Bank credit and trade credit: Evidence from SMEs over the financial crisis

Abstract

This article uses panel data to test the extent to which trade credit acted as a substitute for bank finance in Small and Medium-sized Enterprises (SMEs), in the aftermath of the financial crisis of 2008. It demonstrates that the reduction in the supply of funds to SMEs was compounded by the contraction of net trade credit within the sector. Nevertheless, trade credit played a vital role in the adjustment of the sector by easing the burden of the financial crisis for some SMEs. Thus, the relative importance of trade credit increased for financially “vulnerable” SMEs that were less liquid, highly dependent on short-term bank finance, and with greater levels of intangible assets, when entering the crisis. In terms of a redistribution effect, financially stronger firms extended relatively more trade credit, to financially vulnerable SMEs in the aftermath of the financial crisis. In addition, the analysis demonstrates that the financial position of SMEs entering the crisis was more important in determining the impact of the financial crisis on trade credit use than company characteristics of age and size.

JEL Classification: L26, L14, G20, B41, C33

Keywords

financial crisis panel data redistribution effect Small and Medium-sized Enterprises (SMEs)substitution effect trade credit

Introduction

In the aftermath of the financial crisis in 2008, bank lending to Small and Medium-sized Enterprises(SMEs) declined significantly in most developed economies. Over the period 2008–2011, new bank lending to SMEs (lending of <€1m) declined by 47% in the European Union (EU) varying from 21% in Italy, 66% in Spain and 82% in Ireland.¹ This is a major policy concern as lack of access to finance inhibits investment and economic recovery. A financial crisis not only affects the supply of money to firms for investment proposes, but it also has knock-on effects on the day-to-day operations of firms as they seek alternative sources of finance when overdrafts facilities are reduced or withdrawn (Cosh et al., 2009). Firms adjust by financing their activities from other sources including cash reserves, loans from families and business credit cards (Fraser, 2009). Firms also seek to manage their working capital more efficiently by delaying payments to suppliers and restricting customer credit. This extra credit may be negotiated or taken without agreement as both suppliers and buyers adjust to the new conditions (Garcia-Appendini and Montoriol-Garrig, 2013). It is generally agreed that trade credit has an augmented role for SMEs when access to finance from formal institutions is restricted (Petersen and Rajan, 1997). With a few notable exceptions (see Carbó-Valverde et al., 2012; Vermoesen et al., 2013), the majority of studies that examine the impact of the financial crisis on trade credit use are based on listed firms from emerging market economies (Bastos and Pindado, 2013; Love et al., 2007) rather than upon unlisted SMEs which form the focus of our study. This is an important research agenda since SMEs are (a) more dependent on trade credit to cover their short-term financing needs (Berger and Udell, 1998; Petersen and Rajan, 1997), (b) have less potential to access external financing sources (Berger and Udell, 1998; Cowling et al., 2012; Rodriguez-Rodriguez, 2006), (c) are more likely to be adversely affected by asymmetric information and financing constraints (Beck et al., 2008; Bernanke, 1983; Stiglitz and Weiss, 1981; Tsuruta, 2008) and (d) are influenced by banking market concentration and the level of financial development (Agostino et al., 2012; Ge and Qiu, 2007).

This article addresses a number of important questions in the trade credit literature within the context of SMEs. First, what is the role of trade credit in periods of financial crisis? Is there a redistribution of credit from financially stronger firms through trade credit finance to firms that are financially constrained in accessing bank finance? Second, does trade credit act as a substitute for bank credit in a credit constrained economy and are there sectoral differences associated with any substitution effect? This study contributes to evidence on redistribution and substitution effects by testing the case for unlisted SMEs during the pre-crisis and post-crisis periods. It is the first study to our knowledge, that applies panel data analysis on the use of trade credit finance by SMEs over the financial crisis period using Amadeus data.

Ireland provides a useful setting for this analysis, as SMEs in Ireland account for approximately 98% of all enterprises and 68% of all private sector employment (Central Bank of Ireland, 2012). Following the financial crisis and subsequent recession, Irish SMEs experienced both a dramatic reduction in bank lending and aggregate demand with gross domestic product (GDP) falling for three consecutive years from 2008 to 2010. During the period 2003–2007, bank loans to property-related businesses grew by unprecedented levels from €45 billion in 2003 to a peak of €125 billion in the first quarter of 2008 (Whelan, 2013). Ireland’s banking sector model also changed dramatically from one based on traditional deposit-based lending to being highly dependent upon raising funds from short-term borrowing on international inter-bank and money markets. During the same period, international borrowings of the six main banks in the country rose from €15 billion in 2003 to over €100 billion in 2007, representing over half of the country’s GDP (Whelan, 2013). This type of funding ultimately proved to be less stable than traditional deposit-based lending. While Ireland was very much exposed to a potential crisis, it was not the only European country where banking deposits were insufficient to satisfy the growth in domestic credit expansion. In the 30 European countries, the average ratio of bank deposits to GDP grew from 57% in 1999 to 89% in 2007, whereas the average ratio of private credit to GDP grew much more quickly from 67% in 1999 to 107% in 2007 (Lane and McQuade, 2013).

