Master’s Thesis 15 credits Department of Business Studies Uppsala University
Spring Semester of 2020
Date of Submission: 2020-06-03
The impact of Working Capital Management on Firm Performance in different phases of a business cycle - Evidence from Sweden
Göransson Tobias Lundqvist Victor Svensson Martin
Supervisor: Derya Vural-Meijer
Abstract
The recent financial crisis 2008 had an impact on Swedish firms which led to further focus on how companies manage their financial resources. This study investigates the relationship between working capital management (WCM) and firm performance, and how it’s affected during different phases of a business cycle in Sweden between 2008-2018. Previous studies find inconsistent results of WCM and firm performance, where the relationship can be positive, negative, or concave.
The sample of this study consists of 2,526 firm-year observations from 449 Swedish listed firms over the time period 2008 to 2018, a multiple OLS-regression is conducted to examine the relationship. The findings indicate that companies can enhance firm performance by managing their WCM more efficiently, measured as the cash conversion cycle (CCC). Additional test finds that WCM is of large importance for high-performing firms. Furthermore, WCM is not seen to be different during different phases of business cycles.
Key words: Business cycle, Cash conversion cycle, Firm performance, Tobin's Q, Working capital efficiency, Working capital management
Definitions
Items Explanation
Working Capital (WC) The difference between a company’s current assets and its current liabilities
(Filbeck & Kreuger, 2005)
Working Capital Management (WCM) A definition of how a firm finances its operating business and how much liquidity the organization has available to satisfy the short-term requirements imposed by current liabilities (Sharma & Kumar, 2011)
Tobin’s Q A predominated variable to measure market-
based performance (Wang, 2002)
Cash Conversion Cycle (CCC) A perception of the organization's time lag between the expenditure for the purchase of raw materials and the total sales of finished goods (Deloof, 2003)
Days Receivables Outstanding (DRO) The number of days account receivables (Deloof, 2003)
Days inventory Outstanding (DIO) The number of days inventory outstanding (Deloof, 2003)
Days Payables Outstanding (DPO) The number of days account payables (Deloof, 2003)
Table of Contents
1 Introduction ... 1
2 Literature review ... 4
2.1 Working Capital Management and Firm Performance ... 4
2.2 Cash Conversion Cycle and Working Capital ... 8
2.3 Business Cycles ... 9
2.4 Hypothesis ... 11
3 Methodology ... 12
3.1 Introduction ... 12
3.2 Observed time period ... 12
3.3 Data, population and sample ... 14
3.4 Regression Model and Analysis ... 15
3.4.1 Regression model ... 15
3.4.2 Dependent variables ... 16
3.4.3 Independent variables ... 16
3.4.4 Control variables ... 17
3.4.5 Normality, skewness and kurtosis ... 19
3.4.6 Multicollinearity and singularity ... 19
3.4.7 Homoscedasticity and linearity ... 20
3.4.8 Methodology criticism ... 20
4 Result and analysis ... 21
4.1 Descriptive statistics ... 21
4.2 Correlation matrix ... 22
4.3 Regression result and analysis ... 23
4.4 Additional test ... 28
4.5 Regression model and additional test ... 28
5 Conclusion ... 31
6 Future Research ... 32
References ... 33
1
1 Introduction
This study examines the relationship between working capital management (WCM) and firm performance during different phases of a business cycle in Sweden between 2008-2018. Several studies argue that the latest economic downturn in 2008 brought new light to WCM practices (Scholleova, 2012; Enqvist, Graham & Nikkinen, 2014; Ramiah, Zhao & Moosa, 2014;
Wasiuzzaman, 2015). Where efficient WCM became crucial to enhance firm performance and stay competitive (Baños-Caballero, García-Teruel & Martínez-Solano, 2012; Yazdanfar &
Öhman, 2014).
Prior to the financial crisis 2008, WCM was not given enough attention, it was mostly used to ensure financial stability and not considered a measure to improve liquidity (Enqvist et al., 2014). During 2008, the banks started to tighten their credit standards which led to less borrowing for firms. When firms were unable to borrow to the same extent as before, more emphasis was focused on WCM (Chiou & Cheng, 2006). Scholleova (2012) further show that firms that efficiently managed their WCM during the financial crisis of 2008 withstood their competitors. Furthermore, Enqvist et al. (2014) show the importance of WCM and the impact of different phases of a business cycle. They argue that efficient WCM is more essential to manage during economic downturns compared to booms. Where the argument behind this is that liquidity often comes under pressure during economic downturns. Although Enqvist et al.
(2014) argue that the benefits of an efficiently managed WCM are many, Aktas, Croci, and Petmezas (2015) reveals that efficient WCM, as a potential source of cash to fund growth, is often neglected by firms.
WCM is according to Chiou and Cheng (2006) one of the most crucial parts in the field of corporate finance policy, where capital budgeting and structure are the others. The theory of WCM discloses how to manage WC in terms of efficiency, profitability, liquidity, and solvency to enhance firm performance (Brigham, Gapenski & Ehrhardt, 1999). Furthermore, firms with low levels of cash can, by investing in high levels of inventories, endanger their liquidity. The declining levels of liquidity may result in insolvency and eventually financial distress as the firm can’t fulfill all its requirements (Kortman, Wicks & Ojeda, 2017). Given that WC has an impact on liquidity, there is no surprise that WCM is recognized as a fundamental aspect of financial performance crucial to optimize for all companies to hold a steady course (Padachi, Narasimhan, Durbarry & Howorth, 2008).
2
In acknowledgment of the importance of WCM, previous studies have often examined the impact WCM has on firm performance by measuring WCM as the cash conversion cycle (CCC) (Jose, Lancaster & Stevens, 1996; Wang, 2002; Deloof, 2003; Enqvist et al., 2014; Yazdanfar
& Öhman, 2014). Although CCC is commonly used in research as an expression for companies' efficiency of WCM, there are different opinions on what strategy to implement for a firm to achieve optimal utilization of its WC. The results from previous studies imply three different outcomes. Either there is a negative relationship between WCM and firm performance, meaning that a shorter CCC increases firm performance (Jose et al., 1996; Wang, 2002; Enqvist et al., 2014; Yazdanfar & Öhman, 2014), or a positive relationship meaning a longer CCC increases firm performance (Gill, Biger & Mathur, 2010; Sharma & Kumar, 2011). The third alternative is a concave relationship which implies that there is an optimal level of WC to maximize firm performance (Baños-Caballero, García-Teruel & Martínez-Solano, 2014; Aktas et al., 2015;
Afrifa, 2016).
