have more points of view and to sharpen the author’s model on final. 52 4.3.4. Pooled OLS Regression with Year Dummies and Industry Dummies In this Pooled OLS regression model, confidence level is 95%. R-square (0.2967) is much higher than Pooled OLS regression with year dummies (0.2039). Table 13: Results from Pooled OLS Regression with Year Dummies and Industry Dummies note: Y2013 omitted because of collinearity note: INDUSTRY21 omitted because of collinearity Equation Obs Parms RMSE "R-sq" F P CASH 490 32 .100523 0.2967 6.232182 0.0000 CASH Coef. SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND Y2009 Y2010 Y2011 Y2012 Y2013 INDUSTRY1 INDUSTRY2 INDUSTRY3 INDUSTRY4 INDUSTRY5 INDUSTRY6 INDUSTRY7 INDUSTRY8 INDUSTRY9 INDUSTRY10 INDUSTRY11 INDUSTRY12 INDUSTRY13 INDUSTRY14 INDUSTRY15 INDUSTRY16 INDUSTRY17 INDUSTRY18 INDUSTRY19 INDUSTRY20 INDUSTRY21 _cons -.0014209 -.1288726 -.0070142 .2240204 .0999557 -.0201483 .0353365 -.0190097 -.0210827 -.0141568 -.0165655 0 -.0232596 -.0500787 -.0403213 .0074256 .0575333 -.046778 -.0765262 -.027788 .0434972 .1445922 .0282113 -.0248942 -.0357 .0330441 -.0365133 .0154301 .0272164 .0559656 -.0303139 .0091409 0 .1480997 Std. Err. .0048278 .0278764 .0028099 .0503157 .0890582 .0096816 .0130011 .0145557 .0145286 .0144932 .0144452 (omitted) .0406925 .0558791 .046489 .0365271 .0395472 .0454335 .0565479 .0562289 .0405527 .0475263 .0405886 .0377721 .0577662 .0567568 .0395084 .035758 .0347846 .0404333 .0410518 .045362 (omitted) .1362316 t P>|t| [95% Conf. Interval] -0.29 -4.62 -2.50 4.45 1.12 -2.08 2.72 -1.31 -1.45 -0.98 -1.15 0.769 0.000 0.013 0.000 0.262 0.038 0.007 0.192 0.147 0.329 0.252 -.0109082 -.1836541 -.0125361 .1251422 -.0750576 -.0391742 .0097874 -.0476139 -.0496336 -.0426382 -.0449527 .0080664 -.074091 -.0014922 .3228985 .274969 -.0011224 .0608857 .0095945 .0074683 .0143247 .0118216 -0.57 -0.90 -0.87 0.20 1.45 -1.03 -1.35 -0.49 1.07 3.04 0.70 -0.66 -0.62 0.58 -0.92 0.43 0.78 1.38 -0.74 0.20 0.568 0.371 0.386 0.839 0.146 0.304 0.177 0.621 0.284 0.002 0.487 0.510 0.537 0.561 0.356 0.666 0.434 0.167 0.461 0.840 -.1032268 -.1598899 -.1316795 -.064356 -.0201832 -.1360619 -.1876516 -.1382867 -.0361952 .0511954 -.0515517 -.0991223 -.1492196 -.0784918 -.1141535 -.0548401 -.0411408 -.0234922 -.1109871 -.0800026 .0567076 .0597324 .0510369 .0792072 .1352498 .0425059 .0345993 .0827107 .1231897 .2379889 .1079744 .0493339 .0778196 .1445801 .0411269 .0857002 .0955735 .1354235 .0503592 .0982845 1.09 0.278 -.1196169 .4158162 . The null and alternative hypotheses are: Ho: β = 0 (the linear correlation coefficient is zero) and H1: β ≠ 0 (the linear correlation coefficient is different from zero). Area in each tail of the t distribution equal 0.025 and degrees of freedome df are 15,188. From the t distribution table 53 of Merrington (1941), the critical values of t are -1.960 and 1.960. This result is consistent with result from Fama-Macbeth model. In table 10, we can see that level of cash holding has positive relation with cash flow, cash flow volatility, and dividend. And level of cash holding has negative relation with are size of business, leverage, bank debt, liquidity. This result is consistent with result of Ozkan and Ozkan (2004); Bates et al (2009) and Meggison and Wei (2012). Year dummies and industry dummies were not accounted to the result, because of rejecting Ho hypothesis. Sizes of firm and cash flow volatility have no relation with cash holding. This result is opposed to with result of Ozkan and Ozkan (2004). The correlation between cash holding and size of firm, leverage, bank debt and liquidity shows that this model is consistent with previous results of Ferreira and Vilela (2004); Custodio et al. (2005). This result is supported by theories of trade-off and pecking order, but oppose to free cash flow. Note that dummy variable year 2013 and industry 21 are omitted because of multicollinearity. Industry21 is a group of construction companies, refer to table 02. Construction company had specific relation to level of cash holding because it requires large amount of cash in short period and sometimes it operates on real estate. That is the reason why the author had paid attention to remove construction companies from the beginning. However, the risk is still being because of the nature of construction industry. The author uses the Stata command VIF to check for multicollinearity. Table 14:VIF Test for Pooled OLS with Year and Industry Dummies 54 . vif Variable VIF 1/VIF INDUSTRY17 INDUSTRY4 INDUSTRY16 INDUSTRY12 INDUSTRY5 INDUSTRY15 INDUSTRY19 INDUSTRY1 INDUSTRY18 INDUSTRY9 INDUSTRY11 INDUSTRY21 INDUSTRY20 INDUSTRY3 INDUSTRY6 INDUSTRY10 INDUSTRY8 INDUSTRY14 INDUSTRY13 INDUSTRY2 LEV Y2009 Y2010 Y2011 Y2012 SIZE CFLOVAT CFLOW DIVIDEND LIQ BANKDEBT 20.84 11.58 10.47 8.81 6.33 6.27 5.34 5.26 5.25 5.23 5.16 3.10 3.08 3.06 3.02 3.02 2.12 2.09 2.04 2.03 1.67 1.64 1.64 1.63 1.62 1.58 1.54 1.51 1.23 1.19 1.06 0.047989 0.086321 0.095511 0.113468 0.158081 0.159574 0.187255 0.190286 0.190314 0.191214 0.193774 0.322592 0.324830 0.327019 0.330887 0.331071 0.470747 0.479009 0.490936 0.493075 0.597936 0.608345 0.610615 0.613598 0.617683 0.631652 0.650467 0.661112 0.813703 0.839914 0.947463 Mean VIF 4.21 The mean vif (4.21) shows the medium possibility of multicollinearity (10) happening for model. The result also points out that Chemicals – Pharmaceutical (industry4); Electrical equipment - Electronics – Telecommunications (industry16); and Food - Beverage - Tobacco (industry17) are highly correlated with other variables, however the author cannot find out reasons within this test. This issue were not mentioned in previous empirical tests. 55 4.3.5. Fixed Effects Regression Model Confidence level is 95%. The result is shown on below: Table 15: Results from Fixed Regression Model . xtreg CASH SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND, fe vce(robust) Fixed-effects (within) regression Group variable: COMP Number of obs Number of groups = = 490 98 R-sq: Obs per group: min = avg = max = 5 5.0 5 within = 0.0794 between = 0.0193 overall = 0.0256 corr(u_i, Xb) F(7,97) Prob > F = -0.4899 = = 2.77 0.0114 (Std. Err. adjusted for 98 clusters in COMP) Robust Std. Err. CASH Coef. t SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND _cons .0344116 -.000475 -.0015692 .1411554 .8079611 -.0046942 .0394675 -.8802116 .0479829 .102256 .0009881 .0990445 .7699556 .0083442 .0163286 1.300362 sigma_u sigma_e rho .09862284 .08568744 .56983871 (fraction of variance due to u_i) 0.72 -0.00 -1.59 1.43 1.05 -0.56 2.42 -0.68 P>|t| 0.475 0.996 0.116 0.157 0.297 0.575 0.018 0.500 [95% Conf. Interval] -.0608212 -.203425 -.0035302 -.0554206 -.7201876 -.0212551 .0070597 -3.461071 .1296444 .2024749 .0003919 .3377313 2.33611 .0118667 .0718753 1.700648 R-square overall is quite low 0.0256. The null and alternative hypotheses are: Ho: β = 0 (the linear correlation coefficient is zero) and H1: β ≠ 0 (the linear correlation coefficient is different from zero). Area in each tail of the t distribution equal 0.025 and degrees of freedome df are 40,018. From the t distribution table of Merrington (1941), the critical values of t are 1.960 and 1.960. This result is much different with result of other models used above. We reject Ho with correlation cefficient of dividend variable (positive t 2.42). We can see that level of cash 56 holding has only positive relation with dividend. This result is consistent with result of Custodio et al. (2005). 4.4. Result from Supporting Tests for Regression Models 4.4.1. Hausman Test Result from Hausman above helps the author choose between fixed effect model or random effect model. This test is put in the end of analysis part just in order to make structure of study balanced, however this is done before deciding what model. Table 16: Results from Hausman Test . hausman fixed random Coefficients (b) (B) fixed random SIZE LEV BANKDEBT CFLOW CFLOVAT LIQ DIVIDEND .0344116 -.000475 -.0015692 .1411554 .8079611 -.0046942 .0394675 -.0015901 -.1035369 -.0038318 .2132835 .1519073 -.0153829 .0360273 (b-B) Difference .0360016 .1030619 .0022626 -.0721282 .6560538 .0106888 .0034402 sqrt(diag(V_b-V_B)) S.E. .0260898 .038036 .0009444 .0258646 .2651012 .0025606 .0075381 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 37.76 Prob>chi2 = 0.0000 It basically tests whether the unique errors (ui) are correlated with the regressors, the null hypothesis is they are not. It means that the null hypothesis Ho is: the coefficient of the panel model is zero (the preferred model is random effects) and the alternative hypothesis H1 is: the coefficient of the panel model is different from zero (the alternative is fixed effects). The confidence level is 95%. With the prob (chi2) value less then 0.05, then we reject Ho. It means 57 that fixed effects model is preferred to use in this study. People can continue with fixed effect model in part 4.3.5 above. 