• No results found

The impact of cash management on corporate credit ratings

N/A
N/A
Protected

Academic year: 2021

Share "The impact of cash management on corporate credit ratings"

Copied!
39
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The Impact of Cash Management on Corporate Credit Ratings

Adéla Ščepková Master Thesis University of Amsterdam MSc Business Economics-Finance track

Supervisor: Vladimir Vladimirov July 2016

Abstract

This study analyses the impact of cash management on firm credit ratings elaborating on how capital structure affects firm’s creditworthiness. Several articles have addressed the impact of working capital management and capital structure on a firm’s performance. However, based on my research and to my knowledge no study looking into the impact of cash management on firm’s credit ratings has been widely published. I believe this thesis can contribute to the existing literature by adding a different perspective to the current study. The results of this thesis are valuable to financial managers, treasurers, analysts and investors as they provide insight into additional factors in the balance sheet that impact the credit rating assessments. The findings suggest that a higher leverage ratio negatively impacts the credit rating and is positively affected by an increasing cash conversion cycle and higher liquidity. Excessive amount of leverage decreases the ability a firm can repay its debt obligations. Liquidity, on the other hand, provides the firm with a buffer in case of economic downturn. Lastly, the positive relationship between cash conversion cycle and credit ratings implies that higher rated firms that have additional cash or other sources of funding are not willing to shorten their cycle as it takes a lot of effort and may disrupt their customers. In addition, the supply of trade credit triggers the slowdown of the cash conversion cycle. The models used for testing the hypothesis proved valid through several robustness checks and tests to control for unobserved heterogeneity and reverse causality. Hereby, the results suggest that cash managers and treasurers should take their firms current credit rating into account when setting the strategy for cash management and capital structure changes.

JEL Classification: G02, G11, G32

(2)

Statement of Originality

This document is written by Adéla Ščepková 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 mentioned in the text and its references have been used in creating it.

The Amsterdam Business School is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Section 1 Introduction

The objectives of cash management are to manage all activities in the most economical and efficient manner, maximize revenues, reduce costs and optimize sources and uses of cash. Sufficient amount of liquidity gives a firm the possibility to pursue growth opportunities and acquire new customers (Reider, 2005, pp. 7-12). In many industries accessing debt and equity funding can be quite challenging. Firms often concentrate on accessing their financing externally, disregarding the large component of a firm’s capital that is locked up in the operational processes. To be able to unlock this additional source of funding, firms need to optimize their cash management strategy. By doing this, the firm improves its overall balance sheet structure and financial stability by being less dependent on external funding and providing greater returns to the shareholders. Furthermore, by efficiently managing its cash, the firm is able to exercise greater control over its cash flows and manage its operations smoothly (Reider, 2005, p.2).

Credit ratings are important for managers when making capital structure decisions, due to the discrete benefits and costs connected to various rating levels (Kisgen, 2006, pp.1035-1037). Higher-rated firms can issue debt at lower yield and vice versa. Overall, firms’ credit analysis provides a benchmark of the creditworthiness of a firm that is then useful in both financial and commercial purposes (Fight, 2006, p.41). Credit ratings help firms to compete, achieve enhanced transparency and name recognition. In addition, credit ratings facilitate the access to the domestic and foreign capital markets and thereby reduce firms’ reliance on banks (Judge & Korzhenitskaya, 2012, p.29). Thus, by obtaining higher ratings, firms can gain international visibility, increased stability and become less reliant on the external financing. Credit rating’s methodology takes into account many factors, quantitative as well as qualitative. The main objective of this study is to perform an empirical assessment of the impact of balance sheet management by means of cash management on firms’ credit ratings. Although, credit ratings are often crucial for firm’s financial health, so far not many academic papers have addressed this area (Maung & Mehrotra, 2010). Despite a number of studies concentrating on the importance of the cash management, to my knowledge, there is no other research focusing on the effect of cash management strategies on firms’ credit ratings.

Following previously conducted studies, this paper employs a quantitative approach. More specifically, the ordinal logit regression is utilized to perform empirical analysis on the ordered dependent variable, the credit ratings. Two regression models test the hypotheses. The analysis of this paper yields the following key results: i) the cash conversion cycle (CCC) and liquidity are positively and significantly related to the firms’ credit ratings and ii) the leverage ratio has a significant negative impact on the firm’s creditworthiness. The results suggest that treasurers and cash managers should consider credit ratings when deciding upon new cash management strategies. Despite the benefits that higher levels of leveraging offer, increased levels of debt expose the firm to higher risk of not being able to meet its debt obligations. Secondly, liquidity has been found to have positive effects on credit ratings as it offers the firm a buffer in case of economic downturn. Lastly, in

(4)

the case of the CCC the results reveal opposite to what was initially expected. Previously conducted studies show that by lowering their CCC firms can access additional cash for future investments and hence pursue more growth opportunities. However, the availability of trade credit triggers slowdown of the CCC. Firms provide trade credit to its suppliers in order to access the market and provide competitive terms. Shortening the CCC can be a difficult task and many firms lack the incentive to improve it, not realizing the overall impact on credit ratings and ultimately financing costs and decreased shareholders’ returns (Kling et al., 2014, p.128). The thesis further addresses the endogeneity issues arising from the two-way causal relationship between the credit ratings and the CCC. Even after controlling for the endogeneity and unobserved heterogeneity the models yield same results implying validity of the analysis.

This study offers a valuable contribution to the theory as well as to the practice. The findings of this thesis are of interest to financial managers, treasurers and investors as they give a clear picture on what kind of strategies affect firms’ perceived creditworthiness. Lastly, this thesis will form a base for further research by evaluating gaps in the existing literature.

The content of this paper is as follows: Section 2 reviews the theoretical background by expanding on previous academic studies. Section 3 forms the hypotheses which are based on the related literature and findings. The same section includes the formulation of the empirical models further used in order to test the relationship between cash management and the firms’ credit ratings. Description of the sample, the presentation of the data sources and descriptive statistics are in Section 4. Sections 5 and 6 show the results of the regression analysis further elaborating on the robustness of these results. Section 7 concludes, stating limitation of the study and offering suggestions for further research.

Section 2 Literature Review

2.1 Cash Management

An extended global economic slowdown has resulted in difficult business conditions and increased emphasis by executives to optimize their operational processes. Efficient cash management does not only help a firm to improve its free cash flow, but also to create a stable competitive position in the market. Effective management strives to maximize firms’ cash generation and make sure that a firm has cash when needed (Reider, 2005, p.11). One of the main objectives of cash management is the collection and disbursement of cash. It is important for both new and mature businesses since it concentrates on areas such as level of liquidity and management of cash balances (Davidson, 1992; in Mauchi et al, 2011, p.1300). Effective cash management is crucial for ensuring that companies’ finances are in a good position. This affects the functioning of a business as a whole. Having enough cash, within a reasonable limit, gives a firm access to a wide range of possibilities such as developing new products and concentrating on growth opportunities. A positive cash buffer not only allows the company to take advantage of opportunities but also serves as a safety guard in a heightened risk

(5)

environment. Not having a sufficient amount of liquidity restricts the company from growing and increases the chance of the firm going bankrupt (Reider, 2005, pp. 7-12).

