The Effect of Cash Holdings on Firm Value
with Regard to the Relation between
Cash Flow and Investment Opportunities
and a Natural Experiment with State Credit Ratings
__________________________________________________________________________
June 2018
Abstract
This paper explores the impact of cash holdings on firm value when internal- or external availability of funding for future investments increases. A setting is studied in which the availability of internal funding is measured by the correlation between cash flow and investment opportunities, whereas the availability of external funding is measures by the credit rating of the state in which the firm is incorporated. The main results show that when these measures for internal- and external availability for funding increase, an increase in cash holdings has a negative effect on firm value. For high growth firms, the impact of availability of internal funding for future investments is stronger. These results imply that it could not be optimal for a firm to increase its cash holdings when the availability of internal- or external funding has increased.
Name: Anthonie Niklaas Schipper Student Number: 10748032
Programme: MSc Finance Specialization: Corporate Finance Supervisor: Dr. S. Terovitis Coordinator: Dr. J.E. Ligterink
Statement of Originality
This document is written by Student Anne Schipper who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
1 Introduction . . . 4
2 Literature Review . . . 6 2.1 Cash Holdings
2.2 Cash Flow & Investment Opportunities 2.3 Availability of External funding
2.4 High Growth Firms
3 Data . . . 9 3.1 Data collection and preparation
3.2 Summary Statistics
4 Methodology . . . 13 4.1 Cash Flow & Investment Opportunities
4.2 State Credit Ratings 4.3 High Growth Firms 4.4 Remaining
5 Results . . . 20 5.1 Cash Flow & Investment Opportunities
5.2 State Credit Ratings 5.3 High Growth Firms
5.4 Remaining
5.5 Robustness checks
6 Conclusion and Discussion . . . 27
7 References . . . 28
1 Introduction
There are various reasons for firms to retain their earnings instead of reinvesting it or distributing it to shareholders. While reinvesting or distribution over shareholders might increase shareholder wealth, it might even be in the best interest of shareholder for a firm to hold on to a certain amount of cash. One important reason is holding on to cash as a precautionary saving. Firms need to be able to capitalize on possible future investment opportunities. Cash flow from that period, internal funding, might not always be sufficient to do this, or can be unpredictable. External funding can be costly, not only due to interest payments, but also due to asymmetrical information (Myers and Majluf, 1984). However, holding on to excessive amounts of cash can also be very costly. One example for this is due opportunity costs. When firms hold on to too much cash, which they can also be using for positive net present value investments, profitability will decrease (Almeida, Campello and Weisbach, (2004). This creates a trade-off problem on deciding the optimal amount of cash holdings for a firm. Because of the direct effect on shareholder wealth and firm profitability, the optimal amount of cash holdings is a topic that affects the entire economy. Research to better understand the effects of the availability of internal- and external funding on the optimal amount of cash holdings for a firm can provide crucial insights on this topic. Because of these reasons, this research is constructed to try to find an answer on the following research question: Do managers of firms, for which it has become easier to get internal or external funding for future investments, destroy firm value when they increase the firm’s cash holdings? To find an answer on this research question, empirical research with the use of OLS-regressions with interaction variables will be performed. This will be done in a setting where the correlation between cash flow and investment opportunities will be used to measure the availability of internal funding for a firm’s future investments. This correlation describes the alignment of the demand and supply of cash within a firm. As a measure for the availability of external funding for future investments, the credit rating of the state in which the firm is incorporated are used. An increase in this credit rating will make it easier and cheaper for a firm to acquire this form of funding. The states credit rating is used, because its fluctuation is assumed to be independent of a single firm’s actions, therefore providing an exogenous shock. For both the measure of internal- and external availability it is tested if an increase in availability has a positive effect on the value of a firm. Afterwards, it is tested if an increase in cash holdings while there is also an increase in the availability of internal- or external funding
has a negative effect on the value of a firm. In addition to these tests, is will also be tested if these effects of availability of internal- and external funding have a bigger effect on high growth firms. These firms should be more sensitive to their ability to capitalize on investment opportunities.
Results of this research show that both the increase in correlation between cash flow and investment opportunities as an increase in state credit rating have a positive effect on firm value. More importantly, results show that when these measures for internal- and external availability for funding increase, an increase in cash holdings has a negative effect on firm value. For high growth firms, the results indicate that the impact of availability of internal funding is stronger, whereas there is no significant difference in the impact of the availability of external funding. These results imply that it could not be optimal for a firm to increase its cash holdings when the availability of internal- or external funding has increased. When this is the case, firms can allow themselves to hold on to lower amounts of cash and still be able to capitalize on possible future investment opportunities. Because cash holdings are costly, increasing them in these scenarios will be bad for the firm’s value.
The remaining content of this paper will be structured in the following way. In section 2, the related literature to the topics that are discussed in this paper will be discussed, along with the place this paper will have within the existing literature. Afterwards, in section 3, the collecting, preparation and summary statistics of the data that is needed to perform the statistical tests of the research in this paper will be discussed. In section 4, the methodology of the statistical tests in this paper will be discussed for a better understanding and replicability of the results in this paper. The results of the empirical research in this paper will be displayed and discussed in section 5. In section 6, the conclusion and discussion will be provided, in the form of a brief summary of the paper and recommendations for further research on this topic. Finally, section 7 and 8 will provide the references and the appendices respectively.
2 Literature Review
In this section the related literature to the topics in this research, as well as the place of this paper within the existing literature, will be discussed. Firstly, there will be an introduction to existing literature on cash holdings in general. Secondly, previous research about the relation between cash flow and investment opportunities will be discussed. After that, research on the topic of availability of external monetary funding is briefly touched upon, which is related to the change in a state’s credit rating in this paper. Finally, existing literature on the topic of the possible different effect cash holdings have on high growth firms will be discussed.
2.1 Cash Holdings
The topic of cash holdings is one on which a lot of research has been done. Research on the costs of cash holdings is especially related to this research. An important reason to have a sufficient amount of cash as a firm is to be able to capitalize on possible future investment opportunities as is shown in for example the paper by Han and Qui (2007). These kind of cash holdings are called precautionary savings and are the most important form of cash holdings regarding the research in this paper. The results in the paper by Campello, Graham and Harvey (2010) show that these precautionary savings have become even more important to managers since the recent financial crisis. The results of the research by Foley et al. show that firms hold onto cash due to taxation reasons. Multiple reasons to hold on to cash, often in the interest of shareholders, have been causing cash holdings of firms to grow significantly over the last thirty years, as is shown in the research of Bates, Kahle and Stulz (2009). However, there are also costs connected to having cash holdings. This should make managers decide to not hold on to more cash than necessary. One example of a paper that mentions opportunity costs as an important factor of costs of cash holdings is the one by Almeida, Campello and Weisbach (2004). They state that higher cash holdings make a firm less profitable than it could be, due to it having to pass on valuable projects to retain its cash. The results of the research by Harford (1999) show that excess cash holdings can be costly, because they might cause a firm to engage into investments with a negative net present value. Alltogether, this is called the trade-off problem, Lin and Chiu (2017). Another view on this, as shown in the paper by Opler et al. (1999) is that the amount of cash holdings does not matter for firm value and there is no optimal amount for this. Regarding cash holdings, this paper will differ from the mentioned
papers by researching the effect of cash holdings on firm’s value, especially after internal or external funding for future investments has become better available.