58 4.5. Summary of Results Combining above regressions, the author summary onto table 11 as followings: Table 17: Result Summary of Regression Models Independent Variables SIZE LEV BANKDEBT CFLOW CFVOLAT LIQ DIVIDEND CONS_ N R-square Coef. P>|t| Coef. P>|t| Pooled OLS Regression with Dummies Fixed Effects Model Year and Industry Year Dummies Dummies Coef. P>|t| Coef. P>|t| Coef. P>|t| 0.00048108 -0.15406476 -0.00698628 0.26045296 0.02095928 -0.0280311 0.03325728 0.11173998 0.038 -3.46 -2.754 2.802 0.196 -1.974 2.878 0.802 0.0014847 -0.2104057 -0.0162702 0.3317434 -0.1428694 -0.0780228 0.0287456 0.1130525 0.826 0.000 0.024 0.003 0.279 0.005 0.268 0.516 -0.00018 -0.1530947 -0.0072192 0.2629966 0.0213117 -0.0279361 0.0346445 0.1344688 Fama-MacBeth Regression Model 5 0.1999 Crossectional Regression Model 98 0.4182 490 0.2039 0.968 0.000 0.012 0.000 0.798 0.005 0.006 0.255 -0.0014209 -0.1288726 -0.0070142 0.2240204 0.0999557 -0.0201483 0.0353365 0.1480997 490 0.2967 0.769 0.000 0.013 0.000 0.262 0.038 0.007 0.278 0.0344116 -0.000475 -0.001569 0.1411554 0.8079611 -0.004694 0.0394675 -0.880212 0.72 0 -1.59 1.43 1.05 -0.56 2.42 -0.68 490 0.0256 From the synthetical result of four regression methods above, six factor (variables), which affects to level of cash holding of companies listed in Vietnam stock market, are size of firm, leverage, bank debt, cash flow, liquidity, and dividend. In general, most of figures show the united correlations of cash ratio (dependent variable) and other independent variables by four regression models. However, there are three variances on result among regression models. Firstly, size of firm indicates contrary correlation between Pooled OLS regression with dummies and FamaMacBeth regression model. Pooled OLS point out positive relation between cash holding and size of firm. Moreover, high value of p-value indicated that this linear correlation coefficient is zero. This leads to insignificance impact between them. The result is not consistent with pecking order theory, cash flow theory, and results of authors Pinkowitz and Williamson (2001); and Ozkan and Ozkan (2004), and it is not as expectation from the beginning of the author. Secondly, cross-sectional regression model implies that cash holding has positive relation with cash flow volatility, however the results are not consistent among regression models. Finally, positive correlation between dividend and cash flow is consistent with the result of authors Opler et al. 59 (1999) and Custodio et al. (2005), however it is not as expectation from the beginning of the author. Moreover, with the confidence level 95%, cross-sectional regression model indicates linear correlation coefficient between cash holding and dividend is zero. Cross-sectional regression model implies that there is no significance impact between cash holding and dividend. Factors, which have positive relation with cash holding, are leverage, bank debt, cash flow volatility, and liquidity. Correlation between cash holding and leverage is consistent with pecking order theory and most result of authors Kim et al. (1998); Opler et al. (1999); Pinkowitz and Williamson (2001); Ferreira and Vilela (2004); Ozkan and Ozkan (2004); Custodio et al. (2005); Bates et al. (2009); Meggison and Wei (2012); and Phuong (2013). Correlation between cash holding and bank debt is consistent with result of authors Opler et al. (1999); Pinkowitz and Williamson (2001); Ferreira and Vilela (2004); Ozkan and Ozkan (2004); Custodio et al. (2005); Bates et al. (2009); Meggison and Wei (2012) ; and Phuong (2013). Correlation between cash holding and liquidity is consistent with trade off theory and most result of authors Opler et al. (1999); Pinkowitz and Williamson (2001); Ferreira and Vilela (2004); Ozkan and Ozkan (2004); Custodio et al. (2005); Bates et al. (2009); Meggison and Wei (2012); and Phuong (2013). Note that dummy variable year 2013 and industry 21 are omitted in regression result because of multicollinearity. Construction firms, in their nature of industry, could be related to one another in sample size. For easy view, table 12 shows results of what factors and related theories affect level of cash holding from the beginning: Table 18: Thesis’s Result Summary Independent Expect Result Variables by from Empirical Studies In relation to related theory author analysis 60 SIZE (+) (N/A) (+): Pinkowitz and Williamson (+):Pecking order No (2001); Ozkan and Ozkan (2004), theory and Cash significance Trinh (2012), and Phuong (2013) flow theory impact LEV (-) (-) (-): Kim et al. (1998), Opler et al (-):Pecking order (1999), Pinkowitz and Williamson theory and Cash (2001), Ferreira and Vilela (2004), flow theory Ozkan and Ozkan (2004), Custodio et al. (2005), Bates et al. (2009), Meggison and Wei (2012), Trinh (2012), and Phuong (2013) BANKDEBT (-) (-) (-): Ferreira and Vilela (2004), Ozkan and Ozkan (2004), and Phuong (2013) CFLOW (+) (+) (+): Opler et al. (1999), Pinkowitz (+):Pecking order and Williamson (2001), Ferreira and theory Vilela (2004), Ozkan and Ozkan (2004), Custodio et al. (2005), Bates et al. (2009), and Meggison and Wei (2012), and Phuong (2013). LIQ (-) (-) (-): Opler et al.(1999), Pinkowitz and (-):Trade off Williamson (2001), Ferreira and theory Vilela (2004), Ozkan and Ozkan (2004), Bates et al. (2009), Meggison 61 and Wei (2012), and Phuong (2013) DIVIDEND (-) (+) (+): Opler et al. (1999), Custodio, et No al. (2005), and Phuong (2013) significance impact in Crosssectional Regression Model. Note that cash flow and cash low volatility are grouped into cash flow. 62 Chapter 5. Research Conclusion 5.1. Conclusion and Implication This thesis is expected to supplement evidences for factors that affect level of cash holding, based on theories and empirical researches. 98 samples are companies listed in two stock exchanges, Ho Chi Minh and Ha Noi, in period from 2009 to 2013. In summary, the result supports for empirical evidences as well as related theories. Below is summary, which answer for two thesis questions: For the first question, in figure above, five factors (variables), which affects to level of cash holding of companies listed in Vietnam stock market, are leverage, bank debt, cash flow, liquidity, and dividend. Factors, which have negative relation with cash holding, are leverage, bank debt, cash flow volatility, and liquidity. For the second question, three differences in result imply some situations and recommendation for firms listed in Vietnam stock market. Regression models provide contrary results about size of firm and cash flow volatility in relation to level of cash holding. It implies three issues. Firstly, each firm reacted differently to the changes of Vietnam stock market. Secondly, stock market is highly volatile. Thirdly, operation and investment activities of firms are not effective, and are possible to bear loss. The remaining difference is in relation between dividend payout and level of cash holding. At the beginning, the author expected that this was negative correlation, but actually it was positive correlation. In years, which firms paid out dividend, they kept a lot of cash. The figure 04 shows that the amount of cash reserved at the end of years is double amount of dividend paid out. The direct correlation of cash holding and dividend here is not strong for opinion of using excess cash to pay dividend. However, this is consistent with results of Opler et al (1999) and Custodio et al. (2005). Another observation is that, firms held low level of bank debt and liquidity. It means that firms try to pay their bank debt 63 to avoid interest expenses and also found long-term debt to finance for operations. Firms try to balance between short-term and long-term operation. The consideration between costs and benefits of holding cash is necessary to find the optimized level of cash holding. If holding cash on operations brings effectiveness for business, firm needs to accumulate level of cash for potential projects. Conversely, if there is no integrated strategy for cash management, firm will face to many difficulties. From the result of this study in the summary, the author proposes three main recommendations as below:  Firm needs to reduce accumulating cash for field which is not their main industries. With this manner, firm will avoid wasting its cash on ineffective projects. In period 20092013s, many firms invested in projects which are not their industries such as real esta or stock market…etc. Instead of earning profit, the return on investment was negative. With the cash accumulated, firm should concentrate on their industry, their strength to maximize business operation and profit.  