Working capital management and financing : the impact of debt and equity on the cash conversion cycle in Western Europe

Hannah Sophie de Rooij

EXECUTIVE MASTER OF FINANCE AND CONTROL AMSTERDAM BUSINESS SCHOOL – UVA

SUPERVISOR: JOOST BERGEN VERSION: FINAL VERSION 1.0 DATE: 11 JUNE 2016

Working Capital Management and

Financing

THE IMPACT OF DEBT AND EQUITY ON THE CASH

CONVERSION CYCLE IN WESTERN EUROPE

1

THESIS ACKNOWLEDGEMENT

I am using this opportunity to express my gratitude to everyone who supported me throughout the course of the Executive Master of Finance and Control. In particular, I would like to thank my thesis supervisor Mr. Bergen for his guidance, advice and reviews during the writing of this thesis. I would also like to thank Rick van Dommelen, Gladys Bos-Deunig and Danny Siemes from PwC for giving me this opportunity and their support!

Further, I would like to express my warm thanks to my partner in life Frederik van Dalfsen for his ongoing support, confidence and patience during the past 3 years of studying. Without his support and help I would not have been able to graduate. In addition I would like to thank my family and friends for their encouraging and understanding throughout the course of my study.

Thank you all,

Hannah Sophie de Rooij 11 June 2016

2

TABLE OF CONTENTS

1.0 INTRODUCTION 4

1.1 Research objective and central question 4

1.2 Relevance of the research 5

1.3 Structure of the report 7

2.0 THEORETICAL FRAMEWORK 8

2.1 Introduction 8

2.2 Working capital and the CCC 8

2.3 Elements of the Cash Conversion Cycle 10

2.3 Internal and external working capital determinants 11

2.3.2 Profitability and the length of the CCC 12

2.3.3 Growth opportunity and the length of the CCC 13 2.3.4 Industry and geographical location, and the length of the CCC 13 2.4 Capital structure as working capital management determinant 15

2.5 Private equity versus public equity 19

2.6 Bank debt versus public debt 21

3.0 RESEARCH METHODOLOGY 23

3.2 Data set and sample 23

3.3 Analyses used in this study 24

3.4 Definition of variables used in this study 25

4.0 QUANTITATIVE DATA ANALYSIS AND DISCUSSION 28

4.1 Descriptive statistics 28

4.2 Pearson correlation analysis 30

4.3 One-way ANOVA analysis 32

4.4 Regression analysis 32

5.0 QUALITATIVE DATA ANALYSIS AND DISCUSSION 38

5.1 Introduction to qualitative analysis 38

5.2 PE portfolio companies and WC optimization 38

5.3 The influence of WC optimization on the level of WC 40

5.4 Levers of WC improvement 42

6.0 CONCLUSION, LIMITATIONS AND SUGGESTIONS FOR FURTHER RESEARCH 44

3

6.2Discussion and conclusion 44

6.3Limitations and suggestions for further research 46

6.4Reflection 47

7.0BIBLIOGRAPHY 49

4

1.0 INTRODUCTION

Working capital performance is often considered as an important indicator of ‘good management’ and an instrument to measure how well a company is actually managed. Working capital

management (‘WCM’) is next to capital budget and capital structure an important component in a firm’s financial decision management. On a daily basis financial managers and controllers are confronted with important working capital decisions which may highly affect firm value and short-term liquidity (Jose et. al, 1996).

Although the importance of working capital is broadly recognized, in literature WCM is receiving less attention than longer-term investment and financing decisions. This while companies who are not managing their working capital position efficiently may face financial distress. Those companies can then even go bankrupt, despite having healthy operations and profits. This is why in practice

controllers spend relatively a lot of time on managing the firm’s working capital processes and liquidity position. A company should be able to pay-off its short-term debt in time, while at the same time making short-term investments to maintain the required inventory level for future sales

(Richards and Laughlin, 1980).

In 2006 Kieschnick et al. conducted a research to the relationship between corporate WCM and shareholders’ wealth. They found evidence that an extra dollar invested in net operating working capital is worth less than the incremental dollar held in cash. The latter increases the firm’s free cash flow (‘FCF’), and therefore firm value (sum of all discounted future free cash flows is equal to firm value). Therefore the level working capital can be considered as an important determinant of firm value. This makes it relevant to investigate what actually does affect and determine the level of working capital.

The last three decades several researchers have studied both internal and external determinants of working capital. Internal determinants are firm characteristic specific factors, such as firm size, while external factors consist of e.g. macroeconomics factors as the economic climate (Zariyawati et al., 2010).

1.1 Research objective and central question

Prior empirical studies have mainly focused on working capital determinants such as firm size, profitability, revenue growth and industry (Zariyawati et al, 2010; Chiou and Cheng, 2006). The results of those studies were mixed due to time differences and different measurements of variables. This study investigates whether

financing – a firm’s capital structure and type of ownership – is a determinant of working capital as well. Capital structure is usually defined as how a firm does finance its assets, overall operations and growth

5

opportunities by using different sources of financing funds, and various types of debt and equity (Frank and Goyal, 2009). However, in literature researchers refer mostly to a firm’s debt-to-equity ratio (‘debt ratio’). This is an important indicator for a company’s risk profile, as in general highly leveraged firms entail greater risks to become financially distressed and to go bankrupt, than firms with a relatively low leverage ratio. In particular for highly leveraged companies it may be rather important to free up cash by shortening the working capital cycle. Or in other words, to lower the Cash Conversion Cycle (‘CCC’), which is the required period of time to convert cash disbursements back into a cash inflow from a firm’s regular course of operations (Richards and Laughlin, 1980). In general, companies favor internal funds above attracting more debt or equity as financing source (Myers, 1984). However, for relatively highly leveraged firms it can be more difficult or not even possible to attract (expensive) external financing. In this way companies may be forced to be creative in unlocking cash by improving working capital processes. This leads to the following main research question:

Into what extent is a firm’s ‘capital structure’ and ‘ownership’ a determinant of WCM?

The following sub questions are derived from this central question:

(1) What is working capital and why is the CCC a common measure of working capital performance? (2) What are determinants of WCM?

(3) How may a firm’s debt ratio affect the length of the CCC?

(4) How do different sources of financing – public debt versus bank debt and public equity versus private equity – may impact the length of the CCC?

The research questions lead to the following research design:

Figure 1 – Research design

1.2 Relevance of the research

As mentioned before relatively a lot of research has been done to specific determinants of WCM (e.g. firm size, industry, geographical location). Though, those studies neglect the influence of how and by whom the firm is financed on the level of working capital. As shareholders and debtholders try to

To assess whether capital structure and type of ownership is a

determinant for the length of the CCC Quantitative and

qualitative analysis results

Set of hypotheses Private and public

equity

Cash Conversion Cycle

Private and public debt Theory regarding ownership and monitoring of management Theory regarding debt as corporate governance mechanism Theory regarding determinants of CCC Quantitative and qualitative analysis results Quantitative and qualitative analysis results

6

minimize agency costs by aligning their interests with those of (executive) management (Armstrong et al., 2010) it is relevant to investigate whether specific shareholders and debtholders influence explicitly or implicitly (top-level) management to be more cash-efficient.