One of the crucial parts of the firm’s cash management is the cash flow forecasting. Using the profit planning techniques, such as developing an operating plan for the near future, is important in order to control the financial aspects of the functioning of a business. This offers many advantages to the firm, for example the possibility of evaluation, anticipating the need for additional resources and the increase in confidence of investors (Wielen & Alphen, 2006, pp.211-215). The balance sheet and income statements are of high importance, however, they show only a part of the picture. It is important to note that even a firm with very stable financial statements can find itself in a difficult situation due to the lack of liquidity. The proper management of cash requires the right timing. The suppliers, creditors and employees are expected to be paid on time and the business needs to make sure it has enough cash to meets its obligations when due (Scarborough, 2008, p.234). Therefore, a firm needs to gain control over its cash inflows and outflows. In this way, firms are able to avoid holding unnecessary levels of liquidity and are able to increase the amount of interest earned on the invested cash (Scarborough, 2008, p. 235).

Managing liquidity is another aspect of the cash management that makes sure a firm has enough liquidity to meet its payments. Enough cash and near cash assets offer a firm the possibility to earn additional revenue by knowing where to invest and to know, which accounts to use in order to earn highest possible interests. Therefore, through this channel a firm can boost its cash flows (Kim, Mauer & Sherman, 1998, pp.335-336). 2.1.1 Working Capital Management

The objective of the working capital management (WCM) is to make sure that a firm is able to pay all of its short-term obligations when due. Mismanagement may adversely affect the profitability of a firm and in severe cases might lead to liquidity crisis (Ukaegbu, 2013, p.2). By efficiently managing their working capital, firms can become less dependent on the external financing and thus boosting their financial flexibility. The importance of WCM has been a topic of many papers in finance literature, investigated from different perspectives. Both studies of Howorth and Westhead (2003) and Deloof (2003) concentrate on optimal levels of working capital in order to maximize firm value. Deloof (2003) investigates the relation between WCM and corporate profitability. He bases his sample on 1,009 large Belgian non-financial firms over a period of 1992-1996. He uses the CCC as a proxy for WCM. The findings suggest that managers can increase firms’ profitability and create value for their shareholders by decreasing the receivable and inventory periods to a reasonable level. In line with Deloof (2003), Ukaegbu (2013) also investigates the relationship between profitability and WCM. In addition, he accounts for differences across countries with differential industrial levels among developing economies in Africa. The results reveal that there is a strong negative relationship between net operating profit, as a proxy for firm’s profitability, and the CCC. The study also proves a positive relationship between profitability and a firm’s size, implying that larger firms are more likely to have

(6)

higher-skilled employees, enjoy economies of scale and diversification. Shin and Soenen (1998) support these claims and investigate the relationship between WCM and the value creation for shareholders. They use net trade cycle as an alternative measure of WCM. By performing correlation and regression analysis, their results show a significant impact of working capital on both the liquidity as well as profitability of a firm. They conclude that a shorter net trade cycle coincides with higher-risk adjusted stock returns.

The traditional view on the relationship between CCC and the firm’s profitability is that by decreasing the cycle a firm can perform better. In contrast, Nobanee (2009) provides evidence that decreasing cash conversion does not necessarily lead to solely positive effects. He claims that by reducing its inventory levels a firm can find itself facing inventory shortages that result in lost sales and by shortening its receivable period a firm could lose its better paying credit customers. Lastly, a firm runs the risk of reputational damage by prolonging its payable period (p.1). Nobanee (2009) suggests a different approach to test the effect of WCM. He uses the optimal inventory, receivable and payable levels to investigate the effect on a firm’s profitability. The author concludes that optimal CCC is a more accurate measure of working capital that maximizes sales, profitability and market value of firms.

2.1.2 Cash Management and Capital Structure

There are two main theories that reflect how firms determine their capital structure: the trade-off theory and the pecking order theory. The idea of the trade-off theory is that managers consider both costs and benefit when deciding on the mix of their debt and equity financing. In the trade-off framework, a firm sets its target debt-to-value ratio and gradually moves towards the target. On the other hand, the pecking order theory suggests that a firm has a preference to use internal financing over debt but prefers debt rather than issuing equity. In the pure form of this theory, the firm does not have a target debt-to-value ratio (Myers, 1984, p.576). Donaldson (1961) claims that due to the pecking order firms tend to accumulate retained earnings. As such, they become more levered when their profitability is low and vice versa. In contrast, Hovakimian, Opler and Titman (2001) find that although firms take into account pecking order theory when making capital structure decision in the short-run, in the long-term they tend to move towards a structure consistent with the trade-off theory framework. The decided amount of a firm’s mix between debt and equity financing depends on many factors and varies across entities, industries and credit cycles. In general, higher profitability brings lower expected costs of financial distress and therefore, this type of firm values the interest tax shields more. Furthermore, previously conducted research suggests that more mature firms have, on average, greater access to debt and are more attractive to investors due to factors such as higher diversification, lower default risk and better reputation (Hovakimian et al, 2001; Frank & Goyal, 2009, p.7). Frank and Goyal (2009) investigate a sample of publicly traded United States (US) companies between 1950 and 2003 in order to determine, which factors have a reliable association with the market-based leverage. They find a positive relation between tangible assets, size,

(7)

inflation and the amount of a firm’s leverage. On the other hand, leverage is negatively affected by the market-to-book ratio and the amount of profits.

Faulkender and Petersen (2005) also examine how firms choose their capital structure. More specifically, they empirically investigate whether firms with the access to the public bond markets have significantly more debt. They indeed conclude that less constrained firms have higher leverage ratios by more than 50% than firms without an access. Consistent to the findings of Hovakimian et al. (2001), they reach a conclusion that firms that are more profitable have lower levels of debt.

Eljelly (2004) empirically assesses the relationship between profitability and liquidity-measured by the current ratio and CCC. He investigates a sample of 29 joint stock companies in the Saudi Arabia between the period of 1996 and 2000. His results suggest that there is a negative relationship between profitability and the liquidity. By having excessive amounts of liquidity, the firms experience unnecessary costs and losses. Eljelly (2004) suggests that active liquidity management could minimize these negative effects. The findings are in line with the paper of Acharya, Davydenko and Strebulaev (2012). The authors claim that the fact that common intuition suggests more liquid firms are safer, is a naïve prediction. Furthermore, they show that larger cash holdings are associated with higher levels of credit risk. More specifically, although firms with higher cash reserves are less likely to default in the near future, cash holdings might have a positive relationship with the default probabilities in the long-term. Compared to Eljelly (2004) and Acharya et al. (2012), Reider (2005) finds dissimilar results. He claims that liquidity represents the possibility to make necessary payments when needed and to ensure the continuity of the business operations (p.15). Also, low levels of liquidity may result in increased financial costs affecting the firm’s ability to meet its obligations (Maness & Zietlow, 2005, p.25). Chukwunweike (2014) in his study concludes that there is a significant positive correlation between the current ratio and the profitability. However, he states the amount of liquidity should not be too low nor too excessive. Not having enough liquidity has adverse effects on the firm’s creditworthiness. Overly high levels of liquidity, on the other hand, indicate accumulated idle cash, on which a firm is potentially losing any additional profits through lack of further investment into the business (Chukwunweike, 2014, p.82)

2.2 Credit Ratings

Credit ratings have a role of evaluating a firm’s credit risk and indicating default probabilities of the issuers of various debt securities. There are in general two types of credit ratings, the short-term and long-term. Whereas the short-term ratings respond to the market fluctuations that occur only over a short period of time, the latter concentrate on long-term projections and tend to remain stable over a business cycle. The credit ratings agencies (CRA) take into account both quantitative and qualitative information, such as management expertise or future development and planning. The credit ratings assigned by the agencies provide valuable insight to the investors, thus reducing the investors costs of information (Dittrich, 2007, p.9).