WCM and firm performance have to our knowledge only been investigated on the Swedish market by Yazdanfar and Öhman (2014). According to Kortman et al. (2017), Swedish companies are in general more inefficient in managing their WC than most companies in the rest of Europe. For example, the average company in Europe has a CCC of 42 days. At the same time, the average company in Sweden has a CCC of 69 days (ibid). This proves that Swedish companies have a larger amount of cash tied up in WC, which according to Yazdanfar and Öhman (2014) making this an important topic for Swedish companies to consider staying competitive. Another reason for examining the relationship between WCM and firm performance on the Swedish market is that most previous research has investigated this relationship in larger countries (Aktas et al., 2015; Afrifa, 2016; Filbeck, Zhao & Knoll, 2017).
Yazdanfar and Öhman (2014) argue that the business environment and capital structure in Sweden are different compared to larger countries, which affect the WC. The differences in WC efficiency, economic structure, and business environment in Sweden compared to other countries further motivate the investigation of WCM and firm performance on the Swedish market.
3
This study extends the current literature on the relationship between WCM and firm performance in Sweden. Firstly, by studying the relationship over different phases of a business cycle, and secondly, using the market-based measure Tobin’s Q to identify firm performance.
The aim of this study is to generate an understanding for Swedish management regarding the effects WCM has on firm performance during different phases of a business cycle in Sweden. Thus, the thesis aims to answer the question:
How does working capital management affect firm performance over different phases of a business cycle in Sweden?
4
2 Literature review
In the following section, the study will describe working capital management and firm performance, where the relationship between WCM and firm performance will be investigated.
The definition of working capital management, with a focus on CCC, will be processed and compared to different phases of a business cycle.
2.1 Working Capital Management and Firm Performance
Working capital refers to the difference between current assets and current liabilities (Deloof, 2003), where firms manage their WC to finance its operating business. Firms usually invest large amounts of cash in WC and are therefore seen as a crucial function within the firm (Shin
& Soenen, 1998; Deloof, 2003; Gill et al., 2010; Sharma & Kumar, 2011; Enqvist et al., 2014).
Yazdanfar and Öhman (2014) state that a part of a firm's strategy is to manage its WC to satisfy the short-term requirements imposed by current liabilities. Hence, the management needs to do a trade-off between opportunity cost and carrying cost when optimizing the WC investments (ibid). The trade-off is a typical example when firms’ have a risk-return nature of financial decision making (Sharma & Kumar, 2011). The way firms choose to manage their WC depends on the nature of the business and the firm's strategy (Yazdanfar & Öhman, 2014; Sawarni, Narayanasamy, & Ayyalusamy, 2020). Furthermore, Filbeck et al. (2017) argue how firms manage their WC is vital to enhance growth and sustainability, where the WC can be seen as the senior management's responsibility to maneuver.
The chosen WC strategy by the management can be defined as WCM (Deloof, 2003).
Depending on the WCM, firms can optimize the level of internal and external capital sources, where the goal is to free up additional internal funds. To use internally generated resources is normally cheaper than external funds (Sharma & Kumar, 2011). The strategy sets requirements on management when evaluating different WC investments, where they can avoid over- or under investment costs (Yazdanfar & Öhman, 2014). When an organization has optimized the WCM short-term financial strategy, organizations can establish a long-term financial strategy for growth through, for example, long-term investments (Sharma & Kumar, 2011; Yazdanfar
& Öhman, 2014). However, too much investment in WC can lead to minimized value for the shareholders since larger investments require more financing, and firms need to compare the opportunity cost. On the other hand, if a company has low levels of WC, additional investments can lead to growth by the possibility to meet an unexpected rise in demand (Aktas et al., 2015).
5
Furthermore, there is a risk of having too much capital tied up in unnecessary WC since it decreases the liquidity in the company. Too low levels of liquidity can lead to financial distress due to the difficulties to fulfill all requirements for the firm (Sawarni et al., 2020).
There are many studies investigating the relationship between WCM and firm performance, and with closer inspection, the studies have different ways of measuring firm performance (Wang, 2002; Wu, 2011; Enqvist et al., 2014; Altaf & Shah, 2017). The most common way to examine firm performance is either by market-based or accounting-based measures. Enqvist et al. (2014) use the accounting-based measure return on assets (ROA) as a firm performance measure, while Wang (2002), Wu (2011), and Altaf and Shah (2017) use the market-based measure Tobin’s Q. The market-based measures take future expectations into account when valuing a business, which allows capturing the expected future progress of firms (Plenborg, 2002; Fernandez, 2007; Jennergren, 2008). Tobin's Q is considered to be a credible market- based measure, due to its ability to capture the underlying risk which further enables a fair view of a firm's current and future performance (Wu, 2011; Baños-Caballero et al., 2014).
Previous studies (Wang, 2002; Kieschnick, Laplante & Moussawi, 2013; Baños-Caballero et al., 2014; Aktas et al., 2015; Afrifa, 2016; Sawarni et al., 2020) also use different measures when they interpreter the efficiency of WCM. The most common way to examine WCM efficiency is either by net working capital (NWC), net trade cycle (NTC), or cash conversion cycle (CCC). As shown in table 1, different relationships between WCM and firm performance are presented depending on the country, time period, and measures.
6
Table 1. Literature review
Study Country Context Period Market-based firm
performance Significant result: WCM
& Performance Afrifa, (2016) UK Non-financial
listed firms
2004-2013 Tobin’s Q NWC: Concave
CCC: Convex Aktas et al., (2015) US Non-financial
listed firms
1982-2011 Excess stock return NWC: Concave
Baños-Caballero et al., (2014)
UK Non-financial quoted firms
2001-2007 Tobin’s Q NTC: Concave
Kieschnick et al., (2013)
US Non-financial listed firms
1990-2006 Excess stock return NWC ↓ Sawarni et al., (2020) India Non-financial
listed firms
2012-2018 Tobin’s Q NTC ↓
Wang, (2002) Japan &
Taiwan
Non-financial listed firms
1985-1996 Tobin`s Q CCC ↓
Shin & Soenen,
(1998) Global Non-financial
non-listed and listed firms
1975-1994 Excess stock return NTC ↓
Accounting-based firm performance Enqvist et al., (2014) Finland Non-financial
listed firms 1990-2008 ROA CCC ↓
Jose et al., (1996) US Listed and non-listed firms
1974-1993 ROA
ROE CCC ↓
Yazdanfar & Öhman,
(2014) Sweden Non-financial
listed firms 2008-2011 ROA CCC ↓
Deloof, (2003) Belgium Non-financial listed firms
1992-1996 GOI CCC →
Gill et al., (2010) US Non-financial
listed firms 2005-2007 ROA CCC ↑
Table 1 notes: CCC: cash conversion cycle, NWC: net working capital, NTC: net trade cycle, Tobin’s Q: Market-based measure for firm performance, ROA: Return on assets ↓ denotes a significant negative relationship. GOI: gross operating income
Reading table 1, previous studies find contradictory results when examining the relationship between WCM and firm performance. Three different relationships are recurring, that is either a negative, positive, or concave relationship.
