The effect of NWC on firm performance: A comparison of manufacturing firms within
the European Union
Lisa Sikking - S2355817
University of Groningen, Faculty of Economics and Business MSc International Financial Management
Supervisor: dr. J.H. von Eije
June 9, 2017
Abstract
This paper examines the effect of net working capital on firm performance for a sample of listed manufacturing firms for the 28 countries within the European Union. Based on previous research this study hypothesizes that 1) NWC ratio has an effect on firm performance and 2) There is an optimal ratio of NWC. The results show that there exists a significant linear relation between NWC and firm performance for some developed countries and for emerging countries. The results demonstrate that this linear relation may be positive or negative. The results show the existence of an optimal ratio of NWC, a minimum, has been found. Overall the findings show, that the relation between NWC and firm performance is unique for every country and possibly for every firm.
2 1.Introduction
This paper analyses the influence of the ratio of net working capital (NWC) on firm performance for countries within the European Union (EU). In this study NWC is defined as inventories plus accounts receivables minus accounts payables to sales at the end of each year (Aktas, Croci and Petmezas, 2015).
NWC ratio insights are very valuable for firms. According to the latest report of EY (2016), there is a tremendous amount of unnecessary cash tied up in NWC. The leading 1.000 European companies hold between €280 billion and €480 billion of cash unnecessarily tied up in NWC in 2015, which is between 4% and 7% of their aggregate sales. The reduction of NWC increases the amount of free cash flow, which could be useful for internal financing. This is in particular an advantage for managers and shareholders when external financing is expensive or when access to the capital market is difficult (Kieschnick et al. 2011). From the agency cost point of view excess NWC is an unprofitable investment, which may give managers slack, but reduces the pay-outs to shareholders (Eckbo and Kisser, 2013).
In the literature scholars find consensus on the idea that NWC has an effect on firm performance (Baños-Caballero et al. 2012). However, there are two contradicting views on this relationship. One of the views focusses on the positive effect of a higher ratio of NWC on firm performance. These positive effects, amongst other effects, include higher sales and earnings (Deloof, 2003), a reduced risk of price fluctuations (Blinder and Maccini, 1991), larger inventories result in a lower risk of stock-outs (Corsten and Gruen, 2004) and by keeping accounts receivables low the firm can increase firm performance by gaining early payment discount provided by suppliers (Deloof, 2003). However, the other view focuses on a negative relationship between NWC and firm performance, as the problem of overinvestment may occur. Additional investment in inventories raises costs such as warehouse rent, security and insurance expenses (Kim and Chung, 1990). The negative effects may also be the result of extra costs related to additional financing, as this involves transaction costs, asymmetric information (Opler, Pinkowitz, Stulz and Williamson, 1999) and agency costs (Stiglitz and Weiss, 1981). These contradicting views led to the idea that a nonlinear relation between NWC and firm performance might exist, shaped as an inverted U-shape relation. As firms may increase their firm performance by moving closer to the optimal ratio of NWC (Baños-Caballero et al. 2012;2014; Aktas et al. 2015).
3 and emerging countries. In literature, there is hardly any information on which ratio of NWC generates the best firm performance. So far, an optimal ratio of NWC, which is researched in three articles (Baños-Caballero, Garcia-Teruel, Martinez-Solano, 2012; 2014; Aktas et al. 2015).
From a managerial point of view, this paper may be valuable, first, for firms that want to improve their firm performance. This paper may be used as reference material, to compare their NWC ratio with the country average, as firms may have too much or too little NWC. Second, the optimal ratio of NWC of a country may be an incentive for managers to not internationalize to a particular country, such as when the optimal NWC ratio in the target country is high while attracting external finances is difficult.
In this paper I will make use of manufacturing firms listed on the stock exchange in the EU. The focus is on countries within the EU to be able to make relevant comparisons. Countries within the European Union have to oblige to the EU objective. First of all, members should be democracies that respect the rule of law and human rights. Second, these countries must conform to the open-market with the free competition principle. Lastly, all EU members have to adopt the EU acquis, which embodies the accumulated legislation. So, members of the European Union have to accept the political, economic and monetary aims (Schimmelfennig, 2001). These similar aims make it is easier to compare the countries within the sample. The sample is then divided on the basis of the gross domestic product (GDP) per capita. Based on the benchmark, the sample is divided into 11 developed and 17 emerging countries. Furthermore, the dataset will cover a six year period from 2011 to 2016. The methodology part of this paper makes use of the panel data methodology to estimate the models.
The results indicate that there exists a linear relation between NWC and firm performance within the EU. However, the results show the existence of a negative relation (Finland, France, UK, Greece and Poland) as well as a positive relation (Ireland and Italy) in both developed and emerging countries. The robustness check demonstrates that there exists an optimal ratio of NWC, a minimum, upon the linear relation in Finland and France. Overall, the results indicate that the relation between NWC and firm performance is unique for every country and possibly every firm.
The remainder of this paper is structured as follows. Section 2 will discuss the relevant findings within the literature on the relationship between NWC and firm performance and the influence of the differences between developed and emerging countries on this relation. In section 3, the empirical model and the data will be described. Section 4 provides the results and the robustness checks. Finally, Section 5 concludes the paper.
2.Literature Review
4 performance. Moreover, this section sheds light on the optimal ratio of NWC. The positive-, negative- and optimal relation will be applied to developed and emerging countries.
2.1 The relation between NWC and firm performance
Under perfect financial markets, NWC management would have no influence on firm performance, as obtaining external finance would be no problem and therefore new investments would not depend upon the availability of internal finance (Modigliani and Miller, 1958; Lewellen, McConnell and Scott, 1980). However, in reality perfect financial markets do not exist, hence, managing NWC efficiently is an important aspect of the overall firm strategy to increase firm performance (Shin and Soenen, 1998). NWC management involves the management of accounts receivables, inventories and accounts payables. The literature proposes arguments for three types of relations between NWC and firm performance. First, a positive relation between NWC and firm performance (e.g. Mathuva, 2009). Secondly, a negative relation (e.g. Shin and Soenen, 1988; Wang, 2002; Deloof, 2003; Kieschnick et al. 2013). The majority of papers indicate a negative relationship between NWC and firm performance. The most recent articles mention the existence of an optimal ratio of NWC in the shape of an inverted U-shape relation between NWC and firm performance (Baños-Caballero et al. 2012;2014; Aktas et al. 2015).
2.1.1 Positive relation between NWC and firm performance
The positive relation between NWC and firm performance may be the result of positive effects associated with an increase in NWC, such as an increase in sales resulting in a higher firm value (Deloof, 2003). There are several explanations for this positive relation. One of these explanations is that an additional investment in inventories has several benefits for firms with a low ratio of NWC (Aktas et al. 2015). Larger inventories may also help for smoothening production, such as with fluctuating sales in the form of loss of sales due to stock-outs (Mathuva, 2009). Additionally, larger inventories help to speculate on, and hedge against, potential price fluctuations and are held to reduce supply costs by buying more inventory at the same time (Blinder and Maccini, 1991). Furthermore, according to Schiff and Lieber (1974) increasing inventories enable firms to provide better services to their customers and reduce productions costs resulting from fluctuations in production.
