1 Introduction
The impact of the COVID-19 pandemic was particularly strong, and on Small and Medium Enterprises (SMEs) stronger than that of the global financial crisis in 2008. In normal credit conditions, banks tend to consider SMEs as high-risk borrowers, because they are subject to liquidity shortfalls and bankruptcies reflecting conditions such as asymmetric information and agency problems, liquidity and profitability conditions and their ownership structure. Although relationship banking prevents errors in SME lending and in evaluating the creditworthiness of firms (Baas and Schrooten 2006), external shocks can severely disrupt this form of lending.
At the same time, especially since the 2008 crisis, the demand for a new economic and sustainable recovery has been recognized as increasingly important. European Union (EU) policy has placed great emphasis on a green economic growth (European Commission 2018, 2019). Greening the economy will have big impacts on many industries such as agriculture, transport, extractive industries, manufacturing, construction, and services, etc., in which SMEs prevail at local level. SMEs are in fact crucial for most national economies in the EU-28: they account for more than 99.8% of all businesses and generate more than 56% of value added and about 66.6% of employment.
Advertisement
In this context we explore the impact of being “green” for SMEs as a response to variability in the environment of corporate business. Because SMEs do not often have an ESG rating and because there are significant divergences in the ESG ratings published by six important rating agencies (Berg et al. 2022), we focus on “green” companies in industries characterized by a particularly high green component. We ask two main questions. First, did green SMEs have better access to external financial credit than non-green SMEs during the COVID-19 pandemic? Second, during the COVID-19 pandemic, did green SMEs rely more heavily on trade credit than banking and financial credit compared to non-green SMEs?
We concentrate on Italian SMEs, because in Italy the shares of employment and value added generated by SMEs are particularly high (78.1 and 66.9% respectively) compared to the EU average. We focus on manufacturing firms for two reasons. First, firms in manufacturing sectors have a greater concentration of tangible assets (e.g., higher liquidation value) and have better access to debt financing (both trade and bank credit). Secondly we wish to avoid industry-specific firms, such as farms, in agriculture, and industries which may not be sufficiently clearly defined, such as services. Our results if, from one side, appear to exclude the existence, during the COVID-19 pandemic, of a better access to financial credit for green SMEs than for non-green ones, from the other side, confirm a higher importance of trade credit for green SMEs than for non-green ones. The reasons of this relationship are probably complex and can only partially be explained by poor access to bank credit for marginal firms, since on average green SMEs seem to be more profitable than non-green ones. On the contrary, this relationship could also be explained according to the literature which sees trade credit as a component of a long-term portfolio management strategy, i.e., as a tool for consolidating relationships with clients, for price discrimination and/or for increasing firm profitability in variable demand conditions.
This study makes a significant contribution to the literature in two ways. First, it investigates the relationship between the “green” component and the degree of economic and financial resilience of Italian manufacturing SMEs during the COVID-19 pandemic. Second, it aims to verify whether an orientation to the green economy during a complex economic shock like COVID-19 increases or decreases the importance of trade credit relative to bank/financial credit in financing the firm.
The rest of the paper is structured as follows. Section 2 reviews the literature and describes the hypotheses. Section 3 and 4 describe our sample, the variables used, the methodologies we used in our analysis and the results. Section 5 concludes.
Advertisement
2 Literature review and hypotheses development
The impact of the COVID-19 pandemic has been investigated in many countries (Kuckertz et al. 2020) and in many industries (Demirguc-Kunt et al. 2021; Donthu and Gustaffson 2020; García-Carbonell et al. 2021; OECD 2020). Although the global financial crisis in 2008 severely affected financial risk of SMEs (Cowling et al. 2018), the effects of COVID-19 on SMEs are thought to be even more severe (Baker et al. 2020; Bartik et al. 2020; International Trade Centre 2020; Howell et al. 2020). For example, social venture crowdfunding faced numerous difficulties during the crisis (Farhoud et al. 2021), start-ups and small businesses in China (Brown et al. 2020) and in England (Brown and Rocha 2020) underwent a big decline in equity investments in the first quarter of 2020 compared with same quarter in 2019. Although SMEs enjoy greater flexibility and adaptive capacities (Bartz and Winkler 2016; Battisti and Deakins 2012; Burns 2016; Gilmore et al. 2013) than bigger firms, at the same time they have a higher probability of failure than large and established firms (Berger and Udell 1998; Davidson and Gordon 2016; Doern et al. 2019; Herbane 2013, 2019; May and Lixl 2019; Beck and Demirguc-Kunt 2006; Juergensen et al. 2020), partly due to their ownership structure (Martin et al. 2019).