Sawarni et al. (2020) find a negative relationship between WCM and firm performance. They explain that the negative relationship is derived from an inverse relationship between firms with long inventory days and firm performance. They state that it’s negative to have capital blocked
7
in the form of inventory. Furthermore, firms with an efficient WCM experience higher Tobin’s Q, which reveals that the market responds positively to WCM efficiency (ibid). Kieschnick et al. (2013) conclude that firms' performance can be enhanced if organizations manage WC efficiently. This is also supported by Enqvist et al. (2014) who state that firms’ can increase their performance by managing inventories efficiently and lower account receivable collection times.
Gill et al. (2010) find a positive relationship between WCM and firm performance arguing that lower levels of WC could possibly result in a lower level of sales. They argue that higher levels of WC will enable firms to meet unexpected increases in demand by having a larger amount of investment in inventory. This is in line with Sharma and Kumar (2011) who argue that longer CCC can help companies to meet unexpected rises in demand and therefore create a higher firm performance.
Baños-Caballero et al. (2014), Aktas et al. (2015), and Afrifa (2016) find that there is a concave relationship between WCM and firm performance. This means that the non-linear relationship is positive when a company holds a low level of investments in WC and negative when companies hold higher levels of WC. A firm can, by optimizing the level of WC, increase firm performance. This can be done either by reducing or increasing investments in WC since an incremental dollar invested in net WC is worth less than an incremental dollar held in cash (ibid). Furthermore, Kieschnick et al. (2013) consider that it’s several financial constraints that influence the value of additional investments in WC, such as a firm’s future sales expectations, debt loans, and bankruptcy risk.
8
2.2 Cash Conversion Cycle and Working Capital
The cash conversion cycle (CCC) expresses the time it takes for a firm to convert its investments in inventory and other resources into cash flows from sales. CCC measures how long each dollar is tied up in the production and sales process before it gets converted into cash received.
In detail, the CCC takes into account how much time the firm needs to sell its inventory, how much time it has to pay its bills, and how much time it takes to collect receivables (Mathur, 2003).
Figure 1: Components of the cash conversion cycle Source: Mathur, (2003)
As previously mentioned, firms can choose to have either a short or long CCC. If a firm has a longer CCC, it has made large investments in WC. This strategy can be beneficial since a longer CCC might increase firm performance due to a greater opportunity to meet unexpected rises of demand (Sharma & Kumar, 2011). On the other hand, if the organization chooses to have a shorter CCC, it will increase the amount of cash in the company at the expense of not being able to meet unexpected rises in demand. However, the increased amount of cash can be invested in growth activities which in turn increases the firm performance (Sawarni et al., 2020).
9
Another way to streamline the WC is by shortening the maturity payment date of account receivables, like in figure 1 (Mathur, 2003; Aktas et al., 2015). Managing account receivables is a way to increase the liquidity in the company as well as ensuring that customers pay their invoices on time (Aktas et al., 2015). Yazdanfar and Öhman (2014) state that firms that strive for a short CCC can reduce the account receivables period, extend the supplier credit term, and lower the inventory. By utilizing a short CCC strategy, the firm performance can be improved.
A short CCC includes a lower inventory and extended supplier credit term which leads to a higher operating risk. It is, therefore, a delimitation between firm performance and risk (ibid).
Wang (2002) examines how different companies manage their CCC by separating top- performing firms and bottom performing firms. By assessing top-performing firms as firms with Tobin’s Q higher than 1 and bottom performing firms as firms with Tobin’s Q lower than 1, the result shows that top-performing firms have shorter CCC than bottom performing firms.
This result is in line with Filbeck et al. (2017) finding that top-performing firms have shorter CCC than bottom performing firms. Filbeck et al. (2017) separate the top and bottom performing firms by categorizing top-performing firms as the 25 percent firms with the highest profitability and the 25 percent with the lowest profitability as the bottom performing firms (ibid). However, higher-performing firms are often superior compared to their competition in many aspects of their business. For example, top-performing firms could have better products, business plans, suppliers, or financing which leads to better terms of payment with both customers and suppliers (Newbert, 2008).
2.3 Business Cycles
Business cycles have different phases, downturn, slump, recovery, and boom, which can be described as fluctuations in economic activity in the long-term development of the economy (McInnes, 2000). Enqvist et al. (2014) state that the different phases in a business cycle affect WCM practices differently depending on which financial planning a company has established.
Companies need to take into account the different phases when forming a WCM strategy and determining their WCM practices to stay competitive (ibid). However, there is a difference of opinion between researchers whether efficient WCM is more essential during economic booms or downturns (Abuzayed, 2012) where one group of researchers believe that WC efficiency is more essential for firms during the booming economic periods. Further, other researchers believe firms can by strategically managing the WC improve their competitive position and
10 firm performance (ibid). Other researchers emphasized that WCM is more important for firms to withstand the impacts of economic turbulence (Enqvist et al., 2014).
Einarsson and Marquis (2001) investigate to what extent firms use external financing to finance their WC requirements during different phases of business cycles. They find that firm's external financing is countercyclical, and it increases in economic downturns. This means that firms are more willing to lend capital during economic downturns since capital is usually tied up in inventory and firms need to cover its liquidity shortage (ibid). Braun and Larrain (2005) support the findings that external financing is more pronounced and important during an economic downturn. They further show that firms’ composition of external respectively internally generated financing is an important factor to be able to meet future obligations during an economic downturn, wheres firms dependent on external financing will be more affected during a downturn. Further, Roberts (2003) argues that companies with the possibility to invest during economic downturns often benefit strongly compared to their competition in the long-run, which reinforces the previous authors' arguments.