Firm needs to pay more attention to cash flow volatility in order to have best holistic level of cash holding. As the study’s supposition, if cash flow volatility increases, firm will hold more cash. However, the regression with mean value shows negative correlation, it means that firms sometimes do not pay attention to cash flow volatility in last 5 years. Ferreira and Vilela (2004) showed that firms which have high volatility in cash flow, will be possible to face the cash shortage. So that, firms need to focus on cash flow volatility to avoid business risks and also get business stable.  Firms should have a holistic cash management to accumulate cash in business. The management should include current account, deposit account, receivables management, payables management, liquidity management and technology management. The integrated receivables solutions help to convert firm’s receivables to cash faster and gain more control over liquidity to take business to new opportunity of profitability. In figure 64 03 in descriptive statistic, normally firms hold cash around 11% over the total assets. If firm does not have good management on cash, firm could bear costs of holding cash, such as interest and opportunity cost. The author proposes three approaches to manage cash as follow. Firstly, firm needs to carefully select strategic partners for external finances. This will help business to decrease risk of liquidity and shortage on cash. Moreovers, firms could have a low rate of interest payment. Then, firm can use its cash for operations or investment instead of limiting it. A notice is that strategic partner selection is the key for business’s stability. Secondly, firm needs to build and develop models for forcasting fluctuation of cash holding. So that, firm can enhance its flexibility on business’s crisis and then increase business’s value. These models can built professionals internally, or consultant for external partners. Thirdly, firm should find and select investment channels with low costs. This is good for business in unstable economy. In practice, high yield projects will have high costs of investment. Firms should ultilize cash by increasing cash cycle faster such as overnight rate. In case of having surplus cash, firm could pay dividend or buyback shares. This will enhance investor’s confidence. In short, firms should have a backup plan for bad situations and carefully considerate a cash policy, in which company can have sales growth and maximized profitability. Since there are limited researches in cash holding in Vietnam, the empirical analysis of this research expect to provide a new understanding for management of cash. 5.2. Limitation of the Study This research was carried out by an individual and research’s issues were quite deep and new in Vietnam, therefore, the research was foreseen that will undergo some difficulties. Firstly, with the limitation of time and capabilities of collecting data, the author selects 98 companies listed in Vietnam stock market, instead of expanding for non-listed ones. The 65 listed companies can provide audited financial statements, which are supervised by Stock Exchange Committee. Those financial data are available for the author. However, because of many fluctuations on listed companies, only 98 companies are consistent with predefined criteria. Some companies are not listed in examined period 2009-2013. Secondly, the author has chosen to compare and evaluate companies listed on the exchanges with the exception of those operating within the finance and real estate sectors since these sectors are differently and more heavily regulated in regard to the investigated variables than other sectors. Companies were also excluded on the basis of outlier values as it would also distort the result due to a large impact of a few observations. Finally, many micro-economic and macro-economic factors, which are not used in this research, can be relevant and input to research’s theories. It is limited by the author’s working experience and academic knowledge on those fields. Those macro-economic factors are Revenue, salary, right of investor, the level of investor protection, asymmetric information, equity turnover, R&D expenses, expenditures, and merge & acquisition. Those micro-economic factors are economic crisis, inflation, unemployment ratio, business environment, taxation, and supply of capital market. 5.3. Future Research In future work, it would be appropriate to focus on following aspects:  Expand sample size and studied period of companies listed in Vietnam stock market.  Study and compare with other countries to provide the difference in cash management.  Differentiating between long term and short term debts.  Construction firms should be carefully considered to be in sample size or to be dropped from the regression equation. 66  Micro-economic and macro-economic factors should be included. Those macroeconomic factors are Revenue, salary, right of investor, the level of investor protection, asymmetric information, equity turnover, R&D expenses, expenditures, and merge & acquisition. Those micro-economic factors are economic crisis, inflation, unemployment ratio, business environment, taxation, and supply of capital market. 67 References 1) Atrill, P. & McLaney, E. (2004). Accounting: An Introduction, (4th Eds). United Kingdom, UK. Pearson Publication. 2) Bates, W. T., Kahle, M. K., & Stulz, M. R. (2007). Why do US companies hold so much more cash than they used to? The Journal of Finance. 64(5), pp.1985–2012. 3) Berk, J. K. & DeMarzo, P. E. (2011). Corporate Finance, (4th Eds). United Kingdom UK. Pearson Publication. 4) Bigelli, M. & Sanchez, F. J. (2010). Cash holding in private firms. Working Paper. Retrieved from University of Bologna. 5) Bover, O. & Watson, N. (2005). Are there economies of scale in the demand for money by firms? Some panel data estimates. Journal of Monetary Economics, 52, pp. 1569-1589. 6) Bryman, A., & Bell, E. (Eds.). (2011). Business Research Methods. New York: Oxford University Press. 7) Chie, M. S. (2011). The determinants of Cash Holdings: Evidence from Dutch Listed Companies. Master Thesis. Retrieved from University of Twente. 8) Custodio, C., Ferreira, M. A., & Raposo, C. (2005). Cash holdings and business conditions. Working paper. Retrieved from London, United Kingdom: London School of Economics and Political Science (LSE). Lisbon, Portugal: Universidade Nova de Lisboa and Instituto Superior de Cinêcias do Trabalho e da Empresa (ISCTE) – School of Business. 9) Damodaran, A. (2005). Dealing with cash, cross holdings and other non-operating assets: Approaches and implications. Working paper. New York, United State: New York University, Department of Finance. 68 10) Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. The Academy of Management Review. Vol.14. 11) Fama, E. F. (1963). Mandelbrot and the stable Paretian hypothesis. Journal of Business 36, pp.420-429. 12) Fama, E. F. (1965a). The behavior of stock market prices. Journal of Business 38, pp.34– 105. 13) Fama, E. F. (1965b). Random walks in stock market prices. Financial Analysts Journal 21, pp.55–59. 14) Fama, E. F. & MacBeth, J. D. (1973). Risk, return, and equilibrium: empirical tests. Journal of Political Economy, 81, pp.607–636. 15) Farrar, D. E. & Glauber, R. R. (1967). Multicollinearity in Regression Analysis: The Problem Revisited. Review of Economics and Statistics 49. pp.92–107 16) Faulkender, M., (2002). Cash holdings among small businesses. Working paper. Retrieved from SSRN. 17) Ferreira, M. A. & Vilela, A. S. (2004). Why do companies hold cash? Evidence from EMU countries. European Financial Management, 10(2), pp.295 - 319. 18) Harris, M. & Raviv, A. (1978) Optimal Incentive Contracts with Imperfect Information. Retrieved from Carnegie Mellon University, Mimeo. 