I am currently working at PwC NL within ‘Advisory – Deals practice’. I am among others providing advice to companies how they can free up cash by e.g. improving their working capital processes. In cooperation with PwC Europe, PwC UK does publish annually the ‘PwC Annual Global Working Capital Survey’. Within this survey they consider the following four factors have an impact on a company’s working capital requirements and relative performance to other companies (Annual Global Working Capital Survey by PwC, 2015):

1. Company size 2. Type of business

3. Economic maturity of the region 4. Management focus

Factor number 4 makes it even more important for shareholders and debtholders to influence management that they will make lower management layers more cash aware (e.g. stimulating lower-level management and controllers to improve working capital processes). Moreover, when a specific type of shareholder put more pressure on its management to follow a strict cash policy, by for instance requiring management to have proper working capital mechanisms in place, controllers should be aware of the consequences in case a company is acquired by those type of owners. This can suddenly change the daily work activities of the (business) controller hugely. Their focus may then shift from budgeting and forecasting oriented activities towards more working capital process and cash improvement activities.

A private equity party is potentially more cash focused than public shareholders. PE shareholders may encourage management to continuously improve the level of working capital by optimizing certain operational and collecting processes. The same counts for private versus public debtholders. In general private debtholders are more in the position to monitor management and to check whether the company complies to certain debt covenants (e.g. specific financial ratios).

In addition to the more practical relevance of this research, I would like to provide scientific evidence to an important corporate governance theory that considers debt as an important governance

mechanism for shareholders to discipline its management (Armstrong et al, 2010). By doing this research I hope to find out that firms close to an optimal debt-to-equity ratio (Miao, 2005) will have a

7

lower CCC in comparison to over- or underleveraged companies. This as management from relatively highly leveraged companies have less cash at their disposal, forcing them to do less (unnecessary) investments in inventory or to unlock cash by extending supplier terms and shortening payment terms from their customers (Bergen, 2006). However, companies which are overleveraged and have debt levels beyond the more optimal debt-ratio are potentially in financial distress. Recent research has found out that those companies are actually worse cash management performers with an increased CCC.

1.3 Structure of the report

The second chapter of this study will provide a theoretical framework around WCM, touching upon the working capital determinants and the potential relation between the level of working capital and a firm’s capital structure. This chapter will form the basis for forming the hypotheses, which will also be presented in chapter 2. In chapter 3 the research method is presented, describing the research design, the research sample, the data collection process and statistical tests for the quantitative part of this study. This chapter will also discuss how the qualitative research part of this study will be executed. The subsequent 2 chapters will provide both the quantitative and qualitative results, followed by a discussion linked to the theoretical framework of chapter 2. The last chapter will conclude, present this study’s limitations and provide suggestions for further research.

8

2.0 THEORETICAL FRAMEWORK

2.1 Introduction

This chapter will introduce the CCC as useful measure of firm’s effective WCM and especially cash management (Attari and Raza, 2012). This is followed by a comprehensive description of the previously researched determinants of working capital management. The rest of the chapter will focus on literature around capital structure and firm ownership in relation to WCM.

2.2 Working capital and the CCC

The ‘current ratio’ (also called ‘liquidity ratio’ or ‘cash ratio’) has traditionally been the key indicator of a firm’s liquidity position and working capital position (Richards and Laughlin, 1980; Preve and Sarria-Allende, 2010). It indicates whether a company has enough current (short-term) assets to fulfill its short-term debt obligations. Anything below 1 indicates a negative working capital. The current ratio is calculated by the following formula:

Current ratio = Current Assets / Current Liabilities

There is always much debate about what portion of cash belongs to current assets. Mostly cash is only included for firms that has to maintain a large cash balance for day-to-day operations. This is then considered as operational cash. The investment in operational cash may have a high ‘cost of carry’ (WACC -/- investment), as a company’s WACC does often exceed the interest on the outstanding cash amount. Firms will therefore invest their excess cash in treasury bills, short-term government

securities and commercial paper and bank deposits (current yield environment). This may deliver companies a higher return than the outstanding cash on the bank account will generate. Though, companies should hold a sufficient amount of operational cash. Not merely to finance their daily operations, but as well as kind of insurance that may be needed in times of deteriorated business performance. Operational cash is therefore often considered as part of the company’s working capital position.

Several studies have revealed (Chiou and Cheng, 2006) that companies with a high current ratio tend to show better firm performance. However, those studies did not really evaluate how efficient a company manages its working capital position since the level of efficiency is not measured by those ratios (e.g. current and the quick ratio).

Though, it is important that companies manage their assets as efficient as possible. Managers have to manage continuously the tradeoff between on the one hand holding sufficient liquid funds to finance their daily business, while on the other hand increasing the company’s level of profitability by actually

9

current assets the firm’s Return On Capital Employed (‘ROCE’) will be increased, enhancing firm value (Kieschnick et al, 2013).

Figure 2 - Return on Capital Employed (ROCE)

Contemporary working capital performance literature is more and more referring to the CCC as a measure for WCM instead of the current and quick ratio. The CCC incorporates the time interval between actual expenditures on a firm’s purchase of productive resources and the cash receipts from selling finished products (Richards and Laughlin, 1980; Raheman and Nasr, 2007). The longer this time lag, the larger the investment in working capital (Deloof 2003). What we really like to know is how well a company is able to decrease this time lag, as this is an important indicator to measure efficiency.

Figure 3 - Cash Conversion Cycle (CCC)

Firm buys inventory Firm pays for inventory Firm sells product Firm receives payment

Inventory Accounts receivable

Accounts payable

Cash out

Cash Conversion Cycle

10

2.3 Elements of the Cash Conversion Cycle

The CCC contains the following elements (De Vries and Van der Wielen, 2014):

(1) Days Sales Outstanding (DSO): A measure of the average number of days that a company takes to collect cash after the sales of goods or services have been delivered.

(2) Days Inventories On-hand (DIO): A measure to give an idea of how long it takes for a company to convert its inventory into sales.

(3) Days Payables Outstanding (DPO): DPO is an indicator of how long a company takes to pay its trade creditors.

The Cash Conversion Cycle is the sum of the number of days of sales outstanding (DSO) and days of inventory on-hand (DIO), minus the days of payables outstanding (DPO), making the following formula:

CCC = DSO + DIO -/- DPO

As an effective way to release cash for investments and debt repayment firms focus on lowering the CCC by implementing working capital improvements such as:

1) Optimize the working capital drivers accounts receivables, inventory and payables by adjusting payment terms to suppliers and from customers

2) Optimize the working capital drivers by improving operational working capital processes (OTC, FTF and PTP) by e.g. influencing the way of working, people, improving systems on cash performance, adjusting service levels and introducing strategic procurement.