5 demand trade credit may encourages buyers to acquire products, because of temporary attractive payment terms (Emery, 1987).
Overall, the positive relation is the result of trade credit and additional inventories for firms with low ratios of NWC as they stimulate sales as well as firm performance (Deloof, 2003).
2.1.2 Negative relation between NWC and firm performance
On the other hand, there are some disadvantages concerning the higher ratios of NWC. The majority of finance literature has shown that there exists a negative relation between NWC and firm performance (e.g. Shin and Soenen, 1988; Wang, 2002; Deloof, 2003; Kieschnick et al. 2013). The negative linear relation may be related to the fact that investments in current assets do not necessarily result in an equivalent increase in profitability (Baños-Caballero et al. 2012). Firm performance may decrease when costs outweigh the benefits of higher investments in NWC (Deloof, 2003). These costs may be related to holding extra stock, such as warehouse rent, security and insurance expenses (Kim and Chung, 1990). Furthermore, for increasing the ratio of NWC additional capital is needed. Costs related to attracting external finance include transaction costs, agency costs and asymmetric information. The transaction costs may involve brokerage costs (Opler et al. 1999). Asymmetric information may raise costs as lenders may limit the supply of additional credit, known as credit rationing (Stiglitz and Weiss, 1981). The agency costs are the result of different interests of shareholders and debtholders. Raising additional finance for investments may be beneficial for debtholders but not for shareholders. Therefore, for highly leveraged firms it may be expensive and difficult to obtain external finance (Opler et al. 1999).
The negative linear relation may also exist between NWC and firm performance due to accounts receivables. A firm might increase its firm performance by lowering their accounts receivables by making it attractive for customers to pay their bills on time. Lowering accounts payables increases the amount of available cash, which may be used for valuable investments to increase firm performance (Mathuva, 2009).
6 flexibility by lowering the NWC ratio (Aktas et al. 2015). Based on evidence provided by previous literature I hypothesize that:
Hypothesis 10: NWC ratio has no effect on firm performance.
Hypothesis 1a: NWC ratio has an effect on firm performance.
2.2 NWC and firm performance in developed and emerging countries
The theoretical line of reasoning used in this paper so far showed that the linear relationship between NWC and firm performance may differ between firms as well as countries as a whole. It may even be valuable to distinguish between developed and emerging countries, as there are distinct differences. In emerging countries firms face larger and more market imperfections, such as underdeveloped legal and financial institutions, compared to developed countries (Gibson and Tsakalotos, 1994; Levine, 1997; Zhang, von Eije and Westerman, 2015). La Porta et al. (1997), compare the legal systems of 49 countries and show the differences in legal rules as well as in law enforcement. Considering the protection of shareholders and creditors, common law countries score high on protecting shareholders and creditors, while civil law countries have a lower score. They show that poorer countries have more trouble with enforcing laws than richer countries. Additionally, the differences in legal environments also influence capital markets. The results show that in developed countries with better legal protections, firms have more access to external finance as capital markets are larger and more capital is available (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 1997).
Late payment forces suppliers to provide trade credit to their customers. According to Tewolde and von Eije (2007), in the case of market imperfections suppliers may be more inclined to provide trade credit and customers to accept trade credit in line with the cooperation theory. However, in emerging countries selfish behaviour may also occur, as it is more difficult to attract external finance. Hence, firms may be forced to behave more selfish by receiving the maximum amount of trade credit from suppliers, while restricting trade credit to their customers (Tewolde and von Eije, 2007). This selfish behaviour provides the possibility to use internal finance for investments.
7 external finance) are the main factors influencing the investment in NWC (Hill et al. 2010; Baños et al. 2014).
2.3 Optimal ratio of NWC
So far, the paper considered the positive and negative relation between NWC and firm performance. However, the literature shows that the positive relation is stronger for firms with a lower ratio of NWC and the negative relation is stronger for firms with a higher ratio of NWC (Aktas et al. 2015). Firms with low ratios of NWC may increase their firm performance by increasing their ratio of NWC until a certain optimum as long as costs outweigh the benefits. Beyond this optimum ratio of NWC firm performance will increase by lowering the ratio of NWC, as an additional dollar of cash is then more valuable than an additional dollar in NWC (Kieschnick et al. 2013; Baños-Caballero et al. 2012). Until now only three articles research the existence of an optimal ratio of NWC (Caballero et al. 2012; Baños-Caballero et al. 2014; Aktas et al. 2015). According to Baños-Baños-Caballero (2012), previous literature already described the existence of an optimal ratio for all components of NWC separately, including accounts receivables (Nadiri, 1969; Emery, 1984), inventories (Ouyang et al. 2005) and accounts payables (Nadiri, 1969). For the literature focusing on the optimal ratio of NWC, Baños-Caballero et al. (2012) are the first to mention an inverted U-shape relationship between NWC and firm performance. They study a sample of non-financial small and medium-sized Spanish firms, while making use of the cash conversion cycle as a measure of NWC and profitability as a measure of operating performance. Their results imply that when a firm moves away from the optimal ratio, firm performance decreases. In a later study Baños-Caballero et al. (2014) study the relationship between NWC and firm performance. The sample consists of non-financial UK firms, where NWC is measured by using the net trade cycle measured as (accounts receivables/sales)*365 + (inventories/sales)*365 – (accounts payables/sales)*365 and firm performance is measured by using the market to book value of assets. They provide strong evidence for an inverted U-shape relation, as the optimal firm value will be balanced by costs and benefits of investments in NWC. Aktas et al. (2015) study the relationship between NWC and firm performance by using excess NWC measured as NWC-to-sales ratio minus the median of the NWC-to-sales ratio of an industry and stock performance as the 1-year excess return in year 𝑡. They provide evidence of a negative relationship between excess NWC and firm performance for firms with abnormally high ratios of NWC and a positive relationship for firms with underinvestment in NWC. Their evidence is comparable to the results of Baños-Caballero et al. (2012; 2014), and is extended by also taking financial constraints into account to be able to assess the effect of unnecessary cash tied up in NWC (Aktas et al. 2015). Based on prior research this study hypothesizes that1:
8
Hypothesis 20: There exists no optimal NWC ratio
Hypothesis 2a: There exists an optimal NWC ratio
3. Model and Data 3.1.1 Linear model
According to a few previous studies, the relationship between NWC and firm performance might be linear. In this paper, hypothesis 1 tests whether NWC has an effect on firm performance. For the model tested, this paper will use the model of Banos-Caballero et al. (2014) as an example. In their paper Baños-Caballero et al. (2014) use the net trade cycle to measure NWC management. However, following other previous studies (e.g. Kieschnick et al. 2013; Aktas et al. 2015) on NWC management, this paper makes use of the concept of net working capital (NWC). NWC is the ratio of inventories plus receivables minus payables against sales (Hill et al. 2010, Kieschnick et al. 2013 and Aktas et al. 2015). To test for a linear relation of NWC, firm performance 𝑄𝑖,𝑡 for firm 𝑖 and time t will be regressed against net working capital (𝑁𝑊𝐶𝑖,𝑡−1) for firm 𝑖 and year t-1. All independent variables are lagged by one year against the dependent variable to control for endogeneity (Aktas et al. 2015), therefore over the period 2011 to 2016, the values of variables in year 2011 will be used for predicting the values of 2012. Other variables are controlled for in the regression model as well, as they might have an influence on firm performance. The control variables included in this regression model are firm size (𝑆𝐼𝑍𝐸𝑖,𝑡−1), the leverage of the firm (𝐿𝐸𝑉𝑖,𝑡−1), growth opportunities (𝐺𝑅𝑂𝑊𝑇𝐻𝑖,𝑡−1) and return on assets (𝑅𝑂𝐴𝑖,𝑡−1). These variables are used to estimate the following fixed effects panel model:
𝑄𝑖,𝑡 = 𝛽0+ 𝛽1𝑁𝑊𝐶𝑖,𝑡−1 + 𝛽2𝑆𝐼𝑍𝐸𝑖,𝑡−1 + 𝛽3𝐿𝐸𝑉𝑖,𝑡−1+ 𝛽4𝐺𝑅𝑂𝑊𝑇𝐻𝑖,𝑡−1 + 𝛽5𝑅𝑂𝐴𝑖,𝑡−1 + λ𝑡+ 𝜂𝑖+ 𝜀𝑖,𝑡 (1a)
where 𝑄𝑖,𝑡 is firm performance, measured as the market-to-book ratio, which is calculated as the ratio of the sum of the market value of equity and the book value of debt divided by the book value of total assets at the end of each year2 (e.g. Agrawal and Knoeber, 1996; Thomsen, Pedersen and Kvist, 2006;
Wu, 2011; Baños and Caballero et al. 2014). Net working capital (𝑁𝑊𝐶𝑖,𝑡−1) will be calculated as the ratio of accounts receivables plus inventory minus accounts payables to sales. So, (𝑁𝑊𝐶𝑖,𝑡−1) measures the ratio of NWC at the end of last year to sales at the end of last year (e.g. Hill et al. 2010; Kieschnick et al. 2013; Aktas et al. 2015). The dependent variable 𝑄𝑖,𝑡 and the independent variable 𝑁𝑊𝐶𝑖,𝑡−1 are the most important variables as they are used for estimating the linear relation between NWC and firm
2 According to Smirlock, Gilligan and Marshall (1984), this variable might be improved by using Tobin’s q, as
9 performance. The model also includes control variables where (𝑆𝐼𝑍𝐸𝑖,𝑡−1) is the lagged firm size, which is measured as the natural logarithm of sales. Leverage (𝐿𝐸𝑉𝑖,𝑡−1) will be measured by taking the ratio of total debt to total assets lagged one year. The lagged growth opportunities (𝐺𝑅𝑂𝑊𝑇𝐻𝑖,𝑡−1) are being calculated as the ratio of the change in net assets last year divided by two years lagged net assets. Return on assets (𝑅𝑂𝐴𝑖,𝑡−1) is the ratio of Earnings Before Interest and Taxes (EBIT) divided by total assets both lagged one year. Furthermore, the time dummy variables λ𝑡 are the same for firms in the same period but differ for firms over time. Economic factors may influence firm performance over time while firms cannot control these factors, therefore these influences are controlled for by time dummy variables. To control for firm characteristics 𝜂𝑖 will be used to control for the unobservable heterogeneity or the firm's unobservable individual effects, so I can control for the particular characteristics of each firm. Finally, 𝜀𝑖,𝑡 is the random disturbance for firm 𝑖 in year 𝑡 (Baños-Caballero et al. 2014).
In equation (1a), the linear relation between NWC and firm performance is tested for all countries that are part of the EU. However, this model will be extended by using a country dummy (𝐶𝐷𝑈𝑀𝑖,𝑡). This dummy will make a distinction between developed and emerging countries based on their average GDP level per head in 2015. Therefore, we will test the following model:
𝑄𝑖,𝑡 = 𝛽0+ (𝛽1+ 𝛿1 𝐶𝐷𝑈𝑀𝑖,𝑡) 𝑁𝑊𝐶𝑖,𝑡−1 + 𝛽2𝑆𝐼𝑍𝐸𝑖,𝑡−1 + 𝛽3𝐿𝐸𝑉𝑖,𝑡−1+ 𝛽4𝐺𝑅𝑂𝑊𝑇𝐻𝑖,𝑡−1 + 𝛽5𝑅𝑂𝐴𝑖,𝑡−1 + λ𝑡+ 𝜂𝑖+ 𝜀𝑖,𝑡 (1b)
where all the variables have the same definition as the variables used in equation (1a). However, the country dummy (𝐶𝐷𝑈𝑀𝑖,𝑡) is included and has a value equal to 1 for firms in emerging countries and firms in developed countries will have a country dummy value equal to 0.
3.1.2 Optimal ratio model
Next to estimating the linear relation between NWC and firm performance, the second hypothesis explores whether there exists an optimal ratio of NWC. To test this hypothesis the same model is being used as in equation (1b). However, this model will be extended by using the square of NWC 𝑁𝑊𝐶𝑖,𝑡−12 and its interaction with the country dummy (𝐶𝐷𝑈𝑀𝑖,𝑡). This dummy will distinguish countries based on their average GDP level per head in 2015. Therefore, we will test the following model3:
𝑄𝑖,𝑡 = 𝛽0+ (𝛽1+ 𝛿1 𝐶𝐷𝑈𝑀𝑖,𝑡) 𝑁𝑊𝐶𝑖,𝑡−1+ (𝛽2+ 𝛿2 𝐶𝐷𝑈𝑀𝑖,𝑡) 𝑁𝑊𝐶𝑖,𝑡−12 + 𝛽 3𝑆𝐼𝑍𝐸𝑖,𝑡−1 + 𝛽4𝐿𝐸𝑉𝑖,𝑡−1+ 𝛽5𝐺𝑅𝑂𝑊𝑇𝐻𝑖,𝑡−1+ 𝛽6𝑅𝑂𝐴𝑖,𝑡−1 + λ𝑡+ 𝜂𝑖+ 𝜀𝑖,𝑡 (2) 3 Term 𝛽
10 where all the variables have the same definition as in model 1b. However, the square of NWC is included in the model as (𝑁𝑊𝐶𝑖,𝑡−12 ). Furthermore, this model also includes the interaction effect between the country dummy and the square of NWC. As in equation (1b), the country dummy (𝐶𝐷𝑈𝑀𝑖,𝑡) has a value equal to 1 for firms in emerging countries and a value of 0 for firms in developed countries.