(8)

2.2.1 Importance of Credit Ratings

Credit ratings are considered important by many market participants such as legislators, issuers, investors and regulators (Becker & Milbourn, 2010, p.494). There are several reasons why they are relevant. The assignment of credit ratings is essential for the proper functioning of the financial system (Becker & Milbourn, 2010, p.494). It enables a firm’s access to raise funding in capital markets and serves as a screening device for investors to lower disclosure requirements. Credit ratings also serve as a monitoring device and therefore decrease the chance of moral hazard. They are used as a signal of stability, trustworthiness and projections for the future. Investors and financial intermediaries use them as an indicator of risk and ability to repay debt obligations. Therefore, the higher the ratings the more attractive the firm becomes to the investors (Becker & Milboun, 2010, p.499). Firms have high incentives to obtain higher rating because it directly affects the access and the cost of capital and the ability to make investments. According to Kisgen (2007), the level of firms’ ratings directly affects certain regulations on bond investment. For example, speculative-grade level securities are often subject to strict capital requirements and regulations in many countries restrict the investment of capital in these types of securities.

2.2.2 Credit rating and Capital Structure

Kisgen (2006) was the first to provide empirical evidence to support the claim that credit ratings directly affect capital structure decision making. Due to the discrete costs (benefits) credit ratings are being considered when making capital structure decisions. In line with these findings, Graham and Harvey (2001), in their survey, conclude that credit ratings are the second most important concern of Chief Financial Officers (CFO) and managers when determining the capital structure of a firm. In addition, their conclusions suggest that when issuing equity, managers are often concerned with stock price appreciations and dilution of the earnings per share.

The ratings serve as a signal of quality. Due to their signaling power, each rating category links different conditions such as the cost of capital that further influence a firm’s level of debt. Kisgen (2006), using empirical tests, examines whether credit rating directly affects the capital structure decision of a firm. He reaches the conclusion that managers are indeed concerned with rating-triggered costs and view rating levels as signals of a firm’s quality. The credit rating drop from investment grade to speculative is the most significant since it affects the access to capital and commercial paper markets and the bond liquidity issues are the most severe. Hovakimian et al. (2009) further contribute to this topic. They analyze how firms target their credit ratings and how they influence corporate decisions in order to achieve their desired ratings. The focus of the paper is on larger transactions in contrast to Kisgen (2006) who studies smaller debt and equity issues and repurchases. Hovakimian et al. (2001) reach the conclusion that firms with higher market-to-book ratios and higher selling expenses often receive better ratings. Their findings, however, are not aligned with the trade-off theory assumption regarding the positive relationship between leverage and tangible assets. More specifically,

(9)

they claim that firms of larger size that have more tangible assets than their peers obtain better ratings. Furthermore, they provide evidence that firms form their corporate finance choices to offset market shocks and hence deviate from their targets. They conclude that the deviations from firms’ targeted leverage ratio and credit ratings tend to be asymmetric. In particular, upgraded firms will not adjust their strategies as much compared to the downgraded firms, who tend to adjust their choices accordingly. Overall, they show that managers have a personal preference for better ratings.

Kisgen (2006) investigates the relation between credit rating changes and the capital structure behavior given Credit Rating-Capital Structure (CR-CS) hypothesis. The fundamental idea behind this theory is that the credit ratings directly affect the capital structure decision making. In his paper Kisgen (2006) mentions several implications of the CR-CS theory. Downgraded firms are more likely to reduce their leverage in the subsequent period than firms that have not experienced the rating downgrade. Secondly, even if controlling for changes in leverage and other firm characteristics there is no significant relationship between upgrade and capital structure decisions. Downgrades that result in higher discrete costs increase the likelihood of leverage-reducing behavior. Lastly, if downgraded, firms will undertake more significant adjustments in order to reach their target leverage levels (p.1037). Another study conducted by Kisgen extends the earlier research on several levels. Kisgen (2009) concentrates mainly on the capital structure behavior after rating changes occur and tests whether ex-post managerial behavior is consistent with the ex-ante behavior. He finds that firms that have been downgraded issue approximately 1.5%-2.0% less debt as a percentage of total assets compared to the other firms. Rating upgrades, on the other hand, are not significantly associated with capital structure decisions. Although confirming that firms tend to adjust their capital structure following rating changes, Maung & Mehrotra (2010) find an insignificant impact on the utility industry. Their main conclusion indicates that credit ratings might not always reflect firms’ underlying characteristics.

2.3 Cash Management and Credit Ratings

In order to achieve their desired credit ratings, managers adjust their strategies in terms of financing, hedging and investment activities. The managers consider capital structure and change it accordingly. As Graham and Harvey (2001) claim, managers and CFOs make changes in their capital structure depending on their rating targets. Additionally, they claim that based on traditional capital structure theories, credit ratings play more important roles than other factors.

According to the CR-CS theory mentioned earlier, credit ratings are of high interest due to the discrete costs associated with lower levels of rating. For instance, long-term credit ratings affect access to the commercial paper market and the opportunity to enter into a long-term supply contract. Furthermore, low-rated firms can incur additional costs due to ratings-triggered events such as required repurchase of bonds. The most severe difference in market access is at the threshold between the investment and speculative or otherwise known as

(10)

junk grade levels. During economic downturns and periods of risk-aversion, at certain rating levels, firms might be restricted from raising debt capital (Kisgen, 2006, pp.1037-1040).

The decision as to which investment and financing strategies to implement reveals how capable a firm’s management is and certainly affects the probability of default. I hypothesize, that cash management indicators impact credit ratings. The cash management mechanisms not only strive to implement the most efficient processes to boost a firm’s cash flows but also serve as a tool to monitor firm’s daily operations.

Section 3 Hypotheses and Methodology

3.1 Hypotheses

Based on the previously discussed literature, the following hypotheses are addressed.

Many academic papers concentrate on the topic of working capital management and more specifically how it affects the firm’s profitability. By efficiently managing their working capital, firms will be able to meet their operating expenses and being able to pay off their short-term obligations when due. Additionally, during unexpected market changes those firms that have their working capital under control are able to react quicker (Ukaegbu, 2013, p.2). The CCC is to be used in the analysis as a proxy for the efficiency of WCM. A longer cycle may lead to higher sales and thus having positive effect on the firm. On the other hand, the firm’s profitability can also decrease together with the CCC due to the costs of the investment in working capital, which rise at a more rapid rate than the benefits of having more inventory and granting additional trade credit to customers are realised (Deloof, 2003, pp.574-575). The availability of additional trade credit to customers triggers a slowdown of the CCC (Kling, Paul & Gonis, 2014, p.130). Firms may provide trade credit to their suppliers in order to maintain the sales, remain competitive and to avoid losing their customers. In addition, shortening the CCC takes effort and high profitable firms lack the incentive to improve it (Kling et al., 2014, p.128). However, for the clear benefits of the shorter cycle, the assumption in this paper is that with a decrease in CCC, more cash is accessible to the firm for future investments in the business (Padachi, 2006). This leads to the formulation of the first hypothesis:

Hypothesis 1: CCC has a negative effect on credit ratings.