11

Figure 4 – CCC in relation to cash release management and shareholder value

(Source: PwC Analysis)

2.3 Internal and external working capital determinants

Many researchers have done research to the determinants of WCM. Those studies examined the connection between e.g. firm size, profitability, growth opportunity, industry, geographical location and the length of the CCC. The following sub-chapters will shortly describe those determinants. 2.3.1 Firm size and the length of the CCC

According to several studies firm size is negatively related with working capital management. This means that in general relatively large firms have a lower CCC than small(er) firms (Zariyawati et al., 2010; Attari and Raza, 2012; Annual Global Working Capital Survey PwC, 2015). This can be

explained by the following:

1) Large firms can afford to hire expertise to establish efficient working capital management (Zariyawati et al., 2010).

2) The larger a business’s share of its served market relative to its competitors, the better its position to purchase raw materials (inventory), negotiate delivery schedules that are in line with production requirements and less they need to use credit and quick delivery as selling tools (Kenneth P. Nun, 1981).

Shareholder value

Profit Capital employed

Working capital (CCC) Fixed assets

Receivables Inventory Payables

Payment terms, fast processes, error-free bills, responsibility,

transparency

Planning quality, availability, complexity, responsibility,

transparency

Payment terms, payment process, discount/interest optimisation, responsibility, transparency Working capital drivers Optimise processes

Order to cash (OTC) Forecast to fulfil (FTF) Procure to pay (PTP) Minimise time between order

and cash received date

Minimise time between goods receipt and despatch date

Maximise time between invoice and cash paid date Revenue Costs

Debt repayment Investments

12

3) Large firms have a better cq. more dominant position in the overall supply chain, which means a higher supplier bargaining and negotiating power (Hill et al., 2010).

Though, other studies did indicate that larger firms have actually a higher number of CCC days (Chiou and Cheng, 2006). This as larger firms have relatively better access to the financial and capital markets which makes it easier for them to find financing sources, which are also often less expensive (Jordan et al., 1998). Figure 4 below shows that credit agencies like S&P and Moody’s regularly rate small and medium-sized enterprises (SMEs) between BB+ and B-, reflecting a ‘Non-Investment Grade’ (Financieringsmonitor, 2015). Smaller firms may therefore have fewer opportunities to attract (cheap) external financing. It is therefore for those companies far more difficult to finance working capital investments. Basel II did even strengthen the capital requirements, meaning that more capital (equity and cash) should be maintained in the company. As a result, smaller firms may be more inclined to monitor closely their working capital position than larger firms.

Figure 5 – Debt rating

2.3.2 Profitability and the length of the CCC

Several studies reveal that companies with a shorter CCC are more profitable than firms with a longer CCC (measured in terms of Return on Assets (‘ROA’)) (Shin and Soenen, 1998; Attari and Raza, 2012; Deloof, 2003; Raheman and Nasr, 2007, Hill et al., 2010 and Bhutto et al., 2011). According to Attari and Raza (2012) this might be due to the fact that firms having more investments blocked in (current) assets may have to find alternative (costly) financing sources. This means that firms that are not able to manage their working capital position efficiently may potentially face high financing costs. This will decrease the firm’s overall profitability.

13

Though, the level of a firm’s profitability is more influenced by its customer strategy and its competitive position within it product segment than WCM itself. Working capital choices are therefore following the (customer) strategy of the company, and not the other way around. For instance, when a company executes a cost leadership strategy, it will probably try to minimize its inventory as much as possible in order to have low storage costs. This will influence the level of working capital and length of the CCC positively. In case a company wants to execute a more product leadership and customer intimacy strategy, it will be more inclined to do (long-term) upfront

investments, which increases the length of the CCC. The company’s (customer) strategy can therefore be considered as a strategic determinant of working capital (Kenneth P. Nun, 1981).

2.3.3 Growth opportunity and the length of the CCC

A firm’s growth opportunity does also determine how managers manage a firm’s working capital. Zariyawati et al. (2010) has found out that growth opportunity, measured by sales growth, is inversely related to working capital management. Firms will increase their short term financing to meet future sales demand.

2.3.4 Industry and geographical location, and the length of the CCC

There is a significant industry effect on firm’s investment in working capital (Hawawini et al., 1986; Etiennot et al., 2012). Figure 5 shows that the level of working capital is determined by specific industry dynamics (Annual Global Working Capital Survey PwC, 2015).

Figure 6 – Industry is an important determinant for the CCC

(Source: PwC Analysis)

Industry norms for the level of working capital differ per industry due to the nature of its operating cycle. Firms operating within the same industry are expected to adhere to certain working capital industry norms as, for instance, it is rather difficult to enforce stricter terms of trade than direct competitors do (Hawawini et al., 1986). Though, there is still a significant spread in working capital

14

performance between top and bottom performers (figure 6). The difference within industries may be to a certain extent explained by the working capital determinants we have just discussed, such as size, growth, profitability and geographical location. This research has to find out whether it is also

influenced by a company’s capital structure and its type of ownership.

Figure 7 – Top and bottom performers within industries

(Source: PwC Analysis)

For certain companies, the working capital ratio will even be negative. For them this is an important source of financing (e.g. Telecom and Airline industry). Other companies, like manufacturers, will have a relatively high working capital ratio as usually cash proceeds will be collected after they have paid their suppliers.

Terms and conditions to customers and suppliers are both industry and location specific. The geographical location of companies does influence the length of the CCC due to e.g. local regulations (Annual Global Working Capital Survey PwC, 2015) and country specific payment habits or norms. The figure below provides an overview of the CCC per region from 2010 until 2014 (figure 7). It shows that there are considerable differences in the length of the CCC per region. Though, differences in the length of the CCC within a specific region are only visible within a small bandwidth over a certain period of time.

15

Figure 8 – Regional differences in the length of the CCC

(Source: PwC Analysis)

2.4 Capital structure as working capital management determinant

Modigliani and Miller, the founders of corporate finance, have shown that in perfect capital markets financing decisions do not matter as firm value is not affected by capital structure. Firm value will be created by making profitable investments in the business’ real assets (Modigliani and Miller, 1958 and 1963). However, in practice capital markets are far from perfect and things like taxes,

information asymmetry and transaction costs do highly influence managerial decisions about the firm’s capital structure.

