OPERATING RISK AND FIRM’S PROFITABILITY: EVIDENCE FROM NON-LISTED MANUFACTURING FIRMS FROM V4 COUNTRIES

Purpose . This study contributes to the ongoing debate on determinants of firms’ resilience strategies while facing a crisis, by referring to the operating risk. The purpose of our work is to examine the effects of operating risk on profitability of non-listed firms in manufacturing sector. Therefore, our study responds to the literature gap within investigating this problem at sector level, and among smaller firms. Methodology . Our evidence is based on accounting-based data, covering four countries (Poland, Czechia, Slovakia and Hungary) and large number of non-listed firms. We apply WLS (Weighted Last Squared) regression to test two models of the impacts of operating risk on firm’s profitability. Findings. Our results indicate that firms exposed to higher operating risk, tend to be less profitable. While considering cross-effects of operating and financial risk in our model, we found strong and sound evidence that the interplay between operating and financial risk exposure exerts a negative impact on firm’s profitability. These results have important implications, by demonstrating that during the period of crisis firms need equity financing to sustain resilience.


Introduction
The operating risk (also referred to as business risk) increases dramatically if a firm faces the decrease in sales revenues and/or the increase of operating costs, in particularthe fixed costs.In such conditions, a firm loses its abilities to earn operating profit (Brigham, Ehrhardt, 2011).The COVID-19 pandemic outbreak has shown how disruptive could be the loss of income, that Guided by the above outlined literature gaps, the purpose of our study is to examine the effects of operating risk on the performance of non-listed firms that operate in manufacturing sector, in four emerging European countries of our interest.In methodical context, our work is guided by the recent studies of Grau and Reig (2020) and Chen et al. (2019), who applied regression model to explain profitability (proxied by return on firm's assets) and operating leverage, as the main explanatory variable and the proxy of firm's operating risk (consistently with O'Brien, Vanderheiden, 1987;Houmes et al., 2012;Harjoto, 2017).With a considerable panel of around 7300 firm-year observations, we have found sound evidence on the negative effects of operating risk on firm's performance.
The remainder of this study is organized as follows.In the first section we develop research hypotheses, with reference to the existing literature.In the second section we explain research design and method.Third section presents and discusses the results.In the final section we conclude, by addressing the implications of our work and outlining the avenues for further research.

Literature review
Firm's operating profitability depends on the interplay of its operating costs and sales revenues, which is clearly highlighted in the break-even point analysis (Brigham, Ehrhardt, 2011).If the firm faces the decrease of sales revenues, and/or increase in operating costs, the ability to achieve the operating profit is constrained.The operating risk itself results from the gearing effect of the structure of operating costs, if we consider the fixed and variable costs (also known as operating leverage effect) (Lev, 1974).The costs' structure is predominantly driven by the structure of assets, which remains determined by firm's investment decisions (O'Brien and Vanderheiden (1987), Houmes et al. (2012), Harjoto (2017)).In this regard, the operating risk is determined by the organisation of internal operating processes in the firm.However, there are branches that per se are more exposed to operating risk, as their performance requires high investments in fixed assets (e.g.heavy industry).Also, firms that operate in the branches where high volatility of demand is possible, are regarded as exposed to greater operating risk, due to the instability of sales revenues.The operating leverage effect demonstrates, that if sales revenues decrease, firms' with greater level of operating risk (determined by the interplay of fixed and variable assets), are exposed to greater loss of operating profits.The degree of operating leverage is helpful in capturing this effect and in quantifying operating risk exposure, as it measures of how much operating profits could be lost due to change in sales revenues (Brigham, Ehrhardt, 2011).
Prior literature provides rich evidence on the effects of operating leverage (and operating risk) on profits.Interestingly, Chen et al. (2019) have shown that operating leverage is significantly and positively associated with profitability, in the favourable and expansionary times.However, in accordance to the trade-off theory, the prevalent evidence confirms the negative impact of operating risk (and high degree of operating leverage) on firms' performance, if measured with profitability.In the SMEs context, prior works confirm the relevance of firm's size if the association of operating risk and profitability is considered (Lopez-Valeiras et al., 2016).Thus, in our study that is focused on non-listed (and thus not large) firms operating in manufacturing sector, our first hypothesis is: H1: Operating risk is negatively associated with firm's profitability The association between operating risk and firm's profitability is more complex, if we consider the total risk of a company by considering the impact of financial risk.Following Hamada (1972) observations, the decisions on how to finance firm's assets (structure of debt and equity) determine the financial leverage and the related financial risk, and add to the operating risk exposure.Greater degree of debt financing, relative to equity, adds to the gearing effect, due to the cost of interest rates.Thus, if the sales revenues decline, firms with greater burden of debt and cost of interests to be covered, face more difficulties in achieving profitability (Brigham, Erhardt, 2011).The interplay between operating and financial risk has been widely researched, and in many instances confirmed the negative association between financial risk and profitability and the offset of these two types of risk (Kahl et al., 2014, Cao, 2015).However, Chen et al. (2019) have noted that firms control the level of their operating risk to avoid insolvency (which is the outcome of too high financial risk exposure).They observed that when facing the arrival of a crisis period, firms with higher operating risk exposure tend to minimize the financial risk exposure determined by their capital structure decisions.In this context, it is very important to properly understand the interplay between the increased operating risk and insolvency, that could be amplified by shortening of funds or more restricted conditions of obtaining additional funding.This interplay is critical for avoiding the bankruptcy threat, through conscious and well-thought managerial decision making while facing external difficulties.Following this observation, our second hypothesis we test in this work, with reference to non-listed firms operating in manufacturing sector, is that: H2: Operating risk is negatively associated with firm's relationship between financial risk and profitability.