From a theoretical standpoint, the effect of environmental responsibility on firm financing during the COVID-19 pandemic can be explained in positive terms for two reasons. First, although authorities and academics use differing definitions of the green economy (Caprotti and Bailey 2014; UNEPP 2011),1 empirical investigations in EU have shown that providing solutions to environmental problems can effectively create new sources of growth (Kasztelan 2017; Lavrinenko et al 2019; Šipilova et al. 2017). The COVID-19 pandemic renewed the emphasis in the literature on the importance of green growth in terms of a synergic relationship between economic growth and the environment (Guerin and Suntheim 2021; Kang and Lee 2021; Mol and Sonnenfeld 2000; O'Callaghan and Murdock 2021). In this context, recent studies (D’Amato 2021; D’Amato and Korhonen 2021; Korhonen and Granberg 2020; Palahì et al. 2020; Taherzadeh 2021) focus on the integration of the green economy, and a circular economy and bioeconomy are recognized as essential to sustainable development policies. Second, there is evidence that greater environmental orientation may decrease the likelihood of negative events both at the firm level (Bouslah et al. 2018) and at the economic level, because such firms, enjoying established reputations as being environmentally responsible, are able to access financial resources more easily (Branco and Rodrigues 2006; Godfrey 2005; Zeidan et al. 2015). On the other hand, if financial institutions and markets do not adequately recognize environmentally responsible practices as an important intangible asset in lowering credit risk, firms may suffer competitive disadvantage because investment in environmentally responsible practices can be costly in the short term and lead to low profitability (Goss and Roberts 2011). The environmental performance of firms has been extensively examined by researchers and policymakers (Banerjee et al. 2019; Bragdon and Marlin 1972; Bruna and Nicolo 2020; Lahouel et al. 2020; Muhammad et al. 2015; Porter and Van der Linde 1995; Williamson et al. 2006). Although proactive environmental practices seem to imply low levels of risk (Godfrey et al. 2009; Muhammad et al. 2015), reductions in the cost of debt (Bauer and Hann 2010), easier access to financial markets (Jo and Na 2012), and better conditions on loans (Goss and Roberts 2011; Magnanelli and Izzo 2017; Sharfman and Fernando 2008), other studies demonstrate that financial risk incurred by socially and environmentally responsible firms is higher than other firms (Kiernan 2007; Seeger and Hipfel 2007), and in some cases investigations are inconclusive (Lee and Faff 2009). A recent study (Wellalage and Kumar 2021) on 3915 unlisted firms in developing countries indicates that firm-level environmental performance has a positive impact on the loan size for firms, particularly small firms. In general terms, firms that exhibit proactive environmental practices are expected to carry a lower level of risk (Godfrey et al. 2009), can access the financial market less easily (Farza et al. 2021; Jo and Na 2012), and have greater leverage (Sharfman and Fernando 2008). These differences are thought to be particularly marked for manufacturing firms because there is evidence that firms in manufacturing industries have better access to debt (banking and trade financing) than non-manufacturing firms. This may reflect the greater concentration of tangible assets (Van Der Wijst and Thurik 1993; Jordan et al. 1998) or lower information asymmetries (La Rocca et al. 2010) between manufacturing compared to non-manufacturing firms. Other research on trade credit finds that firms in traditional or manufacturing sectors obtain trade credit more easily than firms in non-manufacturing industries (Mian and Smith 1992; Psillaki and Eleftheriou 2015). These results are confirmed by additional research in Europe (Casey and O’Toole 2014) and Japan (Taketa and Udell 2007).
Although most of the studies outlined above investigate the relationship between environmental performance and financial risk in normal economic conditions, we argue on the basis of these findings that green SMEs are more likely to have been more financially resilient than non-green SMEs during the COVID-19 pandemic. Our first hypothesis is as follows:
Hypothesis 1
Green SMEs show a higher proportion of external funding/total assets than non-green SMEs both during the COVID-19 pandemic and in normal times (where external funding is represented by trade credit + banking credit).
There is overwhelming evidence to support the positive effect of environmental performance on firm performance, but there is as yet little evidence on the impact of a financial crisis on the relationship between social performance and firm risk (Bouslah et al. 2018; Lins et al. 2017, 2019; Marsat et al. 2020). The relationship between environmental performance and financial risk during the COVID-19 pandemic has been recently investigated on a sample of 3356 MSMEs (micro and small-medium sized enterprises) located in southern and eastern Europe (Wellalage and Kumar 2020) and on a sample of 6597 SMEs located in eastern Europe (Wellalage et al. 2021). The findings suggest that green companies showed a lower probability of liquidity shortfall and bankruptcy during the pandemic, and also the existence of a significant and positive relationship between environmental orientation of the firm and access to external finance. Better environmental performance appears to reduce the level of idiosyncratic risk as perceived by stakeholders (Bouslah et al. 2018), increase the access to financial resources (Zeidan et al. 2015) and boost the restoration of stakeholder trust following periods of crisis (Pricewaterhouse Coopers 2013). These results are consistent with research on listed firms (Farza et al. 2021) before the COVID-19 pandemic: green investments enhance resource efficiency and corporate reputation with corresponding effects on financial performance. The theoretical explanation of the relationship between environmental and financial performance could be that a higher level of trust between firm and shareholders leads stakeholders to increase their level of collaboration and reciprocation and enhances the firm’s reputation (Lins et al. 2017, 2019). High trust levels also boost innovation, which is in turn a competitive advantage in times of crisis, allowing firms to resist and even flourish (Huang et al. 2020) and they provide greater advantages than protection against idiosyncratic firm-specific legal risks (Hong and Liskovich 2019). Higher levels of trust can also extend to supplier companies, which can take advantage of more detailed information on their green clients, including lower levels of risk than non-green clients. Thus, our next two hypotheses are as follows:
Hypothesis 2a
Green SMEs rely on higher levels of trade credit than banking/financial credit both during the COVID-19 pandemic and in normal times.
Hypothesis 2b
Green SMEs depend on higher levels of trade credit than non-green SMEs both during the COVID-19 pandemic and in normal times.