2.2 Cash Flow & Investment Opportunities
Results of various research papers on the topic of the relation between cash flow and investment opportunities show that when these two variables are well aligned (high correlation), firms require lower cash holdings than those firms which have a low correlation between these two variables (Acharya, Almeida & Campello, 2007; Almeida Campello & Weisbach, 2004; Duchin, 2010). This is the case because when there is a greater alignment between demand for- and supply of cash, firms can to some extent rely on their future cash flows to be able to capitalize on future investment opportunities and therefor have a lower need for precautionary savings. Results of the research in the papers by Acharya, Almeida and Campello (2007) and Almeida Campello and Weisbach (2004) also show that a low correlation between these two variables can be used as a measure for financial constraints. To measure investment opportunities, the research in this paper will use research and development (R&D) expenditures. In many research papers, like Carpenter and Guariglia (2008), Tobin’s Q is used as a measure for investment opportunities. However, due to possible causality problems when using Tobin’s Q to predict firm value, this is not used in the research of this paper. As for example the paper by Lyandres and Zhdanov (2013) shows, R&D expenditures are a popular measure for investment opportunities as well, because it is likely to be related to a firm’s investment opportunities, but not likely to be related to the valuation of a firm. Regarding the topic of the relation between cash holdings and investment opportunities, this paper will contribute to the existing literature by not only researching if a higher correlation between these two variables is beneficial for a firm, but also if a higher correlation between these two variables will make an increase in cash holdings a value destroying decision. This creates the first two hypotheses to help answer the main research question of this research. The first hypothesis being that an increase in the correlation between cash flow and investment opportunities has a positive effect on firm value. The second hypothesis being that an increase in cash holdings while there is an increase in the correlation between cash flow and investment opportunities has a negative effect on firm value.
2.3 Availability of External funding
Another research topic in this paper is the effect of availability of external funding. In the paper by Becker‐Blease & Paul (2006), inclusion if firms in the S&P 500 index is used as an exogenous shock for stock liquidity and therefore availability of external funding, because firms do not self-select into the S&P 500 index. However, because this is not exogenously determined with regard to the market capitalization of the firm, this shock cannot be used for the purpose of the research in this paper. Instead, the credit rating of states in which firms are incorporated, is used to provide this research with an exogenous shock. This exogenous liquidity shock would reduce the systematic risk of the firm. The results of the research by Tuzel and Zhang (2017) show that a reduction in systematic risk leads to firms getting easier excess to external funding for future investments, thereby also reducing the firm’s financial constraints. The results of the research by Rauh (2006) show that there is a significant effect of credit ratings, in his case firm’s credit ratings, on the precautionary savings of a firm. In contrast, this paper uses state credit ratings as a measure of availability of external funding because this causes an exogenous shock for the firm. This is, assuming that the actions of a single firm do not affect a state’s credit rating, where a firm’s credit rating will not be exogenous of its actions. Furthermore, this paper will contribute to the existing literature by not only researching if better availability of external funding is beneficial for a firm, but also if better availability of external funding will make an increase in cash holdings a value destroying decision. This creates the third and fourth hypothesis to help answer the main research question of this research. The third hypothesis being that an increase in the credit rating of the state in which a firm is incorporated has a positive effect on firm value. The fourth hypothesis being that an increase in cash holdings while there is an increase in the credit rating of the state in which a firm is incorporated has a negative effect on firm value.
2.4 High Growth Firms
Previous research on the effect of cash holdings of high growth firms, like the papers by Donati (2016) and Lee (2014) show in their results that high growth firms are more sensitive to cash flows and therefor the availability of internal or external funding. The results of the research by Lee (2014) show that cash holdings are one of the factors that have an impact a firm’s ability to be high growth. From those results, it can be concluded that a better availability of internal or external funding might have a stronger impact on high growth firms. As an addition
to the previously mentioned four hypotheses, this creates a fifth hypothesis to help answer the main research question of this research. The fifth hypothesis being that the effects that are measured in the research for Hypothesis one to four are larger for firms in a high growth stadium.
3 Data
In order to find statistical results to support the stated hypotheses in this research, the right data needs to be collected. In this section, the process of gathering the data that is needed to perform this paper’s research, will be described. In the first section the data collection will be elaborated on. The second section will show the summary statistics of the final datasets.
3.1 Data Collection and preparation
To study the relationships between investment opportunities, cash flow, state’s credit ratings, cash holdings and firm value, multiple variables are gathered. An overview of the collected data and the corresponding regression variables for which it is needed in this research can be found in Table 1. All data on selected firms are gathered from the Compustat database from WRDS (Wharton).
In this research the market capitalization of a firm is used to measure a firm’s value. The market capitalization is calculated for each fiscal year, multiplying the amount of common shares outstanding with the share price. For the cash holdings of a firm, the data for cash and cash equivalents or short-term investments is collected. Firms are regarded as being in a high growth state in the same way as it is done by the Organisation for Economic Co-operation and Development (OECD). This is the case when the firms experience a growth in net sales or employees by at least 20% in one year.
As elaborated on before, in this research, research and development expenditures are used as a measure for investment opportunities. A firm’s cash flow is calculated by adding income before extraordinary items and depreciation and amortization, as done in the paper by Duchin (2010). A rolling correlation is created between the variables for cash flow and investment opportunities. This is done by, for each given year, calculating the correlation between these two variables over the past five years. Only when at least five previous years of data is available, a correlation is created. Finally, a dummy variable for an increase in the
correlation between these two variables is created which is equal to one if this correlation is at least 0.05 higher than that in the previous year and zero if this correlation has not changes with respect to the previous year, meaning it has stayed within a range of -0.05 and 0.05.
The data on the credit rating of the states in which firms are incorporated is hand collected form S&P ratings. This data is available from 2004 until 2017. An overview of increases in state credit ratings per year can be found in Appendix 1. A dummy variable is created which is equal to one if the credit rating of the state the firm is incorporated in is better than it was in the previous year and zero otherwise.