With the coefficients of NWC and its square it is possible to calculate the optimal ratio, which is the inflection point of the NWC and firm performance relation. The optimal ratio can be calculated from setting the derivative of NWC to firm performance (𝛽1+ 2𝛽2𝑁𝑊𝐶) equal to zero, which gives that the optimum point for developed countries is found with the formula: −𝛽1/ 2𝛽2. Since NWC is expected to have a positive relation with firm performance for lower ratios of NWC and a negative relation for higher ratios of NWC. To have an optimum the coefficient of NWC, 𝛽1 should be positive and the coefficient of NWC2, 𝛽
2 should be negative. For a minimum the coefficient of NWC, 𝛽1 should be negative and the coefficient of NWC2, 𝛽
2 positive.4 Due to the country dummy the optimal ratio of NWC in emerging countries should be calculated as −𝛽1+ 𝛿1/ 2(𝛽2+ 𝛿2) (Baños-Caballero et al. 2014).
3.2 Methodology
For testing the hypotheses panel data methodology will be used, because this methodology has a couple of advantages. First of all, this methodology assumes that the firms are heterogeneous: each firm is allocated its own dummy (𝜂𝑖). So, in particular, panel data methodology reduces the risk of unobservable heterogeneity related to firm characteristics not included in the model (Hsiao, 1985; Baños-Caballero et al. 2014). Secondly, it enables me to use cross-section data as well as time-series data. Thirdly, panel data is more informative, allows for more variability, has less collinearity and is more efficient (García-Teruel and Martinez-Solano, 2007). To analyse the panel data Eviews was used. All tables are included in the appendix. First, the descriptive statistics will be provided for all 28 countries that are part of the EU together (table 1) and the descriptive statistics of the most important variables for a linear relation between NWC and firm performance per EU country (table 2). Second, table 3 displays the correlation among all variables used for testing the linear relationship between NWC and firm performance for the EU as a whole and for developed and emerging countries separately. Third, Table 4 will show the fixed effects panel OLS regression results for the existence of a linear relation. Column 1 shows the existence of a linear relation within the EU, column 2 shows the existence of a linear relation for developed countries, column 3 shows the existence of a linear relation for emerging countries and column 4 shows the difference in the linear relation for developed and emerging countries. Table 5 will show the regression results of the most important variables for developed countries, showing a positive, negative or no relation between NWC and firm performance. Table 6 includes the regression results of emerging countries for the most important variables and a positive, negative or no
11 linear relation. Finally, a robustness check will be conducted, where the optimal ratio of NWC will be tested upon the linear relation. Table 7 provides the OLS regression output for the existence of an optimal ratio. Column 1 includes the whole EU without country dummies, column 2 includes the existence of an optimal level for developed countries, column 3 shows the results for the existence of an optimal level in emerging countries and column 4 shows the results for the difference in the optimal ratio of NWC between developed and emerging countries. Table 8 will show the regression results of the existence of an optimal NWC ratio for the most important variables for developed countries, showing a possible positive linear relation, negative linear relation, minimum ratio of NWC or optimum ratio of NWC. Table 9 includes the regression results of the existence of an optimal level in emerging countries for the most important variables showing a possible positive linear relation, negative linear relation, minimum ratio of NWC or optimum ratio of NWC.
3.3 Sample, data and descriptive statistics 3.3.1 Sample and data
The sample consists of listed manufacturing firms, which are part of the EU for the period 2011-2016. The distinction between developed and emerging countries will be based on economic development by using the current GDP per capita in US$ over the years 2011-2015. This is a simplified version of measuring economic development (Saci and Holden, 2008). To be able to separate the countries in emerging and developed countries, this study will make use of the average GDP per capita for all countries within the European Union. The average GDP per capita in US$ is compared for the years 2011 until 2015. The lowest average GDP is chosen as the benchmark, which is $32.018 in 2015. As a result of the benchmark based on average GDP per capita the sample exists of 11 developed countries (Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Luxembourg, Netherlands, Sweden, United Kingdom) and 17 emerging countries (Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, Greece, Hungary, Italy, Latvia, Lithuania, Malta, Poland, Portugal, Romania, Slovak Republic, Slovenia, Spain).5 Firm-level data are retrieved from Datastream for the period January 1, 2011 until
December 31, 2016. Firms lacking an ISIN or SIC code will be excluded. Manufacturing firms with the Standard Industrial Classification (SIC) code between 2000 and 3999 are chosen. The focus is on manufacturing firms as according to Fazzari and Petersen (1993), NWC is very important for manufacturing firms as it is fifty percent larger than the total amount of fixed capital stock. Furthermore, according to García-Teruel and Martinez-Solano (2007) manufacturing firms need a larger amount of time to generate cash compared to other industries, hence, manufacturing firms need more financial resources for their operations. Cross-listed firms will be excluded, because they are affected by the regulations of multiple countries. To be able to make a comparison between countries the data retrieved in local currencies will be converted to euros, these countries include Denmark, Sweden, United
12 Kingdom, Hungary, Poland, Czech Republic, Bulgaria, Romania, Croatia. Furthermore, three countries joined the Euro during the sample period, these are Estonia (January 1, 2011), Latvia (January 1, 2014) and Lithuania (January 1, 2015). The exchange rates are taken at the end of each year (December 31) and retrieved from http://www.xe.com/currencyconverter/ (Zhang et al. 2015). Finally, each independent variable will be winsorized at the 1% level to limit problems with extreme outliers, as done in other studies (e.g. Kieschnick et al. 2013).
3.2.2 Descriptive statistics
The descriptive statistics for the EU as a whole are presented in table 1 for a sample of 1988 firms and 5723 observations. As all independent variables are lagged for one year and the independent variable (GROWTH) includes two years lagged net assets, the data covers a period of 2013-2016. The dependent variable firm performance, measured as market to book value of assets has an average value of 1.453, with a median of 0.970. Due to winsorizing at 1%, firm performance has a range of 0.166 to 11.232. The mean of NWC is 0.341 with a median of 0.246, so EU firms have on average a positive ratio of NWC with a range of -2.229 to 6.918. On average firms’ total assets are for 2.1% financed with debt. Firm growth is on average 7.3%. However, return on assets is on average equal to zero with a median of 3.7%.
Insert table 1 about here.
In table 2 the average values for the most important variables are displayed per EU member. Sweden has on average the highest firm performance value of 2.222 and Latvia scores on average the lowest value on firm performance 0.488. Furthermore, all countries have on average a NWC ratio of below 1 except Slovakia. Ireland and Lithuania have on average the lowest ratios of NWC 0.109 (Ireland) and 0.208 (Lithuania), which is based on the average of 117 Irish observations and 20 Slovak observations.
Insert table 2 about here.