The literature on trade credit shows that its relationship with bank credit is complex (Mateut et al. 2006; Matias Gama and Van Auken 2015; McGuinness and Hogan 2014; Wilner 2007). It may appear that trade credit is a substitute for bank credit, because it is mainly requested by companies with poor access to bank credit (Fisman and Love 2003; Nilsen 2002) especially in cases of financial crisis (Love et al. 2007) during monetary restrictions (Brechling and Lipsey 1963; Duca 1986; Herbst 1974; Jaffee and Modigliani 1969; Jaffee 1971; Mateut 2005; Meltzer 1960; Wilner 2000) and in countries with a poorly developed local banking system (Alessandrini et al 2009; Benfratello et al. 2008; Bonaccorsi di Patti and Gobbi 2001; Gagliardi 2009; Guiso et al 2004; La Rocca et al 2010; Petersen and Rajan 1995). But while borrowers are likely to view bank and trade credit as substitutes, trade credit can be considered as complementary to financing by financial intermediaries, who lend to suppliers who in turn relend to their clients (Carbo-Valverde et al. 2016; Choi and Yungsan 2005; Cull et al 2009; Demirgüç-Kunt and Maksimovic 2001; García-Teruel and Martínez-Solano 2010; Garcia-Appendini and Montoriol-Garriga 2013; Jain 2001; Love et al 2007; McMillan and Woodruff 1999; Ogawa et al 2013; Tsuruta 2015). Recourse to trade credit can also be explained by a competitive advantage of supplier companies over banks in the exploitation of informal means that guarantee the repayment of the loan. This competitive advantage may derive from better and/or less expensive information on the financial situation of client firms (Biais and Gollier 1997; Petersen and Rajan 1997; Pike et al. 2005), monitoring advantages (Bukart and Ellingsen 2004; Emery 1987; Freixas 1993; Schwartz and Whitcomb 1979), and product market imperfections (Brennan et al 1988). In this context, some studies (Cannari et al. 2004; Ng et al. 1999) demonstrate big differences across industries in trade credit terms but little variation within industries across time, so that trade credit can be viewed as a component of a long-term portfolio management strategy (Emery 1987), a tool for consolidating relationships with clients reflecting guarantee of product quality (Deloof and Jegers 1996), for price discrimination (Bougheas et al. 2009; Lee and Stowe 1993; Long et al. 1993; Petersen and Rajan 1997; Schwartz and Whitcomb 1978, 1979), for increasing firm profitability (Martínez-Sola and García-Tereul 2014) and for dealing with variable demand conditions (Long et al. 1993). Since trade credit can be used to finance purchases (accounts payable) and clients (accounts receivable), studies examining the relationship between accounts payable and receivable conclude that there is sufficient evidence to support the matching hypothesis, which is that trade debt is influenced by trade credit policy (Bastos and Pindado 2013; Fabbri and Klapper 2008; García-Teruel and Martínez-Solano 2010; Mian and Smith 1994; Paul and Wilson 2007) and some evidence that accounts payable in one year depend on those of the previous year (Bussoli 2017). These results suggest that operational conditions, transaction costs and firm business environment influence the demand and supply of trade credit (Summer and Wilson 2000) and could explain persistent time invariant aspects of trade credit within industries and high heterogeneity across industries (Wilson and Summers 2002; Marotta 2005) in this case too particularly in cases of monetary tightening (Dedola and Lippi 2000; Guiso et al. 2000). Even the importance of trade credit in the different stages of the business life cycle of SMEs depends on the sector, which is extremely important in the financial decisions of SMEs (Psillaki and Eleftheriou 2015; Yazdanfar and Öhman 2017).
In this context, a recent empirical analysis (Arcuri and Pisani 2021) of Italian Medium-sized Enterprises (MEs) examines the relationship between trade credit and green-oriented firms. A panel analysis is applied to 101,250 observations over the period 2010–2019. The results show that, although trade credit is more important for younger, smaller, less profitable and less liquid MEs, green MEs rely more on trade credit than non-green MEs. They offer more trade credit to their clients and at the same time they receive more trade credit than non-green. In addition, green MEs tend to rely on trade credit, regardless of the stage of the company’s life cycle, more stably than non-green MEs. The closer appreciation and knowledge of green manufacturing by companies in same industry compared to the appreciation of green ME characteristics by banks and financial intermediaries may explain these results.
3 Data, sample and methodology
3.1 Data and sample
To answer our research questions, we download the entire records of Italian SMEs from the AIDA Bureau van Dijk database. We aim to cover the periods before and during the COVID-19 pandemic, so our sample includes all companies for which data are available for the period 2017–2020. Our final sample consists of 53,891 SMEs. Following European Commission2 definitions, a small enterprise has between 11 and 50 employees, between 2 and 10 million euro annual turnover, and between 2 and 10 million euro balance sheet total. A medium-sized enterprise has between 50 and 250 employees, between 10 million and 50-million euro annual turnover, and between 10 and 43 million euro balance sheet total. The SMEs included in our sample belong to an “industry in the strict sense”. In the ATECO 2007 code, the classification of economic activity used by ISTAT (the Italian National Institute of Statistics), these are found in Sections B-E. Section B is “Extraction of minerals from quarries and mines”, Section C is “Manufacturing activities”, and Sections D and E are “Supply of electricity, gas, steam and air conditioning” and “Supply of water; sewer networks, activities of waste management and remediation”.
We apply the Differences-in-Differences methodology (diff-in-diff) and carry out a panel analysis and OLS regressions with 215,564 observations, of which 6844 refer to firms belonging to the “green” sector. The concept of “green economy” (Barbier 2012) is perceived as a pathway to sustainability by international organizations such as The World Bank (2012), the United Nations Environment Programme (2011), and the European Commission (2018). A study by the Politecnico di Milano and Camera di Commercio di Milano (2012) identifies specific “green” sectors on the basis of the following concepts. First, the green economy has as its pillars resource efficiency and natural capital. And second, an inclusive green economy is associated with economic growth, human development and opportunities, for people—to improve the living environment and jobs—and for businesses—to increase benefits through more efficient production practices that generate savings. Following the study by the Politecnico di Milano and Camera di Commercio di Milano (2012), we select industries which have an important green component, among those in industry “in the strict sense”, using the following ATECO 2007 codes: Collection, Reuse, Recycling of Waste (Codes 38.11.00, 38.12.00, 38.21.01, 38.21.09, 38.22.00, 38.31.10, 38.32.10, 38.32.20, 38.32.30 and 39.00.01), Efficiency of water systems: (Code 36.00.00), Planning, Reclamation and Rehabilitation of the territory (Code 30.00.09), Waste water treatment (Code 37.00.00), Energy Storage (Code 27.20.00). Most of the companies included in the sample (89.97%) are companies at least 10 years old. Consistently with previous literature (Abu Bakar 2011; Ayyagari et al. 2011; Fort et al. 2012), we classify young firms as those of up to 5 years, mature firms from 6 to 10 years and old firms over 10 years. Panels A and B of Table 1 present the sample distribution by type of sector (i.e., green or non-green) and age (i.e., young, mature or old), respectively.