Debt financing provides companies a tax shield. The interest on debt is classified as a tax-deductible expense that reduces the amount of corporate taxes a company should pay its government. This results in higher cash flows to the firm’s investors and hence a higher firm value. The tax shield does increase the after-tax value of the company equal to the present value of the tax shield. So, this means that debt financing does matter as it increases the future cash flows to its investors and lowers the company’s Weighted Average Cost of Capital (WACC). Though, companies are never fully debt-financed due to financial distress costs (e.g. bankruptcy). This means implicitly that there exists an optimal debt-to-equity ratio, which is highly influenced by firm and industry specific determinants (Miao, 2005). 64 36 48 38 33 46 28 CCC 2013 and 2014 73 37 49 42 33 50 30

16

Information asymmetry does affect a company’s capital structure twofold. First of all, when investors are not properly informed and if they feel they are not in the position to have a certain degree of control over (future) firm performance, it is far more costly for management to finance firm activities with external than internal funds. Lenders charge high(er) interest rates on both short- and long term financing in case reliable firm insight lacks. Managers therefore prefer internal generated cash

(internal equity) above debt and equity financing. This is called the ‘pecking order theory’, described by Myers in 1984. Secondly, information asymmetry increases agency costs. Agency costs mostly arise when ownership (shareholders) and control (managers) is separated. There may exist conflict

between shareholders and management as their interests vary. Shareholders want management to increase shareholder value, while management enrich themselves with company money. Therefore investors try to align their interest with that of management by means of certain corporate

governance mechanisms. One of those corporate governance mechanisms is debt, in addition to more direct instruments such as performance bonuses and stock options.

The governance role of debt comes from (i) the threat of bankruptcy, (ii) the reduction of cash flows and firm value and (iii) monitoring by creditors (Armstrong et al., 2010). Debt serves as a bonding or commitment instrument as payments are fixed ex ante and can (usually) not be circumvented by management. This makes it more plausible that management will not expand its empire by reinvesting profits unwisely (Czarnitzki and Kraft, 2009). In other words, debt is assumed to discipline managers in such a way that retained earnings are spend well. This potentially limits overinvestment in unprofitable projects and unnecessary working capital investments (i.e. large inventory).

In this way the effect of debt financing on a firm’s value is twofold: (i) Debt provides a tax shield limiting tax payments to the government

(ii) Debt may have a disciplining effect by creating intentionally scarcity in a firm’s resources and cash availability limiting overinvestment.

17

Figure 9 – Debt financing and firm value

For relatively high leveraged firms financing may not be readily available at a reasonable cost.

Management does then rely heavily on internal resources to finance their daily operations and future investments. In case a firm is financed close or in line with its optimal debt-to-equity ratio it will be more difficult for management to get access to external credit funding. This is why I assume that in such circumstances management is more disciplined and inclined to release cash by e.g. lowering their working capital investments (e.g. less inventory, late payments) then when they are

underleveraged. My first hypotheses is therefore as follows:

Hypothesis 1a: ‘Underleverage’ is positively correlated with the length of the CCC

Companies that are leveraged beyond the optimal capital structure (i.e. debt ratio) are potentially in financial distress. Recent research has found out that those companies are actually worse cash

management performers with a larger CCC. This means that the amount of time cash is tied up within the company’s operational process - inventory, production and sales process - has been increased (Europe Working Capital Survey, 2015). This may even require the company to attract more loans as cash is needed to finance day-to-day operations (e.g. rent, employees). The second hypothesis I would like to test is the following:

Hypothesis 1b: ‘Overleverage’ is positively correlated with the length of the CCC

Firm value Capital structure % Debt % Equity Interests are tax deductible Disciplining effect Lower investments in WC leads to higher ROCE

18

Hence, I expect that companies financed close or in line with its optimal capital structure will have a shorter CCC in comparison to companies that are under- or overleveraged. The figure below shows this relation graphically.

Figure 10 – Debt financing and the length of the CCC

Industry dynamics are considered as an important determinant for a company’s capital structure (Miao, 2005). Industries with high (technology) growth and fixed operating costs are relatively lower leveraged. On the other hand, companies operating in mature industries with high entry costs and a relatively stable cash flow are higher leveraged. Therefore, each industry does have (a different) optimal debt-to-equity ratio minimizing the cost of capital. Though, Miao (2005) found out that there is also inter-industry variation in the extent to what level firms are leveraged. This is explained by firm-specific factors, such as the company’s tax benefits of debt, the firm-specific associated bankruptcy and agency costs, production choices and its growth trend.

The company cost of capital is usually estimated as a WACC, which is the average rate of return demanded by investors in the company’s debt and equity (Principles of Corporate Finance, 2014). The height of the WACC is influenced by the mix of debt and equity, in other words, the company’s capital structure. The optimal mix is determined by the cost of each component weighted by its relative market value. The cost of equity represents the compensation that the market demands in exchange for its investment in the company and the risk it perceives for owning a share of the company financed by its private equity. This is highly influenced by the risk profile of the company’s assets, generating a stable or less stable cash flow for its owners within a specific industry. Though, shareholders can only claim a portion of the cash flow after all the other stakeholders including the debtholders are paid. Hence, the degree to which a company is financed with debt does influence the

CCC

Leverage Optimal capital structure

19

height of the residual cash flow for its shareholders. Up to a certain point, an extra portion of debt does decrease the WACC as a result of the tax-deductible interest. Though, issuing more debt means that more interest is paid to debtholders before shareholders are paid in e.g. the form of dividends. So, the increased interest and debt repayments influences the volatility of the dividend payments to the equity holders. On a certain moment this does increase the cost of equity, and hence the WACC. The optimal debt-to-equity ratio is however not fixed and can vary over time due to industry or other global economic changes (Principles of Corporate Finance, 2014).

The following two sub-chapters will describe each specific funding source – equity and debt – and in particular its relation to working capital management and the length of the CCC.

2.5 Private equity versus public equity

Private equity parties are known for their aggressive financing and employment cuts. They are easy targets for the never-ending debate whether PE ownership is a healthy form of ownership and whether they contribute positively to the economy. Their behavior is often viewed as rather

opportunistic. PE-returns are assumed to be mainly driven by excessive debt levels as newly acquired companies are often levered to the maximum, increasing their return on equity (‘ROE’).

Though, Jensen’s article published in 1989 was one of the first articles claiming that PE is one of the best forms of ownership (Nordstrom and Halvarsson (2014)). It decreases the agency problem between owners and management and the ‘free rider problem’, causing the lack of monitoring within public companies. According to Jensen (1989) those problems are solved through private ownership in combination with high leverage and powerful incentives (e.g. stock options), stimulating

operational efficiency, employee productivity, and in the end firm value.

The relatively high debt levels ensures a large part of the (excess) cash is spend to interests and debt repayments. This may be one of the main reasons that PE shareholders are more inclined than public shareholders to monitor and influence management’s decisions to run the company as efficient as possible (bankruptcy threat). Often PE shareholders force management to reduce costs continuously, for which they are hold accountable on a monthly or even weekly basis (Kiechel, 2007). This reduces management’s discretion and independent decision power, though simultaneously, their ability to waste free cash flows (Weir et al., 2015).

Weir et al. (2015) have further found out that public firms going private show significant

improvements in financial health (e.g. level of working capital and liquidity) in the post deal years relative to the year before going private. This corresponds to the results of a survey conducted by Katz (2007) to the relation between ownership and financial performance. They gave as a plausible

20

explanation that private equity firm’s specialist expertise in monitoring may enable timely corrective actions to improve performance and reduce the likelihood of firm failure. Moreover, in contrast to public company shareholders, private equity shareholders are usually more closely involved in (strategic) management decisions as they take a board seat and/or are able to agree specific contractual restrictions on the behavior of management (Thompson et al., 1992).