19) Hoang, T. (2012). Economic downturn forecast to hit Vietnam again. The Saigon Times Daily. Retrieved from http://english.thesaigontimes.vn/22916/Economic-downturn-forecastto-hit-Vietnam-again.html in 22 August 2014 69 20) Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review, 87, pp.535–555. 21) Khanh, L. P. (2013). The affect of holding cash on operations and business value in Viet Nam, period 2007–2012. Master Thesis. University of Economics Ho Chi Minh City 22) Kim, C. S., Mauer, D. C. & Sherman, A. E. (1998). The determinants of corporate liquidity theory and evidence. Journal of Financial and Quantitative Analysis, 33, pp.305–334. 23) Kumar, T. K. (1975). Multicollinearity in Regression Analysis. Review of Economics and Statistics 57: pp.365–366 24) Krivogorsky, V., & Grudnitski, G., Dick, W. (2009). Bank debt and performance of Continental European firms. Working paper. San Diego State University. 25) Maness, T. S., & Zietlow, J. T. (Eds). (2004). Short-Term Financial Management. Kendallville: South-Western. 26) Megginson, W. L. & Wei, Z. (2012). State Ownership, Soft-Budget Constraint and Cash Holdings: Evidence from China’s Privatized Companies. Working paper. United States: University of Oklahoma, University of Texas at El Paso. 27) Merrington, M. (1941). Table of percentage points of the t-distribution. Biometrika 32, pp.300. 28) Mikkelson, W. H. & Partch, M. M. (2003), Do persistent cash reserves hinder performance? Journal of Financial and Quantitative Analysis 38, 275—294. 29) Myers, S. & Majluf, N. (1984). Corporate financing and investment decisions when firms have information that investor do not have. Journal of Financial Economics. 70 30) Opler, T., Pinkowitz L., Stulz, R. & Williamson, R. (1999). The determinants and implication of Corporate Cash Holdings. Journal of Financial Economics, 52, pp.3–46. 31) Ozkan, A. & Ozkan, N. (2004). Corporate cash holdings: An empirical investigation of UK companies. Journal of Banking & Finance, 28, pp.2103–2134. 32) Pearson, K. (1895). Notes on regression and inheritance in the case of two parents. Proceedings of the Royal Society of London 58 : pp.240–242. 33) Phuong, N. T. N. (2013). The determinants affecting the cash holding of companies listed on HNX and HOSE. Master Thesis. University of Economics Ho Chi Minh City 34) Pinkowitz, L. & Williamson, R. (2001). Bank power and cash holdings: evidence from Japan. Review of Financial Studies, 14(4), pp.1059-1082. 35) Rodion, E. (2012). Vietnam's dangerous economic downturn. Germany's international broadcaster. Retrieved from http://www.dw.de/vietnams-dangerous-economic-downturn/a16406306 in 22 August 2014. 36) Samuelson, P. A. (1965). Proof that properly anticipated prices fluctuate randomly. Industrial Management Review 6, pp.41–49. 37) Smith, C. W. & Stulz, R. (1985). The determinant of firm’s hedging policies. Journal of Financial and Quantitative Analysis 20, pp.391-405 38) Stock, J. H. & Watson, M. W. (2nd Eds). (2007). Introduction to econometrics. Boston: Pearson Addison Wesley. 39) Trinh, D. L. V. (2012). Cash management in companies listed in HNX and HOSE, period 2006-2012. Master Thesis. University of Economics Ho Chi Minh City. 71 Appendix Appendix 1: 98 Listed Companies Chosen Final Listed List Stock in Company name website Exchange AAM HOSE Công ty cổ phần Thủy sản Mekong http://www.mekongfish.vn /vn/index.aspx ABT HOSE Công ty cổ phần Xuất nhập khẩu thủy www.aquatexbentre.com sản Bến Tre ACL HOSE Công ty cổ phần Xuất nhập khẩu Thủy www.clfish.com sản Cửu Long An Giang AGF HOSE Công ty cổ phần Xuất nhập khẩu thủy www.agifish.com.vn sản An Giang AMV HASTC Công ty cổ phần Sản xuất kinh doanh www.amvibiotech.com dược và Trang thiết bị y tế Việt Mỹ BBC HOSE Công ty Cổ phần Bibica BBS HASTC Công ty cổ phần VICEM Bao bì Bút www.baobibutson.com.vn www.bibica.com.vn Sơn BHS HOSE Công ty Cổ phần Đường Biên Hòa www.bhs.vn BKC HASTC Công ty cổ phần Kho ng sản Bắc Kạn www.backanco.com BLF HASTC Công ty cổ phần Thủy sản Bạc Liêu www.baclieufis.vn/index.p hp BMC HOSE Công ty cổ phần Kho ng sản Bình Định http://www.bimico.vn BPC HASTC Công ty cổ phần Vicem Bao bì Bỉm sơn www.baobibimson.vn 72 BTH HASTC Công ty cổ phần Chế tạo biến thế và Vật www.ctbt.vn liệu điện Hà Nội BXH HASTC Công ty cổ phần Vicem Bao bì Hải www.hcpc.vn Phòng CAN HASTC Công ty Cổ phần Đồ hộp Hạ Long CAP HASTC Công ty cổ phần Lâm nông sản thực www.yfatuf.com.vn www.canfoco.com.vn phẩm Yên B i CLC HOSE Công ty Cổ phần C t Lợi CSM HOSE Công ty cổ phần Công nghiệp Cao su www.casumina.com.vn www.catloi.com.vn Miền Nam CTB HASTC Công ty cổ phần Chế tạo Bơm Hải www.hpmc.com.vn Dương DBT HASTC Công ty cổ phần Dược phẩm Bến Tre www.bepharco.com DCL HOSE Công ty cổ phần Dược phẩm Cửu Long www.pharimexco.com.vn DHC HOSE Công ty cổ phần Đông Hải Bến Tre www.dohacobentre.com DHG HOSE Công ty cổ phần Dược Hậu Giang www.dhgpharma.com.vn DHT HASTC Công ty cổ phần Dược phẩm Hà Tây www.hataphar.com.vn DMC HOSE Công ty cổ phần Xuất nhập khẩu Y tế www.domesco.com Domesco DQC HOSE Công ty cổ phần Bóng đèn Điện Quang www.dienquang.com DZM HASTC Công ty cổ phần Chế tạo m y Dzĩ An www.vietgen.com FMC HOSE Công ty Cổ phần Thực phẩm Sao Ta www.fimexvn.com GDT HOSE Công ty cổ phần Chế biến Gỗ Đức www.dtwoodvn.com Thành GLT HASTC Công ty cổ phần Kỹ thuật điện toàn cầu www.glt.com.vn 73 GTA HOSE Công ty Cổ phần chế biến gỗ Thuận An HAD HASTC Công ty cổ phần Bia Hà Nội - Hải www.hadubeco.com.vn www.tac.com.vn Dương HAI HOSE Công ty Cổ phần Nông dược H.A.I http://www.congtyhai.com /vn HHC HASTC Công ty cổ phần B nh kẹo Hải Hà www.haihaco.com.vn HRC HOSE Công Ty Cổ Phần Cao su Hòa Bình www.horuco.com.vn HSI HOSE Công ty cổ phần Vật tư Tổng hợp và www.hsi.com.vn Phân bón Hóa sinh HTV HOSE Công ty cổ phần vận tải Hà Tiên http://www.vantaihatien.co m.vn HVT HASTC Công ty cổ phần Hóa chất Việt Trì www.viettrichem.com.vn KDC HOSE Công ty cổ phần Kinh Đô www.kinhdofood.com KMR HOSE Công ty cổ phần Mirae www.miraejsc.com LAF HOSE Công ty Cổ phần Chế biến hàng xuất www.lafooco.vn khẩu Long An LIX HOSE Công ty cổ phần Bột giặt Lix MCP HOSE Công Ty Cổ Phần In và Bao bì Mỹ www.mychau.com.vn www.lixco.com Châu MHL HASTC Công ty cổ phần Minh Hữu Liên www.minghuulien.com MKV HASTC Công ty cổ phần Dược thú y Cai Lậy www.cailayvet.com.vn MMC HASTC Công ty cổ phần Kho ng sản Mangan www.mitraco.com.vn NGC HASTC Công ty cổ phần Chế biến thủy sản Xuất www.ngoprexco.com.vn khẩu Ngô Quyền NPS HASTC Công ty cổ phần May Phú Thịnh - Nhà www.phuthinhnb.com 74 Bè NSC HOSE Công ty Cổ phần Giống cây trồng Trung www.vinaseed.com.vn Ương PAC HOSE Công ty Cổ phần Pin Ắc quy miền Nam PMC HASTC Công ty cổ phần Dược phẩm dược liệu www.pharmedic.com.vn www.pinaco.com Pharmedic POT HASTC Công ty cổ phần Thiết bị Bưu Điện PTC HOSE Công ty cổ phần Đầu tư và Xây dựng www.pticjsc.com www.postef.com Bưu điện PTM HASTC Công ty cổ phần Sản xuất, Thương mại www.ptm.vn và Dịch vụ ôtô PTM RAL HOSE Công ty cổ phần Bóng đèn Phích nước www.rangdongvn.com Rạng Đông SAF HASTC Công ty cổ phần Lương thực Thực phẩm www.safocofood.com SAFOCO SAV HOSE Công ty cổ phần Hợp t c kinh tế và www.savimex.com Xuất nhập khẩu SAVIMEX SDN HASTC Công ty cổ phần Sơn Đồng Nai www.dongnaipaint.vn SFI HOSE Công ty cổ phần Đại lý Vận tải SAFI www.safi.com.vn SFN HASTC Công Ty Cổ Phần Dệt lưới Sài Gòn www.sfn.vn SGC HASTC Công ty cổ phần Xuất nhập khẩu Sa 0 Giang SGT HOSE Công ty cổ phần Công nghệ Viễn thông www.saigontel.com Sài Gòn SJ1 HASTC Công ty cổ phần Thủy sản số 1 http://www.seajoco.vn 75 SPP HASTC Công ty cổ phần Bao bì Nhựa Sài Gòn www.saplastic.com.vn SRC HOSE Công ty cổ phần Cao su Sao vàng www.src.com.vn SRF HOSE Công ty cổ phần Kỹ Nghệ Lạnh www.searefico.com SSC HOSE Công ty cổ phần Giống cây trồng miền www.ssc.com.vn Nam SSM HASTC Công ty cổ phần Chế tạo kết cấu thép www.ssm.com.vn VNECO.SSM STP HASTC Công ty Cổ phần công nghiệp thương www.stp.com.vn mại Sông Đà SVI HOSE Công ty cổ phần Bao bì Biên Hòa TAC HOSE Công ty cổ phần Dầu thực vật Tường www.tuongan.com.vn www.sovi.com.vn An TCL HOSE Công ty cổ phần Đại lý giao nhận Vận www.tancanglogistics.com tải xếp dỡ Tân Cảng .vn http://www.thbeco.vn THB HASTC Công ty cổ phần Bia Thanh Hóa TJC HASTC Công ty cổ phần Dịch vụ Vận tải và www.transco.com.vn Thương mại TMS HOSE Công ty cổ phần Kho vận Giao nhận www.transimexsaigon.co Ngoại thương TP.HCM m TNC HOSE Công ty cổ phần Cao su Thống Nhất www.trcbrvt.com TPC HOSE Công ty cổ phần Nhựa Tân Đại Hưng www.tandaihungplastic.co m TRA HOSE Công ty cổ phần Traphaco www.traphaco.com.vn