Table 1
Observations by type of sector and firm age
Type of sector | No. of observations | % of total |
|---|---|---|
Panel A: observations by type of firm sector | ||
Green | 6844 | 3.17 |
Non-green | 208,720 | 96.83 |
Total | 215,564 | 100 |
Panel B: observations by firm age | ||
Young | 4588 | 2.13 |
Mature | 17,032 | 7.90 |
Old | 193,944 | 89.97 |
Total | 215,564 | 100 |
In order to study differences between green and non-green SMEs in terms of economic and financial performance, the degree of financial resilience and the size, age of the firm and the orientation towards trade or bank credit before or during the COVID-19, we collect a set of information that includes age, sector and financial ratio for each SME. Our data source is the AIDA Bureau van Dijk database. Consistently with much existing literature (Canto-Cuevas et al. 2019; Cuñat 2007; Fukuda et al. 2007; García-Teruel and Martínez-Solano 2010; Petersen and Rajan 1997; Yang 2011), we use “Trade credit” (i.e., trade payable/total assets) and “Full credit” (i.e., trade credit + bank credit/total assets),3 as dependent variables. We observe the use of trade and full credit during the firm’s life cycle by considering among the independent variables “Age”, which is the number of years of a firm’s activity. In particular, we use the logarithm of (1 + age). We also consider firm “Size” (i.e., the logarithm of the total assets); “Profitability”, which is the ratio of Earnings Before Interest, Taxes, Depreciation and Amortization (EBITDA) to total assets; “Liquidity” or “current ratio” (i.e., short-term assets/short-term liabilities); trade credit in the previous year (“Trade creditt-1”), which is the ratio of trade payable to total assets of the previous year; “Account receivable” (i.e., trade receivable/total assets). In order to verify whether a potential substitution effect between trade and bank credit exists, we consider “Short-term bank credit” (following AIDA, its bank debts within 12 months, divided by total assets) and “Long-term bank credit” (following AIDA, its bank debts over 12 months, divided by total assets). Table 2 summarizes the descriptive statistics for the entire period 2017–2020. The descriptive statistics for the sub-periods 2017–2019 and 2020 are reported in Appendix (see Table 10).
Table 2
Descriptive statistics—2017–2020
Variables | Mean | Median | SD | Min | Max |
|---|---|---|---|---|---|
Full credit | 0.2728 | 0.2236 | 0.2943 | 0.000 | 0.998 |
Trade credit | 0.1754 | 0.1557 | 0.1549 | 0.000 | 0.995 |
Age | 29.536 | 28.000 | 15.899 | 2.000 | 142.000 |
Size | 8.595 | 8.460 | 1.026 | 1.908 | 14.976 |
Profitability | 0.077 | 0.072 | 0.151 | − 4.533 | 0.998 |
Liquidity | 1.487 | 1.110 | 1.255 | 0.000 | 10.000 |
Trade creditt − 1 | 0.1830 | 0.1639 | 0.1619 | 0.000 | 0.975 |
Account receivable | 0.2469 | 0.2329 | 0.1900 | 0.000 | 0.959 |
Short-term bank credit | 0.084 | 0.024 | 0.146 | 0.000 | 0.999 |
Long-term bank credit | 0.070 | 0.012 | 0.096 | 0.000 | 0.988 |
Short-term bank creditt − 1 | 0.075 | 0.001 | 0.194 | 0.000 | 0.985 |
Long-term bank creditt − 1 | 0.053 | 0.001 | 0.174 | 0.000 | 0.981 |
Table 3 presents the correlations between all the variables used in the analysis. We also run a test of Variance Inflation Factors (VIF). Table 4 shows that no explanatory variables show VIF values above 10, so multicollinearity is not a concern in our models (Yang et al. 2019).