In general, firms with higher cash holdings in their portfolios are considered safer. A firm may use the excess cash, or the buffer, to respond to future cash flow shortfalls. Acharya et al. (2012) show in their research, however, that the opposite is true. They claim that firm with high levels of cash is likely to find itself in danger of distress and therefore, large cash holdings can be associated with increased levels of credit risk. There is a trade-off between investing its cash long-term and hence receiving higher cash flows in the future, and using it as a buffer in case of unfavorable conditions lowering its probability of default (Acharya et al, 2012, p. 3573). In contrast to Acharya et al.(2012), Odders-White and Ready (2006) take an opposing stance. Their findings

(11)

suggest that more liquid firms find it easier to issue new equity in order to avoid financial distress and thus lowering the riskiness of their outstanding debt (p.126). This study’s initial expectation, as common intuition may suggest, is that more liquid firms will have higher credit ratings due to their ability to settle obligations and thus reflecting lower probability of default. Therefore, the second hypothesis states:

Hypothesis 2: Liquidity has a positive impact on the credit ratings.

Higher leverage is closely connected to the increased risk given that high-levered firms have, in general, more difficulty in settling debt obligations. In the trade-off model, however, firms decide on their optimal level of leverage by taking into account benefits and costs of an additional dollar of debt (Fama & French, 2001, p.1). By increasing leverage, firms can receive additional tax benefit that directly increases firms value (Carlson & Lazrak, 2010, p.2333). On the other hand, the costs represent the agency conflicts that may arise between stockholders and bondholders and potential bankruptcy costs. Furthermore, at the optimal leverage levels the costs are offset by the benefits (Fama & French, 2001, p.1). Danis, Rettl & Whited (2014) conclude that firms that are close to their optimal level of leverage experience positive effect of leverage on their profitability. In contrast, Fama and French (2002) and Graham and Harvey (2001) both find a negative relationship between leverage and profitability. It can be expected, that high leverage will lead to poor credit ratings that negatively impact firms’ reputational standing and investor perception of the firm. The last hypothesis is formed:

Hypothesis 3: Credit ratings are negatively affected by leverage.

3.2 Methodology

To test the relationship between the credit ratings and the cash management and more specifically the above-mentioned hypotheses the study uses a quantitative approach. The Standard and Poor’s (S&P) long-term issuer credit ratings represent the dependent variable. In line with the previous research, alphabetical ratings are converted into seven numerical rating categories (see Table A1 in the appendix). To empirically test the hypotheses, similarly as in Ashbaugh-Skaife et al. (2006), the ordered logit model is used. This method is the most appropriate since the dependent variable conveys ordinal risk assessments. When testing for the impact of cash management on credit ratings, one has to be careful that there is only a one-way relationship. To make sure that change in capital structure and cash management does not affect the dependent variable, the lagged values of independent and control variables are used. This is to make sure that changes in cash management arise only after changes in credit ratings. Other studies use this approach to control for the effect of endogeneity The testing of the hypotheses is done using two ordered logit models. Model 1 examines the impact of the working capital management and leverage on firms’ credit ratings. Whereas the second one investigates the effect of liquidity.

(12)

The regression equations state: Model 1

a) 𝐶𝑅𝑡= 𝛼 + 𝛽1 𝐶𝐶𝐶𝑡−1+ 𝛽2𝐿𝐸𝑉𝑡−1+ 𝛽4𝐶𝑎𝑠ℎ_ℎ𝑜𝑙𝑑𝑡−1+ 𝛽5𝐼𝑛𝑡𝑐𝑜𝑣𝑡−1+ 𝛽6𝐶𝑎𝑝𝑖𝑛𝑡−1+𝛽7𝑆𝐼𝑍𝐸𝑡−1+

𝛽8𝑅𝑂𝐸𝑡−1+ 𝛽9𝐵𝑒𝑡𝑎𝑡−1+ 𝜀𝑖

The dependent variable, CR, stands for the S&P long-term issuer credit rating categories. The first independent variable the CCC is important for every business as it shows the time lag of cash moving in and out of the company. The cycle comprises all steps needed for purchasing, production and sales. It starts when money paid for an invoice for a certain good reach the supplier, the cash outflow, and ends with a cash inflow from the supplier for selling the good or service. In short, it is viewed as the amount of days needed to transform a firm’s operating activities into cash returns (Wielen & Alphen, 2006, p.244). Generally, the longer the cycle, the larger the amount of funds trapped in the working capital. By keeping its CCC as short as possible, firms can gain access to the capital needed for investment (Padachi, 2006, p.51). To obtain the CCC the following formula is used:

Equation 1

CCC = Operating Cycle - Days Account Payable

where Operating Cycle = Days Account Receivable + Days Account Inventory

The operating cycle, receivables and inventory, represent the cash inflows. The last item of the equation, on the other hand, stands for the cash outflows. The objective of decreasing the CCC can be achieved by collecting accounts receivable in the shortest amount of time without negatively impacting future sales. Firms may offer use of cash discounts to their suppliers to achieve this objective. Secondly, firms should minimize inventory-on-hand days as far as possible. The optimal situation is receive inventory as it needs to be delivered. This is the so-called “just-in-time approach” that is idealistic but challenging to achieve. Many cash managers do not manage the inventory properly although it represents a large part of firm’s investments. Too much inventory can bring a firm into a danger of running out of cash resources. Additionally, costs of carrying inventory are expensive (Scarborough, 2008, p.258). Lastly, the firm should prolong the accounts payable in order to obtain the shortest CCC. However, there is a trade-off between paying as late as possible and in the meantime earning additional interest, and making sure to pay early enough in order not to damage the firm’s reputation,

trustworthiness and in general, the credit ratings. Paying late could cause suppliers to demand prepayments, which would have a negative effect on company’s cash flows (Scarborough, 2008, p.255). As stated by Kling et al. (2014) for better performing firms that have additional cash or external forms of funding the incentive to decrease their CCC is lower (p.129). To see how the WCM affects the credit ratings, the study divides CCC into separate actions to determine, which part has the strongest influence. This paper concentrates on the CCC as a whole, but it will put more emphasis on days account payable given the fact that this is the part, where cash

(13)

managers and treasurers can make the most impact. Therefore, the first model divides the CCC into two components as in Equation 1 above. The regression is hence:

b) 𝐶𝑅𝑡= 𝛼 + 𝛽1 𝐷𝑃𝑂𝑡−1+ 𝛽2 𝑂𝑝_𝑐𝑦𝑐𝑙𝑒𝑡−1+ 𝛽3𝐿𝐸𝑉𝑡−1+ 𝛽4𝐶𝑎𝑠ℎ_ℎ𝑜𝑙𝑑𝑡−1+ 𝛽5𝐼𝑛𝑡𝑐𝑜𝑣𝑡−1+

𝛽6𝐶𝑎𝑝𝑖𝑛𝑡−1+𝛽7𝑆𝐼𝑍𝐸𝑡−1+ 𝛽8𝑅𝑂𝐸𝑡−1+ 𝛽9𝐵𝑒𝑡𝑎𝑡−1+ 𝜀𝑖

The following independent variable, LEV, stand for the debt-to-equity ratio that is used in order to quantify the effect of leverage. There are perceived costs and benefits of having higher level of debt. Leverage represents a possibility of obtaining additional tax benefits. However, high levels bring riskiness into the business in terms of not being able to pay off its debt and thus increasing its probability of defaulting. The last explanatory variable, Cash_hold, represents the firm’s cash holdings obtained by dividing the cash and short-term investments by the total assets. This ratio tells us how much cash a firm is holding as a percentage of its total assets. Khieu and Pyles (2012) find that downgraded firms tend to increase their cash holdings. However, they do not find any significant relationship with the credit rating increases. The most visible change is if a firm experiences a downgrade from investment to speculative grade. These types of firms hoard the most cash. Model 2

