2. Literature review

sector's share in GDP is 32% [2]. The share of the industry sector in exports is around 92% [4]. In this case, the industry sector will remain important for the Turkish economy in the future. The industry sector, which plays a key role in the Turkish economy, is faced with many problems such as lack of qualified workforce, inadequacy of infrastructure and technology, weak competition power, and difficulties in marketing and financing. The financing problem is one of the most important problems faced by these firms. These firms need to be able to use their existing resources effectively and be self-sufficient because of the scarcity of funding resources and the insufficient accumulation of capital. Working capital, which is seen as the lifeblood of a business, has an important role in the return of the owner's reckoning, and has a

Firms need working capital to begin its business operations, carry on its activities efficiently, and meet its short run obligations [6]. Working capital management is concerned with the dayto-day activities rather than long-term investment decisions [7]. Working capital is a part of firm's current assets, which are converted into cash within a year or less [8]. In this sense, working capital components (WCC) are cash, cash equivalents, inventories, accounts receiv-

Investment in the working capital components is important for all industrial enterprises to be powerful financially. A firm can collect its receivables in a short time and restrict credit sales to reduce account receivables and increase cash inflows. However, rigid sales policies and low credit sales would lead to loss of sales, thus causing profits to fall [6]. On the other hand, high inventory levels and flexible credit sales policy can contribute to increased sales. Because sales on credit allow the customer to examine the product before paying, it may increase sales [9]. There are some advantages to work with high inventory levels such as preventing customer losses caused by not having enough stock level and protecting against price volatilities [10]. However, the high inventory and loose trade credit policies lead to the locking of the money to the working capital [9]. In this context, firms that invest heavily in inventory and accounts receivables may be exposed to low profits [11]. Another component that has an impact on the working capital requirement is accounts payables. Deferment of payments to suppliers can enable the firm to evaluate the product bought and may be a cheap and flexible funding source. But, postponing payments can be expensive, if the firm has got a discount for early payment [9]. In this case, the level of accounts payables of the firm may affect the firm's profitability.

The style of WCM may have a considerable influence on the profitability, risk, and liquidity of the firm [12]. The firm that invests more in current assets is more liquid than a firm that does not invest. This will reduce the firm's liquidity risk, while decreasing overall rate of return, because the return of current assets is less than the return of other assets [13]. While lower investment in the working capital expressed as aggressive working capital policy is associated with higher returns and higher risk, more investment in the working capital expressed as conservative working capital policy is associated with lower return and lower risk [14]. The firm has to choose between aggressive and conservative working capital policies depending on its purpose [15].

Effective WCM is a significant factor affecting the survival of the firm, the continuity of its activities, and the maintenance of liquidity and profitability [16]. Excessive working capital

decisive influence on liquidity [5], is important at this point.

206 Financial Management from an Emerging Market Perspective

ables, and accounts payables.

There are studies in the literature that examine WCC-firm's profitability tradeoff in terms of different countries and different sectors. The findings obtained from these studies vary depending on the method and data set used. Some of these studies are summarized in Table 1.




Table 1. Overview of the studies about WCC-firm's profitability tradeoff.

#### 3. Methodology

Authors Sampling Variables Method Results

Independent variables: Debtors days, inventory days, creditors days, cash velocity, working capital policy, net working capital leverage, size, current

Dependent variables: Return on asset

Independent variables: Receivables Collection Period, inventory turnover period, payable deferral period, cash conversion cycle, current ratio and

Dependent variables: Gross profit ratio Independent variables: Inventory turnover ratio, receivables turnover ratio, payable deferral period, net trade

Control variables: Ratio of financial fixed assets, firm size, financial

Dependent variables: Profitability Independent variables: Cash management, account receivables, inventory and account payables

Dependent variables: Return on assets,

Dependent variables: Net operating

Independent variables: Number of days accounts payable, number of days inventories, the number of days accounts receivables and cash

Control variables: firm size board size.

Dependent variables: Gross profit ratio Independent variables: Inventory turnover ratio, receivables turnover ratio, payable deferral period, net trade

Dependent variables: Firm's

Independent variables: Cash

Control variables: Sales growth, depth

Independent variables: Cash conversion cycle, inventory conversion period, account receivable period, accounts payable period and current

and net profit margin

regression analysis

Single regression analysis

Panel regression analysis

Regression analysis

Panel regression analysis

Panel regression analysis

Panel regression analysis

Multivariate regression model

and debtors days with the

The meaningful relationship exists between the firms' profitability and the working capital components

The relationship between gross profit ratio and independent variables is negative

Cash, account receivables and inventory except accounts payables have a positive relationship with profitability.

Return on assets has a negative relationship with account receivable period and cash conversion cycle while having a positive relationship with

Cash conversion cycles have a negative relationship with net

current ratio.

operating profit

The existence of firms' profitability- working capital components tradeoff is invalid.

Cash conversion cycle is negatively related to firm's

profitability

profitability.