Enqvist et al. (2014) also point out that the economic condition has an effect on an organization’s focus. During economic booms, firms do not focus on CCC to the same extent as during economic downturns. Their focus is instead on revenue and earnings growth. By establishing a WCM strategy that shortens the CCC, especially during an economic downturn, firms can enhance their profitability. This view is confirmed by Scholleova (2012) who find that companies shortening their CCC during the recession of 2008 withstood the financial crisis more successfully. Shortening the CCC releases more liquidity available for the company to fulfill its financial obligations and enhance firm performance (ibid). Further, large investments in WC mean that firms have capital tied up in inventory, and during a recession is liquidity preferred since an economic downturn can reduce the demand for the firm's products (Braun &
Larrain, 2005). During the recession of 2008, one industry that started to improve its CCC efficiency was the auto industry. The auto industry’s average CCC changed from 106 days, during the time period 2006 to 2008, to 35 days, 2012 to 2014. This shows that the economic downturn created a new insight into how the CCC can be handled to create an optimal WC efficiency (Schoar & Zou, 2017).
11 2.4 Hypothesis
Previous studies show the importance of WCM and the link to firm performance in different markets (Wang, 2002; Deloof, 2003; Gill et al., 2010; Aktas, et al., 2015; Altaf & Shah, 2017).
As mentioned before, previous studies show conflicting results when examining CCC and firm performance based on both accounting-based and market-based measures. Enqvist et al. (2014) state that firm performance depends on different phases of a business cycle, Jose et al. (1996) imply a shorter CCC to improve profitability while Gill et al. (2010) argue for a longer CCC.
When examining Tobin's Q as a firm performance instead of accounting-based measures, shorter CCC has been shown to have a positive effect on firm performance (Wang, 2002;
Sawarni et al., 2020). On the other hand, this has never been examined in the Swedish market.
Overall, Swedish companies tend to be less efficient in their WCM than other European firms (Kortman et al., 2017). This may imply that those Swedish firms which have a higher CCC efficiency will result in higher firm performance. Hence, a negative relationship is assumed between CCC and firm performance on the Swedish market which, result in hypothesis 1:
H1a: There is a negative relationship between CCC and firm performance on the Swedish market.
Enqvist et al. (2014) investigate the relationship between WCM and firm performance during different phases of a business cycle. They find that during an economic downturn, the importance of efficient WC is more pronounced than during an economic boom (ibid). There are many ways to estimate firm performance (Jose et al., 1996; Shin & Soenen, 1998; Wang, 2002; Lazaridis & Tryfonidis, 2006; Enqvist et al., 2014) but there is a lack of studies examining Tobin's Q as a firm performance measure during different phases of a business cycle. Tobin's Q enables estimations of the present and expected future stage of firm performance, compared to accounting-based firm performance measures (Campbell & Mínguez-Vera, 2008).
Therefore, the study will investigate whether firm performance in terms of Tobin's Q will follow the same trend as Enqvist et al. (2014) findings who use an accounting-based performance measure. Hence, the following hypothesis is established:
H1b: The relationship between the CCC and firm performance is different during economic downturns and economic booms on the Swedish market.
12 3 Methodology
In the following section, the study first will present the observed time period and sample for the OLS regression. Secondly, the variables in the chosen model are presented followed with an explanation of how to interpret the variables in the regression analysis. Thirdly, the study explains the tests that are done for multicollinearity, normality, and homoscedasticity. Lastly, the study explains how the additional test is conducted.
3.1 Introduction
This study investigates the relationship between WCM and firm performance, and how different phases of a business cycle affect this relationship. The study follows previous literature (Jose et al., 1996; Wang, 2002; Baños-Caballero et al., 2014; Enqvist et al., 2014; Sawarni et al., 2020), using an OLS-regression to examine the relationship between WCM and firm performance. Furthermore, to enable investigation of the impact of different phases of a business cycle, this study follows the methodology of Enqvist et al. (2014). Hence, the study follows a deductive approach with previously established theories and models as a framework for the hypothesis in the study (Bryman & Bell, 2015). The study uses a quantitative research design using financial data on the Swedish market to enable analysis. The research design refers to the overall strategy of the method to ensure that the study will effectively address the research problem. In this study, the financial development of the firms is analysed over several years, using a longitudinal design (ibid).
3.2 Observed time period
The study investigates the relationship between WCM and firm performance during different time periods. The first time period investigated is between 2008 – 2018. Furthermore, to investigate the same relationship between different phases of a business cycle, the study divides the sample period into economic booms and downturns, with further clarification in the following text below. Hence, to enable an investigation during different time periods, three regressions are examined based on arguments of Figure 2 which will be further argued in the text below.
The starting point for the time period is 2008. According to SCB (2019), the negative development of GDP started in 2008 as a consequence of the financial crisis, hence 2008 act as the starting point. Further, the study uses 2018 as the ending point for the study because it’s the
13 most recent year with sufficient data from the Swedish Central Bank. The reference to different phases of a business cycle relates to the irregular fluctuations in economic activity, measured by real GDP, in the long-term development of the economy. To investigate WCM and firm performance during a business cycle, this study uses annual GDP changes as indicators of different phases. Where economic downturns are categorized as the years with negative GDP and vice versa (Enqvist et al., 2014). The sample period for the economic downturn is between 2008-2009. Furthermore, the sample period for the economic boom in the regression is between 2013-2018. Due to the negative GDP in 2012 and to avoid possible contaminating effects related to the financial crisis, the study excludes 2010-2012. The different sample periods enable comparison between the period after the financial crisis with the period during the financial crisis and therefore evaluate if the relationship between WCM and firm performance differs in economic downturns and booms.
Figure 2: Sweden’s Real GDP-growth 2008 - 2018 (%) Source: Central Bureau of Statistics (SCB), (2019).
14 3.3 Data, population and sample
The selection of data in this study is limited to public firms listed on Nasdaq Stockholm (small-, mid-, and large-cap). The selection is justified by the access of available information for which is required to use the regression model between 2008–2018 to investigate the aim of the study.
All the data is conducted from Thomson Reuters Datastream and the regression is executed in SPSS. From the Thomson Reuters database data on market value, net sales, cost of goods sold, inventory, accounts receivables, accounts payables, total assets, and total debt is collected. The sample size results in 492 firms listed at least one year during 2008-2018. Furthermore, to eliminate possible deceptive results, this study removes all the financial firms in the sample, which is in line with Deloof (2003), Baños-Caballero et al. (2014) and Enqvist et al. (2014).