TS4 HOSE Công ty cổ phần Thủy sản số 4 http://seafoodno4.com TSC HOSE Công ty cổ phần Vật tư kỹ thuật nông www.tsccantho.com.vn 76 nghiệp Cần Thơ TTF HOSE Công ty cổ phần Tập đoàn Kỹ nghệ gỗ www.truongthanh.com Trường Thành TYA HOSE Công ty Cổ phần Dây và C p điện Taya 0 Việt Nam UNI HASTC Công ty cổ phần Viễn Liên www.vienlien.com.vn VBH HASTC Công ty cổ phần Điện tử Bình Hoà www.viettronicsbinhhoa.com VDL HASTC Công ty cổ phần Thực phẩm Lâm Đồng www.dalatwine.com.vn VFG HOSE Công ty cổ phần Khử trùng Việt Nam www.vfc.com.vn VGP HASTC Công ty Cổ phần Cảng rau quả www.vegeport.com VGS HASTC Công ty cổ phần Ống thép Việt Đức VG www.vgpipe.com.vn PIPE VIS HOSE Công ty Cổ phần Thép Việt Ý VNA HOSE Công ty cổ phần vận www.vis.com.vn tải biển http://www.vinaship.com. VINASHIP vn www.vinamilk.com.vn VNM HOSE Công ty Cổ phần Sữa Việt Nam VNT HASTC Công ty cổ phần Giao nhận Vận tải www.vinatranshn.com.vn Ngoại thương VPK HOSE Công ty Cổ phần bao bì dầu thực vật 0 VSC HOSE Công ty cổ phần Container Việt Nam www.viconship.com VST HOSE Công ty cổ phần Vận tải và Thuê tàu www.vitranschart.com.vn biển Việt Nam VTB HOSE Công ty Cổ phần Viettronics Tân Bình www.vtb.com.vn VTC HASTC Công ty cổ phần viễn thông VTC www.vtctelecom.com.vn 77 VTL HASTC Công ty Cổ phần Vang Thăng Long www.vangthanglong.com. vn 78 Appendix 2: Secondary Data for Excel and Stata Input Year Firm Listed Cash and cash equivalents Bank debt Dividends paid Cash flow Outstandin g shares at year end Total assets Net working capital 2009 AAM 128,443 7,703 17,821 39,287 11.34 363,935 100,30 2010 AAM 76,635 6,160 28,350 25,002 11.34 331,336 140,21 4 2011 AAM 49,648 4,620 36,935 35,290 11.34 329,977 164,30 2012 AAM 42,872 - 30,574 (8,497) 10.26 285,761 154,00 2013 AAM 54,687 114,293 4,320 12,109 9.94 302,072 139,64 3 2009 ABT 90,331 44,225 2 97,085 11.34 537,004 188,54 2010 ABT 152,350 61,644 45,616 54,947 13.61 601,925 107,05 2011 ABT 106,047 26,968 95,873 10,894 13.61 478,109 171,52 2012 ABT 111,665 131,629 82,627 (158) 11.02 516,133 189,15 2013 ABT 104,933 229,212 51,670 28,532 11.30 688,269 217,25 2009 ACL 63,657 347,880 13,500 39,984 9.00 613,944 (16,91) 2010 ACL 42,358 429,467 35,500 37,306 18.40 726,085 5,137 2011 ACL 31,541 399,029 38,500 94,569 18.40 793,378 64,330 2012 ACL 9,991 439,567 36,799 (5,613) 18.40 831,753 50,871 2013 ACL 8,575 389,150 8,280 18,883 18.40 718,957 23,150 2009 AGF 18,812 478,495 12,859 43,399 12.86 1,209,944 66,870 2010 AGF 47,609 580,072 25,719 56,125 12.86 1,354,627 (1,927) 2011 AGF 66,099 823,941 25,559 75,663 12.86 1,716,936 92,244 79 2012 AGF 53,830 2013 AGF 235,690 650,300 12,779 60,930 12.86 1,564,982 188,84 12,779 48,647 12.86 2,250,909 166,97 1,059,0 19 2009 AMV 1,856 - 314 1,571 2.10 23,563 7,113 2010 AMV 3,057 2,200 2,149 713 2.10 30,005 5,764 2011 AMV 2,514 5,415 1,219 (2,718) 2.12 28,102 3,107 2012 AMV 1,784 2,577 838 (131) 2.12 23,993 2,667 2013 AMV 2,182 2,013 1,077 214 2.12 27,043 2,748 2009 BBC 204,756 17,659 - 77,525 15.42 736,809 (22,12) 2010 BBC 89,081 4,533 - 77,115 15.42 758,841 60,601 2011 BBC 60,321 - 15,438 61,780 15.42 786,198 152,18 2012 BBC 49,471 - 18,603 48,711 15.42 768,378 143,61 2013 BBC 151,707 - 27,728 57,927 15.42 808,294 76,948 2009 BBS 8,551 30,890 3,450 14,171 3.00 107,619 17,326 2010 BBS 6,569 34,085 1,800 9,736 4.00 122,985 39,373 2011 BBS 5,089 37,101 4,800 9,465 4.00 149,449 44,844 2012 BBS 8,184 58,995 4,800 10,384 4.00 178,442 103,07 2013 BBS 21,597 81,651 4,800 8,468 4.00 217,256 87,054 2009 BHS 86,126 380,173 18,454 125,378 18.53 884,740 111,52 2010 BHS 58,759 339,316 62,595 117,197 18.53 1,015,192 155,95 (54,031 2011 BHS 178,778 466,051 62,674 136,758 30.00 1,281,737 ) 1,102,9 2012 BHS 98,524 (71,918 94,135 78,744 31.50 2,107,835 29 2013 BHS 238,292 885,330 ) 62,988 (2,002) 62.99 2,193,791 (190,47 80 8) 2009 BKC 4,342 6,190 - 13,021 6.03 145,598 24,238 2010 BKC 22,051 - 4,828 10,981 6.03 142,439 4,831 2011 BKC 2,030 450 8,144 4,462 6.03 134,400 7,236 2012 BKC 759 8,432 - 6.03 141,169 (12,093 (14,227 ) ) (13,727 2013 BKC 2,924 8,421 - (20,859 6.03 116,928 ) ) (60,321 2009 BLF 10,078 230,454 - 15,682 5.00 338,283 ) (59,045 2010 BLF 19,207 260,235 - 17,224 5.00 375,916 ) (71,899 2011 BLF 18,617 280,777 - 21,928 5.00 423,349 ) (35,729 2012 BLF 6,025 252,497 - 17,419 5.00 410,409 ) (42,575 2013 BLF 21,035 268,827 - 3,738 5.00 479,689 ) 2009 BMC 20,285 - 14,862 14,920 8.26 151,936 20,425 2010 BMC 40,051 - 18,165 82,343 8.26 169,597 16,270 2011 BMC 42,507 - 27,072 13,648 8.26 257,555 90,164 2012 BMC 15,218 - 61,920 46,063 12.39 271,042 70,995 2013 BMC 41,429 - 60,538 42,641 12.39 272,812 52,466 2009 BPC 5,115 - - 22,175 3.80 100,622 36,172 2010 BPC 12,851 18,000 5,700 11,521 3.80 124,636 22,390 81 2011 BPC 9,071 18,445 5,169 11,018 3.80 146,700 32,699 2012 BPC 14,593 40,000 4,742 12,051 3.80 174,184 29,920 2013 BPC 6,072 46,528 3,802 13,357 3.80 182,524 44,006 2009 BTH 5,566 - 3,960 3,227 3.00 58,527 12,927 2010 BTH 16,923 - 3,600 1,648 3.17 78,490 8,336 2011 BTH 12,618 - 1,557 3,436 3.50 64,603 16,611 2012 BTH 8,195 - 1,358 48,271 3.50 48,104 6,270 2013 BTH 9,077 - - 48,372 3.50 46,989 3,698 2009 BXH 3,419 11,342 2,364 12,036 3.00 71,827 22,811 2010 BXH 12,876 28,513 7,290 2,363 3.00 97,894 13,351 2011 BXH 5,080 27,951 1,200 9,021 3.01 111,774 28,037 928,69 2012 BXH 5,945 22,536 3,012 5,993 3.01 110,379 3 2013 BXH 3,367 32,877 2,401 7,484 3.01 117,205 30,586 2009 CAN 11,425 41,088 4,000 15,037 5.00 165,255 39,185 2010 CAN 11,203 38,754 4,000 17,737 5.00 196,890 36,193 2011 CAN 6,673 59,677 7,500 32,888 5.00 255,207 47,979 2012 CAN 8,717 27,659 10,000 10,836 5.00 228,394 29,135 2013 CAN 22,228 11,226 7,500 13,861 5.00 179,336 27,220 (17,141 2009 CAP 5,260 35,038 2,442 13,318 1.10 66,770 ) (17,458 2010 CAP 8,892 34,029 - 21,248 1.10 78,958 ) 2011 CAP 9,123 19,922 - 27,111 1.10 85,849 (4,960) 2012 CAP 29,690 30,983 12,176 18,466 1.10 112,703 (33,528 82 ) 2013 CAP 13,137 10,675 5,950 20,976 1.10 101,642 (3,882) 2009 CLC 37,913 207,072 15,717 59,769 13.10 469,667 88,493 2010 CLC 67,484 316,111 32,767 35,239 13.10 602,846 83,961 2011 CLC 85,677 276,760 23,587 64,958 13.10 626,168 55,327 2012 CLC 18,226 281,944 32,760 38,685 13.10 580,066 133,08 8 186,37 2013 CLC 22,649 163,543 26,208 70,987 13.10 577,267 0 213,64 2009 CSM 46,047 474,176 29,150 265,629 12.65 1,162,357 1 310,59 2010 CSM 63,239 334,672 172,498 43,496 33.96 1,181,287 3 265,92 2011 CSM 42,504 747,617 84,500 21,610 42.25 1,522,885 9 518,45 2012 CSM 30,416 582,477 62,679 344,098 48.14 1,847,051 3 1,256,9 2013 CSM 35,295 586,31 175,529 370,242 61.49 2,920,797 09 9 2009 CTB 2,734 - 3,943 6,262 1.71 61,445 7,265 2010 CTB 4,828 2,803 2,568 9,019 1.71 71,805 6,953 2011 CTB 11,478 7,767 2,571 9,696 2.25 12,442 17,705 2012 CTB 15,958 19,389 10,205 3,500 2.76 163,084 (16,395 ) 2013 CTB 21,772 28,423 - 15,851 2.76 216,920 6,721 83 2009 DBT 6,165 94,881 5,949 5,843 3.00 268,792 38,184 2010 DBT 10,277 78,727 6,999 8,556 3.00 256,341 34,694 2011 DBT 8,948 108,986 6,678 9,618 3.00 292,490 32,721 2012 DBT 16,906 99,163 6,243 11,078 3.00 342,350 16,736 2013 DBT 14,244 92,234 5,844 9,269 3.00 276,375 22,741 2009 DCL 26,013 251,651 8,097 72,328 9.72 640,720 135,26 3 2010 DCL 24,063 330,322 14,370 2011 DCL 28,140 612,232 19,507 21,320 9.91 792,156 9.91 841,771 (27,619 53,103 (18,093 ) ) 2012 DCL 7,772 395,767 24,344 18,606 10.06 665,899 30,662 2013 DCL 16,470 241,185 - 53,947 10.06 611,669 50,698 2009 DHC 9,870 189,157 - 22,643 15.00 400,440 18,029 2010 DHC 13,138 268,449 13,000 23,584 15.00 547,687 1,780 2011 DHC 3,818 212,303 13,500 7,466 15.00 435,520 (138,84 9) (40,671 2012 DHC 2,712 186,210 - 17,622 15.00 435,520 ) 2013 DHC 5,168 101,825 - 46,605 15.00 399,179 2009 DHG 584,129 73,980 30,018 362,101 26.66 1,521,973 (3,479) 177,46 6 327,95 2010 DHG 642,519 12,802 66,880 357,918 26.91 1,819,735 9 479,58 2011 DHG 467,084 21,116 261,400 211,959 65.17 1,995,707 3 84 445,15 2012 DHG 718,975 19,485 131,268 426,793 65.38 2,378,265 2 589,03 2013 DHG 613,287 127,031 229,640 434,655 65.37 3,080,620 0 2009 DHT 21,833 119,324 7,669 21,760 4.12 244,763 38,545 2010 DHT 20,127 121,834 - 25,683 4.12 265,864 17,696 2011 DHT 22,973 147,306 - 29,099 4.48 304,093 47,574 2012 DHT 21,165 113,985 12,549 12,997 6.28 293,031 65,103 2013 DHT 42,202 103,350 - 42,274 6.28 324,780 53,380 2009 DMC 22,822 82,361 27,034 67,383 17.56 709,977 169,12 0 174,90 2010 DMC 41,393 137,068 17,504 73,523 17.56 776,809 8 172,98 2011 DMC 62,495 129,349 38,508 73,149 17.81 836,267 8 258,69 2012 DMC 4,122 77,420 65,895 53,129 17.81 