Sample and data
For the purposes of this work, we use the EMIS database that provides the accounting-based financial data from firms' financial statements, for firms that perform in Emerging European Economies (EMIS, 2023).The database covers the firms which perform i.a. in the four Visegrad Group countries of our interest (hereafter referred to as the V4 countries), namely: the Czech Republic, Hungary, Slovakia and Poland.EMIS database offers the option to request data for a single sector.Thus, we have filtered out financial data of firms that declared manufacturing sector as the main sector of their performance (following the NAICS codes 31-33).
From the EMIS database we have requested financial data in 2017-2019 time-span, to obtain a complete dataset for the pre-pandemic period.We excluded 2020 and 2021 from our analysis.The major reason was that there is some strong evidence that pandemic has caused serious disturbances in firms' performance, and this effect is visible in the firms' financial results for 2020 and 2021, regardless their sector belonging (Hu, Zhang, 2021;Shen et al., 2020).Furthermore, in numerous instances the EMIS dataset was incomplete for 2021 (due to the lagged process of data collection).
Nevertheless, in our request for financial data in 2017-2019 time span, we have controlled for the maximum records obtainable for manufacturing sector , given the 2017 as the first year of our interest.We have also controlled for the overall number of firms that perform in the manufacturing sector in the period of our interest, given the Eurostat dataset (Eurostat, 2017).In Table 1 we explain on how we derived the final number of firm-year observations, under above specified terms of request.Data provided in Table 1 indicate, that there is a slight discrepancy between the number of observations provided in the EMIS database, and according to Eurostat dataset.The EMIS database covers a relatively higher number of records for Hungary and Poland (c.a.14%), in comparison to Czech Republic and Slovakia (2% and 4.6%, respectively).Source: Own study.
However, the ultimate number of records we have downloaded from the EMIS database was guided by the specifics of the dataset.The EMIS database ranges the observations in accordance to the volume of sales revenues, and the number of missing records increases with the number of downloaded observations (at firm level).We have confirmed this for Poland, where the number of missing accounting based financial data was considerable for a sample exceeding 1500 records.In particular, we observed that although the major data was available (sales revenues or the volume of assets), the detailed accounting-based figures were missing.Thus, we have decided to apply the cut-off point, that offers the same number of entry records for Czech Republic, Hungary and Slovakia (guided by the number of available records for these countries in EMIS database), and slightly greater number of records for Poland, which has a considerable number of firms operating in manufacturing sector, according to the Eurostat, and is also widely covered by the EMIS database.
Further, we have carefully verified the number of missing records of the accounting-based financial data relevant to our examinations.We have also controlled for the potential bias in the dataset, by revising the balance between assets and total liabilities, or the components of assets (fixed and current) and the components of funds (equity and short and long term liabilities).As it can be seen in Table 1, the number of missing or biased records was considerable for Poland.After the data-controlling procedures we have implemented, we received 7338 firm-year observations, suitable for empirical analysis.At country level, the sample is relatively balanced, given the country representation in EMIS database and overall Eurostat statistics.
In Table 2 we provide the structure of our sample (on firm-year observation level), given the distribution of the sub-sectors of manufacturing industry.Transportation equipment manufacturing and food manufacturing firms are prevalent in our sample, with 16.38% and 15.85% share, respectively.Source: own study.