Because the data on state credit ratings in only available from 2004, a change in state credit rating can be measured from 2005 and onwards. Because the rolling correlation between cash flow and investment opportunities requires a minimum of five previous years of data, a change in this correlation can be measured from the seventh year of data and onwards. For these reasons, to be able to test all hypotheses on the same dataset for comparability, data is gathered from the year 1999 and onwards. This allows testing on changes in correlation between cash flow and investment opportunities and changes in state credit ratings from 2005 up until 2017.
Due to data availability, and the use of state credit ratings, data is collected from US firms. Similar as in Almeida et al. (2005) and Duchin (2010), excluded from the data set are firm-years in which the market capitalization is less than ten million, cash holdings exceed the total assets and those in which crucial data is missing. For the main tests, but not for the robustness check, financial firms are also excluded, because of the unusual behaviour of cash holdings in financial firms, which might impact the results. This results in a final data set in which there are 8.995 (3.055) observations in which there is an increase or no change in correlation between cash flow and investment opportunities over 1.727 (802) firms (in a high growth state). And 18.751 (6.998) observations in which there is or is not an increase in the state’s credit rating in which the firm is incorporated over 2.361 (1.606) firms (in a high growth state).
Table 1 – Content on regression variables. Shows the variables which are used in the
empirical research of this paper in the first column. The second column provides a brief explanation of the variable and the third column shows the raw data that is needed to construct this variable. High growth firms is not a variable that is used in any of the regressions in this study, but is included in this table to display the data that is necessary to conclude if a firm is in a high growth state or not.
Regression Variable
Explanation Raw data needed
(Compustat codes)
MarketCap The market capitalization/market value. csho, prcc_f
Ln(Cash) The natural logarithm of Cash and Equivalents. che
RiseCorrelation Dummy
A dummy variable which will be equal to 1 if the correlation between cash flow and investment opportunities is bigger, compared to the year before and equal to 0 if this has not changed (stayed within borders).
Ib, dp, xrd
IncreaseInState CreditRating
A dummy variable which will be equal to 1 if the credit rating of the state in which that firm is incorporated increased in that year.
Hand Collected from S&P credit ratings, incorp
InteractionVariable1 Growth_in_Cash*RiseCorrelationDummy -
InteractionVariable2 Growth_in_Cash*IncreaseInStateCreditRating - High Growth Firms Firms as regarded to be in a high growth state
when in that year net sales or number of employees grows by at least 20%.
sale, emp
3.2 Summary Statistics
This section provides an overview of the final dataset that is used for the empirical research in this paper. Table 2a shows the firm’s characteristics, means and standard deviations, for firms that experienced an increase in the correlation between cash flow and investment opportunities and firms that have experienced no change in this correlation with regard to the previous year. The p-value of difference between the means of these two groups, resulting from a t-test, is displayed in the third column. It is important to see that, there are no significant differences in firm characteristics between the means of the two groups divided by this variable. This means that an increase or no change in correlation between cash flow and
investment opportunities is not caused by firm characteristics that are not in the statistical tests and might have had an unwanted impact on the results. For the observations of only firms in a high growth state, the summary statistics are shown in Table 2b in Appendix 2.
Table 2a – Summary statistic on data from the entire database divided in observations
where there is an increase in the correlation between Cash Flow and Investment Opportunities (column 1) and when there is no change* in this correlation (column 2). Given numbers are the mean values of each represented variable. Values are in millions of USD except from Employees. The p-value of the differences is shown in column 3.
Increase in Correlation (1) No Change in Correlation* (2) p-value of Difference (3) Assets 4732.63 (27028.87) 5684.57 (32226.04) 0.158 Cash and Equivalents 326.30 (727.78) 354.49 (757.57) 0.159 R&D Expenses 160.29 (744.18) 194.30 (843.27) 0.237 Total Debt 1219.23 (10560.92) 1383.16 (12893.74) 0.474 Employees 7940 (28.33) 8890 (30.39) 0.168 N 5,984 2,971
*Defined as no change if the correlation between Cash Flow and Investment Opportunities stayed within a range of -0.05/+0.05 compared to the previous year.
Table 3a shows the firm’s characteristics, means and standard deviations, for firms that experienced an increase in the state credit rating in which they are incorporated and firms that have experienced no increase in this state credit rating with regard to the previous year. The p-value of difference between these two groups, resulting from a t-test, is displayed in the third column. There are again, no significant differences in firm characteristics between the means of the two groups divided by this variable. This means that an increase in the state credit rating is not caused by firm characteristics that are not in the statistical tests and might have had an unwanted impact on the results. Neither is there a significant effect of the possibility that certain firms have incorporated themselves in selected states, which might also have biased the results of the research in this paper. For the observations of only firms in a high growth state, the summary statistics are shown in Table 3b in Appendix 2.
Table 3a – Summary statistic on data from the entire database divided in observations
where there is an increase in the credit rating of the state in which the firm is incorporated (column 1) and firms in states that have no increase in credit rating (column 2). Given numbers are the mean values of each represented variable. Values are in millions of USD except from Employees. The p-value of the differences is shown in column 3.
Increase in State’s credit Rating (1) No Increase in State’s credit Rating (2) p-value of Difference (3) Assets 5351.37 (32788.21) 4385.31 (26147.89) 0.530 Cash and Equivalents 274.00 (693.80) 299.54 (693.80) 0.571 R&D Expenses 217.70 (987.43) 150.63 (766.72) 0.223 Total Debt 1119.41 (10056.90) 1128.28 (10273.34) 0.979 Employees 7170 (30) 7320 (27.22) 0.911 N 425 18,326 4 Methodology
For a better understanding of the final results and the replicability of the empirical research, in this section the methodology that is used in the statistical tests will be elaborated on. The focus of this section will be to explain by which way an answer is to be found on the main research question of this research: Do managers of firms, for which it has become easier to get internal or external funding for future investments, destroy firm value when they increase the firm’s cash holdings?
The five previously stated hypotheses are used to find an answer on the main research question. In the first section the statistical methods for the first two hypotheses will be explained. These two hypotheses are constructed to help find an answer on the sub-research question: Do managers of firms with an increasing correlation between cash flow and investment opportunities destroy firm value by increasing their cash holdings? The statistical methods that are used for the third and fourth hypotheses are explained in the second section. This second pair of hypotheses is constructed to help find an answer to the sub-research question: Do managers of firms that have experienced an increase in the credit rating of the state the firm is incorporated in destroy firm value by increasing their cash holdings? The third section of this section will provide explanation on the statistical methods that are
used to find an answer on the fifth hypotheses. This hypothesis is constructed to find an answer on the complementary research question: Are the supposed effects in hypothesis one to four stronger for firms that are currently high growth? In the fourth and final section, remaining methodology including that of the robustness checks will be discussed.