848,948 9 317,25 2013 DMC 3,570 108,761 65,895 53,129 17.81 868,232 2 499,83 2009 DQC 46,446 701,342 9,026 21,798 18.80 1,603,269 0 594,58 2010 DQC 53,731 625,107 18,797 50,277 24.42 1,733,224 8 557,48 2011 DQC 184,927 499,285 30,257 29,512 24.42 1,832,039 9 85 538,26 2012 DQC 201,163 437,223 43,342 17,658 24.42 1,716,046 6 273,32 2013 DQC 432,689 481,145 42,078 101,552 24.42 1,668,580 6 2009 DZM 14,134 37,923 3,969 17,554 2.48 183,621 38,250 2010 DZM 6,918 48,146 10 8,568 3.11 193,789 16,103 2011 DZM 16,273 49,631 3,419 8,912 3.45 248,132 24,150 2012 DZM 25,759 43,514 - (7,906) 5.40 211,817 21,522 2013 DZM 6,438 39,994 2,698 (1,157) 5.40 173,316 34,415 2009 FMC 282,929 452,874 9,115 19,439 8.00 623,407 (248,04 6) 2010 FMC 40,590 305,850 8,637 32,356 8.00 501,973 2011 FMC 134,547 570,933 16,612 26,479 8.00 778,049 3,913 (88,764 ) 2012 FMC 12,738 252,111 18,057 5,636 8.00 458,960 2013 FMC 151,574 405,571 5,056 48,565 13.00 741,771 14,951 (51,617 ) 2009 GDT 36,426 41,338 20,736 14,528 10.37 195,960 (6,342) 2010 GDT 46,390 22,427 10,372 27,946 10.37 192,721 6,725 2011 GDT 39,611 11,635 20,745 26,606 10.37 206,248 37,989 2012 GDT 34,976 53,733 31,110 11,776 10.37 257,096 43,463 2013 GDT 3,499 37,205 18,666 23,400 10.37 263,767 97,810 2009 GLT 3,776 6,046 16,858 6,231 9.18 202,228 46,767 2010 GLT 10,539 27,661 15,577 20,130 9.18 191,440 53,889 2011 GLT 14,401 5,967 29,489 3,627 9.18 176,820 33,671 86 2012 GLT 63,351 2,677 31,952 2013 GLT 25,107 36,514 55,618 2,107 9.18 215,463 67,059 9.18 168,122 85,591 (46,633 ) 2009 GTA 11,065 - 5,881 14,616 10.40 218,816 60,668 2010 GTA 8,031 - 7,091 16,441 10.40 199,377 75,449 2011 GTA 30,643 - 9,020 13,763 10.40 216,886 51,489 2012 GTA 38,031 - 8,324 15,582 10.40 213,797 51,158 2013 GTA 122,564 80,000 8,884 12,103 10.40 285,577 (30,501 ) (11,108 2009 HAD 20,602 - 1,600 58,316 4.00 119,725 ) 2010 HAD 42,756 - 10,000 57,081 4.00 122,338 14,298 2011 HAD 63,627 - 13,953 38,525 4.00 136,265 15,617 2012 HAD 79,518 - 7,796 37,341 4.00 145,048 8,057 2013 HAD 78,528 - 9,918 30,057 4.00 155,332 18,209 2009 HAI 32,087 144,617 - 50,205 14.50 553,395 220,40 7 230,38 2010 HAI 30,107 51,750 29,578 51,742 14.50 506,998 4 255,30 2011 HAI 17,179 157,817 43,482 10,485 15.80 589,689 6 243,67 2012 HAI 21,140 163,739 34,800 9,916 17.40 654,678 3 247,15 2013 HAI 11,931 222,498 51,388 103 17.40 687,631 9 87 2009 HHC 19,698 - 4,992 31,193 5.48 192,350 26,440 2010 HHC 28,400 - 4,023 29,155 5.48 224,397 34,987 2011 HHC 45,088 - 9,023 25,232 8.21 288,333 22,396 2012 HHC 80,654 - 13,532 23,947 8.21 300,326 240 2013 HHC 58,999 - 18,608 18,899 8.21 315,210 28,926 2009 HRC 54,991 8,112 32,340 66,845 17.26 436,749 (13,559 ) 2010 HRC 25,045 8,112 36,068 67,022 17.26 502,114 7,424 2011 HRC 36,779 35,192 69,031 64,435 17.26 620,452 12,105 2012 HRC 79,644 107,927 41,780 51,795 17.26 656,835 (13,162 ) (143,98 2013 HRC 69,644 161,596 35,905 36,003 17.26 699,624 0) 2009 HSI 13,620 397,825 - 22,961 10.00 759,579 35,211 2010 HSI 11,419 214,466 10,731 16,626 10.00 607,463 57,062 2011 HSI 19,340 475,942 14,292 13,654 10.00 829,209 22,277 2012 HSI 8,657 518,706 12,726 (4,682) 10.00 826,421 (13,876 ) (65,753 2013 HSI 3,957 395,584 - (93,850 10.00 581,803 ) ) 113,93 2009 HTV 39,886 - 11,299 18,959 10.08 218,451 4 144,70 2010 HTV 40,608 - 5,661 15,929 10.08 247,345 2 2011 HTV 10,793 - 9,917 19,382 10.08 256,019 159,74 88 9 137,22 2012 HTV 8,440 - 9,618 40,570 10.08 296,723 9 141,64 2013 HTV 15,438 26,277 19,236 20,999 10.08 350,075 9 2009 HVT 1,377 143,224 - 31,974 2.49 219,075 2010 HVT 2,834 109,383 - 33,639 2.49 221,575 29,069 (30,377 ) (23,550 2011 HVT 1,155 101,725 6,511 30,208 7.23 217,615 ) (19,057 2012 HVT 1,155 101,725 - 36,718 7.23 217,615 ) (23,550 2013 HVT 1,155 101,725 - 36,718 7.23 217,615 ) (90,240 2009 KDC 984,611 526,747 134,947 469,155 79.55 4,247,601 ) 610,92 2010 KDC 672,316 474,343 122,784 545,124 119.52 5,031,920 3 (192,35 2011 KDC 967,330 276,562 - 480,057 119.52 5,809,421 7) 106,86 2012 KDC 829,459 151,089 3,546 569,032 159.92 5,514,704 3 (244,14 2013 KDC 1,958,065 569,825 1,189 721,657 167.63 6,378,246 3) 2009 KMR 10,687 119,497 - 31,414 27.30 520,006 58,284 89 145,06 2010 KMR 14,133 119,493 - 43,595 32.45 613,348 5 204,37 2011 KMR 3,348 119,338 - 11,329 32.45 623,594 4 159,71 2012 KMR 9,117 110,937 - 1,396 33.52 597,453 1 2013 KMR 20,352 101,959 - 40,636 34.40 599,898 74,734 2009 LAF 7,282 85,496 - 26,446 13.39 215,358 39,186 2010 LAF 124,518 6,715 15,191 73,710 13.39 354,368 53,121 2011 LAF 52,031 452,487 15,586 535 13.39 723,638 92,061 2012 LAF 7,731 136,407 - 14.73 235,746 (146,38 (1,410, 4) 639) 2013 LAF 28,649 50,899 - 38,989 14.73 188,794 3,872 2009 LIX 94,475 4,554 3,600 105,137 9.00 296,048 58,243 2010 LIX 52,280 65,270 27,000 52,200 9.00 350,116 138,88 3 2011 LIX 85,063 104,856 45,000 25,715 9.00 406,739 57,767 2012 LIX 84,284 42,261 24,297 44,474 10.80 457,315 64,731 2013 LIX 73,947 44,905 16,200 62,181 21.60 531,654 115,91 0 2009 MCP 25,918 54,095 10,184 15,734 8.20 201,979 54,666 2010 MCP 9,321 58,101 10,728 21,291 8.20 228,091 64,176 2011 MCP 12,900 83,323 13,612 33,296 10.34 305,053 72,047 2012 MCP 5,622 59,065 22,440 13,640 10.34 275,906 69,552 2013 MCP 11,867 87,784 16,415 24,628 10.34 321,038 42,026 90 2009 MHL 11,827 17,495 650 7,856 7.68 57,402 4,662 2010 MHL 39,256 15,430 5,600 (685) 3.98 98,364 (2,716) 2011 MHL 13,093 5,695 2,783 912 3.98 59,852 18,667 2012 MHL 3,631 19,500 - 3,612 3.98 134,607 5,892 2013 MHL 7,494 70,803 - 6,817 3.98 180,680 (1,721) 2009 MKV 1,087 4,380 - 2,039 1.00 20,913 7,437 2010 MKV 2,932 5,986 1,200 992 1.06 23,655 6,465 2011 MKV 1,243 5,028 936 (396) 3.00 19,468 3,696 2012 MKV 2,669 5,066 - (1,713) 3.00 18,420 (864) (20,599 2013 MKV 1,953 33,366 - 1,232 3.00 67,995 ) 2009 MMC 954 - 2,344 4,362 1.20 37,716 7,959 2010 MMC 4,474 - - 7,050 3.16 51,522 28,414 2011 MMC 310 4,902 3,792 (4,543) 3.16 53,064 27,109 2012 MMC 541 8,463 - (1,934) 3.16 50,760 23,818 2013 MMC 66 6,003 - (4,203) 3.16 41,822 17,438 2009 NGC 1,066 37,408 1,000 3,958 1.20 58,426 17,585 2010 NGC 815 60,581 1,800 2,322 1.20 88,006 8,374 (15,329 2011 NGC 664 76,605 960 4,486 1.20 110,441 ) (22,498 2012 NGC 846 69,788 1,440 3,342 1.20 101,679 ) (29,049 2013 NGC 2,179 72,078 840 4,680 1.20 107,345 ) 2009 NPS 24,280 26,966 - 6,102 1.06 106,997 (18,574 91 ) 2010 NPS 4,322 7,831 2,295 3,758 2.17 81,358 2011 NPS 9,004 2,608 3,372 5,035 2.17 85,274 (4,250) (12,451 ) 2012 NPS 2,241 3,901 4,329 4,661 2.17 80,065 2013 NPS 3,884 7,804 4,340 343 2.17 80,930 (5,059) (10,081 ) 2009 NSC 72,741 628 4,982 35,745 10.03 246,942 55,749 2010 NSC 17,038 50,315 8,014 41,460 10.03 302,986 58,348 2011 NSC 19,007 76,486 33,011 35,503 10.03 362,548 164,68 4 2012 NSC 121,295 187 28,894 55,589 10.03 393,154 2013 NSC 112,238 - 30,104 75,925 10.03 439,523 58,488 122,60 5 220,86 2009 PAC 55,636 147,038 35,128 131,528 20.54 669,892 7 2010 PAC 197,242 394,509 43,395 110,278 22.55 1,103,439 23,759 2011 PAC 51,478 371,345 10,156 121,880 26.99 1,187,096 78,801 2012 PAC 83,913 292,690 69,731 46,932 26.63 1,145,491 28,511 2013 PAC 199,972 479,746 53,251 66,158 26.63 1,430,558 (119,13 1) 2009 PMC 12,725 - 7,931 27,713 6.46 104,995 39,936 2010 PMC 25,392 - 9,052 30,607 6.46 124,371 42,312 2011 PMC 28,232 - 12,784 33,909 6.46 151,661 52,732 2012 PMC 47,676 - 22,787 29,265 6.46 168,708 46,012 92 2013 PMC 66,987 - 10,527 52,522 6.46 208,928 2009 POT 81,789 86,282 11,692 14,351 19.43 467,727 58,614 123,09 2 114,84 2010 POT 90,944 128,338 14,643 14,508 19.43 511,894 7 176,33 2011 POT 20,546 159,648 15,410 2,989 19.43 644,605 7 175,13 2012 POT 61,605 76,606 5,792 11,396 19.43 634,411 8 2013 POT 56,819 91,098 7,721 11,365 19.43 640,781 24,838 2009 PTC 13,336 44,078 8,868 (349) 10.00 435,378 90,708 2010 PTC 36,012 41,175 1 8,584 10.00 438,620 63,677 2011 PTC 7,961 38,051 - 8,127 10.00 433,678 88,474 2012 PTC 7,972 30,986 - 10.00 360,175 73,712 (43,604 ) 2013 PTC 12,715 643 - 7,478 10.00 332,218 89,854 2009 PTM 445 6,547 - (369) 2.30 21,422 3,788 2010 PTM 1,840 298 - 348 2.30 20,006 8,566 2011 PTM 16,791 - - 1,097 2.30 43,730 9,208 2012 PTM 18,791 - - 1,097 2.30 43,730 7,208 2013 PTM 20,791 - - 1,097 2.30 43,730 5,208 2009 RAL 129,159 438,217 23,000 44,610 11.50 1,049,313 (7,528) 2010 RAL 39,664 547,352 23,000 56,340 11.50 1,170,010 106,16 2 2011 RAL 161,857 663,641 31,050 76,132 11.50 1,399,951 37,039 93 2012 RAL 244,045 679,459 34,500 140,000 11.50 1,531,241 2013 RAL 503,689 785,436 34,500 127,845 11.50 1,772,437 10,727 (180,10 4) 2009 SAF 31,513 - 5,953 14,085 2.71 92,101 