Variables
We proxy firm's profitability with ROA (return on assets), computed as operating income to assets (consistently with Grau and Reig, 2020).We adjust the measurement of return on assets to the operating level, by applying operating income in the nominator, to capture the direct effects of firm's operating performance.Thus, we avoid the drawbacks of net income based ROA, which is the potential impact of tax policy at country-level, or extraordinary items.Adjusted ROA is also in line with the operating leverage measurement procedure and break-even-point analysis approach (Brigham, Ehrhardt, 2011).The operating risk is our main explanatory variable.Following prior literature, we assume that a good measure of firm's operating risk is its operating leverage.However, measuring operating leverage for non-listed firms is quite challenging, as these firms in majority prepare simplified financial statements, where the structure of the operating costs is not detailed.Thus, for the non-listed firms we are not able to soundly use the cost-based measures (e.g.costs of goods sold (COGS), or sales, general and administrative expenses (SG&A)) to determine the level of operating leverage (e.g.Chan et al., 2019;Novy-Marx, 2011).However, in numerous empirical works the operating leverage is proxied with assets structure based measure, namely the share of fixed assets in total assets.The rationale behind is that high level of fixed assets generate high fixed costs (O'Brien, Vanderheiden, 1987;Houmes et al., 2012;Harjoto, 2017).Thus, our main explanatory variablethe operating riskis proxied by a ratio of fixed assets to total assets ORISK.This approach was also adopted by Grau and Reig (2020), in their agri-food sector study.We also control for operating risk effect with dummy variable ORISK_dum.Firms with ORISK above the average measured at sub-sector level, are considered as those of higher operating risk, the remainingas of lower operating risk.
Following Grau and Reig (2020), we control for a range of additional main explanatory variables.The first is the financial risk (FRISK), by considering financial leverage (debt to assets).Financial leverage is also a good indicator of firm's indebtedness, as it shows the share of debts in capital structure.The prevalent prior evidence indicates a negative regression slope between profitability and financial leverage.We also control for firm's size (SIZE), by applying the natural logarithm of sales revenues as a proxy of firm's size.Sales revenues are also informative in the context of firm's market power (Grau, Reig, 2020).Further, we control also for firm's age (AGE), which is captured by the number of years since firm's performance.In our sample we included only the firms that perform min.for 4 years, to eliminate the very young firms, that have just entered the market.Young firms are regarded as those with good growth prospects and more dynamic on the market, which could stimulate their sales revenues, but on the other hand facing greater costs pressure (Martínez-Sola et al., 2014).
Finally, we incorporate three control variables.We control for firm's financial liquidity (LIQ_CR), as the firms of poor financial liquidity are financially constrained, which may exert an impact on their returns.The cash ratio (CASH) is a measure of the available financial slack, that informs on firm's available cash holdings that could be used as a buffer against the adverse market conditions.In this regard, cash ratio is an indicator of firm's resilience to negative impact of unforeseen external shocks.Finally, we control for operating profit margin (OPM) that explains the firm's earnings power, given the difference between the total operating costs and the level of sales revenues.
The definitions of the variables are provided in Table 3. Overall, the definitions of our control variables (see Table 3) are consistent with prior literature, in particular Rajan and Zingales (1995), Welch (2004), Kayhan and Titman (2007), and Lemmon, Roberts, and Zender (2008).The variables we control in this study are guided by prior empirical evidence, given their explanatory effect on profitability (e.g.Chaddad, Mondelli, 2013;Chen et al., 2019;Lopez-Valeiras et al., 2016;Grau, Reig, 2015;Yazdanfar, Öhman, 2015).In the appendix, we provide the descriptive statistics for the variables, and the detailed specification of means of our dependent and main .In Table 5 we additionally provide a detailed specification of the means of our dependent and main explanatory variable (ROA and ORISK), in cross-country dimension.