Fama and French (1992) argue that financial firms' different capital structure will possibly have a deceptive outcome of the result, hence this study follows the same argument. After removing financial firms, the sample size results in 449 firms.
Table 2. Sample for the study
Sample Nasdaq Stockholm 2008–2018
Total No. firms 492
Financial firms 43
Total No. of. Firms for the study 449
Table 2: The sample for the study. The study considers all the public registered firms on Nasdaq Stockholm between 2008- 2018 and exclude financial firms and firms for which no accounting numbers are available. Firms with no observation for one particular year is only excluded for that specific year.
The study includes all firms, which are active at some point in time between 2008-2018, on the Stockholm Stock Exchange. During the time period, some firms have applied for bankruptcy and some are newly established. To mitigate the risk of survival bias results these firms are included in the study. Fama and French (2000) argue that firms with total assets below 10 million SEK can take the form of extreme values when including them in calculations based on quotations of equity to book value of assets. Furthermore, considering that the model has variables estimated by quote measure by market value plus debt to total assets (TA), following Fama and French (2000), firms with total assets below 10 million SEK are excluded from the sample. Firm year observations with missing data are also excluded. Hence, the total number
15 of firm-year observations is 2,526. To ensure that the data retrieved from Datastream is accurate and reliable in accordance with the actual firm figures, the collected data is randomly checked against the annual reports in the sample.
Table 3. Firm year observation
Firm year observation 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Sum
Economic downturn 222 230 452
Economic boom 221 217 228 239 248 255 1,408
Entire time 227 222 217 2,526
Exclusion TA 3 4 4 4 7 7 1 1 1 1 1 34
Table 3: Represent firm year observation for Downturn and Booms and entire period. The entire time period also includes the observations represented in the table for downturn and booms when examining WCM and firm performance with the OLS regression. Exclusion TA (Total asset) represent firm year observations with TA below 10 million SEK.
3.4 Regression Model and Analysis
The study is investigating the relationship between WCM and firm performance during different phases of a business cycle. To test the two different hypotheses, a multiple regression analysis is conducted. The variables of importance to interpret in the regression model are CCC to enable investigation of WCM and firm performance. In line with Enqvist et al. (2014), this study divides the sample into different time periods representing the economic downturn (2008- 2009) and economic boom (2013-2018), and also test the full-time period (2008-2018).
Thereby, the model enables the investigation of WCM and firm performance over different phases of a business cycle. Furthermore, to interpret the result the study either rejects or accepts the hypothesis on a significance level of 1%, 5%, and 10% (Pallant, 2016).
3.4.1 Regression model
Qit = β0 + β1CCCit + β2Size(Ln)it + β3Debtit + β4ROA+ β5Industry + β6Year + εit
Regression model: Dependent variable Q represent Tobin's Q which is the proxy for corporate performance. CCCexamine the efficiency measure for working capital management. Control variable; Size(Ln)as natural logarithm of sales as size, Debt as total debt divided by total asset as debt ratio, ROA as net income divided by total assets one year before, Dummy variables for industry and years.
16 3.4.2 Dependent variables
Firm performance is the dependent variable in the study measured as Tobin's Q. Tobin's Q is a predominated variable to measure market-based performance (Wang, 2002; Florackis, Kostakis, & Ozkan, 2009; Wu, 2011; Baños-Caballero et al., 2014; Sawarni et al., 2020).
According to Smirlock, Gilligan, and Marshall (1984), Tobin's Q is a more suitable measurement than accounting-based measures, due to the fact that market value captures more information concerning the firm. Accounting-based measures only capture the present stage of firms, while market-based performance measures capture firm risks and investors' expectations of the future progress of the firm (Campbell & Mínguez-Vera, 2008). Hence, market-based performance measures are more suitable since the measure allows consideration about the future potential of firms. Following Baños-Caballero et al. (2014) as well as many others (Wang, 2002; Thomsen, Pedersen & Kvist, 2006; Wu, 2011; Afrifa, 2016) Tobin's Q is calculated by using market capitalization plus the book value of debt divided by total asset:
Variable 1: Firm performance
Tobin's Q = (Market value of equity + book value of debt) / Book value of assets
Following Campbell and Mínguez-Veras (2008) way of interpreting Tobin's Q, values above 1 are interpreted as investors on the market expect that the firm will utilize the current resources in a way that creates higher value. Tobin's Q values below 1 mean the opposite expectation from investors.
3.4.3 Independent variables
The study applies the cash conversion cycle (CCC) as an independent variable to capture WCM.
Many previous studies that investigate the efficiency of WCM apply CCC as a measure (Jose et al., 1996; Wang, 2002; Lazaridis & Tryfonidis, 2006; Enqvist et al., 2014). CCC is used as a way to examine WCM by estimating the time it takes for firms to generate input to output (Jose et al., 1996; Deloof, 2003). In line with Jose et al. (1996) and Enqvist et al. (2014), CCC is calculated as CCC=DIO+DRO-DPO (Equations are presented below), which captures WC efficiency. Equation 2-5 in the following text represents the calculation of CCC, DIO, DRO, and DPO.
17 Variable 2: Cash Conversion Cycle
CCC= (Days of account receivables outstanding + Days of inventory outstanding) - Days of account payables outstanding)
The three components of CCC
Equation 1: Days of account payables
DPO (Days Payables Outstanding) = (accounts payables/Cost of goods sold) * 365.
Equation 2: Days of account receivables
DRO (Days Receivables Outstanding) = (accounts receivables/net sales) * 365
Equation 3: Days of outstanding inventory
DIO (Days Inventory Outstanding) = (inventory/cost of goods sold) * 365
3.4.4 Control variables
Control variables are added to the regression model to control for the potential influences of other variables then between the dependent and independent variable. In line with previous studies, the study controls for firm size, debt, ROA, industry and years when running the regression over the whole period (Deloof, 2003; Lazaridis & Tryfonidis, 2006; Baños-Caballero et al., 2014).
Firm size has by previous research shown a positive relationship with firm performance (Yazdanfar & Öhman, 2014). In line with Enqvist et al. (2014) and Baños-Caballero et al.
(2014) this study, therefore, use firm size (SIZE) based on the natural logarithm of sales as a control variable when investigating the relationship between WCM and firm performance.