Table 3
Correlation coefficient matrix
Variables | Full credit | Trade credit | Age | Size | Profitability | Liquidity | Trade creditt − 1 | Account receivable | Short-term bank credit | Long-term bank credit |
|---|---|---|---|---|---|---|---|---|---|---|
Full credit | 1.000 | |||||||||
Trade credit | 0.7234* | 1.000 | ||||||||
Age | − 0.0555* | − 0.141* | 1.000 | |||||||
Size | 0.1080* | − 0.056* | 0.186* | 1.000 | ||||||
Profitability | − 0.1148* | − 0.032* | − 0.018* | 0.103* | 1.000 | |||||
Liquidity | − 0.3361* | − 0.241* | − 0.100 | 0.027* | 0.196* | 1.000 | ||||
Trade creditt − 1 | 0.5960* | 0.817* | − 0.145* | − 0.063* | − 0.021* | 0.208* | 1.000 | |||
Account receivable | 0.3795* | 0.488* | − 0.070* | − 0.009* | 0.070* | 0.023 | 0.434* | 1.000 | ||
Short-term bank credit | 0.7210* | − 0.245* | − 0.029 | − 0.070* | − 0.145* | − 0.296* | − 0.194* | 0.189* | 1.000 | |
Long-term bank credit | 0.5665* | 0.111* | 0.011 | 0.099* | − 0.056* | − 0.126 | 0.110* | 0.031* | 0.184* | 1.000 |
Table 4
VIF test
Variable | VIF | 1/VIF |
|---|---|---|
Age | 1.09 | 0.9199 |
Size | 1.08 | 0.9295 |
Profitability | 1.08 | 0.9270 |
Liquidity | 1.19 | 0.8416 |
Trade creditt − 1 | 1.34 | 0.7455 |
Account receivable | 1.28 | 0.7807 |
Short-term bank credit | 1.20 | 0.8354 |
Long-term bank credit | 1.06 | 0.9458 |
Short-term bank creditt − 1 | 1.15 | 0.8711 |
Long-term bank creditt − 1 | 1.07 | 0.9360 |
3.2 Methodology
To answer our research questions, we use an OLS regression to compare 2019 with 2020 and a panel data approach. We also compare the three-year period 2017–2019 with the year 2020 (see Appendix). We next apply the diff-in-diff methodology. The equation models include the variables described in Sect. 3.1 and additional time dummies as control variables:where i is the firm, and t is the time period; µi represents the firm-specific effects, and εit represents the measurement errors. As specified in the previous section, “Trade credit” is the dependent variable of Eq. (1) while “Full credit” is the dependent variable of Eq. (2). Following previous studies (e.g., Canto-Cuevas et al. 2016, 2019), “Age” is transformed into the logarithm of (1 + age) in the equations. We estimate Eq. (1) using estimators of fixed and random effects to take the individual effects into account (µi), with clustered standard errors at firm level. The Hausman test is performed to ascertain whether the individual effects are fixed or random. If the null hypothesis is rejected, correlation between the independent variables and the individual unobservable effects does not exist and the random effects model is considered a good estimator. The Hausman test results suggest that fixed effect models are appropriate for our dataset. Tables reported in Sect. 4, therefore, show only the results of estimation with a fixed effect model.
$$\begin{aligned} {\text{Trade credit}}_{it} & = \beta_{0} + \beta_{{1}} {\text{Age}} + \beta_{{2}} {\text{Size}} + \beta_{{3}} {\text{Profitability}} \\ &\quad + \beta_{{4}} {\text{Liquidity}} + \beta_{{5}} {\text{Trade credit}}_{{t - {1}}} + \beta_{{6}} {\text{Account receivable}} \\ &\quad + \beta_{{7}} {\text{Short - term bank credit}} + \beta_{{8}} {\text{Long - term bank credit}} \\ & \quad + {\text{time dummies}} + \mu_{it} + \varepsilon_{it} \\ \end{aligned}$$
(1)
$$\begin{aligned} {\text{Full credit}}_{it} & = \beta_{0} + \beta_{{1}} {\text{Age}} + \beta_{{2}} {\text{Size}} + \beta_{{3}} {\text{Profitability}} \\& \quad + \beta_{{4}} {\text{Liquidity}} + \beta_{{5}} {\text{Trade credit}}_{{t - {1}}} \\ &\quad + \beta_{{6}} {\text{Account receivable}} + \beta_{{7}} {\text{Short - term bank credit}}_{{t - {1}}} \\ &\quad + \beta_{{8}} {\text{Long - term bank credit}}_{{t - {1}}} + {\text{time dummies}} + \mu_{it} + \varepsilon_{it} \\ \end{aligned}$$
(2)
Diff-in-diff methodology makes it possible to estimate the effect of a "treatment" (i.e., COVID-19 pandemic) on subjects. It uses a “treatement group or treated” (green SMEs) relative to a control group not exposed to the treatment (non-green SMEs) (e.g., Bertrand et al. 2004). The two groups are observed in two periods, one before and one after the treatment (i.e., 2019 and 2020, respectively). Diff-in-diff in fact makes it possible to: (1) observe how the treatment group mean changes before and after the treatment; and (2) compare this change with the mean over time of a similar group which did not undergo the treatment (i.e., control group). The regression model used for the estimation is the following:where Y is the outcome variables of interest (i.e., Y = Full credit; Y = Short-term bank credit; Y = Long-term bank credit; Y = Trade credit) of firm i at time t, observed in periods 1 and 2. “Green” is a dummy variable which takes the value of 1 for "treated" subjects (green SME) and 0 otherwise (non-green SME). In other words, the variable “Green” captures the possible a priori differences between the treated group and the control group. “COVID” is a dummy variable which takes value 1 in 2020 (the year of COVID-19 pandemic) and 0 otherwise. The coefficient of this variable tells us how much the external credit (full credit, trade credit, short-term bank credit and long-term bank credit) of the control group changes on average after treatment. The variable “Green*COVID” represents the interaction between “Green” and “COVID” and takes a value of 1 for subjects treated in the second period. The coefficient of the variable (Green*COVID) is the parameter of interest because it expresses the effect of the treatment, or the differential impact of COVID-19 on green SMEs. The diff-in-diff estimator is the difference between the mean differences: it takes the difference in the treatment group before and after the treatment (the treatment effect) and subtracts the difference in the control group before and after the treatment (the trend over time), as in the following formula:“Z” in Eq. (3) refers to a set of controls for credit ratios, which are Age, Size, Profitability and Liquidity as measured in previous equations. We limit the number of controls, and do not regress credit ratios on other credit ratios (or on the lags of the credit ratios) in order to avoid significant autocorrelation of errors and inflated standard errors. \(\varepsilon\)it indicates the error term.