Monitoring Indian Economy

208 Financial Management from an Emerging Market Perspective

corporations listed on Dhaka Stock Exchange in Bangladesh

trade firms listed on BIST in Turkey

industry listed on Karachi Stock Exchange in Pakistan

firms listed on BIST in Turkey

In [22] Manufacturing

In [20] Production and

In [27] Firms in textile

In [29] Manufacturing

In [36] Manufacturing firms in Egypt, Kenya, Nigeria and South Africa

In [19] Firms in the retail sector listed on BIST in Turkey

In [8] Cement companies in Kenya

ratio

quick ratio

cycle

tobin-q

ratio

profit

cycle

profitability

conversion cycle

ratio and current ratio

conversion cycle

leverage ratio

In the study, the impact of WCM on profitability is analyzed via panel data regression model. The panel data models, which allow more consistent estimation results by including both time and cross-sectional properties, are modeled in different ways according to effect of the cross section and time properties [30]. In this context, the models in which both constant and slope parameters are constant with respect to cross section and time unit are called as pooled panel data models and are defined as follows:

$$Y\_{it} = a\_0 + \sum\_{k=1}^{K} \alpha\_k X\_{kit} + e\_{it}; \ \mathbf{i} = 1, 2, \dots \\ \text{N}; \ \mathbf{t} = 1, 2, \dots, \mathbf{T} \tag{1}$$

The subscript i in the model is a cross-sectional unit such as an individual or a firm; t represents the time dimension. Yit is the dependent variable, and Xkit denotes k independent variables with cross sectional unit i and time t. In the model, α<sup>k</sup> is the vector of the (kx1) size parameter that does not vary according to the i cross-section unit and time dimension, and α<sup>0</sup> is also the constant term. eit is the error term that is independent and identically distributed with 0 mean and σ<sup>2</sup> variance for all i cross-section units and t time periods (IID) [31].

If both the time and the cross-section are affecting the model, the panel data model takes the name of the two-way panel data model. The model is called as a one-way panel data model if the effect is only a cross-sectional unit or a time effect.

In the case where unit and/or time effect cause changes in some or all of the parameters of the model, the panel data models are named fixed effects panel data model. If the fixed effects model is one way, model is shown as following:

$$Y\_{it} = \alpha\_i + \sum\_{k=1}^{K} \alpha\_k X\_{kit} + e\_{it} \tag{2}$$

Similar to previous model, Yit is the dependent variable and Xkit denotes k independent variables with cross sectional unit i and time t. α<sup>i</sup> is the individual specific coefficients for the crosssectional unit, while the t time dimension is constant. Similarly, α<sup>k</sup> is the vector of the (kx1) size parameter that does not vary according to the i cross-section unit and t time dimension. The model is also named as covariance model or dummy variables model. Unobserved individual effects are achieved by using specific dummy variables:

$$Y\_{it} = \mu\_i D\_N + X\_{it}' \beta + \varepsilon\_{it} \tag{3}$$

DN is the vector of dummy variables [30]. If the model contains both cross section and time effects, the two-way fixed effect model is determined as the following model:

$$Y\_{it} = \mu\_i + \lambda\_t + X\_{it}'\beta + e\_{it} \tag{4}$$

Xit is the vector of independent variables. In the two-way fixed effects models, μ<sup>i</sup> is the individual specific coefficients, λ<sup>t</sup> is the time effects, and β is also the vector of coefficients [30].

The model in which the cross section and/or time effect is included as a component of the model error term is defined as the random effects model. If the random effects model is one way, model is generally expressed as:

$$Y\_{it} = \alpha\_i + \beta X\_{it} + \mathfrak{e}\_{it} \tag{5}$$

$$
\alpha\_i = \alpha\_0 + \mu\_i \tag{6}
$$

$$
\mu\_{it} = \mu\_i + \mathfrak{e}\_{it} \tag{7}
$$

As explained in the fixed effects model, Yit is dependent variable, and Xit is the vector of independent variables. Individual effects consist of a combination of α0, which does not have unit and time effects, and μi, which contains the specific cross section effects. The cross section effects and eit error term are added to the model as a component of uit error term, and the model is predicted with act of knowledge [31]. If both the specific unit effects μ<sup>i</sup> and the specific time effects λ<sup>t</sup> are expressed as a component of the error term eit, the two-way random effects model is mentioned. The two-way random effects model is determined as,

$$Y\_{it} = \alpha\_0 + \beta X\_{it} + \varepsilon\_{it} \tag{8}$$

$$
\varepsilon\_{it} = \mu\_i + \lambda\_t + u\_{it} \tag{9}
$$

The Hausman (1978) test determines whether the fixed effects model or the random effects model is appropriate for panel data analysis [31]. Hausman suggests that the null hypothesis for the test is an appropriate model of the random effects model, which implies that there is no relationship between cross section and explanatory variables [32]. The alternative hypothesis indicates that the appropriate model is the fixed effect model. Hausman test statistic (H) is estimated by the following formula using the variance covariance matrix:

$$H = \left(\widehat{\boldsymbol{\beta}}\_{FE} - \widehat{\boldsymbol{\beta}}\_{RE}\right)^{\prime} \left(\boldsymbol{V} \left(\widehat{\boldsymbol{\beta}}\_{FE} - \widehat{\boldsymbol{\beta}}\_{RE}\right)\right)^{-1} \left(\widehat{\boldsymbol{\beta}}\_{FE} - \widehat{\boldsymbol{\beta}}\_{RE}\right) \tag{10}$$