Furthermore, a correlation matrix has been calculated to test for relationships between variables within the model. The closer the p-value to 1 the more I expect a positive linear correlation between variables. The correlation matrix (table 3), shows that there are variables with significant correlations at a significance level of 1%, so these variables show no correlation. The highest level of correlation is between return on assets and growth with a correlation of 0.734, this may lead to multicollinearity.
13 4.Empirical evidence
4.1 The linear relation between NWC and firm performance
Table 4 summarizes in 4 columns the results of equation (1a) and equation (1b), which test the existence of a linear relation between NWC and firm performance For the EU, developed and emerging countries. Column 1 shows the regression results of equation (1a), including the existence of a positive or negative linear relation between NWC and firm performance in the EU without a country dummy. In this regression no significant relation has been found between NWC and firm performance for all EU countries together. Column 2 includes the regression results of equation (1a) by testing for the existence of a positive or negative relation between NWC and firm performance in developed countries. This regression shows no significant relation between NWC and firm performance within developed countries. Column 3 shows the regression results of equation (1a), the existence of a positive or a negative linear relation between NWC and firm performance for emerging countries. The results of the regression in column 3 show that there exists a significant negative linear relation between NWC and firm performance for all emerging countries together within the EU. The value of this significant linear relation is -5.3% at a significance level of 10% and with a t-statistic of -1.706. This results indicate that hypothesis 10 may be rejected, the NWC ratio has no effect on firm performance, as the results show
14 Insert table 4 about here.
4.2 The linear relation between NWC and firm performance per EU country
In table 5 the OLS regression output of equation (1a) shows the results for the existence of a linear relation in developed countries. There are significant results for 4 developed countries. Finland shows a significant negative linear relation with a value of -0.965 at a significance level of 1%. The results of France show a significant negative linear relation for the value of -0.088 at a significance level of 10%. The UK shows a significant negative linear relation as well for the value of -0.191. Considering these three countries, Finland shows the strongest negative linear relation between NWC and firm performance. The negative linear relation suggests that managers of firms within these countries should lower their NWC ratio to increase firm performance. If managers intend to lower the NWC ratio they can increase their accounts payables or lower their inventories and accounts receivables. According to Deloof (2003), the best possible explanation for a negative linear relation between NWC and firm performance may be that firms with lower firm performance pay their bills later, accounts payables becomes larger which the firm may use as a source of internal financing for valuable investments to increase their firm performance. In Ireland manufacturing firms show on average a highly significant positive linear relation between NWC and firm performance for the value of 1.844 at a 1% significance level. This positive linear relation suggests that managers can increase their firm performance by increasing their NWC ratio, this may be achieved by increasing accounts receivables and inventory while lowering accounts payable. The results reject hypothesis 10, NWC has no effect on firm
performance, as the results confirm the existence of a linear relation between NWC and firm performance for a few developed countries.
15
Insert table 5 about here.
4.3. Robustness check
4.3.1 The optimal ratio of NWC
As robustness check, the optimal ratio of NWC for the NWC and firm performance relation will be tested upon the linear relation. Hypothesis 20, there exists no optimal ratio of NWC and hypothesis 2a,
there exists an optimal ratio of NWC will be tested. Table 7 summarizes the robustness results of equation (2), which includes the square of NWC for the EU, developed and emerging countries. The interaction effect of NWC2 and the country dummy (CDUM) are also included in equation (2). In table
7, column 1 includes the regression results of equation (2), testing for the existence of an optimal ratio of NWC without the interaction effect of country dummies for all the EU countries. The results show that NWC2 is significant and negative for the value of -0.003 at a significance level of 10%. When NWC2
is significant and NWC is not significant, there exists no optimal ratio of NWC. To find an optimal ratio both NWC and NWC2 should have a significant effect on firm performance. Column 2 provides the
regression results for the existence of an optimal ratio of NWC in developed countries. Column 3 shows the results for the existence of an optimal ratio of NWC in emerging countries. Column 4 shows the results for the difference in the optimal ratio of NWC between developed and emerging countries. The results in column 2, column 3 and column 4 do not provide any significant coefficients of NWC or NWC2. Based on all 4 regressions in table 7, hypothesis 2
0, there exists no optimal ratio of NWC, cannot
be rejected.
4.3.2 The optimal ratio of NWC per country
In table 8, the results of the fixed panel regression per developed country are provided. For some of the developed countries both NWC as well as NWC2 have a significant influence on firm performance. For
these countries the optimal ratio of NWC can be calculated. The results in table 8 show that for Belgium the coefficient of NWC is significant at a value of -1.321 at the significance level of 5% and the coefficient of NWC2 is significant for a value of 0.698 at a significance level of 1%. As both NWC and
NWC2 are significant, one can assume that there exists an optimal ratio for a value of 0.946. This optimal
ratio of NWC is actually a minimum, as NWC is a negative number and NWC2 positive the relation
with firm performance will be U-shaped. The regression results of Finland show a significant NWC coefficient of -1.679 at a significance level of 1% and a significant coefficient of 0.305 at a significance level of 1%. The optimal ratio of NWC is 2.752, which is a minimum as well. The third and last optimal ratio of NWC for developed countries that can be interpreted is France. NWC shows a significant relation with firm performance of -0.110 at a significance level of 5% and NWC2 a significant relation
16 France is 6.875, this is a minimum. The minimum of Belgium is the most likely to reflect a minimum that is feasible in practice, the minima of Finland and France seem to be less feasible as they may be too large. Next to the optimal ratio of NWC there are two countries, Denmark and the UK, that show only a significant relation between NWC2 and firm performance. Denmark has a significant negative relation
between NWC2 and firm performance for a value of -1.055 at a 1% significance level. Furthermore, the
UK also shows a significant negative relation between NWC2 and firm performance for a value of
-0.011 at a 5% significance level. The other developed countries do not show any significant relation for both NWC and NWC2 on firm performance. Overall, the results of developed countries in the EU
indicate the existence of a minimum ratio of NWC instead of the predicted optimal ratio of NWC. This finding is inconsistent with the results of Baños-Caballero et al. (2012; 2014) and Aktas et al. (2015), as they find a significant optimal ratio of NWC resulting in an inverted U-shaped relation. Based on these findings hypothesis 20 may be rejected as there exists an optimal ratio of NWC. However, the findings
show the existence of a minimum instead of an optimum in developed countries.
For the emerging countries, as table 9 indicates, there is no optimal ratio of NWC, as both NWC as well as NWC2 are insignificant. As table 9 indicates, there are three significant linear relations
between NWC and firm performance. These findings are consistent with the regression results of table 7, as both tables find a significant linear relation for Greece, Italy and Poland. However, when comparing the coefficients of both tables, one can see that there are some small differences between the values of these coefficients. This can be explained due to adding NWC2 into equation (1b). So, for emerging
countries we do not reject hypotheses 20, there is no optimal ratio of NWC.