4.1 Cash Flow & Investment Opportunities
In this section the research methodology of the first two hypotheses regarding the question; Do managers of firms with an increasing correlation between cash flow and investment opportunities destroy firm value by increasing their cash holdings?, will be elaborated on. An answer on this research question will provide insight on the relationship between firm value and cash holdings regarding the correlation between cash flow and investment opportunities, which focusses on the excess to internal funding for a firm’s future investments.
Hypothesis 1; An increase in the correlation between cash flow and investment opportunities has a positive effect on firm value, implies a positive relation between this correlation opportunities and firm value. Hypothesis 2, An increase in cash holdings while there is an increase in the correlation between cash flow and investment opportunities has a negative effect on firm value, implies a negative relation between cash holdings and firms value, when this correlation is increasing. Before testing the effects that are stated in hypothesis 1 and 2, a direct effect of the correlation between cash flow and investment opportunities on cash holdings is tested. The value of the bèta-coefficient in this regression output is expected to be negative, which implies a negative effect on cash holdings if there is an increase in the correlation between cash flow and investment opportunities. Hence, a better alignment of cash flow and investment opportunities allows a firm to have lower cash holdings. This is done by using the following OLS-regression (a more detailed explanation on the execution of this regression is given in the following subsection):
𝒍𝒏(𝑪𝒂𝒔𝒉) = 𝜶 + 𝜸𝟏 ∗ 𝑹𝒊𝒔𝒆𝑪𝒐𝒓𝒓𝒆𝒍𝒂𝒕𝒊𝒐𝒏𝑫𝒖𝒎𝒎𝒚 + 𝑪𝒐𝒏𝒕𝒓𝒐𝒍𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝒔 + 𝑭𝒊𝒙𝒆𝒅𝑬𝒇𝒇𝒆𝒄𝒕𝒔 + 𝜺
To test for the effects that are stated in hypothesis 1 and 2, the following OLS-regression with an interaction variable will be used:
𝑴𝒂𝒓𝒌𝒆𝒕𝑪𝒂𝒑 = 𝜶 + 𝜷𝟏 ∗ 𝒍𝒏(𝑪𝒂𝒔𝒉) + 𝜷𝟐 ∗ 𝑹𝒊𝒔𝒆𝑪𝒐𝒓𝒓𝒆𝒍𝒂𝒕𝒊𝒐𝒏𝑫𝒖𝒎𝒎𝒚 + 𝜷𝟑 ∗ 𝑰𝒏𝒕𝒆𝒓𝒂𝒄𝒕𝒊𝒐𝒏𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝟏 + 𝑪𝒐𝒏𝒕𝒓𝒐𝒍𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝒔 + 𝑭𝒊𝒙𝒆𝒅𝑬𝒇𝒇𝒆𝒄𝒕𝒔 + 𝜺
Explanation of the variables that are used in the regressions, along with the data that is collected to construct them, is given in Table 1. Both regression will be executed twice, both times including firm- and industry fixed effects and clustered on firm-level. In one of the two instances, year fixed effects are included in the regressions as well. For the second variant a control dummy variable for crisis years is added. This control variable is added because the 2007 financial crisis is an exogenous shock that most likely affects the firm value of the tested firms. In this second case the year fixed effects are left out of the regression, because the control variable for crisis years would otherwise be one hundred percent omitted.
The bèta-coefficients in the main regression as shown above will measure the following effects: β1 measures the linear effect of cash holdings on the firm value. This effect is expected to be positive. β2 measures the effect of a growth in the correlation between cash flow and investment opportunities, compared to the year before, on the firm value. This effect is expected to be positive, because a higher correlation between these two variables is beneficial. β3 measures the additional effect of growth in cash holdings when a firm has a higher correlation between cash flow and investment opportunities against when there is no change in correlation, compared to the year before. This variable is expected to be negative, because a growth in cash should be worth less to these firms.
The value of the β2 - coefficient will provide an answer on hypothesis 1. This coefficient is expected to take on the value of H1.
H0: β2 = 0, there is no significant effect of a rise in the correlation between cash flow and
Investment opportunities on firm value.
H1: β2 > 0, a rise in the correlation between cash flow and investment opportunities has a
positive effect on firm value.
The value of the β3 - coefficient will provide an answer on hypothesis 2. This coefficient is expected to take on the value of H1.
H0: β3 = 0, there is no significant difference between the effect of growth in cash on firm value
for firms of which the correlation between cash flow and investment opportunities has risen over the past year and firms of which this correlation has stayed the same.
H1: β3 < 0, for firms of which the correlation between cash flow and investment opportunities has risen over the past year, the effect on firm value due growth in cash is smaller (maybe negative) compared to the effect on firm value of firms of which this correlation has stayed the same.
4.2 State Credit Ratings
In this section the research methodology of the third and fourth hypotheses regarding the question; Do managers of firms that have experienced an increase in the credit rating of the state the firm is incorporated in destroy firm value by increasing their cash holdings?, will be elaborated on. An answer on this research question will provide insight on the relationship between firm value and cash holdings regarding state credit ratings and therefore the excess to external funding for a firm’s future investments.
Hypothesis 3; An increase in the credit rating of the state in which a firm is incorporated has a positive effect on firm value, implies a positive relation between this credit rating and firm value. Hypothesis 4; An increase in cash holdings while there is an increase in the credit rating of the state in which a firm is incorporated has a negative effect on firm value, implies a negative relation between cash holdings and firms value, when this credit rating is increasing. Before testing the effects that are stated in hypothesis 3 and 4, a direct effect of an increase in the state’s credit rating on cash holdings is tested. The value of the bèta-coefficient in this regression output is expected to be negative, which implies a negative effect on cash holdings if the credit rating of the state in which a firm is incorporated increases. Hence, a better credit rating of the state will make external funding for future possible future investments easier, allowing firms to have lower cash holdings. This is done by using the following OLS-regression (a more detailed explanation on the execution of this regression is given in the following subsection):
𝒍𝒏(𝑪𝒂𝒔𝒉) = 𝜶 + 𝜸𝟏 ∗ 𝑰𝒏𝒄𝒓𝒆𝒂𝒔𝒆𝑰𝒏𝑺𝒕𝒂𝒕𝒆𝑪𝒓𝒆𝒅𝒊𝒕𝑹𝒂𝒕𝒊𝒏𝒈 + 𝑪𝒐𝒏𝒕𝒓𝒐𝒍𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝒔 + 𝑭𝒊𝒙𝒆𝒅𝑬𝒇𝒇𝒆𝒄𝒕𝒔 + 𝜺
To test for the effects that are stated in hypothesis 3 and 4, the following OLS-regression with an interaction variable will be used:
𝑴𝒂𝒓𝒌𝒆𝒕𝑪𝒂𝒑 = 𝜶 + 𝜷𝟏 ∗ 𝒍𝒏(𝑪𝒂𝒔𝒉) + 𝜷𝟐 ∗ 𝑰𝒏𝒄𝒓𝒆𝒂𝒔𝒆𝑰𝒏𝑺𝒕𝒂𝒕𝒆𝑪𝒓𝒆𝒅𝒊𝒕𝑹𝒂𝒕𝒊𝒏𝒈 + 𝜷𝟑 ∗ 𝑰𝒏𝒕𝒆𝒓𝒂𝒄𝒕𝒊𝒐𝒏𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝟐 + 𝑪𝒐𝒏𝒕𝒓𝒐𝒍𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝒔 + 𝑭𝒊𝒙𝒆𝒅𝑬𝒇𝒇𝒆𝒄𝒕𝒔 + 𝜺
Explanation of the variables that are used in the regressions, along with the data that is collected to construct them, is given in Table 1. Like section 4.1, both regressions will be executed twice, but with different control variables and fixed effect. Details of this are explained in section 4.1.