Model and empirical approach
In methodical context, our work is guided by the sector-oriented study by Grau and Reig (2020).More specifically, we follow their procedure in terms of controlling the effects of operating risk on firm's performance.We estimate and contrast two models.Model 1 includes tests all variablesthe operating risk; financial risk, size and age as three additional main explanatory variables, and the three control variables:  =  0 +  1  +  2  +  3  +  4  +  5   +  6  +  7  +  (1) Model 1 helps to verify the first hypothesis we posit in our investigation (H1).Model 2 adds cross-effect ( * _), to better understand if the effect the main variable on profitability depends on whether a firm is of high or low level of operating risk:  =  0 +  1  +  2 ( * _) +  3  +  4  + + 5   +  6  +  7  + (2) In Model 2, following Grau and Reig (2020) approach, we excluded ORISK.Model 2 is applied to verify our second hypothesis (H2).The cross-effect approach is consistent with prior literature in the field (Kestens et al., 2012;Baños-Caballero et al., 2012, 2014;Cen et al., 2015).
As our empirical investigations use accounting-based financial data, there are some limitations for using OLS (Ordinary Least-Squares) regression in modelling.Financial data are often heteroscedastic, thus OLS regression could provide biased estimations, as pointed by Strutz, 2016).We confirmed this problem for our dataset, by performing Breush-Pagan test (p-value<0.000).To handle this issue, we perform WLS (Weighted Last Squared) regression (Ozkan, Ozkan, 2004).We have also winsorised the data at 1%.
In Table 4 we present the Pearson's correlation between the examined variables.Similarly for Grau and Reig ( 2020) study, we have obtained strong correlation for two sets of variables.The first set are two measures of operating performance -ROA and OPM, as both indicate the cost pressures, relative to sales revenues.The second set are two measures of financial distressfinancial leverage FRISK and financial liquidity LIQ_CR, that are interlinked with the debt level.Thus, while running WLS regression models we controlled for multicollinearity of the variables, by applying VIF (Variance Inflation Factor) test.The VIF test, however, has confirmed no multicollinearity (as VIF was below 5 for each set of variables in the tested models).Thus, the strength of correlation in our models was not biasing the results.

Results and discussion
The regression results are provided in Table 5. Model 1 was designed to test the effects of all variables we consider in this study.With 0,1% of statistical significance, this model shows that operating risk is a fundamental determinant of profitability in non-listed manufacturing companies, operating in V4 countries.The negative regression slope between ROA and ORISK indicates that the higher is the operating risk, the lower is firm's profitability.Thus, our first hypothesis (H1) found strong support.This observation is consistent with Grau and Reig (2020) findings for SMEs operating in agri-food industry.In addition, model 1 indicates that more profitable are the non-listed manufacturing companies that are younger, larger, of better finan-cial liquidity, and have higher operating margin (all significant at 0,1%).We have also confirmed that financial risk is positively associated with firm's profitability in our sample.This suggests that the positive effects of financial leveragemore debt in capital structure, enhance firm's profitability.Model 2 was designed to test the cross-effects of operating leverage and financial risk (one of the main explanatory variables).There is an interplay between the effects of operating leverage, and the effects of financial leverage.Typically, firms with a greater degree of operating leverage are exposed to greater operating risk, and for more risky firms, the required rate on return of debt financing could be higher.In our model the cross effect is statistically significant at 0,1%, and the negative regression slope for the cross effect ( * _) is negative.This suggest that the ultimate effect of operating and financial leverage on firm's profitability is negative.Our results confirm that if we study the financial risk effects in isolation, then there it has a positive impact on profitability (and positive effects of financial leverage).However, by considering the interplay of operating and financial risk in a company, embedded in cross-effects, the joint impact of the operating and financial risk on firm's performance is negative.It gives a strong support for our second hypothesis, stating that operating risk is negatively associated with firm's relationship between financial risk and profitability.