The results of Weir’s research are further supported by additional research papers (e.g. Nordstrom and Halvarsson (2014), Baduneko (2010), Bergstrom (2007), Smith (1990) and Jensen (1989)) showing that the operational and financial results of PE firms are actually increased after they have been acquired. Jensen (1989) and Smith (1990) found evidence that operating changes and better working capital management were important drivers for the improvement in financial results. Furthermore, PE shareholders are considered to be more focused on cash flow than on earnings reported for accounting purposes (Kiechel, 2007). The latter is on the other hand actively managed by publicly held firms and considered as an important tool to inform the market about the firm’s

performance and financial position. Management of public firms may therefore be less focused on cash, and hence on the company’s working capital position and internal processes, than management of PE led firms.

Based on previous research results and the way PE firms control their portfolio companies, I believe firms held by PE will have a shorter CCC than public firms. Therefore the third hypotheses is the following:

Hypothesis 2: Private equity is negatively correlated with the length of the CCC

In the qualitative part of my research I will try to find out whether management of PE portfolio companies are indeed more focused on WC optimization as a result of being PE owned instead of publicly or family owned. And if so, what is the important driver behind this? Is this the result of the company’s relatively high debt position, the importance of cash relatively to other financial metrics such as EBITDA and net income, or just something else?

In addition, I would like to know whether the focus on WC optimization did actually lead to more WC improvement initiatives and, according to the perspective of management, to a lower WC position. Those research questions has led to the following hypotheses tested via a digital questionnaire (see appendix 3 for the total question list) that will be send to a number of CFOs and finance directors of PE portfolio companies:

21

Hypothesis 2a: Management of PE portfolio companies are more inclined to focus on WC

optimization under the ownership of an investment company than if the company is owned by other type of shareholders

Hypothesis 2b:Management of PE portfolio companies have a stronger focus on cash than other financial metrics such as EBITDA and net income as their shareholder is predominantly interested in cash

Hypothesis 2c: Management of PE portfolio companies have a strong focus on WC optimization due to a relatively high leverage ratio

Hypothesis 2d: Management of PE portfolio companies with a strong focus on WC optimization consider the level of WC lower than management of PE portfolio companies with little to no focus on WC optimization

2.6 Bank debt versus public debt

When a firm raises debt capital, it will enter into informal and formal contracts with its creditors. Formal contracts include details such as the amount borrowed, the interest rates, the maturity date of the loan and the covenant thresholds (Armstrong et al., 2010). Lenders put covenants in place to protect themselves against borrowers which will default on their financial obligations or have a high probability to go bankrupt. Covenants between banks as lenders and corporations can be very firm-specific and contain both qualitative agreements like key employees retention, as more quantitative elements such as working capital requirements and the firm’s debt ratio. Loans are often issued based on the agreement that companies adhere to certain cash ratios, such as the current ratio (current assets -/- current liabilities) and quick ratio ((current assets -/- inventory)/current liabilities). This provides the lender some control over its outstanding loans as their debtors are forced to maintain a certain level of cash to pay off its interest and debt obligations. Informal contracts refer to informal relationships between lenders and borrowers which influence management’s operational and strategic decisions due to the reputation the firm has established with respect to financial transparency, corporate governance and risk management (Armstrong et al, 2010).

If debt is more concentrated, the lenders have clear incentives to monitor managers, as given the large amount of capital at risk the return from the monitoring activity is significant. With a large capital stake in play, it can pay-off to put some more effort on monitoring and disciplining

management (Czarnitiski and Kraft (2009), Denis and Mihov (2002)) and moreover, there are less debtholders which diminishes potential free-rider behavior among the lenders. Therefore I expect that firms financed by bank debt subject to bank covenants and bank supervision are more likely to

22

comply to formal and informal contracts than firms financed by public debt. This may potentially lead to a lower working capital level than public financed companies operating within the same industry. In their study to determinants of debt sources as corporate borrowings Denis and Mihov (2002) make a distinction between three types of debt, which are as follows:

1. Bank debt

2. Non-bank private debt (private placement from informal investors) 3. Public debt (bonds)

They found out that the credit quality of the issuer is an important determinant of the debt source a corporation is able to issue. Firms with the highest credit quality borrow from public sources, firms with medium credit quality borrow from banks, and firms with the lowest credit quality borrow from non-bank private lenders. Bank and non-bank private debt are safer and cheaper funding sources than public debt in case firms are confronted with higher levels of information asymmetry between them and the capital markets, and there exist a higher probability of default. This often means that lower profitable and less-efficient companies have fewer opportunities to issue public debt. As described in sub-chapter 2.3.3 profitability is negatively correlated to the length of the CCC, claiming the opposite of what has previously been hypothesized, that firms financed by public debt have a lower instead of larger CCC than firms financed by bank debt.

Based on theoretical discussion above I have formulated the following hypotheses:

Hypothesis 3: There exists a difference in the length of the CCC for companies financed by public or private debt (bank debt)

23

3.0 RESEARCH METHODOLOGY

3.1 Introduction

This chapter will discuss the data set, sample, (control) variables and statistical techniques used in this research. The subsequent chapter will analyse and discuss the results following from different statistical tests.

3.2 Data set and sample

The data used in this study was acquired from the database of S&P Capital IQ, which is a worldwide leading provider of software, data and analytics to the financial services community. The database offers financial information of both public and private companies, operating in different industries all over the world. The period covered by this research extends from 2009 until 2013 (5 subsequent years). Financial data from 2014 was not complete and as such excluded from the sample. The sample includes Western European companies, located in Spain, UK, Switzerland, Denmark, Ireland, Finland, Sweden, Italy, Greece, Portugal, Germany, France, Luxembourg, Norway, The Netherlands, Belgium, Austria, Iceland and Monaco. For simplicity reasons to run statistical tests and due to geographical similarities the companies are clustered in the following regions:

(1) Benelux – The Netherlands, Luxembourg and Belgium (2) Scandinavia – Norway, Sweden, Finland and Denmark (3) UK – Ireland and the UK

(4) Central Europe – Germany, France, Austria and Switzerland (5) South Europe – Italy, Portugal and Spain

Iceland, Greece and Monaco are excluded from the sample due to size and differences in the economic situation over time (91 cases). France is included in ‘Central Europe’ instead of ‘South Europe’ as the length of the CCC, including its components, were more similar to that of Germany, Austria and Switzerland than to that of Italy, Portugal and Spain.

Firms from the following industries are included:

(1) As primary industry Healthcare, including sub-industries Biotechnology, Health Care Technology, Healthcare equipment and Supplies, Healthcare Providers and Services, Life Sciences Tools and Services and Pharmaceuticals.