$$\begin{aligned} Y_{{{\text{i}},{\text{t}}}} & = \beta_{0} + \beta_{1*} {\text{Green}} + \beta_{2} *{\text{COVID}} \\ &\quad + \beta_{3} *{\text{Green}}*{\text{COVID}} + \beta_{i} *Z_{i} + \varepsilon_{it} \\ \end{aligned}$$
(3)
$$\left( {{\text{Treatment}}\_{\text{post }} - {\text{ Treatment}}\_{\text{pre}}} \right) - \left( {{\text{Control}}\_{\text{post }} - {\text{ Control}}\_{\text{pre}}} \right) \, = {\text{ Diff - in - Diff estimate}}$$
(4)
4 Results
4.1 Green orientation and external credit during COVID-19
In order to investigate the relationship between the “green” component and the degree of economic and financial resilience during the COVID-19 pandemic, and to verify whether orientation to the green economy tends to increase or decrease the importance of trade credit relative to bank/financial credit in financing the firm, we carry out our analysis on green and non-green SME sub-samples. Table 5 shows the trend over time (2017–2020) of the mean values of the main variables examined. The first group (green SMEs) appear to be funded differently compared to the second (non-green SMEs). Although trade credit received (from suppliers) is higher than bank credit both for green and non-green firms, the importance of trade credit received is much higher for green companies, and it seems to act as a substitute for short term bank credit.
Table 5
Descriptive statistics green and non-green SMEs—trend of mean values (2017–2020)
Variables | Green | Non-green | ||||||
|---|---|---|---|---|---|---|---|---|
2017 | 2018 | 2019 | 2020 | 2017 | 2018 | 2019 | 2020 | |
Mean | Mean | |||||||
Full credit | 0.3002 | 0.2959 | 0.2800 | 0.2351 | 0.2942 | 0.2854 | 0.2734 | 0.2375 |
Trade credit | 0.2072 | 0.2044 | 0.1947 | 0.1789 | 0.1883 | 0.1807 | 0.1735 | 0.1538 |
Profitability | 0.0842 | 0.0832 | 0.0893 | 0.09471 | 0.0820 | 0.0811 | 0.0778 | 0.0645 |
Liquidity | 1.5006 | 1.5239 | 1.5531 | 1.7527 | 1.3705 | 1.4125 | 1.4654 | 1.7119 |
Trade creditt − 1 | 0.2058 | 0.2073 | 0.2044 | 0.1947 | 0.1865 | 0.1882 | 0.1807 | 0.1735 |
Account receivable | 0.2944 | 0.2964 | 0.2941 | 0.2728 | 0.2577 | 0.2533 | 0.2474 | 0.2206 |
Short-term bank credit | 0.0699 | 0.0689 | 0.0634 | 0.0526 | 0.0921 | 0.0922 | 0.0897 | 0.0644 |
Long-term bank credit | 0.0645 | 0.0679 | 0.0664 | 0.0855 | 0.0616 | 0.0632 | 0.0642 | 0.0955 |
Table 6 summarizes descriptive statistics for green and non-green SMEs for 2019 and 2020. The descriptive statistics over the whole period 2017–2020 and the sub-periods 2017–2019 and 2020 are reported in the Appendix (Tables 11, 12).
Table 6
Descriptive statistics for green and non-green SMEs, 2019 in comparison with 2020
Variables | Green | Non-green | Green | Non-Green | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
2019 | 2020 | |||||||||
Mean | Std.Err | Mean | SE | t Statistic | Mean | SE | Mean | SE | t Statistic | |
Full credit | 0.2800 | 0.0067 | 0.2734 | 0.0013 | 0.9075 | 0.2351 | 0.0060 | 0.2375 | 0.0012 | − 0.3638 |
Trade credit | 0.1947 | 0.0045 | 0.1735 | 0.0007 | 5.2714*** | 0.1789 | 0.0044 | 0.1538 | 0.0007 | 6.4305*** |
Profitability | 0.0894 | 0.0035 | 0.0778 | 0.0007 | 2.7848* | 0.09471 | 0.0032 | 0.0645 | 0.0007 | 7.6590*** |
Liquidity | 1.5531 | 0.0319 | 1.4654 | 0.0061 | 2.5995* | 1.7527 | 0.0382 | 1.7119 | 0.0069 | 1.0287 |
Trade creditt − 1 | 0.2044 | 0.0047 | 0.1808 | 0.0007 | 5.6325*** | 0.1947 | 0.0045 | 0.1735 | 0.0007 | 5.2714*** |
Account receivable | 0.2941 | 0.0059 | 0.2474 | 0.0009 | 9.3540*** | 0.2728 | 0.0057 | 0.2206 | 0.0008 | 10.6757*** |
Short-term bank credit | 0.0634 | 0.0034 | 0.0897 | 0.0007 | − 6.5393*** | 0.0526 | 0.0025 | 0.0644 | 0.0006 | − 3.4356*** |
Long-term bank credit | 0.0664 | 0.0028 | 0.0642 | 0.0006 | 0.6260 | 0.0855 | 0.0033 | 0.0955 | 0.0006 | − 2.7726* |
Table 6 reports a period-by-period comparison between green and non-green SMEs. As might be imagined, the impact of COVID-19 is very significant for both green and non-green companies in terms of full credit, trade credit (accounts payable), accounts receivable and short-term credit decrease. Increases in long-term bank credit in 2020 can be explained by the introduction of special state guarantees on bank loans with maturity over 1 year made to companies during the pandemic,4 and in fact, the growth of bank loans other than short term loans to companies in 2020 is also certified by Bank of Italy statistics.5 Table 6 also shows that liquidity increases for both green and non-green companies. This is consistent with the fact that in 2020, when the pandemic started, there was an increase in liquid capital due to the increase in long-term banking debt, and the uncertainty and fears for the future which led companies to postpone investments. Table 6 also shows that profitability in 2020 for green companies increased compared to 2019, while there was a significant reduction for non-green companies. Green SMEs show higher profitability than non-green in both periods, and higher liquidity than non-green ones in normal times.