Variable 3: Firm Size SIZE = In(Sales)
Firms' financial decisions and ways of financing their business have a relationship with firm performance (Yazdanfar & Öhman, 2014). Hence, the firm debt ratio is the second control variable, following Enqvist et al. (2014) and Baños-Caballero et al. (2014) calculation as total debt/total assets. When previous studies control for debt ratio, they find a negative relationship
18 between debt grade and firm performance, meaning that a higher level of debt leads to lower firm performance (Deloof, 2003; Baños-Caballero et al., 2014; Enqvist et al., 2014).
Variable 4: DEBT
DEBT = Total debt / Total assets
Return on asset (ROA) is a commonly used depictor for firm performance. In line with Baños- Caballero et al. (2014), this study adds ROA as a control variable.
Variable 5: Return on Assets ROA= Net income / Total assets it-1
This study is in line with Baños-Caballero et al. (2014) and Enqvist et al. (2014) which apply industry dummy variables to control for potential effects of industry differences. The firms are categorized and divided into nine industries: technology (D1), services (D2), industrials (D3), telecom, (D4), oil & gas (D5), consumer goods (D6), basic materials (D7), healthcare (D8) and real estate (D9). Furthermore, the study control for yearly effects by including yearly dummies to eliminate the possible influence of external effects.
Table 4. Variable description
Variable Type Definition Formula
Tobin's Q Dependent Market-based firm performance Market value of equity + book value of debt /Book value of assets.
Winsorized on 5% level CCC Independent Cash conversion cycle =Working capital
Efficiency
DIO + DRO - DPO. Winsorized on 5%
level
Debt Control Financial decision Total debt/Total asset
Firm size Control Sales numbers as the firm size Natural logarithm for sales
ROA Control Return on asset Net income/ Total assets t-1
Industry Control Firm specific characteristics Dummies for specific industry Year Control Economic condition characteristics Dummies for each sample year Table 4: DIO =day of inventory turnover, DRO = days of account receivables, DPO= days of account payables
19 3.4.5 Normality, skewness and kurtosis
Multiple regression is sensitive to outliers and requires normality within the distribution to accurately analyze the result. To identify if the data is normally distributed descriptive statistics are run which enables observations of skewness and kurtosis values. Skewed data set will result in a normal distribution curve pointing to the right or left and kurtosis with an abnormal peak.
The variables are assumed to influence the accuracy of the regression when skewness and kurtosis values of above 1 or below -1 are observed. (Pallant, 2016)
When first running the descriptives in SPSS the result indicates skewness and kurtosis outside the acceptance levels for Tobin's Q and CCC, which indicates that there are outliers within the dataset. To eliminate the impact of outliers among the dataset winsorizing is suitable (Pallant, 2016). Therefore, following Shin and Soenen (1998), this study winsorize outliers. This study winsorize Tobin's Q and CCC yearly on a 5% level since they showed the influence of outliers.
3.4.6 Multicollinearity and singularity
The study investigates if the model suffers from a multicollinearity problem by examining a variance inflation factor-test (VIF) and Pearson's correlation matrix. Multicollinearity problem signifies if the independent variables are correlated with each other which may impair the outcome of the regression. Results that imply multicollinearity are the ones with a correlation between two variables above 0,9/-0,9 or VIF value above 10 (Pallant, 2016). Before testing the regression, it is also essential to check if the independent variables will suit the model by measuring the correlation between the dependent and independent variables (ibid). The independent variables suited in the model can be check by a correlation matrix where the variables have to reveal a result above 0,3 as an acceptable level (ibid). When testing for multicollinearity by using collinearity statistics in SPSS the VIF values for the dependent variables as well as independent variables (except the yearly dummies and industry) is approximately 1. According to Pallant (2016) values below 5 indicate no multicollinearity issue, hence the impact of the explanation factor among the chosen independent variables is not a problem for the model.
20 3.4.7 Homoscedasticity and linearity
Homoscedasticity and linearity refer to the identification of the normal distribution among the variables in the dataset. Plotting the standardized residuals against the predicted residuals enables the investigation of homoscedasticity. To reject the possibility of homoscedasticity, the variance between the dependent variable and residuals should be similar to each other in the data set, which can be confirmed. Furthermore, linearity is also checked for in the data set between the dependent and independent variables, assuming a straight-line relationship of the coefficients in the partial plot. (Pallant, 2016)
3.4.8 Methodology criticism
The study uses secondary data and according to Bryman and Bell (2015), in order to construct the analysis correctly, there are some problems needed to be addressed. During the time period for the study, the secondary data is not standardized, and the development of accounting rules has led to different accounting techniques and definitions. The development of new accounting rules during the observed time period leads to discrepancies in the observed financial data, which in turn could result in an unavoidable degree of errors in the data. However, by controlling for year and industry in the regression model, the model takes this problem into consideration, which is further discussed within the selection of control variables.
Previous studies concerning WCM and firm performance on the Nordic market have examined CCC (Enqvist et al., 2014; Yazdanfar & Öhman, 2014). In line with Enqvist et al. (2014), this study uses CCC to examine the WCM, but there are other ways to capture WCM as the net trade cycle (NTC) examined by Shin and Soenen (1998). Both NTC and CCC capture the duration of the cash tied in the operating cycles and use account receivables, inventory, and payables to investigate WC efficiency. The difference between the two WCM measures is that NTC uses sales in the denominator compared to CCC which uses both sales and cost of goods.
Hence, when comparing the results in this study with other literature, which uses NTC, the indication of the results should be interpreted similarly since both variables capture WCM but the actual numbers are difficult to compare.
21 4 Result and analysis
In the following section, the result of the study will be presented. First, the descriptive statistic is presented. Secondly, the result of the correlation matrix is described. Third, the results of the regression are presented and analysed.
4.1 Descriptive statistics
Table 5 represents the variables of the essence to be able to answer the hypotheses in the study.
Those control variables excluded are dummies for years and industry. The exclusion is due to the fact that they are binary variables and thus not have any valuable descriptive interpretation.
All the variables are assumed to be normally distributed and have been tested in SPSS to confirm normality.
Table 5. Descriptive statistics
Variable n Mean Median Min Max Std
Q 2,526 2.0147 1.1948 0.3551 13.4135 2.3051
CCC 2,526 90.2868 80.9692 -105.1142 457.4488 76.8688
Size 2,526 14.3292 14.1993 2.6390 19.6288 2.1519
Debt 2,526 0.1952 0.1686 0 1.1605 0.1719
ROA 2,526 0.0381 0.0577 -1.0481 1.2842 0.1662
Table 5: The sample consist of descriptive statistic for the chosen variables in the OLS-model, n represent firm year observation.