Conclusions
In this work, we have revised the interplay between firm's profitability and operating risk.The problem is timely, as refers to the discussion on firm's resilience capabilities, in response to a shock (such as the pandemic outbreak and lockdown).The disturbances in firm's performance, that result in inability to generate sales revenues, and at the same time followed by the increased operating costs, are particularly difficult for firm's profitability, and in the longer run impact firm's ability to survive.In this regard, to develop the efficient resilience strategies, the effects of greater operating risk on firm's profitability, in terms of other determinants of financial performance, are critical.
Our empirical work contributes to this challenging endeavor, by revising a sample of nonlisted manufacturing firms, that operate in four European emerging economies (the V4 coun-tries).Our work fills in the important gap, by providing evidence for smaller firms (as nonlisted companies are usually those who operate on smaller scale), which is rare.The prevalent evidence exists for large, listed firms.Moreover, we provide sound evidence at sector level, by investigating single isolated sector (manufacturing sector).Finally, in our study we provide evidence for the cross-effects of operating and financial risk, to capture the major determinants of firm's profitability.
Our findings indicate that firms exposed to higher operating risk tend to be less profitable (in line with our first hypothesis).We have also confirmed that higher operating risk is negatively associated with the relationship between firms financial risk and profitability.In other words, high operating risk is negatively associated with the potential positive effects of financial leverage.
Our study has several important implications.First of all, it confirms that the external shocks that result in the increase of firm's operating risk level (due to lower sales revenues and/or higher operating costs), threaten firm's profitability significantly, in two dimensions.The first dimension one is due to the direct impact of firm's operating risk on profitability, due to the effects of financial leverage.The second dimension confirms the interplay between operating risk and financial risk, that could be even more devastating for firm's profitability.Once the shock increases the operating risk and the first tensions, a company may seek financial support, by applying more debt financing.The potential positive financial leverage effects, however, are neutralized and reversed by the pressure of operating risk.Thus, the increase of debt financing is not leading to the improvement of firm's performance.This observation indicates that smaller firms, that perform in manufacturing sector, need strong equity financing support, to sustain resilient to external shocks.
Our study is limited to one sector (manufacturing), as this sector is the largest one.However, further investigations should be performed, as the operating risk depends strongly on the type of sector.Our study is also limited in regional context, as it focuses on four countries, which we treat as homogenous in terms of some features of their economic performance, being the non-core European countries.However, further inquires shall revise if similar observations stem from e.g.core European Union countries, to detect the possible impacts of country or cultural related variables.Our study is also limited to the smaller firms, that are not listed on the stock exchange.On one hand, this exerts further limitations in data availability and simplified empirical approach as regards the selection of variables.However, the existing literature evidence is scarce for smaller firms, thus adds value and relevance to our findings.To avoid biased results, we extracted a wide range of data, and adjusted the empirical procedure to obtain statistically sound results (by applying WLS instead of OLS regression, and controlling heteroscedasticity and multicollinearity of variables).
Our study is also limited by time horizon of the analysis.Our intention was to cover a period of relatively undisturbed firms' performance, with systemic shocks.Thus, we adopted the backward looking approach, by collecting data until 2019, as COVID-19 was very disruptive to firms' performance.Further studies could address the interplay between operating risk and firm performance during the shock, in line with Chen, et al. 2019, who covered the crisis 2008+ to confirm if the observed effects are stronger in unfavorable times.Also, the study could be replicated for post-COVID-19 period (starting from 2022 onwards), to confirm the impacts of operating risk on unlisted firms' performance.

Table 1 .
The procedure of sample composition

Table 2 .
Sample composition: structure of observations in sub-sector dimension

Table 3 .
Definitions of the variables Source: Own study.

Table 4 .
Pearson's correlation coefficients between the examined variables

Table 5 .
WLS regression results for ROA

Table 1B .
Means of ROA and ORisk, at country and sub-sector level Source: Own study.RYZYKO OPERACYJNE A