5. Conclusion
The aim of this paper is to examine the effect of the NWC on firm performance for listed manufacturing firms in the 28 European Union countries with panel data from 2011-2016. As methodology a cross-section fixed effects panel data regression has been used. This paper analyses a possible positive, negative or no linear relation between the NWC ratio and firm performance. The main contribution of this paper is that it tests the possible linear relation for 28 EU countries and makes a distinction between developed and emerging countries.
17 rejected as these findings suggest the existence of a negative linear relation between NWC and firm performance for a few developed countries separately and for the emerging countries together and for a few separately. However, the results also suggest that there exists a positive linear relation in some developed and emerging countries.
The robustness check demonstrates that there may exist an optimal ratio of NWC upon the linear relation, in this paper a minimum. For two countries, Finland and France, the significant linear relation may be replaced by an optimal NWC ratio, which is a minimum, to optimise firm performance. This indicates that hypothesis 20, there is no optimal ratio of NWC, may be rejected, as we do find an optimal
ratio, which is a minimum.
These results have some firm policy implications. From a managerial point of view, managers of manufacturing firms in the EU should strive for a lower NWC ratio to improve firm performance, based on a negative linear relation, but in other countries a higher NWC may improve firm performance. However, in some countries there might be an optimal ratio of NWC, which is a minimum. These implications are important for managers planning to move or expand to another country. It may be very disadvantageous to enter a country where a positive linear relation exists between NWC and firm performance, when obtaining external finance is very difficult in that particular country. To optimise firm performance managers of a manufacturing firm should consider a unique strategy for each country within the EU and possibly for each firm separately.
Overall, this paper shows the existence of a positive linear relation, a negative linear relation and an optimal minimum of NWC for a few countries. These results indicate that we cannot have the same strategy for all EU countries, as there is no unity within the EU.
Suggestions for future research are researching the distinction between firms with higher NWC ratios and lower NWC ratios, as the current level may have a direct influence on the linear relation. Future research might also consider focusing on other industries as there might exist differences between industries.
18 6. References
Agrawal, A., & Knoeber, C. R. (1996). Firm performance and mechanisms to control agency problems between managers and shareholders. Journal of financial and quantitative analysis, 31(03), 377-397.
Aktas, N., Croci, E., & Petmezas, D. (2015). Is working capital management value-enhancing? Evidence from firm performance and investments. Journal of Corporate Finance, 30, 98-113.
Baños-Caballero, S., García-Teruel, P. J., & Martínez-Solano, P. (2012). How does working capital management affect the profitability of Spanish SMEs?. Small Business Economics, 39(2), 517-529.
Baños-Caballero, S., García-Teruel, P. J., & Martínez-Solano, P. (2014). Working capital management, corporate performance, and financial constraints. Journal of Business Research, 67(3), 332-338. Blinder, A. S., & Maccini, L. J. (1991). The resurgence of inventory research: what have we learned?.
Journal of Economic Surveys, 5(4), 291-328.
19 Deloof, M. (2003). Does working capital management affect profitability of Belgian firms?. Journal of
business finance & Accounting, 30(3‐4), 573-588.
Deloof, M., & Jegers, M. (1996). Trade credit, product quality, and intragroup trade: some European evidence. Financial management, 33-43.
Eckbo, B. E., & Kisser, M. (2013). Corporate funding: who finances externally. Unpublished working
paper. Tuck School of Business at Darmouth.
Emery, G. W. (1984). A pure financial explanation for trade credit. Journal of Financial and
Quantitative Analysis, 19(03), 271-285.
Emery, G. W. (1987). An optimal financial response to variable demand. Journal of Financial and
Quantitative Analysis, 22(02), 209-225.
Ernst, Young (2016). All tied up: working capital management report. Available at: www.ey.com. Fazzari, S. M., Hubbard, R. G., Petersen, B. C., Blinder, A. S., & Poterba, J. M. (1988). Financing
constraints and corporate investment. Brookings papers on economic activity, 1988(1), 141-206. Fazzari, S. M., & Petersen, B. C. (1993). Working capital and fixed investment: new evidence on
financing constraints. The RAND Journal of Economics, 328-342.
Gibson, H. D., & Tsakalotos, E. (1994). The scope and limits of financial liberalisation in developing countries: A critical survey. The Journal of Development Studies, 30(3), 578-628.
Hsiao, C. (1985). Benefits and limitations of panel data. Econometric Reviews, 4(1), 121-174.
Hill, M. D., Kelly, G. W., & Highfield, M. J. (2010). Net operating working capital behavior: a first look. Financial management, 39(2), 783-805.
Juan García-Teruel, P., & Martinez-Solano, P. (2007). Effects of working capital management on SME profitability. International Journal of managerial finance, 3(2), 164-177.
Kieschnick, R., Laplante, M., & Moussawi, R. (2013). Working capital management and shareholders’ wealth. Review of Finance, 17(5), 1827-1852.
Kim, Y. H., & Chung, K. H. (1990). An integrated evaluation of investment in inventory and credit: a cash flow approach. Journal of Business Finance & Accounting, 17(3), 381-389.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1997). Legal determinants of external finance. Journal of finance, 1131-1150.
Levine, R. (1997). Financial development and economic growth: views and agenda. Journal of
economic literature, 35(2), 688-726.
Lewellen, W. G., McConnell, J. J., & Scott, J. A. (1980). Capital market influences on trade credit policies. Journal of Financial Research, 3(2), 105-113.
Long, M. S., Malitz, I. B., & Ravid, S. A. (1993). Trade credit, quality guarantees, and product marketability. Financial management, 117-127.
20 Mathuva D, (2009). The influence of working capital management components on corporate profitability: a survey on Kenyan listed firms. Research Journal of Business Management, 3, 1-11.
Nadiri, M. I. (1969). The determinants of trade credit in the US total manufacturing sector.
Econometrica: Journal of the Econometric Society, 408-423.
Ng, C. K., Smith, J. K., & Smith, R. L. (1999). Evidence on the determinants of credit terms used in interfirm trade. The Journal of Finance, 54(3), 1109-1129.
Ouyang, L. Y., Teng, J. T., Chuang, K. W., & Chuang, B. R. (2005). Optimal inventory policy with noninstantaneous receipt under trade credit. International Journal of Production Economics, 98(3), 290-300.
Perfect, S. B., & Wiles, K. W. (1994). Alternative constructions of Tobin's q: An empirical comparison. Journal of empirical finance, 1(3-4), 313-341.
Petersen, M. A., & Rajan, R. G. (1997). Trade credit: theories and evidence. Review of financial studies,
10(3), 661-691.
Saci, K., & Holden, K. (2008). Evidence on growth and financial development using principal components. Applied Financial Economics, 18(19), 1549-1560.
Schiff, M., & Lieber, Z. (1974). A model for the integration of credit and inventory management. The
Journal of finance, 29(1), 133-140.
Schimmelfennig, F. (2001). The community trap: Liberal norms, rhetorical action, and the eastern enlargement of the European Union. International organization, 55(01), 47-80.
Shin, H. H., & Soenen, L. (1998). Efficiency of working capital management and corporate profitability.