Hausman test statistics fits the asymptotic χ<sup>2</sup> distribution with parameter k. V is the variance covariance matrix of the difference between the estimators. βbFE and βbRE are the fixed effects and random effects estimators, respectively. As a result of the analysis, it is determined whether the predicted model is a fixed effects model or a random effects model [30].

### 4. Data and variables

variables with cross sectional unit i and time t. In the model, α<sup>k</sup> is the vector of the (kx1) size parameter that does not vary according to the i cross-section unit and time dimension, and α<sup>0</sup> is also the constant term. eit is the error term that is independent and identically distributed

If both the time and the cross-section are affecting the model, the panel data model takes the name of the two-way panel data model. The model is called as a one-way panel data model if

In the case where unit and/or time effect cause changes in some or all of the parameters of the model, the panel data models are named fixed effects panel data model. If the fixed effects

K

k¼1

Similar to previous model, Yit is the dependent variable and Xkit denotes k independent variables with cross sectional unit i and time t. α<sup>i</sup> is the individual specific coefficients for the crosssectional unit, while the t time dimension is constant. Similarly, α<sup>k</sup> is the vector of the (kx1) size parameter that does not vary according to the i cross-section unit and t time dimension. The model is also named as covariance model or dummy variables model. Unobserved indi-

DN þ X<sup>0</sup>

DN is the vector of dummy variables [30]. If the model contains both cross section and time

Xit is the vector of independent variables. In the two-way fixed effects models, μ<sup>i</sup> is the individual specific coefficients, λ<sup>t</sup> is the time effects, and β is also the vector of coefficients [30]. The model in which the cross section and/or time effect is included as a component of the model error term is defined as the random effects model. If the random effects model is one

As explained in the fixed effects model, Yit is dependent variable, and Xit is the vector of independent variables. Individual effects consist of a combination of α0, which does not have unit and time effects, and μi, which contains the specific cross section effects. The cross section effects and eit error term are added to the model as a component of uit error term, and the

Yit ¼ μ<sup>i</sup> þ λ<sup>t</sup> þ X<sup>0</sup>

αkXkit þ eit (2)

itβ þ eit (3)

itβ þ eit (4)

Yit ¼ α<sup>i</sup> þ βXit þ eit (5)

α<sup>i</sup> ¼ α<sup>0</sup> þ μ<sup>i</sup> (6)

uit ¼ μ<sup>i</sup> þ eit (7)

with 0 mean and σ<sup>2</sup> variance for all i cross-section units and t time periods (IID) [31].

Yit <sup>¼</sup> <sup>α</sup><sup>i</sup> <sup>þ</sup><sup>X</sup>

Yit ¼ μ<sup>i</sup>

effects, the two-way fixed effect model is determined as the following model:

the effect is only a cross-sectional unit or a time effect.

vidual effects are achieved by using specific dummy variables:

model is one way, model is shown as following:

210 Financial Management from an Emerging Market Perspective

way, model is generally expressed as:

In this study, the tradeoff between WCC and firm's profitability is examined via the annual data for the period 2005–2016 of 41 firms listed on BIST Industrial Index in Turkey. In order to examine WCC firm's profitability tradeoff, dependent variable is defined as return on assets (ROA); independent variables are cash conversion cycle (CCC), inventory conversion period (ICP), payables deferral period (PDP), and control variables are sales growth (SG), the ratio of short-term financial debts to short-term debts (FDSD), and the ratio of fixed assets to total assets (FATA).

ROA widely used and accepted as measure of profitability [23] indicates the rate of return provided by firm's assets [13]. CCC measures the effectiveness of the working capital [9, 22]. CCC expresses the time spent between the expenses for purchasing raw materials and the collection of sales [9, 11, 12]. Longer CCC means the more investment in the working capital [9, 11], in other words, the more current asset financing needs [8]. CCC consists of three components: receivables collection period, ICP, and PDP. ICP refers to the time required for the conversion of raw materials to finished goods and then the sale of these products. PDP is the average time firm's suppliers give it to pay for its purchases [33]. The other component of CCC, receivables collection period, was not included in the study, since this variable was not statistically significant in the models formed. SG, FDSD, and FATA as control variables were also used to increase the reliability level of the models established in the study [34]. All variables and its formulations in the study are shown in Table 2.


Table 2. Descriptions of the variables.