The bèta-coefficients in the regression as shown above will measure the following effects: β1 measures the linear effect of cash holdings on the firm value. This effect is expected to be positive. β2 measures the effect of an increase in the credit rating of the state in which a firm is incorporated, compared to the year before, on firm value. This effect is expected to be positive, because a higher state credit rating is beneficial. β3 measures the additional effect of growth in cash holdings when the credit rating of the state in which the firm is incorporated has increased, compared to the year before. This variable is expected to be negative, because a growth in cash should be worth less to these firms.
The value of the β2 - coefficient will provide an answer on hypothesis 3. This coefficient is expected to take on the value of H1.
H0: β2 = 0, there is no significant effect of a rise in the credit rating of the state in which the
firm is incorporated on firm value.
H1: β2 > 0, a rise in the credit rating of the state in which the firm is incorporated has a positive
effect on firm value.
The value of the β3 - coefficient will provide an answer on Hypothesis 4. This coefficient is expected to take on the value of H1.
H0: β3 = 0, there is no significant difference between the effect of growth in cash on firm value
for firms of which the credit rating of the state they are incorporated in has increased over the past year and firms of which this credit rating has stayed the same.
H1: β3 < 0, for firms of which the credit rating of the state in which the firm is incorporated has
risen over the past year, the effect on firm value due growth in cash is smaller (maybe negative) compared to the effect firm value of firms of which this credit rating has not increased.
4.3 High Growth Firms
In this section the research methodology of the fifth hypothesis will be elaborated on. This hypothesis is an extension on the first four hypotheses. Results on the tests for this hypothesis will provide additional insight on the main research question, especially regarding firms that are in a state of high growth. These high growth firms are expected to be more sensitive to the availability of internal or external funding then their non-high growth counterparts, therefor experiencing greater fluctuation in cash holdings and firm value when there are changes in the correlation between cash flow and investment opportunities or the state’s credit rating.
To test hypothesis 5; Effects that are measured in the research for Hypothesis 1-4 are larger for firms in a high growth stadium, the same regressions as the ones used to test the first two will be executed in exactly the same way as described in previous sections of this section. The difference will be that for the regressions to test hypothesis 5, only the firm-year observations will be included which are defined as being in a high growth state (as defined in section 3.1). As previously stated, summary statistics for the datasets as used in these regressions can be found in Appendix 2. Hypothesis 5 implies that the effects for the β2- and β3-coefficients in the main regressions are bigger (further away from zero) when tested on the dataset with only firms-years in a high growth state compared to when they are tested on the entire dataset. Therefor coefficients are expected to take on the values as suggested in the H1 cases as stated bellow.
H0: β2-High Growth = β2-whole dataset, there is no greater effect of an increase in the correlation
between cash flow and investment opportunities or a rise in the credit rating of the state in which the firm is incorporated on firm value for firms that are in a state of high growth.
H1: β2-High Growth > β2-whole dataset, there is a greater effect of an increase in the correlation
between cash flow and investment opportunities or a rise in the credit rating of the state in which the firm is incorporated on firm value for firms that are in a state of high growth.
H0: β3-High Growth = β3-whole dataset, there is no greater difference between the effect of growth in
cash on firm value for firms of which the correlation between cash flow and investment opportunities or the credit rating of the state they are incorporated in have increased over the past year and firms of which this correlation or credit rating has stayed the same.
H1: β3-High Growth < β3-whole dataset, there is a greater difference (because β3 is expected to be
between cash flow and investment opportunities or the credit rating of the state they are incorporated in have increased over the past year and firms of which this correlation or credit rating has stayed the same.
4.4 Remaining
As an additional test, regressions are executed of which the results will help to conclude if there is a relation between the external- (increase in the credit rating of the state in which a firm is incorporated in) and the internal availability (increase in correlation between cash flow and investment opportunities) of funding for possible future investments. If there is statistical evidence of a relation, the two parts of this research will be less of two different researches on their own. It is not probable that the correlation between cash flow and investment opportunities of a firm will affect the state’s credit rating in which the firm is incorporated, so this test will look for a relation in the form of the state’s credit rating affecting a firm’s correlation between cash flow and investment opportunities. For this test, the following OLS-regression is used:
𝑹𝒊𝒔𝒆𝑪𝒐𝒓𝒓𝒆𝒍𝒂𝒕𝒊𝒐𝒏𝑫𝒖𝒎𝒎𝒚 = 𝜶 + 𝜷𝟏 ∗ 𝑰𝒏𝒄𝒓𝒆𝒂𝒔𝒆𝑰𝒏𝑺𝒕𝒂𝒕𝒆𝑪𝒓𝒆𝒅𝒊𝒕𝑹𝒂𝒕𝒊𝒏𝒈
=+ 𝑪𝒐𝒏𝒕𝒓𝒐𝒍𝑽𝒂𝒓𝒊𝒂𝒃𝒍𝒆𝒔 + 𝑭𝒊𝒙𝒆𝒅𝑬𝒇𝒇𝒆𝒄𝒕𝒔 + 𝜺
A variant of this regression, but with a lagged variable of IncreaseInStateCreditRating, by one year, is also executed, to find a possible delayed effect of an increase in the state’s credit rating on a firm’s correlation between cash flow and investment opportunities. Explanation of the variables that are used in the regressions, along with the data that is collected to construct them, is given in Table 1. The regressions will be executed twice, once with firm-, industry- and year-fixed effects and once including a control dummy variable for crisis years and excluding the year-fixed effects, for reasons that are already explained. For there to be a relation between these two variables, the bèta-coefficients need to be significant and are expected to be positive, in which case a better availability of external funding will cause a firm to be better able to manage the internal availability of funding for possible future investments.