612 2010 SAF 19,941 - 2,776 18,677 4.38 105,959 20,413 2011 SAF 27,849 - - 27,611 4.55 127,976 27,740 2012 SAF 16,004 - 11,364 14,660 4.55 126,483 42,796 2013 SAF 27,822 - 15,683 16,688 4.55 134,213 37,934 2009 SAV 127,168 127,157 19,068 3,106 9.96 585,547 105,65 2 153,62 2010 SAV 93,850 137,855 509 16,712 9.96 536,541 8 181,75 2011 SAV 72,040 154,501 4,819 11,367 9.96 631,632 8 168,40 2012 SAV 39,625 114,746 6,686 10,107 9.96 682,250 8 158,31 2013 SAV 29,013 155,474 7,074 5,620 9.96 653,948 6 2009 SDN 5,285 7,908 945 4,946 1.52 40,550 5,683 2010 SDN 3,621 6,646 2,817 3,543 1.52 40,177 7,521 2011 SDN 4,192 7,379 3,036 3,053 1.52 44,396 7,227 2012 SDN 1,731 1,875 3,492 5,246 1.52 42,003 12,490 2013 SDN 10,633 11,313 3,796 4,446 1.52 53,505 5,180 2009 SFI 88,588 - 5,424 33,981 8.29 326,360 (79,745 ) 94 (103,23 2010 SFI 128,180 - 5,862 32,827 8.29 397,670 1) (153,01 2011 SFI 182,385 - 12,435 23,810 8.29 444,383 3) (143,65 2012 SFI 168,123 2,045 12,435 37,622 8.70 491,417 4) (102,53 2013 SFI 152,342 949 12,794 25,020 8.70 517,824 5) 2009 SFN 2,112 4,364 5,032 11,127 3.00 57,519 23,806 2010 SFN 8,633 10,405 3,790 5,865 3.00 61,525 16,721 2011 SFN 2,500 23,745 4,073 9,780 3.00 71,076 15,198 2012 SFN 3,867 22,063 4,583 6,956 3.00 66,057 14,634 2013 SFN 1,531 18,014 4,625 10,365 3.00 70,557 15,746 2009 SGC 13,946 18,898 7,148 17,575 7.15 112,644 59,180 2010 SGC 8,695 10,000 5,957 9,263 7.15 117,012 33,889 2011 SGC 9,413 20,750 7,149 46,056 7.15 142,282 57,829 2012 SGC 10,128 - 21,443 1,897 7.15 127,726 39,859 2013 SGC 14,215 - 17,869 11,050 7.15 130,189 37,723 2009 SGT 449,394 518,737 - 85,503 67.27 2,018,043 (80,648 ) 114,34 2010 SGT 6,436 999,365 6 36,216 67.27 2,334,870 2 1,067,9 2011 SGT 54,437 (131,97 - 51 2012 SGT 10,750 771,822 (672,00 67.27 1,463,032 4) - (256,54 6) 67.27 1,902,710 (547,45 95 8) 9) (332,38 2013 SGT 13,381 744,764 - 3,751 67.27 1,813,085 8) 2009 SJ1 44,386 5,485 2,685 9,588 3.85 106,427 4,769 2010 SJ1 5,657 33,525 7,383 4,858 3.85 154,830 8,725 2011 SJ1 6,298 2,823 3,500 14,554 3.85 111,291 6,011 2012 SJ1 1,098 65,948 3,833 13,150 3.85 190,213 (18,509 ) (59,325 2013 SJ1 1,916 104,093 3,834 10,050 3.85 234,498 ) 2009 SPP 4,532 154,416 2,328 15,405 3.50 319,544 68,482 2010 SPP 2,791 269,454 6,000 31,302 3.50 453,572 62,793 2011 SPP 2,896 325,693 8,808 27,276 3.50 589,181 109,82 6 2012 SPP 1,348 397,910 6,152 23,987 3.50 659,152 94,070 2013 SPP 1,314 416,068 5,980 30,004 3.50 718,522 93,228 2009 SRC 41,960 153,175 9,359 148,534 10.80 581,816 5,676 2010 SRC 15,252 56,091 12,756 49,337 10.80 570,723 48,504 2011 SRC 27,681 213,042 24,393 23,450 10.80 660,513 29,525 2012 SRC 29,877 129,791 8,105 80,291 10.80 534,620 87,523 2013 SRC 42,218 115,846 20,168 67,008 10.80 525,288 141,37 1 200,42 2009 SRF 57,099 85,803 5,775 38,095 8.02 680,457 8 2010 SRF 170,747 69,222 32,606 19,369 8.02 591,040 40,755 96 2011 SRF 168,654 75,222 29,262 33,009 8.13 718,478 59,628 2012 SRF 129,861 118,941 39,228 269,162 8.13 703,502 90,677 2013 SRF 246,711 109,426 14,878 357,368 16.25 729,106 55,377 2009 SSC 87,824 24,700 10,869 44,059 9.94 232,893 30,517 2010 SSC 62,417 4,055 14,918 39,179 14.99 260,110 87,133 2011 SSC 70,262 2,018 25,050 36,746 14.99 315,581 85,962 156,82 2012 SSC 43,084 13,047 29,585 42,515 14.99 424,017 3 172,01 2013 SSC 25,952 32,251 14,946 66,347 14.99 468,093 9 2009 SSM 29,274 28,686 2,888 36,265 5.50 180,921 24,026 2010 SSM 65,557 55,242 11,063 3,346 5.50 150,936 11,220 2011 SSM 4,753 43,193 8,000 (1) 5.50 186,514 51,834 2012 SSM 21,416 45,143 6,925 (620) 5.50 146,377 33,997 2013 SSM 35,459 58,879 4,171 7,944 5.50 165,302 24,190 2009 STP 4,323 - 10,500 9,445 3.50 87,573 45,389 2010 STP 35,224 - 7,000 22,833 7.00 177,202 65,287 2011 STP 12,632 - - 18,061 7.00 182,117 64,984 2012 STP 17,743 4,184 9,961 4,444 7.00 183,530 63,448 2013 STP 7,596 10,438 - 11,458 6.64 190,514 89,378 2009 SVI 25,450 - 3,873 28,596 3.90 182,012 6,495 2010 SVI 21,702 - 11,765 36,139 3.90 290,500 3,476 2011 SVI 87,365 - 11,627 59,234 5.92 434,995 (9,555) (33,653 2012 SVI 43,695 5,000 22,237 64,419 8.92 549,662 ) 97 2013 SVI 73,768 - 8,895 111,982 10.70 610,872 2009 TAC 136,448 110,278 37,960 17,725 18.98 651,956 13,929 (34,821 ) 2010 TAC 122,473 223,860 26,564 89,169 18.98 944,175 39,264 2011 TAC 103,698 305,790 37,946 15,941 18.98 1,031,008 18,386 2012 TAC 200,553 175,641 30,404 63,687 18.98 1,001,871 (59,489 ) (207,71 2013 TAC 406,466 298,052 30,363 70,225 18.98 1,222,589 6) 2009 TCL 94,100 25,502 - 115,856 17.00 447,332 35,452 2010 TCL 89,122 62,167 41,381 99,287 18.70 612,167 34,017 2011 TCL 75,285 237,834 41,037 115,173 18.70 903,262 56,059 2012 TCL 90,081 233,978 59,668 95,584 20.94 806,858 7,204 2013 TCL 110,719 172,880 19,410 127,440 20.94 795,131 62,855 2009 THB 20,564 76,578 12,868 35,537 11.42 351,698 (72,254 ) (59,244 2010 THB 25,940 42,954 13,542 54,261 11.42 325,422 ) (35,546 2011 THB 23,006 34,581 17,214 43,242 11.42 304,321 ) (18,730 2012 THB 42,139 - 19,422 38,771 11.42 301,348 ) (33,469 2013 THB 89,141 - 19,624 40,442 11.42 343,299 ) 2009 TJC 11,306 164,752 1,094 16,451 3.00 251,923 (35,586 98 ) (20,678 2010 TJC 10,656 127,350 1,806 21,946 6.00 248,724 ) (31,668 2011 TJC 21,055 123,843 3,000 15,010 6.00 248,923 ) (20,862 2012 TJC 2,688 104,949 - 9,274 6.00 227,802 ) (84,535 2013 TJC 14,496 75,328 - 18,470 6.00 190,025 ) 2009 TMS 53,270 77,191 10,107 47,106 10.10 459,226 36,379 2010 TMS 46,474 59,679 4 73,606 16.52 601,335 42,824 2011 TMS 56,486 76,998 16,512 63,711 18.28 648,299 (4,313) 2012 TMS 87,794 98,986 23,974 66,180 23.07 755,678 (40,112 ) (89,315 2013 TMS 90,512 126,082 21,595 107,653 23.07 836,511 ) 2009 TNC 24,216 - 5,847 33,038 19.25 268,526 82,917 2010 TNC 71,656 - 19,266 41,702 19.25 308,375 58,627 2011 TNC 140,096 - 28,843 51,861 19.25 343,609 35,684 2012 TNC 197,672 - 38,453 41,017 19.25 380,977 6,815 2013 TNC 139,520 - 38,453 3,094 19.25 348,066 47,564 2009 TPC 27,494 12,100 - 67,547 20.55 366,347 204,90 0 205,36 2010 TPC 45,253 5,680 4 46,783 24.43 374,506 4 99 (103,24 2011 TPC 374,470 602,339 30,842 17,960 24.43 991,359 6) 239,67 2012 TPC 111,581 235,645 42,983 (3,067) 24.43 613,666 8 321,35 2013 TPC 21,014 291,073 29,777 2,263 24.43 659,919 8 192,71 2009 TRA 5,434 39,651 16,517 43,147 10.20 406,152 0 233,92 2010 TRA 29,068 63,321 217 74,098 12.34 578,868 6 208,60 2011 TRA 43,241 75,188 24,540 76,799 12.34 838,443 4 158,52 2012 TRA 104,329 155,867 48,973 99,281 12.34 968,484 9 184,88 2013 TRA 256,585 82,361 27,078 166,271 12.34 1,087,715 6 2009 TS4 13,207 179,923 7,772 21,372 8.47 366,746 74,255 2010 TS4 2,479 239,255 12,830 25,448 11.50 549,277 71,629 (34,513 2011 TS4 5,138 350,870 20,459 22,955 11.50 707,228 ) (42,862 2012 TS4 1,785 505,514 6,996 26,746 11.50 915,690 ) (54,706 2013 TS4 461 495,455 6,552 26,066 11.50 986,537 ) 2009 TSC 76,881 650,138 9,975 27,050 8.31 930,258 (653,44 100 1) (66,539 2010 TSC 96,753 565,791 16,626 12,688 8.21 830,881 ) (40,894 2011 TSC 138,452 652,681 8,105 35,168 8.01 963,159 ) (59,129 2012 TSC 14,685 - 8,709 (44,206 8.31 550,123 ) ) (20,050 2013 TSC 28,915 - 4 10,681 8.31 331,988 ) 1,035,7 2009 TTF 6,930 183,64 21,549 14,089 18.97 2,176,303 85 9 1,613,9 2010 2011 2012 2013 TTF TTF TTF TTF 17,157 217,23 23,424 48,762 31.24 2,634,824 91 5 1,921,4 115,01 24,463 - 48,509 31.24 3,337,028 44 9 1,841,3 214,43 6,318 - 33,636 33.95 3,297,738 92 3 1,981,7 302,70 65,513 - 31,354 51.01 3,518,037 11 4 (133,89 2009 TYA 123,482 365,195 - 37,135 27.90 611,203 3) (158,04 2010 TYA 188,700 357,305 - 52,885 27.90 785,069 7) (49,385 2011 TYA 125,051 312,152 - 57,707 27.90 671,370 ) 101 2012 TYA 127,947 197,876 - 58,125 27.90 633,818 4,063 2013 TYA 157,514 113,016 - 53,173 27.90 637,197 24,035 2009 UNI 2,354 - 3 18,266 6.76 122,090 91,400 2010 UNI 52,540 - 4,672 8,288 9.25 173,634 99,244 2011 UNI 8,454 - 4,676 9.48 170,325 (11,814 130,79 ) 2 2012 UNI 3,907 - 8 13,094 9.47 184,275 63,713 2013 UNI 2,354 - 9,357 (8,485) 13.74 179,342 41,200 2009 VBH 5,554 - 870 3,712 2.90 45,244 5,643 2010 VBH 2,332 - 1,740 4,227 2.90 42,730 21,273 2011 VBH 958 - 2,900 2,655 2.90 36,559 27,443 2012 VBH 11,719 - 2,900 899 2.90 36,421 16,942 2013 VBH 8,971 - 2,030 (578) 2.90 35,257 18,988 2009 VDL 42,235 - 3,444 12,939 2.15 93,906 17,201 2010 VDL 51,080 3,819 2,519 19,631 2.15 119,220 10,760 2011 VDL 23,160 10,466 6,829 12,849 3.13 124,355 47,439 2012 VDL 49,762 9,254 6,599 18,562 3.13 139,517 30,522 2013 VDL 44,241 - 3,730 22,946 6.26 141,381 53,848 2009 VFG 64,420 74,224 17,872 89,978 8.13 639,992 84,112 2010 VFG 77,177 100,148 16,247 