Q represent the proxy for corporate performance. CCC examine the efficiency measure for Working Capital Management.
Control variable; Sizeas natural logarithm of sales, Debt as total debt divided by total asset as debt ratio, Dummy variables for industry and years.
The descriptive statistics in table 5 state that the average CCC for Swedish listed firms on Nasdaq is approximately 90 days. This result implies that in general, it takes 90 days for Swedish firms to turn input into output, in terms of cash. The standard deviation is approximately 77 days, which shows that it is a variety of the examined firm’s CCC in the data sample. The data sample has a wide interval between the minimum and maximum CCC, where the minimum is approximately -105 days and the maximum is approximately 457 days.
Negative CCC results, such as the minimum -105, reveals that firms conduct the payment 105 days before they have to pay for the goods sold.
The mean debt ratio in the data sample is approximately 0,195 (19,5%). This means that the average debt ratio of the examined data sample is 19,5%. The minimum and maximum interval
22 of the data sample’s debt ratio is between 0 and 1,1605. The standard deviation is approximately 0,17 (17%), meaning that there is a variety of firm's debt ratios when financing their business.
Furthermore, ROA has a mean value of approximately 0,0381 (3,81%) with a standard deviation of 0,1662 (16,62%). The minimum and maximum interval of the examined data sample are approximately between -1,048 and 1,2842.
4.2 Correlation matrix
The following result of the correlation matrix is presented in table 6. The correlation matrix, Pearson Correlation, presents the correlation between the model’s variables and measures the strength and direction between pairs of continuous variables. In addition, the correlation evaluates if there is a linear relationship between the statistical evidence (Pallant, 2016). The model’s dummy variables, years, and industry are excluded from table 6. Furthermore, a correlation of -1 is perfectly negative, meaning that the dependent variable moving in opposite directions as the independent variable. A correlation of zero means that there is no correlation.
A correlation of +1 means that the correlation is perfectly positive, which means that the dependent variable moves in the same direction as the independent.
Table 6. Pearson Correlation Matrix
Variable Q CCC Size Debt ROA
Q 1
CCC -0,013 1
Size -0,151*** -0,075*** 1
Debt -0,108*** -0,050*** 0,205*** 1
ROA 0,035 -0,111*** 0,323*** -0,076*** 1
Table 6 shows the correlation between the variables in the model. The study examines non-financial Swedish listed firms on Nasdaq Stockholm during between 2008-2018. *, **, *** refers to significant on a 10 %, 5 % or 1 % level. The Q and CCC are yearly winsorized on 5%.
Table 6 shows the correlation between the variables for the time period 2008 to 2018. CCC has a negative correlation at the 1 % level with ROA (-0,111), and a weak linear relationship with size (-0,075) and debt (-0,05). Furthermore, the variable size has a positive correlation with debt (0,205) and ROA (0,323). The correlation between debt and ROA is less than +0,1 and
23 more than -0,1, which implies that there is a weak linear relationship. The strongest correlation is between the independent variables size and debt (0,205), and size and ROA (0,323). Since the correlation is below 0,9 there is no sign of multicollinearity between the variables.
Furthermore, the VIF value is below 5 which according to Pallant (2016) is the highest acceptable level. Therefore, the VIF test also confirmed that there is no sign of multicollinearity.
4.3 Regression result and analysis
The following result of the regression model is presented in Table 7 below. The regression model (1) represents the relationship between WCM and firm performance during the whole time period. Model (2) and (3) represent the same relationship for downturn respectively boom.
Table 7. Regression result
Model 1 Model 2 Model 3
Expected direction 2008-2018 Downturn Boom
(Intercept)
CCCit - -0,061***
(0,001)
-0,064 (0,002)
-0,046*
(0,001)
Size - -0,114***
(0,024)
-0,141***
(0,066)
-0,120***
(0,030)
Debt - -0,083***
(0,271)
0,064 (0,753)
-0,146***
(0,337)
ROA + 0,076***
(0,003)
0,009 (0,007)
0,119***
(0,004)
Adj. R2 8,2% 5,8% 11,6%
Yearly dummy Yes Yes Yes
Industry Dummy Yes Yes Yes
P-value 0,000 0,000 0,000
Observations 2,526 452 1,408
Table 7: Represent the result from regression during the whole sample period (2008-2018), downturn (2008-2009) boom (2013- 2018) on non-financial firms listed on Nasdaq Stockholm. The result in the patentees represent the standard error, *, **, *** refers to significant on a 10 %, 5 % or 1 % level. CCC and Q are winsorized on 5% level. Yearly and industry effects are controlled for respective firm year.
The results in model (1), representing the time period between 2008-2018, show a statistically significant negative relationship between firm performance measured as Tobin’s Q and CCC.
24 The CCC variable is -0,061, meaning that if the CCC will increase with 1 day, Tobin's Q will on average decrease with 0,061 if the other variables stay the same. The regression model (1) is significant at a 1 percent level with an adjusted R2 at 8,2%. All control variables, size, debt, and ROA are also statistically significant at the 1% level. Size has a negative beta of -0,114, debt has a negative beta of -0,083, and ROA has a positive beta of 0,076. Due to the significant negative relationship between CCC and firm performance hypothesis H1a is accepted.
H1a: There is a negative relationship between CCC and firm performance on the Swedish market. – Accepted
The result of this study, as previously mentioned, show a negative relationship between CCC and firm performance. The negative relationship between CCC and firm performance implies that firms with shorter CCC, on average, add more value to their shareholders compared to firms with longer CCC, which is in line with what Sawarni et al. (2020) state. Furthermore, the negative relationship may be explained by many reasons. One reason can be that shorter CCC increases the liquidity in the company (Yazdanfar & Öhman, 2014). The higher amount of liquidity can result in more value-adding through relocation of resources to more profitable investments since less financial resources are tied up in e.g inventory. The previous argument is supported by Sawarni et al. (2020) who argue that it’s negative to have capital blocked in the form of inventory. This is further argued by Afrifa (2016) who states that firms can enhance shareholder value by using an optimal level of cash invested in WC. Another interesting consideration is that Yazdanfar and Öhman (2014) find a negative relationship between CCC and firm performance by using an accounting-based measure (ROA) on the Swedish market.