𝐶𝑅𝑡 = 𝛼 + 𝛽1 𝐿𝐼𝑄𝑡−1+ 𝛽2 𝑆𝐼𝑍𝐸𝑡−1+ 𝛽3𝑅𝑂𝐸𝑡−1+ 𝛽4𝐵𝐵𝑒𝑡𝑎𝑡−1+ 𝜀𝑖

Model 2 examines the second hypothesis by concentrating on the level of liquidity a firm holds on its balance sheet. To measure the liquidity one can look at the proportion of current assets with respect to the current liabilities. Calculating one of the cash flow ratios reveals how quickly a firm can repay its debt obligation. The independent variable, LIQ, is measured by the quick ratio, which represents the ability of a firm to fulfill its obligations with the most liquid current assets and thus excludes inventories from the equation.

Quick Ratio = Current Assets - Inventories Current Liabilities Control Variables

Additional firm-specific variables that in prior studies proved to be significant are added to the models. Int_cov is the interest coverage ratio which is dividing operating income before depreciation by interest expense (Ashbaugh-Skaife, 2006; Maung & Mehrotra, 2010). This ratio shows how quickly a firm is able to pay off the interest on its outstanding debt and therefore, serves as a proxy for firm’s default risk. Next, Cap_in represents the capital intensity ratio. It is measured by gross property, plant and equipment divided by the total assets. The

Cap_in ratio accounts for differences in the asset structure of various firms. As mentioned in Maung &

Mehrotra (2010), high capital intensity represent lower risk since tangible assets serve as a better collateral. Therefore, these firms expose debt providers to lower risk and hence expect to obtain higher credit ratings (Ashbaugh-Skaife, 2006, p.220). However, the S&P rating criteria state that moderate capital intensity is optimal since high ratios come with the loss of a firm’s operational flexibility (S&P rating criteria; in Maung &

(14)

Mehrotra, 2010, p.12). Additionally, following Deloof (2003) and Padachi (2006), this study computes the natural logarithm of sales as a proxy for a firm’s size. The coefficient of this variable is expected to be positively related to the credit ratings. This could be explained by the fact that larger size often indicates higher age and thus more established products and increased diversification leading to lower probability of default (Maung & Mehrotra, 2010). Furthermore, the return on equity, ROE, which is the net income divided by the total shareholders’ equity, represents a measure of a firm’s profitability. Lastly, to control for company’s riskiness, firm’s beta is used (Blume et al, 1998; Bhojraj & Sengupta, 2003). This variable stands for the firm’s volatility compare to the market. High values are connected to higher risk but also to potential increased return.

Section 4 Data and Descriptive Statistics

4.1 Data

The data used for further analysis is retrieved from the COMPUSTAT database. Issuer credit ratings are readily available starting from 1985. As this study concentrates on the US market, only the US firms that have registered credit ratings in the North America database are taken into consideration. The time span covers the period starting from January 1985 until December 2015. The credit ratings available in the database are the S&P long-term domestic issuer credit ratings. As stated by the agency, these ratings represent a forward-looking opinion about obligor’s creditworthiness. The assigned ratings in the sample range from AAA to the lowest rate of D/SD. The ratings are in general composed of two categories: the investment and the speculative grade. The investment grade corresponds to the higher rated securities that have ratings ranging from triple-A to the BBB-, the latter represents higher risk securities from BB+ and below (see Table A1 in the appendix for more detailed classification). In the appendix, Figure A1 depicts the average of the credit ratings across the whole sample period. As evidenced in the figure, the average was decreasing over the years and became approximately stable around 2002 reaching an average BB- mark just after the financial crisis. It is known that CRAs are being associated with the causes of the crisis. They were accused of being too slow in updating the information and of being too aggressive in assigning overly optimistic grades. In 2007 and 2008 there was a dramatic drop in the creditworthiness of the structured finance securities and by the end of 2009, more than half of the securities have been downgraded from their original triple A to junk bonds (Benmelech & Dlugosz, 2010, pp.161, 190).

To be consistent with prior research the financial firms (SIC 6000-6999) and utilities (SIC 4900-4999) are excluded from the sample (Fama & French, 2002; Hovakimian et al., 2008; Faulkender & Petersen, 2006; Kisgen, 2006; Flannery & Rangan, 2006). Financials are known for their complex and distinctive balance sheets (Hovakimian et al., 2001, p.4). In addition, they are often highly leveraged, which in case of non-financials would indicate high probability of distress, and thus leading to biased results (Fama & French, 2002, p.8). Utilities are taken out due to their subjectivity to regulatory supervision (Bates, Kahle & Stulz, 2009).

(15)

Additionally, only firms that have consecutive data for minimum 3 years and have been rated for at the least the same period of time are part of the sample. Due to the filtering procedure the final sample consists of 3,852 firms corresponding to the 58,551 firm-year observations.

As the cash management is known to vary across industries the industry fixed effects are taken into account when performing the regression analysis. Too see how the sample is distributed across sectors see Figure A2 in the appendix.

4.2 Descriptive Statistics

Table 1 below shows the descriptive statistics of various cash management attributes and firm characteristics implemented in the models. All variables have been studied in detail and those that revealed outliers were winsorized accordingly.

Table 1. Descriptive Statistics

Mean SD p25 Median p75 Count

Cash management

CR 3.430 1.351 2 3 4 34,841

CCC 80.111 64.641 32.620 70.924 118.235 33,615

Days Account Payable 46.034 25.636 26.180 39.088 59.181 39,151

Operating Cycle 125.219 67.151 76.026 115.513 163.333 34,191

Leverage 1.801 1.671 0.647 1.348 2.528 57,835

Liquidity 1.281 0.671 0.759 1.102 1.646 55,594

Cash Holdings 0.090 0.089 0.018 0.055 0.138 57,851

Firm Characteristics

Interest Coverage Ratio 8.334 8.273 2.251 5.083 11.188 55,446

Capital Intensity Ratio 0.608 0.425 0.275 0.532 0.872 57,918

Firm Size (ln) 6.917 1.910 5.709 6.922 8.191 57,660

Return on Equity 0.092 0.175 0.006 0.107 0.195 57,835

Beta 1.062 0.584 0.672 1.010 1.392 24,203

Observations 58,551

Similar as in Ashbaugh-Skaiffe et al. (2006), the average company in the sample has a rating at the edge of speculative and investment grade (BBB-/BB+). It’s CCC is equivalent to 80.111 days, days payable to 46.034 and mean operating cycle is 125.219 days. This means that it takes approximately 46 days for a company to pay its suppliers. Operating cycle of 125.219 days indicates the time lag between selling inventory and receiving cash from sale. The leverage ratio is equal to approximately 1.801 and liquidity to 1.281. The mean debt to equity ratio of 1.801 suggests that an average firm in the sample has 1.8 times more debt than its shareholders’ equity. The proxy for liquidity, the quick ratio, measures the amount of the most liquid assets available for the current liabilities. A value of 1.281 means that an average firm has $1.281 of its most liquid assets to match with $1 of its current liabilities. As these ratios vary greatly across industries, it is difficult to make a conclusion on what is the appropriate value. Furthermore, the mean cash holding ratio is 0.090, which suggests