24

(2) As primary industry Materials, including the sub-industries Chemicals, Construction materials, Containers and Packaging, Household Products, Machinery, Metals and Mining, Paper and Forest products

(3) As primary industry Telco, including the sub-industries Diversified and Wireless Telecommunication Services

(4) As primary industry Industries, including the sub-industries Construction and Engineering, and Transportation and Logistics

(5) As primary industry Consumer goods, including the sub-industries Durables and Apparel, and Retail

The dataset contained originally 5,248 firms. Firms with incomplete data and outliers are excluded from the sample which is mandatory to run regression analyses. 3,834 companies with incomplete data and 648 companies with (extreme) outliers were eliminated. As a result the sample included 766 firms. The number of eliminations were so high as this study includes 9 different numerical variables taken over a period of 5 subsequent years. Only a complete data set without any outliers of missing data for both the public and PE companies were considered to be useful to run the statistical tests. Individual cases were considered to be an outlier based on their Z-score. This is a statistical

measurement demonstrating the relationship to the mean in a group of Z-scores. This enables researchers to identify whether a case is typical or atypical within a particular set of data. Cases with Z-scores above or below 3 were excluded from the sample of this research.

3.3 Analyses used in this study

In this study we have conducted three types of data analysis:

(1) Descriptive analysis, providing detailed information about the dependent, independent and control variables for the complete sample, as well as clustered per industry and region. This analysis provides insight in the relevant aspects of the CCC and detailed information about the variables used in this study.

(2) Quantitative analysis, using the following statistical tests:

(i) the Pearson correlation analysis to measure the degree of association between the variables. This test does not yet prove any causal relationship between the variables. I have used the Pearson correlation analysis for the variables that are measured on a ratio scale. (ii) One-Way ANOVA analysis (ANalysis Of VAriance) to investigate significant differences

25

among industries and countries in terms of length of the CCC. Industries and countries are seen as different ‘ groups’ and are categorical variables.

(iii) Hierarchical Multiple Regression analysis to find out if there exists a causal relationship between the variables used in this study. Hierarchical Multiple Regression analysis is a variant of the basic Multiple Regression analysis, enabling the researcher to find out whether the addition of multiple variables does actually explain more of the variance in the dependent variable. Moreover, it makes it possible to find out whether there are any moderating effects affecting the influence of the independent variables on the dependent variable. For this study I would like to moderate for the variable ‘Industry’.

The following multivariate regressions are executed:

Hypothesis 1a: CCC = α + β1Underleverage + β2Profitability + β3Firm size + β4Growth +

β5Industry + β6Geographical location + ε

Hypothesis 1b: CCC = α + β1Overleverage + β2Profitability + β3Firm size + β4Growth +

β5Industry + β6Geographical location + ε

Hypothesis 2: CCC = α + β1Ownership + β2Profitability + β3Firm size + β4Growth +

β5Industry + β6Geographical location + ε

Hypothesis 3: CCC = α + β1Bank debt + β2Profitability + β3Firm size + β4Growth +

β5Industry + β6Geographical location + ε

(3) Qualitative analysis in the form of a digital questionnaire which has been sent to management of a number of PE portfolio companies. This research should further

substantiate the quantitative analysis testing hypothesis 2. I would like to find out in practice whether management of PE portfolio companies are indeed more focused on WC optimization as a result of being PE owned instead of publicly or family owned. And if this is the case, what is then the important driver behind this (e.g. debt position, focus on cash rather than other financial metrics)? Appendix 3.0 contains the question list.

3.4 Definition of variables used in this study

The variables used in this study consist of a dependent variable, 4 independent and 5 control variables.

26

Cash Conversion Cycle (CCC): CCC = DIO (Days Inventory Outstanding) + DSO (Days Sales

Outstanding) – DPO (Days Payables Outstanding).

(1) DIO is calculated as follows: Average inventory year x / (Cost Of Goods Sold year x /365). (2) DSO is calculated as follows: Average accounts receivables year x / (Sales year x / 365). (3) DPO is calculated as follows: Average accounts payables year x / (Cost Of Goods Sold year x / 365).

The CCC, DIO, DSO and DPO are calculated in the number of days. Independent variables:

Overleverage and underleverage: A company is considered to be overleveraged, respectively

underleveraged when its capital structure deviates from the capital structure industry mean. Capital structure is calculated as ‘Total Debt’ / ‘Total Equity’. For instance, when the capital structure industry mean is 20%, then a company is underleveraged with 5% in case its debt-to-equity ratio is 15%, and overleveraged with 5% in case of a debt-t0-equity ratio of 25%. Debt includes e.g. short-term borrowings, current portion of long-short-term debt, current portion of capital leases, long-short-term debt, capital leases and financial current and non-current debt. Total equity includes e.g. total preferred equity, total common equity, total minority interest. Both variables are measuring the relative distance from the average debt-to-equity ratio industry wise.

Bank debt: Includes the portion of bank debt as percentage of the total debt. Bank debt is defined as

outstanding balance for revolving credit plus outstanding balance for term loans plus outstanding balance for federal home loan bank borrowings. The variable measures in percentages into what extent the company is levered with bank debt.

Ownership: This variable makes a distinction between private and public equity. Private equity

companies are owned by financial sponsors such as NPM Capital and HAL Investments. Public companies are listed on stock exchanges such as the AEX, Copenhagen Stock Exchange and the London Stock Exchange. The variable is transformed into a dummy variable, which is set at 1 when firms are backed by a public investor and 0 otherwise.

Control variables:

The definition of the control variables are based on previous literature described in chapter 2.

Profitability: Measured as Return on Assets (ROA) in percentages, calculated by ‘Net Income / Total

Assets’. Net income is calculated as net revenues minus operational expenditures, depreciation, interest, taxes and other expenses. Total assets include total current assets, net PPE, long-term investments, goodwill, other intangibles, loans and leases, and other assets.

27

Firm size: Firm size is measured by the natural logarithm of assets, expressed in millions of euros. Growth: Growth refers to the revenue growth and is therefore calculated as (this year’s sales -/-

previous year’s sales)/ previous year’s sales, which is also expressed in percentages.

Both Industry and Geographical location are described above in chapter 3.2. The variables are transformed into several dummy variables. Each dummy is equal to 1 if the firms operates in the corresponding sector or industry, and 0 otherwise. For the variable ‘Industry’, the Healthcare sector is selected as the reference category. For the variable ‘Country’, the Benelux serves as the reference category.

28

4.0 QUANTITATIVE DATA ANALYSIS AND DISCUSSION

4.1 Descriptive statistics

Table 1 shows the descriptive statistics for the total sample of this research for a period of 5 years from 2009 until 2013 (3,830 year observations). The overview presents the mean, the standard deviation, minimum and maximum of the different variables of interest in this study.