Financial practice and education, 8, 37-45.
Smirlock, M., Gilligan, T., & Marshall, W. (1984). Tobin's q and the Structure-Performance Relationship. The American Economic Review, 74(5), 1051-1060.
Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. The American
economic review, 71(3), 393-410.
Studenmund, A. H. (1997). Using econometrics: A practical guide. New York Addison-Wesley. Thomsen, S., Pedersen, T., & Kvist, H. K. (2006). Blockholder ownership: Effects on firm value in
market and control based governance systems. Journal of Corporate finance, 12(2), 246-269. Wang, Y. J. (2002). Liquidity management, operating performance, and corporate value: evidence from
Japan and Taiwan. Journal of Multinational Financial Management, 12(2), 159-169.
Wu, H. L. (2011). Can minority state ownership influence firm value? Universal and contingency views of its governance effects. Journal of Business Research, 64(8), 839-845.
Zhang, W., von Eije, H., & Westerman, W. (2015). Does Cash Contribute to Value? A Comparison of Constrained and Unconstrained Firms in China and Germany. Eurasian Journal of Economics
21 7. Appendix
Table 1: Descriptive statistics of 1988 EU firms, 2013-2016: 5426 observations
Variable Mean Standard
Deviation
Median Minimum Maximum
Q 1.453 1.549 0.970 0.166 11.232 NWC 0.341 0.704 0.246 -2.229 6.198 SIZE 11.690 2.678 11.766 2.996 17.345 LEV 0.021 0.033 0.013 -0.059 0.249 GROWTH 0.073 0.368 0.019 -0.573 3.521 ROA 0.000 0.181 0.037 -1.241 0.316
The variable data are retrieved from Datastream including 1988 listed manufacturing firms and 5723 observations within the European Union, for the period 2013-2016. All variables are lagged one year. Information regarding mean, median, standard deviation, minimum and maximum is provided for the dependent variable firm performance (Q) measured as the market to book value of assets. The independent variables are net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as the ratio of change in net assets to one year lagged net assets and return on assets (ROA) measured as the ratio of EBIT to total assets. Data are winsorized at a 1% level.
22 Table 2: Descriptive statistics whole EU
Countries Mean Q Mean NWC Firms Obs.
23 Portugal 0.856 0.254 16 70 Romania 0.594 0.575 61 217 Slovakia 0.580 1.247 6 20 Slovenia 0.545 0.272 11 47 Spain 1.553 0.363 47 202 Sweden 2.222 0.343 197 719 UK 1.954 0.313 336 1361 Total EU 1.434 0.353 1988 7900 Total developed 1.657 0.320 1332 5276 Total emerging 0.987 0.420 656 2624
The variable data are retrieved from Datastream including 1988 listed manufacturing firms and 7900 observations within the EU, 1332 firms and 5276 observations of developed countries within the EU and 656 firms and 2624 observations of emerging countries within the EU, for the period 2013-2016. All variables are lagged one year. Information regarding the mean per country is provided for the dependent variable firm performance (Q) measured as the market to book value of assets. The independent variable net working capital (NWC) is measured as the ratio of accounts receivables plus inventory minus accounts payables to sales. Annual binary variables (not reported) control for time effects. Data are winsorized at a 1% level.
Table 3. Correlation matrix
Q NWC SIZE LEV GROWTH ROA
Q 1.000 NWC 0.110*** 1.000 SIZE -0.329*** -0.255*** 1.000 LEV 0.304*** 0.085*** -0.021a 1.000 GROWTH 0.155*** 0.133*** -0.172*** -0.013b 1.000 ROA -0.345*** -0.170*** 0.623*** -0.138*** 0.011c 1.000
Correlation matrix where the variable data are retrieved from Datastream including 1988 listed manufacturing firms and 5723 observations within the European Union, for the period 2013-2016. All variables are lagged one year. Dependent variable is firm performance (Q) measured as the market to book value of assets, net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as the ratio of change in net assets to one year lagged net assets and return on assets (ROA) measured as the ratio of EBIT to total assets. Data are winsorized at a 1% level.* indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
24
Table 4: Regressions for the EU, developed and emerging countries
Variables Models Whole EU without country dummy effect Developed countries Emerging countries Whole EU with country dummy effect 1 2 3 4 CONSTANT 3.200*** 4.013*** 1.901*** 3.206*** (8.109) (6.864) (5.225) (-0.736) NWC -0.034 -0.033 -0.053* -0.022 (-0.125) (-0.919) (-1.706) (-0.736) CDUM*NWC -0.045 (-0.758) SIZE -0.148*** -0.193*** -0.080** -0.148*** (-4.408) (-3.987) (-2.456) (-4.417) LEV -1.596*** -2.096*** 0.698 -1.595*** (-2.917) (-3.037) (0.894) (-2.916) GROWTH -0.093** -0.101** -0.079 -0.092** (-2.546) (-2.229) (-1.342) (-2.516) ROA 0.355*** 0.349** 0.404** 0.357*** (2.793) (2.128) (2.436) (2.806)
t YES YES YES YES
R-squared 0.784 0.757 0.861 0.784
25
VIF 6.633 5.904 10.352 6.634
Observations 5723 3836 1887 5723
Firms 1988 1332 656 1988
Panel data cross-section fixed effects. The variable data are retrieved from Datastream including 1988 listed manufacturing firms, which includes 1332 firms from developed countries and 656 firms from emerging countries. There are 5723 observations within the EU, 3836 observations in developed countries and 1887 observations in emerging countries, for the period 2013-2016. This table provides the OLS regression output for the existence of a linear relation. Column 1 shows the existence of a linear relation within the EU, column 2 shows the existence of a linear relation for developed countries, column 3 shows the existence of a linear relation for emerging countries and column 4 shows the difference in the linear relation for developed and emerging countries. All variables are lagged one year. Dependent variable is firm performance (Q) measured as the market to book value of assets, net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as the ratio of change in net assets to one year lagged net assets and return on assets (ROA) measured as the ratio of EBIT to total assets. Annual binary variables control for time effects. Data are winsorized at a 1% level. The t-statistics are given in parenthesis below the coefficient. Reported F-statistic corresponds to a Wald test of the hypothesis that all coefficients excluding the constant are equal to zero. Adjusted R-squared is included as goodness of fit. * indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
Table 5. Fixed panel regressions for developed countries Developed
Countries
NWC Linear
relation
R-Squared F-statistic Firms Obs.