In order to deepen the analysis on changes in levels of trade credit as a component of full credit, as well as short-term bank credit and long-term bank credit, we also apply diff-in-diff methodology. Table 7 shows the results (Table 13 in the Appendix reports the results considering 2017–2019 as the period before the treatment, i.e., 2020).
Table 7
Results of diff-in-diff analysis (dependent variables: full credit, Trade credit, Short-term bank credit, long-term bank credit)
Dependent variable (outcome variables) | ||||
|---|---|---|---|---|
Full credit | Trade credit | Short-term bank credit | Long-term bank credit | |
Constant | 0.2307*** | 0.2056*** | 0.0326*** | − 0.0094** |
(0.0087) | (0.0046) | (0.0043) | (0.0041) | |
Green | − 0.0023 | 0.01706*** | − 0.0221*** | 0.0033 |
(0.0070) | (0.0037) | (0.0035) | (0.0033) | |
COVID | − 0.0036* | − 0.0148*** | − 0.0213*** | 0.0331*** |
(0.0019) | (0.0009) | (0.0009) | (0.0009) | |
Green*COVID | 0.0048 | 0.0019*** | 0.0148** | − 0.0128** |
(0.0103) | (0.0005) | (0.0052) | (0.0048) | |
Age | − 0.0218*** | − 0.0311*** | 0.0086*** | 0.0001 |
(0.0015) | (0.0008) | (0.0008) | (0.0007) | |
Size | 0.0327*** | 0.0128*** | 0.0091*** | 0.0110*** |
(0.0009) | (0.0005) | (0.0005) | (0.0004) | |
Profitability | − 0.1524*** | − 0.0002 | − 0.1010*** | − 0.0328*** |
(0.0063) | (0.0034) | (0.0032) | (0.0030) | |
Liquidity | − 0.0664*** | − 0.0260*** | − 0.0277*** | − 0.0127*** |
(0.0007) | (0.0004) | (0.0003) | (0.0003) | |
N. observations | 84,564 | 84,465 | 84,556 | 84,564 |
R squared | 0.0429 | 0.0882 | 0.1020 | 0.0429 |
Results presented in Table 7 confirm that there is a differential impact of COVID-19 on green SMEs, probably due to policy changes in green firms compared to non-green firms during the peak period of the pandemic (2020). We find that COVID-19 significantly reduced the difference in long-term bank credit (-0.0128 basis points). On the other hand, COVID-19 increased the difference in trade credit and short-term bank credit of green SMEs (0.0019 and 0.0148 basis points, respectively). After COVID-19, in fact, the levels of both trade credit and short-term bank credit fell, but for green SMEs less sharply than for non-green ones.
4.2 Green orientation and determinats of trade credit during COVID-19
Tables 6 and 7 show that the level of full credit between green and non-green companies is similar both before and after the COVID-19 pandemic, whereas significant differences exist for trade credit and short-term bank credit. In trade credit, green SMEs have a higher level than non-green SMEs, and in short-term bank credit, non-green SMEs have a higher level than green SMEs. Since green SMEs are more profitable both before and after the COVID-19 pandemic and the difference in trade credit level between green and non-green companies increases during the pandemic, trade credit might be considered as a component of a long-term portfolio management strategy, consolidating relationships with clients and increasing firm profitability in facing variable demand conditions. We also investigate a possible substitution effect between trade credit and short-term bank credit, using the main determinants of trade credit for green and non-green SMEs over the periods 2017–2020 and the two-year period 2019–2020 (see Table 8), according to OLS regression methodology (see Eq. 1). We also examine 2019 in comparison with 2020 (see Table 9), according to the diff-in-diff analysis (see Eq. 3). Our models include the following control variables: age, size, profitability, liquidity, trade credit in the previous year, account receivable, and short and long-term bank credit. The dependent variable is trade credit.