The methodology for the empirical tests that are performed as a robustness check differs very little from the main methodology that is already discussed in this section. Every regression that is executed in order to find the main results of this research is also executed
for the robustness check, but on a dataset that does not exclude financial firms. Expected is that the coefficients of interest will not show different directions or lose their significance, in which case there is stronger evidence for the proposed hypotheses and a better argument can be made when providing an answer on the research questions in this paper.
5 Results
In this section the results of the empirical research are displayed and elaborated on. For all Tables that show results counts that the first and the third column are discussed in section 5.1 or 5.2, because these results are the outcomes of the regressions on the entire dataset. The results in the second and the fourth column will be discussed in section 5.3, because these results show the outcome of the same regressions, but on the subset with only observation of firms in a high growth state.
5.1 Cash Flow & Investment Opportunities
In this section, the results of the statistical tests that are executed in order to find an answer to the sub-research question; Do managers of firms with an increasing correlation between cash flow and investment opportunities destroy firm value by increasing their cash holdings?, are displayed. These results will help to confirm or reject the first two hypotheses.
As stated in section 4.1, before testing the effects that are stated in hypothesis 1 and 2, a direct effect of the correlation between cash flow and investment opportunities on cash holdings is tested. Results of these tests can be found in Table 4a in Appendix 3. The results show that an increase in a firm’s correlation between cash flow and investment opportunities results in lower cash holdings. In the first regression by an estimated 3.5% on average and in the second regression by an estimated 3.9% on average, both being significant at the 10%-level. These results are as expected, because a better alignment of cash flow and investment opportunities allows a firm to have lower cash holdings. The negative coefficient of the control variable for crisis years can be interpreted as firms having lower cash holdings when they are in crisis years.
Table 5a shows the output of the main regressions regarding the relationship between firm value and cash holdings with respect to the correlation between cash flow and investment opportunities. The coefficients in the first row show the effect of cash holdings on
firm value and are for both regressions positive and significant. For regression one and two these coefficients estimate that on average the firm value increases by 394.92 and 582.90 per percentage increase in cash holdings respectively. The coefficient for in the fourth row for the control dummy variable of crisis years is negative and significant and estimates an average decrease of firm value of 716.20 when a firm is in a crisis year.
The coefficients in the second row show the effect of an increase in the correlation between cash flow and investment opportunities on firm value. These are the coefficients of interest for hypothesis 1. Both coefficients for the first and second regression are positive and significant at the 5%-level, estimating an average increase in firm value of 264.68 and 297.60 respectively, when the correlation between cash flow and investment opportunities have increased compared to the previous year. These results are as expected and confirm hypothesis 1; An increase in the correlation between cash flow and investment opportunities has a positive effect on firm value. The coefficients in the third row show the effect of cash holdings while firms experience an increase in the correlation between cash flow and investment opportunities. These are the coefficients of interest for hypothesis 2. Both coefficients for the first and second regression are negative and significant at the 5%-level, estimating an average decrease in firm value of 95.23 and 104.76 per percentage increase in cash holdings respectively, when the correlation between cash flow and investment opportunities have increased compared to the previous year. These results are as expected and confirm hypothesis 2: An increase in cash holdings while there is an increase in the correlation between cash flow and investment opportunities has a negative effect on firm value.
Results in this section imply that a better alignment between a firm’s cash flow and investment opportunities is beneficial for the firm value. Investors notice that firms are able to capitalize on positive NPV investment opportunities, using internal funding. Another implication from this section is that managers do destroy firm value by increasing their cash holdings when there is a better alignment between these two variables. Investors punish firms for holding on to excess cash, because excess cash holdings are costly.
Table 5a – Shows the regression output (and p-values) on the main regression regarding the
availability of internal funding for possible future investment opportunities (correlation between cash flow and investment opportunities). Includes the r-squared and amount of observations in the used dataset. Output in the first and third column are on the complete dataset. Output in the second and fourth column are on the dataset of firms in a high growth state. The significance level of the coefficients of 1%, 5% and 10% is displayed with ***, ** and *, respectively. Dependent Variable: Market Capitalization Regression 1 All Regression 1 High Growth Regression 2 All Regression 2 High Growth ln(Cash) 394.92*** (0.000) 532.88*** (0.000) 582.90*** (0.000) 889.63*** (0.000) Increase in Correlation Dummy 264.68** (0.027) 492.99 (0.132) 297.60** (0.016) 739.23** (0.040) Interaction Variable* -95.23** (0.014) -182.49* (0.052) -104.76** (0.011) -236.28** (0.026)
Control Variable Crisis Dummy
- - -716.20***
(0.000)
-606.97*** (0.001)
Firm FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Year FE Yes Yes No No
R-Squared 0.935 0.906 0.927 0.893
N 8554 2424 8554 2424
*ln(Cash) times Increase in Correlation Dummy. 5.2 State Credit Ratings
In this section, the results of the statistical tests that are executed in order to find an answer to the sub-research question; Do managers of firms that have experienced an increase in the credit rating of the state the firm is incorporated in destroy firm value by increasing their cash holdings?, are displayed. These results will help to confirm or reject the third and fourth hypotheses.
As stated in section 4.2, before testing the effects that are stated in hypothesis 3 and 4, a direct effect of an increase in the state credit rating in which a firm in incorporated on cash holdings is tested. Results of these tests can be found in Table 6a in Appendix 4. The results show that an increase in the state credit rating results in lower cash holdings. In the first regression by an estimated 2.0% on average, but not statistically significant and in the second regression by an estimated 9.8% on average, being significant at the 5%-level. These results are as expected, because a better credit rating of the state will make external funding for future possible future investments easier, allowing firms to have lower cash holdings. The
negative coefficient of the control variable for crisis years can be interpreted as firms having lower cash holdings when they are in crisis years.
Table 7a shows the output of the main regressions regarding the relationship between firm value and cash holdings with respect to the state credit rating. The coefficients in the first row show the effect of cash holdings on firm value and are for both regressions positive and significant. For regression one and two these coefficients estimate that on average the firm value increases by 225.46 and 353.51 per percentage increase in cash holdings respectively. The coefficient for in the fourth row for the control dummy variable of crisis years is negative and significant and estimates an average decrease of firm value of 644.76 when a firm is in a crisis year.