26
UK -0.191*** - 0.735 10.170*** 336 1005
(-2.849)
Total -0.033 0.757 11.239*** 1332 3836
Panel data cross-section fixed effects. The variable data are retrieved from Datastream including 1332 listed manufacturing firms and includes 3836 observations of developed countries within the EU, for the period 2013-2016. This table provides the OLS regression output of equation (1a) for the existence of a linear relation per developed country. All variables are lagged one year. Dependent variable is firm performance (Q) measured as the market to book value of assets, net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as the ratio of change in net assets to one year lagged net assets and return on assets (ROA) measured as the ratio of EBIT to total assets. Annual binary variables control for time effects. Data are winsorized at a 1% level. The t-statistics are given in parenthesis below the coefficient. Reported F-statistic corresponds to a Wald test of the hypothesis that all coefficients excluding the constant are equal to zero. Adjusted R-squared is included as goodness of fit. * indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
Table 6. Fixed panel regressions for emerging countries
Countries NWC Relation R-Squared F-statistic Firms Obs.
27 Poland -0.205** - 0.880 23.556*** 146 (-2.072) Portugal -0.052 0.781 9.777*** 16 55 (-0.152) Romania -0.069 0.655 6.162*** 61 148 (-1.060) Slovakia 0.531 0.873 8.505*** 6 13 (0.241) Slovenia -0.585 0.793 8.463*** 11 36 (-0.873) Spain 0.034* 0.898 28.129*** 47 149 (0.143) Total -0.053* - 0.861 21.216*** 656 1887
Panel data cross-section fixed effects. The variable data are retrieved from Datastream including 656 listed manufacturing firms and 1887 observations of emerging countries within the EU, for the period 2013-2016. This table provides the OLS regression output of equation (1a) for the existence of a linear relation per emerging country. All variables are lagged one year. Dependent variable is firm performance (Q) measured as the market to book value of assets, net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, net working capital its square (NWC2), firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as ratio of change in net assets to one year lagged net assets and return on assets (ROA) measured as ratio EBIT to total assets. Annual binary variables control for time effects. Data are winsorized at a 1% level. The t-statistics are given in parenthesis below the coefficient. Reported F-statistic corresponds to a Wald test of the hypothesis that all coefficients excluding the constant are equal to zero. Adjusted R-squared is included as goodness of fit. * indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
Table 7: Robustness regressions for EU, developed and emerging countries
28
t YES YES YES YES
R-squared 0.784 0.757 0.861 0.784 F-Value 12.944*** 11.236*** 21.166*** 12.926*** Optimum6 -1.667 0.021 2.559 0.75 VIF 6.638 5.909 10.353 6.639 Firms 1988 1332 656 1988 Observations 5723 3836 1887 5723
Panel data cross-section fixed effects. The variable data are retrieved from Datastream including 1988 listed manufacturing firms, which includes 1332 firms from developed countries and 656 firms from emerging countries. There are 5723 observations within the EU, 3836 observations for developed countries and 1887 observations for emerging countries, for the period 2013-2016. This table provides the OLS regression output for the existence of an optimal ratio. Column 1 includes the whole EU without country dummies, column 2 includes the existence of an optimal level for developed countries, column 3 shows the results for the existence of an optimal level in emerging countries and column 4 shows the results for the difference in the optimal ratio of NWC between developed and emerging countries. All variables are lagged one year. Dependent variable is firm performance (Q) is the market to book value of assets, net working capital (NWC) is the ratio of accounts receivables plus inventory minus accounts payables to sales, net working capital its square (NWC2), firm size (SIZE) is the natural logarithm of sales, leverage (LEV) is the ratio of total debt to total assets, growth opportunities (GROWTH) is ratio change in net assets to one year lagged net assets and return on assets (ROA) is ratio EBIT to total assets. Annual binary variables control for time effects. Data are winsorized at a 1% level. The t-statistics are given in parenthesis below the coefficient. Reported F-statistic corresponds to a Wald test of the hypothesis that all coefficients excluding the constant are equal to zero. Adjusted R-squared is included as goodness of fit. * indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
Table 8. Robustness fixed panel regressions for developed countries
Countries NWC NWC2 Optimum R-Squared F-statistic Firms Obs.
Austria 0.112 -0.010 5.600 0.807 12.651*** 30 104 (0.552) (-0.783) Belgium -1.321** 0.698*** 0.946 0.848 17.457*** 53 160 (-5.393) (5.454) Denmark -0.498 -1.055*** -0.236 0.907 31.082*** 50 170 (-0.895) (-2.750) Finland -1.679*** 0.305*** 2.752 0.875 22.948*** 65 220 (-7.035) (4.347) France -0.110** 0.008* 6.875 0.813 14.061*** 267 702 (-2.020) (1.823) Germany 0.237 0.003 -39.500 0.882 24.673*** 235 690 (1.627) (0.575) Ireland 0.383 -0.597 0.321 0.769 10.040*** 26 88 (0.320) (-1.367) Luxembourg 0.700 -1.187 0.295 0.840 10.751*** 16 42 (0.551) (-1.345) Netherlands -0.541 -0.193 -1.402 0.783 10.935*** 57 158 (-0.255) (-0.166) Sweden 0.251 -0.015 8.367 0.656 6.806*** 197 497 (1.411) (-1.410) UK -0.093 -0.011** -4.227 0.736 10.192*** 336 1005
29
(-1.121) (-1.985)
Total 1.650e-04 -0.004 0.021 0.757 11.236*** 1332 3836
Panel data cross-section fixed effects. The variable data are retrieved from Datastream including 1332 listed manufacturing firms and 3836 observations from developed countries within the EU, for the period 2013-2016. This table provides the OLS regression output of equation (2) for the existence of an optimal ratio of NWC upon a linear relation in developed countries. All variables are lagged one year. Dependent variable is firm performance (Q) is the market to book value of assets. The independent variables are net working capital (NWC) measured as the ratio of accounts receivables plus inventory minus accounts payables to sales, net working capital its square (NWC2), firm size (SIZE) measured as the natural logarithm of sales, leverage (LEV) measured as the ratio of total debt to total assets, growth opportunities (GROWTH) measured as ratio change in net assets to one year lagged net assets and return on assets (ROA) measured as the ratio EBIT to total assets. Annual binary variables control for time effects. Data are winsorized at a 1% level. The t-statistics are given in parenthesis below the coefficient. Reported F-statistic corresponds to a Wald test of the hypothesis that all coefficients excluding the constant are equal to zero. Adjusted R-squared is included as goodness of fit. * indicates significance at the 10% level, ** indicates significance at the 5% level, *** indicates significance at the 1% level.
Table 9. Robustness fixed panel regressions for emerging countries
Countries NWC NWC2 Optimum R-Squared F-statistic Firms Obs.
30 Poland -0.346** 0.013 14.286 0.880 23.497*** 146 417 (-2.518) (1.476) Portugal 0.478 -0.732 0.290 0.793 10.005*** 16 55 (1.045) (-1.678) Romania -0.097 0.007 9.438 0.652 6.016*** 61 148 (-1.217) (0.612) Slovakia x x x x x 6 x Slovenia 1.361 -3.032 0.216 0.789 7.875*** 11 36 (0.534) (-0.793) Spain 0.081 0.015 -0.844 0.897 27.298*** 149 (0.268) (0.256) 47 Total -0.048 1.723e-04 2.559 0.861 21.166*** 656 1887