(16)

that on average 9% of firm’s total assets are cash or short-term investments. Regarding the firm’s characteristics, the average firm has an interest coverage ratio equal to 8.334. This indicates that companies in the sample are able to pay off the interest on their outstanding debt as the value of the ratio indicates that they earn approximately 8 times as much as they need to pay out. Capital intensity ratio of 0.608 implies that 60.8% of total assets are property, plant or equipment. Moreover, the standard firm in the sample has size of 6.917, calculated as a natural logarithm of sales, which corresponds to approximately 1 billion. Return on equity equals to 0.092. Lastly, beta of approximately 1 means that on average firms in the sample are nor more nor less volatile or risky than the rest of the market.

Table 2 below depicts the correlations between the variables used in each model. The correlation between the dependent variable, CR, and the independent and control variables is in most cases significant at 1% level and in the direction as predicted. However, a few variables show opposing results. The CCC has insignificant correlation with the dependent variable close to 0. Additionally, the coefficient of LIQ is not in line with the expectations. The negative coefficient of 0.107 predicts that liquidity levels and the credit ratings are negatively correlated. This is supported by the findings of Eljelly (2004) claiming that high liquidity is not always optimal due to the unnecessary costs and losses associated with it. Also, Op_cycle is not in line with the expectations. The operating cycle’s correlation is significant at 1% level, however, it is positively correlated to the credit ratings. This would mean that higher rated firms would have longer days inventory and receivables outstanding. This result is unexpected nonetheless, the explanation for this could be that firms with higher ratings tend to offer more trade credit to their suppliers, increasing their days accounts receivable. In addition, lowering CCC takes effort and better performing firms do not value the benefits that come with it as much. The coefficient of

Op_cycle is equivalent to 0.051 and hence the effect is not as significant. Other correlations are as predicted.

Specifically, DPO, Int_cov, Cap_in, ROE and SIZE are significantly positively related to the dependent variable. There is also a significant inverse correlation between LEV, Cash_hold, Beta and the credit ratings. On average, the correlation between various independent and control variables falls below 0.30 with the exception to the correlation between capital intensity and CCC. The coefficient of -0.380 is significant at 1% level implying a significant negative correlation between Cap_in and CCC. Same appears with one of the CCC’s components, the operating cycle, that shows a negative correlation with the capital intensity ratio of 0.353. Earlier, it has been mentioned that capital intensity is measured by property, plant and equipment as a proportion of total assets. The higher the ratio the lower the risk for the debt issuers. This in turn can lead to better payment conditions for the firm and hence affecting the length of the CCC.

(17)

Table 2. Pearson Correlation Matrix

Model 1a CR CCC LEV Cash_hold Int_cov Cap_in SIZE ROE Beta

CR 1 CCC 0.001 1 LEV -0.166*** -0.111*** 1 Cash_hold -0.039*** 0.061*** -0.209*** 1 Int_cov 0.507*** 0.028*** -0.327*** 0.187*** 1 Cap_in 0.120*** -0.380*** -0.023** -0.276*** -0.005 1 SIZE 0.572*** -0.208*** 0.118*** -0.207*** 0.209*** -0.051*** 1 Roe 0.276*** -0.035*** -0.102*** 0.021*** 0.298*** -0.007 0.167*** 1 Beta -0.194*** 0.013* 0.003 0.086*** 0.011 -0.035*** 0.096*** -0.031*** 1

Model 1b CR DPO Op_cycle LEV Cash_hold Int_cov Cap_in SIZE ROE Beta

CR 1 DPO 0.068*** 1 Op_cycle 0.051*** 0.276*** 1 LEV -0.166*** 0.038*** -0.093*** 1 Cash_hold -0.039*** 0.060*** 0.088*** -0.209*** 1 Int_cov 0.507*** -0.002 0.033*** -0.327*** 0.187*** 1 Cap_in 0.120*** 0.090*** -0.353*** -0.023** -0.276*** -0.005 1 SIZE 0.572*** 0.046*** -0.195*** 0.118*** -0.207*** 0.209*** 0.051*** 1 Roe 0.276*** 0.031*** -0.041*** -0.102*** -0.021*** 0.298*** 0.007 0.167*** 1 Beta -0.194*** 0.012*** 0.020*** 0.003 0.086*** -0.011 -0.035*** 0.096*** -0.031*** 1

Model 2 CR LIQ SIZE ROE Beta

CR 1 LIQ -0.107*** 1 SIZE 0.551*** -0.284*** 1 ROE 0.331*** -0.026*** 0.214*** 1 Beta -0.181*** 0.112*** 0.022** -0.046*** 1 * p < 0.1, ** p < 0.05, *** p < 0.01

(18)

Section 5 Results

This section presents and discusses the empirical results of the two regression models and the constructed hypotheses stated in Section 4. It comprises of two parts elaborating on Model 1 and 2, respectively.

5.1 Regression Results of Model 1

Table 3 presents regression results of the first model that measures the impact of working capital management, leverage levels and several control variables on the credit ratings. The results in general include coefficient estimates, robust standard error in the parentheses adjusted for clustering across firms and the McFadden's pseudo R-squared coefficient representing the overall fit of the model at the bottom of the table. Column 1 depicts the results of a regression including only CCC and Cash_hold as independent variables. The second column looks at the different components of CCC to examine, which one has the most impact. Column 3 concentrates on the effect of leverage on firm’s creditworthiness. In column 4 and 5, previous columns are combined together to see the merged effect of CCC and leverage. Finally, in the last two columns the regression models control for the year and industry differences within the sample.

The result from columns 1,4 and 6 suggest, that CCC is positively related to the credit ratings highly significant at 1% level. For a one additional day to the CCC, the log-odds of receiving higher rating increase by 0.004 in column 1 and stay approximately the same in column 4 and 6. More specifically, despite the coefficient close to zero, credit ratings tend to increase with CCC. This is not in line with the previous research (Deloof, 2003; Padachi, 2006). To investigate the result more closely, the coefficients of DPO and Op_cycle from columns 2, 5 and 7 are examined. All three columns yield the same results. DPO and Op_cycle reveal similar result as in the case of CCC. Both are positively and significantly related to the credit ratings with their coefficients close to 0. The coefficient of the DPO is in the direction as predicted, however, the Op_cycle shows the opposite. Especially during the economic downturns, firms had to provide trade credit to the customer in order to remain competitive, maintain sales and keep good relationships. As trade credit slows down the collection of cash it triggers an increase of CCC (Kling, Paul & Gonis, 2014, p.129). Furthermore, as Kling et al. (2014) argue, firms with sufficient funding and access to external markets do not improve their CCC. This is true even if an increase would create more shareholders’ value. Also, better performing firms are often not willing to reduce their CCC as it takes a lot of effort and it could disrupt customers (pp.129-130). This finding is consistent with Nobanee (2009) that provides evidence that lower CCC can result in inventory shortages, losing of good credit customers or leading to bad reputation due to delayed payments.