The coefficients in the second row show the effect of an increase in the state credit rating on firm value. These are the coefficients of interest for hypothesis 3. Both coefficients for the first and second regression are positive and significant at the 1%- and 10%-level respectively, estimating an average increase in firm value of 273.98 and 203.23 respectively, when the state credit rating has increased compared to the previous year. These results are as expected and confirm hypothesis 3; An increase in the credit rating of the state in which a firm is incorporated has a positive effect on firm value. The coefficients in the third row show the effect of cash holdings while firms experience an increase in state credit rating. These are the coefficients of interest for hypothesis 4. Both coefficients for the first and second regression are negative, being only significant at the 5%-level for the latter one, estimating an average decrease in firm value of 89.90 and 118.39 per percentage increase in cash holdings respectively, when the correlation between cash flow and investment opportunities have increased compared to the previous year. The sighs of these results are as expected, but for only regression 2 significant. Because of this, these results partly confirm hypothesis 2: An increase in cash holdings while there is an increase in the correlation between cash flow and investment opportunities has a negative effect on firm value.
Results in this section imply that an increase in the states credit rating is beneficial for the firm value. Investors notice that firms are better able to capitalize on positive NPV investment opportunities, using cheaper external funding. Some of the results imply that managers do destroy firm value by increasing their cash holdings when the state credit rating increased. Investors punish firms for holding on to excess cash when it has become cheaper for that firm to attract external funding, because excess cash holdings are costly.
Table 7a – Shows the regression output (and p-values) on the main regression regarding the
availability of external funding for possible future investment opportunities (credit rating increase in state in which the firm is incorporated). Includes the r-squared and amount of observations in the used dataset. Output in the first and third column are on the complete dataset. Output in the second and fourth column are on the dataset of firms in a high growth state. The significance level of the coefficients of 1%, 5% and 10% is displayed with ***, ** and *, respectively. Dependent Variable: Market Capitalization Regression 1 All Regression 1 High Growth Regression 2 All Regression 2 High Growth ln(Cash) 225.46*** (0.000) 218.37*** (0.000) 353.51*** (0.000) 404.43*** (0.000) Increase in State CR Dummy 273.98*** (0.007) 266.60* (0.073) 203.23* (0.052) 96.11 (0.539) Interaction Variable* -89.90 (0.124) -72.94 (0.519) -118.39** (0.050) -71.60 (0.546)
Control Variable Crisis Dummy
- - -644.76***
(0.000)
-498.96*** (0.000)
Firm FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
Year FE Yes Yes No No
R-Squared 0.934 0.899 0.927 0.889
N 18559 6478 18559 6478
* ln(Cash) times Increase in State Credit Rating Dummy. 5.3 High Growth Firms
In this section the results regarding the sub-research question; Are the supposed effects in hypothesis one to four stronger for firms that are currently high growth?, will be discussed. These results will help to confirm or reject hypothesis 5; Effects that are measured in the research for Hypothesis 1-4 are larger for firms in a high growth stadium.
Like in section 5.1 and 5.2, first the results that show the direct effect of an increase in a correlation between cash flow and investment opportunities and an increase in the state credit rating in which the firm is incorporated on cash holdings. These results can be found in Table 4a and Table 6a in Appendix 3. The results show that for high growth firms an increase in a firm’s correlation between cash flow and investment opportunities has a bigger (more negative) effect on cash holdings. In the first regression the effect is estimated to be 3.5% on average and in the second regression 3.9% on average. These coefficients are 3.11 and 2.49 times bigger than for the entire sample and also increase in significance. This implies that cash holdings of firms in a high growth state are more sensitive to an increase in correlation between cash flow and investment opportunities. The results also show that an increase in
the state credit rating has a bigger (more negative) effect on cash holdings. In the first regression the effect is estimated to be 16.9% on average and in the second regression 30.0% on average. These coefficients are 8.45 and 3.06 times bigger than for the entire sample and also increase in significance. This implies that cash holdings of firms in a high growth state are more sensitive to an increase in state credit rating.
The coefficients regarding hypothesis 1 can be found in Table 5a. For the first regression the coefficient for high growth firms increases to 492.99, which is 1.86 times bigger than that for all observations. However, this coefficient loses in significance, possibly due to a smaller sample size. For the second regression the coefficient for high growth firms increases to 739.23 which is 2.48 times bigger than the coefficient for all observations. These results confirm hypothesis 5 regarding the stated effects in the first hypothesis, there is a stronger effect for firms in a high growth state.
The coefficients regarding hypothesis 2 can also be found in Table 5a. For the first regression the coefficient for high growth firms increases to -182,49, which is 1.92 times bigger than that for all observations. However, this coefficient loses a bit in significance, possibly due to a smaller sample size. For the second regression the coefficient for high growth firms increases to -236,28 which is 2.26 times bigger than that for all observations. These results confirm hypothesis 5 regarding the stated effects in the second hypothesis, there is a stronger effect for firms in a high growth state.
The coefficients regarding hypothesis 3 and 4 can be found in Table 7a. For both hypotheses and both regressions the coefficients of high growth firms are smaller and with less significance than the ones for all observations. However, the signs of the coefficients stay as expected. These results reject hypothesis 5 regarding the stated effects in the third and fourth hypotheses, there is no stronger effect for firms in a high growth state. The lesser effects for high growth firms regarding these two hypotheses might be due to what is discussed in the paper by Donati (2016). He finds that for small firms, which is often related to high growth possibilities, it is hard to get external funding, which makes them more dependent on internal funding like retained earnings.
5.4 Remaining
In this section, the results of additional test to find a relation between the internal- and external availability, which form the two parts of the research in this paper, will be discussed.
The results can be found in Table 8a in Appendix 4. The results show that for both regressions and for the full dataset as well as for only high growth firms, there is a positive significant effect of the lagged increase in state credit rating on the change in the correlation between cash flow and investment opportunities. Coefficients for this lagged variable are all at least significant at the 5%-level and all exceed 0.05, which is the minimum change that is needed for it to be regarded as growth in the correlation between cash flow and investment opportunities in this research. This means that when an increase in the state’s credit rating occurs, the correlation between cash flow and investment opportunities of the firms that are incorporated in that state will grow in the year after. This can be interpreted as an increase in availability of external funding will cause a better alignment between cash flows and investment opportunities and therefor increase the availability for internal funding in the year after. This connects the two parts of this research. For both regressions, the coefficients on the database with only observations for high growth firms are at least twice as big.
5.5 Robustness Checks
The results of the robustness checks are al displayed in Appendix 5. The tests that are performed for the robustness checks are the same tests as the ones for the main results, for all observations and for high growth firm-years, but with financial firms included in the dataset.
The results of the tests to find a direct relation between cash holdings and the correlation between cash flow and investment opportunities and a direct relation between cash holdings and the state’s credit rating are displayed in Table 4b and Table 6b respectively. The results of the main tests regarding hypotheses 1 and 2 can be found in Table 5b, where the results of the main tests regarding hypotheses 3 and 4 can be found in Table 7b.