The average CCC is 74 days, with a standard deviation of 62 days. On average, companies are paid within 54 days, which is stretched out to a maximum of 149 days. Companies take on average 69 days before they will pay their suppliers. The relatively high average number of inventory days (90 days) is mainly influenced by high inventory levels in the Consumer goods sector, as well as the Healthcare and Materials industries (see below table 2).

The average DSO and DPO of this research sample are more or less in accordance with the DSO and DPO of the Annual Global Working Capital Survey PwC (2015), being respectively 47 and 68 days for Western European companies. Yet, on overall the CCC differs with 38 days (74 versus 36 days), as the DIO of this sample counts 36 days higher to a total of 90 days.

29

The descriptive statistics per Industry can be found in table 2. The CCC of the industries Healthcare and Consumer goods are driving the relatively high average CCC within this sample due to a high DIO ratio (respectively 95 and 109 days). The sector specific results of the Annual Global Working Capital Survey PwC (2015) differ with the results of this survey. This may be caused by a few reasons:

1) This research sample focuses on Western European companies, while the PwC survey includes companies worldwide

2) The industries and sub-industries examined in this study do not completely correspond to the industries within the PwC survey (potentially another categorization of sub-industries) The industries Materials, Industries and Telco are materially higher leveraged than the Healthcare and Consumer goods industry. This can be explained by the fact that those industries are relatively mature, with high entry costs and are defined as ‘asset heavy’, serving as collateral for the outstanding loans.

Table 3 provides an overview of the descriptive statistics per Region for the period 2009 until 2013. Central Europe has the largest CCC (82 days). This is mainly influenced by a high DIO, while

companies operating within the Benelux have a materially lower CCC with an average duration of 62 days. South European companies are significantly higher leveraged compared to companies within

N PE Public CCC DSO DPO DIO ROA (%)

Sales growth (%)

Capital

structure (%) Bank debt (%) Assets (€m)

Healthcare 140 57 83 87 60 68 95 6.7% 7.9% 41.3% 36.5% 384 Materials 220 131 89 74 61 78 92 4.3% 4.0% 54.5% 29.6% 337 Consumer goods 295 139 156 86 44 66 109 5.4% 6.4% 43.8% 33.9% 303 Industries 98 47 51 34 60 59 32 3.9% 5.4% 55.7% 34.7% 555 Telco 13 2 11 -10 55 71 7 5.4% 2.1% 52.8% 56.6% 836 N = 766 N PE = 376 N Public = 390

Table 2: Descriptive Statistics per Industry

30

the other regions. Beforehand I did expect that South European companies will have the largest CCC, as that specific region has been affected the most during crisis time, resulting in high DSO and DIO levels. In comparison to the other regions the DSO and DIO are indeed high. Though, the CCC is relatively low (70 days) due to a very high DPO ratio, counting to 114 days. This is significantly higher than the DPO of higher regions.

The Telco industry did only count 13 companies after the elimination of outliers from its original sample (N = 33). As a result the size became too small to get any reliable quantitative results for this industry. I will therefore not incorporate Telco companies in the following quantitative analyses.

4.2 Pearson correlation analysis

Table 4 presents the Pearson correlation coefficients of all dependent, independent and control (scale) variables for the period 2009 until 2013. The individual WCM measures of the CCC are included as additional dependent variables (DSO, DPO and DIO). The analysis contains 753 companies for the following industries:

1. Healthcare 2. Materials

3. Consumer goods 4. Industries

A smaller sample has been used to analyse the relation between ‘Bank debt’ and the other variables (N = 379). This sample includes merely financial data of public companies as the full sample,

31

including PE companies, contains no to limited information about the percentage of outstanding bank debt in PE companies. For public companies, on the other hand, the information was readily available and therefore useful for this analysis.

The degree of correlation between two variables depends on its size represented by its coefficient (r). When r > 0.5 the effect is considered to be large. For this specific sample this only counts for the relation between the CCC and DIO, Firm size and Ownership (public companies are larger in terms of assets), and Firm size and Bank debt (larger companies are more heavily financed with bank debt than other sources of debt).

The results show a significant relation between the measures of the CCC (DSO, DPO and DIO) and the CCC. This is not unexpected as the individual WCM measures in- or decrease the CCC. The DIO and DSO are therefore both positively correlated with the CCC, while the DPO is negatively

correlated.

The results show further that overleverage is significantly negatively correlated to the CCC and DPO. This means that, against my expectations, an increase in leverage beyond the optimal capital

structure ratio may decrease the DPO and therefore shortens the length of the CCC. On the other hand, the variable underleverage is significantly negatively correlated to the CCC. Hence, the less a company is financed with debt the higher the CCC. From both observations we can carefully draw the preliminary conclusion that leverage itself may be negatively correlated with the CCC. Though, this hypothesis should be tested per separate industry as the extent to which companies are leveraged is highly influenced by the industry itself. However, the current sample per industry used in this study is unfortunately not large enough to get reliable results.

There is a positive relationship between ownership and the CCC, which may be consistent with my view that public companies have a larger CCC than PE companies. PE companies seem to shorten the CCC by increasing the payment terms to its suppliers as there is a negative correlation between DPO and ownership. On the other hand public companies are confronted with a larger DIO (r = .133 between DIO and ownership). Ownership is further negatively correlated with the variable

overleverage, which seems plausible as in general public companies are less (over)leveraged than PE companies.

There is a significant, though no strong positive relation between the ROA and the CCC, meaning that more profitable companies have a longer CCC. This could be explained by the significant negative relation between DPO and ROA (r = -.137), which indicates that less profitable companies wait longer to pay their creditors.

32

4.3 One-way ANOVA analysis

The one-way ANOVA analysis with Duncan test from Post-Hoc tests was conducted to find out whether there exists significant differences among industries and countries in terms of length of the CCC. Though, this test was merely applicable to use to investigate differences in the length of the CCC for the variable ‘Country’ as the test ‘Homogeneity of Variances’ meets the required homogeneity assumption (all groups have the same or similar variance).

The ANOVA results show a relatively low F-value, though > 1 (2,217; see table 5). This means that we can conclude that among the countries studied in this sample there appear differences in terms of the CCC length. However, this result is only significant at the 0.1 level, which means that there is a 10% chance that the differences found are driven by coincidence.

The Post-Hoc test shows further that only the length of the CCC of companies within the Benelux compared to Central European companies significantly differ at the 0.05 level (see table appendix 1). This probably mean that the variance in the CCC will hardly be explained by the differences in

countries used in this sample. This is not surprising as only Western European countries are included in this study. Those countries may share payment similarities due to, among others, extensive trade relations within the European Union.

As for the variable ‘Industry’ the homogeneity assumption requisite has not been met to use the one-way ANOVA test, I had to conduct the Kruskal-Wallis test. This is the nonparametric alternative to the one-way ANOVA test. Based on those results the null hypothesis claiming that the industry samples are equal to each other can be rejected (level of sig. > 0.05). Therefore the CCC of the different industries are not equal to each other, which means that the industries used in this sample will probably explain a part of the variance in the length of the CCC (see table appendix 2).