(19)

Table 3. The regression output of the effects of cash management and leverage on firms’ credit ratings

The dependent variable used is the corporate credit ratings and the various columns correspond to the following models:

(1) CR = f (cash management, firm characteristics)

(2) CR = f (cash management components, firm characteristics) (3) CR = f (leverage, firm characteristics)

(4) CR = f (cash management and leverage, firm characteristics)

(5) CR = f (cash management components and leverage, firm characteristics) (6) CR = f (cash management and leverage, firm characteristics) + Fixed Effects

(7) CR = f (cash management components and leverage, firm characteristics) + Fixed Effects

CCC stands for the cash conversion cycle. For the leverage ratio (the debt-to-equity) the variable LEV is used. Cash_hold represents the cash holdings ratio of a firm. Int_cov stands for the interest coverage ratio, whereas Cap_in is the capital intensity ratio. A firm’s size and return on equity ratio are depicted by SIZE and ROE,

respectively. Beta is the firm’s equity beta. Last two items are the components of CCC: DPO and the Op_cycle, that represent days payable outstanding and operating cycle. Robust standard errors adjusted for clustering across firms are reported in the parentheses. The significance levels are reported with ***, **, * corresponding to 0.01, 0.05 and 0.1, respectively. (1) (2) (3) (4) (5) (6) (7) CCC 0.004*** 0.005*** 0.004*** (0.001) (0.001) (0.001) LEV -0.257*** -0.268*** -0.289*** -0.287*** -0.305*** (0.028) (0.033) (0.031) (0.036) (0.035) Cash_hold -0.434 -0.562 -2.659*** -2.864*** -1.781*** -1.939*** (0.425) (0.416) (0.487) (0.482) (0.521) (0.516) Int_cov 0.086*** 0.099*** 0.097*** 0.140*** 0.139*** (0.007) (0.009) (0.009) (0.009) (0.009) Cap_in 0.171* 0.624*** 0.592*** 0.290* 0.330** (0.097) (0.136) (0.135) (0.156) (0.156) SIZE 0.784*** 0.789*** 0.761*** 0.800*** 0.810*** 1.176*** 1.178*** (0.041) (0.041) (0.033) (0.043) (0.042) (0.052) (0.051) ROE 3.546*** 3.518*** 2.225*** 2.326*** 2.339*** 2.429*** 2.417*** (0.202) (0.198) (0.192) (0.224) (0.223) (0.242) (0.243) Beta -0.836*** -0.848*** -0.873*** -0.841*** -0.851*** -0.564*** -0.573*** (0.056) (0.056) (0.044) (0.052) (0.052) (0.063) (0.063) DPO 0.005** 0.005** 0.005** (0.002) (0.002) (0.002) Op_cycle 0.004*** 0.006*** 0.005*** (0.001) (0.001) (0.001) Year Fixed Effects Industry Fixed Effects No No No No No No No No No No Yes Yes Yes Yes Observations Pseudo R-squared 12,738 0.1439 12,738 0.1484 16,761 0.2011 12,605 0.2058 12,605 0.2108 12,605 0.2988 12,605 0.3028

(20)

The coefficient of leverage is negatively related to the credit ratings, statistically significant at 1% level. The results from column 2 to 7 suggest that for one unit increase in LEV, the dependent variable is expected to decrease on average by 0.281 on the scale from 1 to 7, ceteris paribus. This finding was expected as highly leveraged firms have harder time repaying their debt obligations and are therefore perceived riskier. The coefficient of Cash_hold is in line with the findings of Acharya et al. (2012). They claim that firms with high levels of cash are likely to find themselves in danger of distress and hence larger cash holdings ratio is associated with increased credit risk. By holding excessive amounts of cash, firms are losing potential revenue that could have been earned by investing surplus cash in a profitable manner. In this way, the shareholders’ confidence in the company’s management is reduced.

Following Ashbaugh-Skaife et al. (2006) and Maung & Mehrotra (2010), Int_cov and Cap_in serve as additional control variables in this model. In line with the prior research, both Int_cov as well as Cap_in have a positive impact on credit ratings. The coefficients of the interest coverage ratio in columns 3 to 7 suggest that a one unit increase in the control variable leads to a rise of the log-odds of having better rating on average by 0.112, ceteris paribus. This coefficient is significant at 1% level. Recall, that the interest coverage ratio shows how quickly a firm is able to repay its debt obligations and thus, the higher the ratio the lower the default risk. Next, the capital intensity coefficient is also found to be positive and significant at 1%. Capital intensity represents the proportion of property, plant and equipment compared to its total assets. The direction of Cap_in coefficient is equivalent to the findings of Hovakimian et al. (2001) and suggests that the higher the capital intensity the better since tangible assets can serve as a better collateral. The debt providers in this case are exposed to lower risk and firms are therefore expected to receive higher ratings.

Furthermore, the coefficients of the three control variables, SIZE, ROE and Beta are consistent with prior research. As in Hovakimian et al. (2001), the coefficient on firm size is positively related to the credit ratings, significant at 1% level. This indicates that firms of a bigger size are less likely to default due to several reasons such as increased diversification, economies of scale, reputation and stability. Next, profitability, depicted by

ROE, is as expected, positively related to the dependent variable. The coefficient in column 1 suggests that for

a one unit increase in the return on equity ratio credit ratings are expected to change by 3.546 in the log-odds scale, ceteris paribus. Profitability represents the capability of a firm to generate cash flows and is related to lower expected costs of financial distress. Lastly, Beta has a negative and significant coefficient at 1% level (Blume et al, 2008; Bhojraj and Sengupta, 2003). This coefficient suggests that the firm’s increased equity risk has an inverse effect on the dependent variable. As evidenced from columns 6 and 7, the results are robust to the inclusion of year and industry fixed effects.

(21)

5.1 Regression Results of Model 2

The results of Model 2 are reported in the Table 5 below. This model concentrates most importantly on the impact of liquidity on the perceived creditworthiness. Column 2 differs from 1 solely by the fixed effects inclusion.

Table 4. The regression output of the effects of liquidity on firms’ credit ratings

The table presents results from the analysis of the ordered logit regression of Model 2 examining the impact of liquidity and various control variables on the dependent variable, firm’s credit ratings. LIQ represents the liquidity ratio. SIZE stands for the natural logarithm of firm sales. Next, ROE is the return on equity ratio and firm’s equity beta is depicted by Beta. Robust standard errors adjusted for clustering across firms are reported in the parentheses. The significance levels are reported with ***, **, * corresponding to 0.01, 0.05 and 0.1, respectively. (1) (2) LIQ 0.284*** 0.552*** (0.062) (0.067) SIZE 0.819*** 1.168*** (0.032) (0.041) ROE 3.317*** 3.880*** (0.186) (0.204) Beta -0.814*** -0.644***

Year Fixed Effects Industry Fixed Effects

(0.050) No No (0.058) Yes Yes Observations Pseudo R-squared 16,097 0.1530 16,097 0.2242

From the results above it can be concluded that there is significant positive relationship between firm’s liquidity and its credit ratings. The coefficient from column 1 and 2 states that for a one unit increase in the liquidity the ratings are expected to increase by 0.284 or 0.552, respectively, on the log-odds scale, holding everything else constant. The finding suggests that the higher the liquidity, the lower the risk of defaulting. This is explained by the statement that higher liquidity provides firms with an ability to repay its debt obligations when due. This then lowers the firm’s risk and increases its credibility and trustworthiness. Reider (2005) finds similar results. In his study, he claims that enough liquidity is crucial in order to make necessary payments and ensure continuity of business operations. The coefficients of the control variables are closely comparable to the results from Model 1. Both, SIZE and ROE, are significantly positively related to the dependent variable, whereas the firm’s equity risk, Beta, has a negative coefficient significant at 1% level. This indicates that firms of larger size, higher profitability and lower risk are expected to obtain better ratings. By including industry and fixed effects in column 2, results prove to be robust.