Finally, the results for the additional tests that were conducted to find a relation between the correlations between cash flow and investment opportunities and the state’s credit ratings can be are displayed in Table 8b.
For none of the results from the robustness checks are notably different from the results of the main tests. This implies that there is stronger evidence for the proposed hypotheses and a better argument can be made when providing an answer on the research questions in this paper.
6 Conclusion and Discussion
This paper tries to find an answer to the question: Do managers of firms, for which it has become easier to get internal or external funding for future investments, destroy firm value when they increase the firm’s cash holdings? This is a relevant question because of the trade-off problem that managers face when deciding the optimal amount of cash holdings for firms. Firm need to have sufficient cash holdings to be able to capitalize on possible future investment opportunities, but cash holdings are also costly. When internal- or external funding becomes better available for a firm, it is less dependent on precautionary savings for future investments, which should allow cash holdings to decrease.
To find an answer on the main research question, empirical research with the use of OLS-regressions with interaction variables is performed. The correlation between cash flow and investment opportunities will be used to measure the availability of internal funding for a firm’s future investments. The credit rating of the state in which the firm is incorporated is used as a measure for availability of external funding for future investments. Results of this research show that both the increase in correlation between cash flow and investment opportunities as an increase in state credit rating have a positive effect on firm value. More importantly, the results show that when these measures for internal- and external availability for funding increase, an increase in cash holdings has a negative effect on firm value. The results also indicate that the impact of availability of internal funding is stronger for high growth firms, whereas there is no significant difference in the impact of the availability of external funding. These results imply that it could not be optimal for a firm to increase its cash holdings when the availability of internal- or external funding has increased. When this is the case, firms can allow themselves to hold on to lower amounts of cash and still be able to capitalize on possible future investment opportunities. Because cash holdings are costly, increasing them in these scenarios will be bad for the firm’s value. Due to the direct effect on shareholder wealth and firm profitability, the implications of these results on the optimal amount of cash holdings can have a large effect on the entire economy.
Future research on this topic could go further in depth to find the impact of cash holdings on firm value, regarding the availability of internal- or external funding for future investments. Expanding the observations of state credit ratings would allow for the use of a bigger dataset overall, which would improve the research. Also, entirely different measures for availability of internal- or external funding could be used to tested if the results of this
research could be replicated. For example, changes in the interest rates could be used as a measure for availability of external funding. Including a dummy variable for high growth firms and multiple interaction variables to the regressions used in this research could make for a stronger statistical comparison between effects of high and non-high growth firms. Lastly, even though R&D expenditures are used as a measure for investment opportunities instead of Tobin’s Q, the regressions in this research still might suffer from endogeneity issues. For example, a form of reversed causality between the correlation between cash flow and investment opportunities and firm value. Due a limited time and recourses for this thesis, these recommendations could not be implemented in this research.
7 References
Acharya, V. V., Almeida, H., & Campello, M. (2007). Is cash negative debt? A hedging
perspective on corporate financial policies. Journal of Financial Intermediation, 16(4), 515-554.
Almeida, H., Campello, M., & Weisbach, M. S. (2004). The cash flow sensitivity of cash. The Journal of Finance, 59(4), 1777-1804.
Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009). Why do US firms hold so much more cash than they used to? The journal of finance, 64(5), 1985-2021.
Becker‐Blease, J. R., & Paul, D. L. (2006). Stock liquidity and investment opportunities: Evidence from index additions. Financial Management, 35(3), 35-51.
Campello, M., Graham, J. R., & Harvey, C. R. (2010). The real effects of financial constraints: Evidence from a financial crisis. Journal of financial Economics, 97(3), 470-487.
Carpenter, R. E., & Guariglia, A. (2008). Cash flow, investment, and investment opportunities: New tests using UK panel data. Journal of Banking & Finance, 32(9), 1894-1906. Donati, C. (2016). Firm growth and liquidity constraints: evidence from the manufacturing and
service sectors in Italy. Applied Economics, 48(20), 1881-1892.
Duchin, R. (2010). Cash holdings and corporate diversification. The Journal of Finance, 65(3), 955-992.
Foley, C. F., Hartzell, J. C., Titman, S., & Twite, G. (2007). Why do firms hold so much cash? A tax-based explanation. Journal of Financial Economics, 86(3), 579-607.
Han, S., & Qiu, J. (2007). Corporate precautionary cash holdings. Journal of Corporate Finance, 13(1), 43-57.
Harford, J. (1999). Corporate cash reserves and acquisitions. The Journal of Finance, 54(6), 1969-1997.
Lee, N. (2014). What holds back high-growth firms? Evidence from UK SMEs. Small Business Economics, 43(1), 183-195.
Lin, H. C., & Chiu, S. C. (2017). Tradeoff on corporate cash holdings: a theoretical and empirical analysis. Review of Quantitative Finance and Accounting, 49(3), 727-763. Lyandres, E., & Zhdanov, A. (2013). Investment opportunities and bankruptcy
prediction. Journal of Financial Markets, 16(3), 439-476.
Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of financial economics, 13(2), 187-221.
Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of financial economics, 52(1), 3-46.
Rauh, J. D. (2006). Investment and financing constraints: Evidence from the funding of corporate pension plans. The Journal of Finance, 61(1), 33-71.
Tuzel, S., & Zhang, M. B. (2017). Local risk, local factors, and asset prices. The Journal of Finance, 72(1), 325-370.
8 Appendices
Appendix 1 31
- An overview of increases in state credit ratings per year
Appendix 2 32 - Table 2b - Table 3b Appendix 3 33 - Table 4a - Table 6a Appendix 4 34 - Table 8a Appendix 5 35 Robustness Checks - Table 4b - Table 5b - Table 6b - Table 7b - Table 8b
Appendix 1
Overview of increases in state credit ratings per year – In states that
are not included in the table, no increase in that state’s credit rating has occurred in the period 2005-2017.
State Years in which an Increase in Credit Rating Occurred
Alaska 2008, 2012 Arizona 2015 California 2013, 2015 Colorado 2007 Florida 2005 Hawaii 2007, 2016 Idaho 2011 Indiana 2006, 2008 Iowa 2008 Kentucky 2015 Louisiana 2008, 2009, 2011 Maine 2007 Massachusetts 2005, 2012 Montana 2008 Nebraska 2011 Nevada 2006 New Jersey 2005 New York 2015 North Dakota 2005, 2009, 2013 Oklahoma 2008 Oregon 2007, 2011 Rhode Island 2005 South Dakota 2011, 2015 Tennessee 2006, 2017 Texas 2009, 2013 Washington 2007 West Virginia 2009 Wisconsin 2008 Wyoming 2008, 2011