Table 8
Results of analysis—green and non-green SMEs—periods 2017–2020 and 2019–2020
Variables | Period 2017–2020 | Period 2019–2020 | ||
|---|---|---|---|---|
Green | Non-green | Green | Non-green | |
Constant | 0.165*** | 0.228*** | 0.0724** | 0.0296*** |
(0.033) | (0.007) | (0.0239) | (0.0034) | |
Age | − 0.013*** | − 0.010*** | − 0.0186*** | − 0.0086*** |
(0.003) | (0.001) | (0.0044) | (0.0006) | |
Size | − 0.002 | − 0.014*** | 0.0056** | 0.0037*** |
(0.003) | (0.001) | (0.0024) | (0.0003) | |
Profitability | − 0.107*** | − 0.027*** | − 0.1166*** | − 0.0383*** |
(0.014) | (0.002) | (0.0164) | (0.0021) | |
Liquidity | − 0.027*** | − 0.019*** | − 0.0135*** | − 0.0103*** |
(0.002) | (0.001) | (0.0019) | (0.0003) | |
Trade creditt − 1 | 0.078*** | 0.081*** | 0.5088*** | 0.5707*** |
(0.014) | (0.002) | (0.0141) | (0.0024) | |
Account receivable | 0.269*** | 0.277*** | 0.1918*** | 0.1835*** |
(0.011) | (0.002) | (0.0112) | (0.0020) | |
Short-term bank credit | − 0.151*** | − 0.105*** | − 0.0403*** | − 0.0406*** |
(0.021) | (0.003) | (0.0024) | (0.0025) | |
Long-term bank credit | − 0.055*** | − 0.069*** | 0.0091 | − 0.0007 |
(0.021) | (0.003) | (0.0203) | (0.0028) | |
Year dummies | Included | Included | Included | Included |
R squared | 0.383 | 0.376 | 0.6618 | 0.7033 |
Table 9
Results of analysis—green and non-green SMEs, 2019 in comparison with 2020
Variables | Green | Non-green | ||
|---|---|---|---|---|
2019 | 2020 | 2019 | 2020 | |
Constant | 0.0288 | 0.0730** | 0.0132** | 0.0237*** |
(0.0260) | (0.0259) | (0.0039) | (0.0037) | |
Age | − 0.0165** | − 0.0171*** | − 0.0063*** | − 0.0049*** |
(0.0049) | (0.0045) | (0.0007) | (0.0006) | |
Size | 0.0071* | − 0.0033 | 0.0037*** | 0.0014*** |
(0.0026) | (0.0025) | (0.0004) | (0.0003) | |
Profitability | − 0.1219*** | − 0.0941*** | − 0.0594*** | − 0.0252*** |
(0.0197) | (0.0242) | (0.0027) | (0.0029) | |
Liquidity | − 0.0078** | − 0.0098*** | − 0.00764*** | − 0.0073*** |
(0.0024) | (0.0020) | (0.0003) | (0.0002) | |
Trade credit − 1 | 0.6751*** | 0.6510*** | 0.709*** | 0.6756*** |
(0.0165) | (0.0176) | (0.0029) | (0.0029) | |
Account receivable | 0.1411*** | 0.1157*** | 0.1210*** | 0.1334*** |
(0.0128) | (0.0143) | (0.0024) | (0.0024) | |
Short-term bank credit | 0.0260 | − 0.0502* | 0.0476*** | − 0.0216*** |
(0.0277) | (0.0301) | (0.0029) | (0.0035) | |
Long-term bank credit | 0.0134 | − 0.0036* | 0.0215*** | − 0.0157*** |
(0.0257) | (0.0221) | (0.0039) | (0.0031) | |
Year dummies | Included | Included | Included | Included |
R squared | 0.6806 | 0.6730 | 0.7120 | 0.7146 |
For trade credit, Table 8 shows:
-
a negative relationship with age, liquidity, and particularly profitability and short-term bank credit for both green and non-green SMEs;
-
a positive relationship with trade credit-1, and with account receivable, which confirms the matching hypotheses, for both green and non-green SMEs.
The year-over-year comparison reported in Table 9 shows that during the COVID-19 pandemic a substitution effect between trade credit and short-term bank credit is verified, particularly for green companies.
5 Conclusions and policy implications
This paper examines the financial resilience of green manufacturing compared to non-green manufacturing SMEs in Italy during the COVID-19 pandemic. Since the AIDA Bureau van Dijk database contains no specific data on green or ESG ratings, we consider as “green” those SMEs in industries characterized by higher green component than other industries, following the study by the Politecnico di Milano and Camera di Commercio di Milano (2012).
We verify the three hypotheses formulated in Sect. 2. As shown in Tables 6 and 7, the proportion of full credit is not higher for green firms than non-green firms both before and after the COVID-19 pandemic, so Hypothesis 1 is not proven. On the other hand, Hypotheses 2a and 2b are fully proven: the difference between green and non-green SMEs is shown in the importance of trade credit for the first group. Accounts payable and receivable are both significantly more important for green than for non-green SMEs, while short term bank credit is higher for non-green than for green SMEs (see Tables 6 and 7, 10, 11, 12). The overall analysis on green orientation and external credit shows that the COVID-19 pandemic affected the level and the differences in trade credit between green and non-green SMEs. Before the pandemic, green companies relied on a higher level of trade credit than non-green companies, and in 2020 the green company trade credit level stayed higher than the non-green company level. In other words, the difference in trade credit between green and non-green companies remained positive during the pandemic. This leads us to believe that a substitution effect between trade and short-term bank credit exists, in particular for green SMEs during the COVID-19 pandemic. In other words, trade credit received compensated for the reduction in short-term bank credit for green SMEs. Tables 8 and 9 in fact report that trade credit received is much more strongly negatively correlated to short-term bank credit for green SMEs than for non-green ones.
At the same time, descriptive statistics on the entire period as well as statistics for the years up to 2020, before the pandemic, show that, on average, green SMEs are younger and smaller but more profitable, and in better financial condition, than non-green SMEs. They also take up and offer more trade credit than non-green SMEs.
To conclude, the importance of trade credit for green SMEs could be explained by two complementary trends. A substitution effect, demonstrated by previous literature and recent research on Italian firms [Arcuri and Pisani (2021) on Italian green middle-sized companies] appears to occur on one segment of green SMEs (presumably the youngest/smallest and less profitable) not only in the pandemic period, but also before 2020. Meanwhile, for the segment of green SMEs which make average profitability of green SMEs higher than that of non-green SMEs, trade credit appears to be a component of long-term commercial strategy with clients (Emery 1987; Deloof et al. 1986) and for increasing firm profitability (Martínez-Sola and García-Tereul 2014), which, as noted above, is higher for green SMEs (Table 6). This ‘successful’ segment of green SMEs could be seen as a subclass of the green firms considered, in which trade credit is an important instrument of competition. This conclusion is in fact consistent with previous literature showing that the use of trade credit varies significantly across industries (Cannari et al. 2004; Ng et al. 1999; Marotta 2005; Wilson and Summers 2002).
Declarations
Conflict of interest
None.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