(22)

Section 6 Robustness Tests

In this section, several robustness tests are performed to examine how the regression coefficients behave when the models are slightly modified, by adding, substituting or removing one or more variables. If the resulting coefficients conclude robustness, then structural validity of the models can be assumed.

Firstly, the ordered dependent variable is substituted. Secondly, the return on equity ratio is excluded from the regressions and the return on assets and return on sales are used as alternative measures for a firm’s profitability. Thirdly, an additional variable, the R&D intensity (RDI), is added that has proven in previously conducted studies to have a significant impact on firms’ credit ratings. As a last test, an exogenous shock is added to the Model 1 and fixed effects model is implemented in order to rule out any unobserved heterogeneity and endogeneity between dependent and explanatory variables.

6.1 Robustness Test 1: Investment versus Speculative grade

As mentioned earlier, credit ratings follow a sequential pattern. The dependent variable with multiple categories makes it difficult to measure the marginal effects of changes in the cash management attributes and other firm characteristics. Therefore, following Ashbaugh-Skaife et al. (2006), an alternative dependent variable is used. In general, credit rating levels are classified into two categories- the investment grade and the speculative grade. Thus, the ordered dependent variable is replaced with the binary for the investment grade and the logit regression is performed. In the appendix, Table A3 displays the results from the logit regression using investment grade dummy variable as the dependent variable. This robustness check is performed on both models. The resulting coefficients are very closely related to the original models. There is a significant positive effect of CCC, LIQ, Int_Cov, Cap_in, SIZE and ROE on the investment grade binary variable. On the other hand, coefficients of LEV, Cash_hold, and Beta have a negative impact on the dependent variable. The coefficient of Cap_in loses its significance when controlling for fixed effects. This suggests that the effect is smaller than previously anticipated.

6.2 Robustness Test 2: Profitability Ratio

Secondly, as an additional robustness test the proxy for profitability is substituted. New ordered logit regressions are performed using return on assets and return on sales, respectively. Both variables as well as the original return on equity measure a firm’s profitability. Return on assets, calculated as the net income over the average total assets, measures how effective a firm’s management is in employing its asset base. The higher the ratio the better. Furthermore, return on sales provides insight into firm’s operational efficiency. The ratio is calculated by dividing the net income by the total sales. Increase in the ratio suggest higher growth, whereas a decrease serves as a negative signal of a firm’s performance. In the appendix, Table A4 presents the results from the different regressions. All the variables show similar results to the original models with a majority still keeping the significance levels at 1%. The dependent variable is positively affected by CCC, LIQ, Int_cov,

(23)

Cap_in and SIZE significant at 1% level. Whereas LEV, Cash_hold and Beta show a negative relation with the

credit ratings, but also with a majority significant at 1% (with exception to LIQ that in columns 1 and 3 in Panel B shows significance levels at 5%). Both ROA and ROS coefficients are positive and highly significant at 1% suggesting that more profitable firms receive a better rating.

6.3 Robustness Test 3: Additional Explanatory Variable

Third robustness rest is performed by adding an extra explanatory variable in the model (see Table A5 in the appendix). Liao (2013) provides evidence that RDI has a significant impact on the credit ratings. RDI, calculated by R&D expenditures divided by total assets, measures how much a firm invests in research and development compared to its assets. By investing in this area, a firm can experience future growth and hence having a positive impact on the creditworthiness. Therefore, the R&D expenditure can be considered an internal investment and be included as part of cash management. The ratio in the sample ranges from 0 to 0.15 with a mean of 0.034. In the appendix, Table A3 displays the results. As evidenced in the table, most of the variables depict similar results. Also, the RDI coefficient is in line with the expectations. Both models show a strongly positive and highly significant coefficient at 1% suggesting RDI to be beneficial for the firm’s credit ratings. The coefficient from column 1 in Panel A suggests that with a one point increase in the RDI the log-odds of having higher credit ratings increase by 6.701, on a scale from 1 to 7, holding everything else constant. This coefficient might seem high, however, taking into account the RDI average ratio of 0.034, a 1 point increase in the R&D expense compare to its total assets does not seem rational. The coefficient of RDI in column 1 and 2 from Panel B is equivalent to 10.305 and 5.441, respectively, highly significant at 1% level. 6.4 Robustness Test 4: Endogeneity

The issue of endogeneity was partly addressed in the previous section. The evidence provided earlier suggests that cash management and capital structure of a firm have a significant impact on credit ratings. However, as credit ratings serve as a screening device and a signal of a firm’s creditworthiness they have an effect on the business itself. Maung and Mehrotra (2010) provide evidence that credit ratings affect a firm’s cost of debt and its overall cost of capital (p.1). In addition, lower rated firms are perceived to have a higher probability of default and hence, lower ability to repay its debt obligations. Debt providers take this into account and require riskier types of firms to pay earlier. Also, the possibility of receiving trade credit is lower. On the other hand, better performing firms with higher average ratings have lower cost of capital and their payment conditions are more favorable. All these factors have an impact on CCC. This is where the problem of endogeneity arises. Reverse causality, in this case, occurs due to two-way causal relationship between the credit ratings and the CCC. One cannot perfectly control for endogeneity issues but there are a few methods that can be employed in order to reduce the range of this effect.

As previously completed in the original models, independent and control variables are lagged in order to make sure that changes in cash management and capital structure arise only after the credit ratings are modified. This

Referenties

GERELATEERDE DOCUMENTEN

Besides this, I also found that independence of the board of the parent has no significant effect on the relationship between foreign subsidiaries financial reporting quality and

The positive coefficient on DLOSSRDQ means that firm with negative earnings have a stronger relationship between credit ratings and risk disclosure quality compared to firms

The results confirmed the expected relation between the market value (measured using the market price to book ratio) and the credit rating, as well as relations between the CR

Rating estimated credit rating to capture Standard &amp; Poor’s credit ratings equivalent on a scale of 2–27 (AAA to D) at the end of year t; Review, an indicator variable that

In doing so, the answer is sought to the question of whether investors in the bond market have changed their focus towards Long Term Issue Credit Ratings (LTRs)

Positive and significant coefficients are acquired for the boundary dummy, which implies that firms with a BBB- rating get assigned lower credit ratings (equivalent of a

To test the cannibalism hypothesis, the researchers analyzed a sample of the human and animal (cattle) remains to compare butchery/trauma evidence. Following

Een alternatieve verklaring zou kunnen zijn dat in de eerste vijf jaar de snelheid van cognitieve achteruitgang omhoog gaat door leeftijd, maar dat dit niet zichtbaar is door