61,995 9.76 714,974 99,081 193,97 2011 VFG 40,460 34,737 14,622 86,813 12.68 931,061 5 141,82 2012 VFG 87,903 217,164 38,308 60,194 12.77 1,110,479 2 2013 VFG 142,037 76,834 32,709 56,546 13.20 1,108,128 91,674 102 120,47 2009 VGP 7,622 13,967 8,001 18,069 6.20 171,263 6 146,58 2010 VGP 3,475 25,008 7,241 22,553 8.21 207,598 6 139,02 2011 VGP 5,498 57,201 24,081 567 8.21 242,432 0 2012 VGP 59,352 38,184 15,823 4,757 8.21 208,590 86,255 2013 VGP 10,538 121,112 12,658 7,418 8.21 299,483 33,123 2009 VGS 364,894 524,085 7,307 57,691 13.46 1,128,991 (73,021 ) 117,30 2010 VGS 59,382 498,201 22,522 25,513 37.20 1,102,118 0 2011 VGS 25,102 419,330 14,436 7,470 36.00 1,133,689 2012 VGS 31,503 512,042 - 34,636 36.00 1,272,009 42,575 (27,186 ) 2013 VGS 19,608 485,864 3,954 33,166 36.00 1,075,956 16,749 2009 VIS 180,125 870,215 40,460 218,332 15.00 1,497,518 (8,646) 2010 VIS 226,782 533,232 - 146,859 30.00 1,657,855 37,516 2011 VIS 400,130 704,197 58,757 4,111 30.00 1,318,453 (96,605 ) 1,996,0 2012 2013 VIS VIS 457,724 (607,89 29,993 27,062 41.27 2,809,733 80 3) 1,789,0 (501,66 347,230 - 94,359 49.22 2,568,778 81 2009 VNA 77,549 576,767 2) 29,417 39,574 20.00 1,073,170 (253,05 103 2) (202,37 2010 VNA 20,082 664,983 97 152,992 20.00 1,179,413 4) (250,06 2011 VNA 24,001 641,416 42,186 106,659 20.00 1,392,014 2) (225,65 2012 VNA 8,446 142,680 132 7,889 20.00 1,311,008 1) (255,65 2013 VNA 4,537 171,974 9 (8,436) 20.00 1,204,826 7) 2,614,1 2009 VNM 426,135 25,737 351,281 2,908,1 35.13 8,482,036 56 1,765,20 2010 VNM 613,472 2012 2013 2009 VNM VNM VNM VNT 3,156,515 1,252,120 2,140,4 567,960 - 10,773,032 23 18 3,891,3 3,364,6 741,428 55.61 15,582,672 44 31 2,222,99 4,131,9 5,713,5 4 13 00 5,307,0 3,167,23 4,153,3 5,316,8 61 5 05 - 8,056 16,103 - 83.40 2,745,645 48,473 2,661,3 35.31 0 2011 53 83.40 19,697,868 22,875,414 87 5.47 177,405 (7,076) (15,299 2010 VNT 66,635 - 10,301 12,914 5.47 212,429 ) (32,290 2011 VNT 81,543 - 7,405 18,599 5.47 254,254 ) 2012 VNT 85,438 - 18,552 12,531 5.47 263,167 (30,150 104 ) (52,692 2013 VNT 110,643 - 10,075 21,585 5.47 264,465 ) (32,707 2009 VPK 3,932 66,253 - 18,218 8.00 149,756 ) (22,423 2010 VPK 6,985 43,163 - 19,739 8.00 155,464 ) 2011 VPK 10,847 31,516 - 34,554 8.00 155,768 4,471 2012 VPK 39,239 19,816 11,997 45,553 8.00 186,976 18,833 2013 VPK 54,832 86,689 15,967 41,150 8.00 216,721 18,599 2009 VSC 58,985 58,883 32,581 156,843 12.03 647,348 110,29 6 177,32 2010 VSC 88,632 24,781 35,895 187,823 12.03 811,576 7 322,07 2011 VSC 59,052 846 70,769 161,788 23.89 856,939 2 233,18 2012 VSC 46,856 40,381 95,079 177,977 24.04 1,054,559 5 (27,636 2013 VSC 281,725 8,399 89,557 214,941 28.81 1,132,563 ) 1,625,5 2009 2010 VST VST 64,285 (424,51 27,006 158,792 40.00 2,789,420 70 4) 1,862,5 (376,66 32,485 124 339,342 43.17 3,325,870 93 2011 VST 176,584 1,668,3 7) 71,451 95,686 59.00 3,255,640 (621,60 105 38 2) 2,084,7 2012 VST 44,281 (58,560 2,520 64 (534,90 59.00 3,019,255 ) 1) 2,029,5 2013 VST 53,859 (350,55 - 2,754 59.00 2,761,616 99 3) 2009 VTB 20,303 54,626 - 31,848 11.98 316,737 2010 VTB 37,423 28,673 9,063 20,969 11.98 293,614 71,409 100,50 7 2011 VTB 76,480 23,901 16,713 4,630 11.98 278,599 2012 VTB 20,686 14,131 22,288 (5,560) 11.98 265,584 58,449 113,52 9 130,55 2013 VTB 7,515 24,140 7,657 12,154 11.98 268,507 9 2009 VTC 14,384 11,098 - 8,740 4.05 120,377 21,195 2010 VTC 3,158 6,783 - 4,702 4.05 95,186 32,204 2011 VTC 3,169 8,188 - (3,870) 4.05 93,778 25,508 2012 VTC 2,199 2,346 - 3,856 4.05 84,853 24,941 2013 VTC 10,540 1,612 - 4,141 4.05 86,360 18,368 2009 VTL 19,756 29,338 1,852 6,939 1.80 104,624 (4,299) 2010 VTL 6,928 50,393 1,979 5,199 1.80 105,409 9,590 2011 VTL 5,470 42,650 1,979 3,173 1.80 92,297 13,257 2012 VTL 6,378 39,228 - 1,043 1.80 121,887 5,472 2013 VTL 7,868 61,306 - 4,569 1.80 102,201 5,060 106 [...]... affecting the level of cash holding in order to bring general insights of Vietnam stock market to help individual investors make decisions Keywords: Vietnam stock market, descriptive, univariate, regression, level of cash holding XIII This page is intentionally left blank XIV Chapter 1 Introduction The aim of this chapter is to provide the problem background and then to understand the purpose behind the. .. by companies listed in Vietnam stock market The objectives are:  To identify factors that affect the level of cash holding and then measure on their impact to cash management of companies listed in Vietnam stock market  To propose recommendations for the management 3 1.3 Research Questions This research will provide a practice of 98 companies listed on Vietnam stock market resulted from related theories... outlines the practical steps which need to be undertaken to use above methods on companies listed in Vietnam stock market A framework for each methodology is drawn up The emphasis is on analysis trend of factors affecting the level of cash holding when being put in consideration and explanation with related theories The main goal of this study is to present the result of each method and factors affecting. .. affect level of cash holding, by using financial theories, as well as presents academic results of previous authors 2.1 Definitions, Time Value and Reasons for Cash Holding 2.1.1 Cash Definitions Everyone is actually familiar with the term cash daily, and more generally, the Oxford online dictionary defines cash as “coins and notes” According to financial theories, they definite cash that include cash, ... order to face these challenges, firm must reserve cash available to cover investment needs in the future If not, they run into the risk of reject the project and investment value The author mainly uses the first two reasons in this study Two main theories used to explain factors are the trade-off theory and the pecking order theory 2.2 Information Asymmetry Theory and Related Theories In the 1960s, both... questions:  What firm factors affect the level of cash holding of companies listed on Vietnam stock market?  What are the implications for cash management on companies listed on Vietnam stock market? 1.4 Research Scope Due to time limitation and data availability, this research focuses on companies listed on two stock exchanges in Vietnam: Ho Chi Minh and Hanoi Stock Exchange Most of them are supervised... Figure 1: Theoretical Framework to Be Applied in This Research 1.7 Contribution and Implication of Thesis to Business and Society This research provides a review of Vietnam stock market based on examining the level of cash holding and cash management It suggests an approach allowing local individual investors to make informed decisions in stock market A suitable model for analyzing cash holding is really... arise in a shareholder-manager situation, concerns among other things the optimal level of cash holdings Jensen (1986) worried that managers have motives for having higher levels of corporate cash holdings than what is optimal for shareholders 10 Trade-off theory implies that businesses will set up an objective of cash level by balancing between marginal cost and marginal benefit of holding cash The. .. just a theoretical perspective as around the new millennium increasingly many companies were striving for decreased cash balance; financial managers globally advocated the benefits of small cash 1 holdings, Mintz (2000) stated Holding excess cash might lead to a loss of an opportunity that could have increased the company’s value With holding high level of cash, it will decrease earning yield of business... allocation of resources, they will accumulation of cash Damodaran (2008) gave his opinion that it is not for shareholders but for the purpose of opening wide funds The management is favorite cash holdings, because they are afraid of risks According to Opler et al (1999), the management holds cash to get flexibility in pursuing their own goals The final one is for investing in the future If the capital market ... participate in the final thesis defense Approved by (Signed) PhD Duong Nhu Hung II FACTORS AFFECTING THE LEVEL OF CASH HOLDING OF COMPANIES LISTED IN VIETNAM STOCK MARKET In Partial Fulfillment of the. .. related theories The main goal of this study is to present the result of each method and factors affecting the level of cash holding in order to bring general insights of Vietnam stock market. .. MBA02002 Title of Thesis: Factors affecting the level of cash holding of companies listed in Vietnam stock market Advisor: PhD Duong Nhu Hung I assure that the content of this thesis has been