4.4 Regression analysis

To identify the determinants of the CCC I have conducted two hierarchical multiple regression analyses. Table 6 presents the results of the first regression. Model 1 (the baseline model) contains the moderator and control variables. Model 2 includes the independent variables ‘Overleverage’, ‘Underleverage’ and ‘Ownership’. The independent variable ‘Bank debt’ is excluded from the analysis

33

in table 6, as this sample contains both public as PE companies. Table 8 presents the outcome of the hierarchical multiple regression analysis for ‘Bank debt’ and the CCC. In both regressions model 3 investigates the moderating effect of ‘Industry’.

The R Square (coefficient of multiple determinations) is showing the percentage of the variance in the dependent variable (CCC) explained by the independent variables. The control and moderator

variables in model 1 explains 10% of the variance in the CCC, which is significant at the 1% level. The addition of the independent variables enriches the model with 4%, explaining 14% of the variance (significant at 1% level). This percentage is consistent with other studies to WCM (Chiou and Cheng, 2006; Deloof, 2003 and Hill et al., 2010) varying between the 5% and 20%.

The third model does not explain any further the variance in the CCC, contributing only 1% to 15, though, not significant at a 1%, 5% nor 10% level. Therefore the results following from this model are not sufficiently reliable to draw generic conclusions which are applicable for the whole population. Hence, the interaction effect of ‘Industry’ is not significantly related to the independent variables ‘Overleverage’, ‘Underleverage’ and ‘Ownership’, meaning that the expected influence of the independent variables on the CCC is not stronger in one specific industry.

The proposed positive effect of ‘Underleverage’ and ‘Overleverage’ on the CCC (respectively

hypothesis 1a and 1b) is not supported by model 2 as their coefficients (b = -4.53 and 6.88) are not significantly related to the CCC at 1%, 5% or 10% level. The results may therefore not be used to draw a generic conclusion, which leaves only room to discuss here the results for this specific sample. Contrary to my expectation the direction of the ‘Overleverage’ coefficient is negative meaning that overleveraged companies within this sample do have a shorter instead of longer CCC. The CCC of overleveraged companies is approximately 4.53 days shorter counting to 69.5 days compared to the average CCC of 74 days (see descriptive statistics of table 1). As mentioned before, this could mean that ‘leverage’ as such may shorten the CCC, even when companies may be in a distressed situation by overleveraging their balance sheet. This may suggest that in those situations managers still focus, or even strengthen their focus on increasing their DPO and decreasing their DIO and DSO. Though, the direction of the ‘Underleverage’ coefficient does correspond with my expectation that underleveraged companies do have a larger CCC, counting to 81 days (b = 6.88).

Following from these results, figure 6 on page [x] is adjusted towards the found relationship between the CCC and ‘leverage’ described above (see below).

34

Figure 11 - Debt financing and the length of the CCC

Hypothesis 2, proposing a negative relationship between the CCC and PE is significantly supported by model 2 (b = 25.34, p < 0.01). I therefore accept hypothesis 2 that PE owned companies have a

shorter CCC than public companies. The information in table 6 shows us that the CCC of public companies is approximately 25 days longer than PE owned portfolio companies. Table 7 supports this finding by showing the difference in the length of the CCC per industry.

CCC

Leverage Optimal capital structure

CCC

Leverage Optimal capital structure

35

The information from table 7 shows further that the public companies within this sample are larger than the PE portfolio companies, measured in terms of assets. Though this will not explain our finding that PE owned companies have a shorter CCC since the control variable ‘Firm size’ is not sufficiently reliable explaining a portion of the variance in the length of the CCC ( p > 0.001, 0.05 and 0.1). This result is supported by previous research, which was also not conclusive about the effect of firm size on the CCC (Zariyawati et al., 2010; Attari and Raza, 2012; Chiou and Cheng, 2006).

Model 1 Model 2 Model 3

Main effects Overleverage _{-4.53 -3.95} Underleverage _{6.88 13.19 *} Ownership _{25.34 *** 24.20 ***} Interaction effects Overleverage * Materials _1.93

Overleverage * Consumer goods _4.70

Overleverage * Industries _-8.02

Underleverage * Materials _-17.72

Underleverage * Consumer goods _-23.53

Underleverage * Industries _-47.25

Ownership * Materials _4.63

Ownership * Consumer goods _-9.07

Ownership * Industries _-11.45

Moderator and control variables

ROA _{85.76 ** 97.83 ** 99.61 **} Firm size _{2.43 * -2.20 -1.75} Growth _{-36.91 * -34.06 * -38.16 **} Materials _{-10.25 -6.61 -7.79} Consumer goods _{-0.03 0.87 -0.88} Industries _{-52.36 *** -48.96 *** -50.07 ***} Scandinavia _{14.63 * 10.26 9.28} UK _{8.99 10.00 8.65} Central Europe _{16.87 ** 15.57 ** 15.11 **} South Europe _{3.78 7.09 6.90} R Square _{0.10 *** 0.14 *** 0.15} ∆ R Square _{0.041 ** 0.01} CCC

***. Correlation is significant at the 0.01 level (2-tailed). **. Correlation is significant at the 0.05 level (2-tailed). *. Correlation is significant at the 0.1 level (2-tailed).

Table 6: Hierarchical regression analyses

The table below provides an overview of the effects of Over-, Underleverage, Ownership and Industry on the CCC.

36

In addition, table 7 provides the capital structure ratio for both PE and public companies. In the Healthcare and Consumer goods industry the PE owned portfolio companies are relatively highly leveraged (46% versus 47%; 38% versus 49%) considering their firm size in proportion to those of the public owned companies. Hence, as previously suggested, the extent to which a company is leveraged may probably influence management to have a strong focus on WCM for cash reasons. The qualitative results should further find out whether this explanation holds.

Hypothesis 3 states that there exists a difference in the length of the CCC for companies financed by public or private debt. Based on the regression results provided in table 8, this hypothesis must be rejected. Both the R squared and the coefficient of ‘Bank debt’ is nog significant at the 1%, 5% and 10% level, following from model 2. Therefore it is not possible to demonstrate statistically significant that a specific type of debtholders, influence the length of the CCC, both negatively or positively. For this particular sample it holds that the CCC of companies financed with bank debt is larger with

Industry Variable PE Public

Healthcare CCC 67 100 DSO 24 64 DPO 45 64 DIO 59 100 Capital structure 46% 47% Firm size 99 595 Materials CCC 62 92 DSO 64 56 DPO 86 66 DIO 85 102 Capital structure 45% 69% Firm size 106 678 Consumer goods CCC 76 94 DSO 45 42 DPO 67 66 DIO 98 118 Capital structure 38% 49% Firm size 74 508 Industries CCC 28 39 DSO 60 61 DPO 65 54 DIO 32 32 Capital structure 44% 67% Firm size _{96 978}

Table 7: Descriptive statistics

The table below provides an overview of the Capital structure, Firm size and CCC per Industry.