This study investigates the same relationship by using a market-based measure (Tobins’Q).
Hence, efficient CCC seems to have an impact on both market and accounting-based firm performance measures on the Swedish market.
It is noteworthy that previous studies (Deloof, 2003; Lazaridis & Tryfonidis, 2006; Gill et al., 2010; Abuzayed, 2012; Kieschnick et al., 2013; Enqvist et al., 2014; Sawarni et al., 2020) examine WCM and firm performance in different markets. According to Abuzayed (2012), the contradictory results from previous studies could be explained by the different characteristics between countries. Abuzayed (2012) states that it’s more difficult for companies in developing countries to access external financing and therefore the focus on shorter CCC is more crucial.
They further argue that CCC would be of less importance in developed countries. However, the
25 negative results in this study between CCC and firm performance could indicate contradiction compared to Abuzayed (2012) since Sweden is categorized as a developed country (Swedish Central Bank, 2019). This indication could also be further supported by Kieschnick et al. (2013) who show a negative relationship between NTC and Tobin’s Q on the US market, which can be regarded as a world-leading economy.
The adjusted R2 is 8,2% in the model (1) and shows a significant relationship. This result implies that the model (1) explains 8,2% of the variation for Tobin's Q, which is considered low compared to Sawarni et al. (2020) showing an adjusted R2 by 16,77% and Afrifa (2016) showing an adjusted R2 by 14,21%. Since this study's adjusted R2 is considerably lower than Afrifa (2016) and Sawarni et al. (2020) the result in this study should, therefore, be discussed with caution. Furthermore, debt shows a negative relationship with Tobin's Q which is in line with Baños-Caballero et al. (2014), and the negative result of size is in line with Sawarni et al.
(2020). Thus, the result for size and debt implies that higher-performing firms are smaller and have less debt to total assets. Furthermore, ROA shows a positive significant relationship with Tobin's Q. This result implies that firms with higher accounting-based performance also have a higher market-based performance.
Moreover, another interesting consideration highlighted by Newbert (2008) is that higher- performing firms are superior compared to their competition in many aspects of their business.
For example, top-performing firms could have better products, business plans, suppliers, or financing which leads to better terms of payment with both customers and suppliers (ibid). One line of thought, based on these arguments, is that the result in the model (1) possible could be explained by the opposite direction of the relationship between CCC and firm performance.
Hence, from the previous argument, it could be argued that shorter CCC doesn't directly lead to higher firm performance, but instead, that the competitive advantage that top-performing firms often have, leads to a shorter CCC. The shorter CCC could occur due to top-performing firms' ability to optimize trade credits which shortening CCC.
The regression model (2) representing the economic downturn shows an adjusted R2 of 5,8%
at a 1% significance level. The CCC shows a non-significant negative relationship on -0,064.
The control variable size shows a negative relationship of -0,141, debt shows a positive relationship of 0,064, and ROA shows a positive relationship of 0,009. However, the only observed significant control variable is size, which is significant at the 1 percent level. The
26 regression representing the economic boom, model (3), shows a significant negative relationship on -0,046 between firm performance and CCC at the 10 percent level. Model (3) shows a 1% significant adjusted R2 on 11,6%. The control variable size shows a negative relationship of -0,120 with a significant level at 1 percent. Debt shows a negative relationship of -0,146 with a significant level of 1 percent and ROA shows a positive relationship of 0,119 with a significant level of 1 percent. The regression representing the economic downturn (2) shows indications of a greater negative relationship between firm performance and CCC compared to the period representing the economic boom (3). However, due to the non- significant result for the relationship between CCC and firm performance during the economic downturn, hypothesis H1b is rejected.
H1b: The relationship between the CCC and firm performance is different during economic downturns and economic booms on the Swedish market. – Rejected
The result from the regression shows a trend of a greater negative result during the economic downturn compared to the economic boom which is in line with Enqvist et al. (2014) and Filbeck et al. (2017). However, the result of this study is non-significant compared to Enqvist et al. (2014) and Filbeck et al. (2017) findings. To elaborate, one possible explanation for the different results can be explained by the different firm performance measures each literature use. Enqvist et al. (2014) use an accounting-based measure derived from the actual outcome of firm performance during a fixed time period, as the past performance. This study instead uses a market-based measure, Tobin's Q, which enables capturing the expectation of the future progress of the firm (Campbell & Mínguez-Vera, 2008), regardless of business cycles. Hence, market-based performance measures consider a longer time horizon while accounting-based performance measures are limited to a fixed time. Therefore, it is possible that the non- significant result in this study occurs due to the fact that CCC efficiency doesn't matter for the market-based performance during an economic downturn.
Notably, Filbeck et al. (2017) argue that a possible explanation to a more pronounced negative relationship during economic downturns is that the liquidity for many firms comes under pressure. This condition, which often occurs during economic downturns, emphasizes the importance of efficient WCM practices according to Scholleova (2012). This study shows a negative CCC during the downturns but since the result is non-significant, the results of this study can’t support the previous argument on the Swedish market.
27 The result from the downturn (model 2) and the boom (model 3) show an adjusted R2 of 5%
respectively 11,6%. The result explains 5% respectively 11,6% variation for Tobin's Q in each model. This study's adjusted R2 is therefore considerably lower than Enqvist et al. (2014) who show approximately adjusted R2 of 23%. This study's result should, therefore, be discussed with caution. However, since this study shows a higher Adjusted R2 in the model (3) than (2) and a negative result for CCC on a 10% significance level during the boom period (3) but not during the downturn (2), the result may reveal some interesting insight about firms' CCC strategy. To elaborate, Swedish firms may have begun to adopt a WCM strategy after the financial crisis of 2008, where attention towards efficient CCC becomes essential for firms to improve and stay competitive. The previous argument is supported by Singh and Kumar (2014) who argue that the financial crises during 2008 have brought attention towards a WCM strategy for firms. Schoar and Zou (2017) also show the change in WCM strategy after the financial crisis of 2008, where firms began to improve WCM by shortening CCC. Furthermore, Arbuzayed (2012) argues that WCM during an economic boom is important for firms to improve because it enhances competitive advantage. Hence, the possible explanation of the opposite result in this study compared to Enqvist et al. (2014) can be the different time frames each study investigates where it becomes more essential with improved WCM the last years regardless of the phase of the business cycle. However, as mentioned earlier, the result in this study is not statically proven during the economic downturn and the previous elaboration is just speculations.