In the immediate aftermath of the crisis in 2008, the Irish government began a period of severe austerity measures coinciding with the ‘troika bailout’ and the introduction of a banking guarantee. However, despite these measures and the guarantee to bank deposits, international investors continued to withdraw their funding. Research has found that, in particular, deposits of non-residents at the Irish banks covered by the guarantee declined from €162 billion in August 2010 to €116 billion by November 2010. Not surprisingly, coinciding with the reduction in official bank funding itself, there was a substantial reduction in financial institutional lines of credit to SMEs. Figures show that the outstanding stock of credit to non-financial, non-property-related private sector fell by over 30% from its peak in Q4 of 2008 to mid-2011 (Central Bank of Ireland, 2012).

This study shows that net trade credit has declined over the crisis period indicating an overall reduction in inter-firm financing in the aftermath of the financial crisis. It provides evidence of the substitution of bank finance for trade credit when firms are constrained in accessing bank credit, consistent with Carbó-Valverde et al. (2012), Fishman and Love (2003) and Petersen and Rajan (1997). The findings demonstrate a financial ‘adjustment process’, whereby financially vulnerable SMEs adapt to the restriction in lending in the immediate aftermath of the crisis, by substituting bank finance for trade credit which is most likely extended by financially stronger SMEs. The panel analysis also indicates that the financial position of an SME entering the crisis is a more important determinant of who redistributes credit within the sector than firm age or size.

To explore such trade credit finance issues, this article is structured as follows: The next section provides a review of the literature on trade credit financing during times of credit restriction, focusing on the theories and evidence relevant to the hypotheses tested in this study. The following section describes the data source, methodology and variables used in the empirical analysis. This is followed by the presentation and discussion of findings. The limitations of the research and avenues for further research are presented prior to the conclusion

Literature review and hypotheses development

The role of trade credit in financing SMEs in periods of credit restriction

The importance of trade credit in financing economic activity is well documented in the literature; so, for example, Ng et al. (1999) find that the amount of trade credit exceeds the primary money stock M1 by a factor of 1.5 in the United States. In general, trade credit contracts differ for firms depending on a number of factors including the industry setting and the duration of the business relationship (Fishman and Love, 2003). This is important, particularly for SMEs heavily reliant on internal funds and bank lending, especially bank overdrafts as a means of short-term financing. While trade credit is viewed as a short-term method of financing (Nilsen, 2002), it plays an important role in the everyday organisation and decision-making of firms (Rodriguez-Rodriguez, 2006; Martínez-Solo et al, 2013).

Theory on trade credit is not novel and its role has been examined from a number of different perspectives including transaction costs, redistribution, substitution, market power and relationship lending. One of the key benefits to suppliers in extending trade credit is the knowledge they have regarding the credit worthiness of the firms with whom they do business. This is a result of ongoing monitoring of orders, repayment schedules and the ability to enforce repayment or cut off future supplies (Love and Zaidi, 2010). Obtaining favourable trade credit terms allows firms to reduce their overall borrowing costs, especially by obtaining discounts for early payment (Aktas et al., 2012; Giannetti et al., 2011). Trade credit has been shown to sustain the sales and profitability of SMEs over periods of financial crisis (Banos-Caballero et al., 2012; Coulibaly et al., 2013). However, trade credit can be a very expensive form of finance if firms do not avail of the early discount facility (Nilsen, 2002; Petersen and Rajan, 1997). Therefore, it may be beneficial for firms to hold cash as a precautionary motive to avoid postponing and incurring the costs of late payment for goods (Wu et al., 2011).

The financing of firms during times of credit shortages has received increased attention (Drakos, 2013; Love and Zaidi, 2010; Vermoesen et al., 2013). Redistribution theory, developed by Meltzer (1960), suggests that large liquid firms are net suppliers of credit to smaller firms because they have better access to bank finance. Empirical evidence of this theory has been shown in periods of ‘tight’ money (Calomiris et al., 1995; Nilsen, 2002). Calomiris et al. (1995) showed that large firms use commercial paper to fund trade credit during periods of monetary tightening and can provide a cushion during a credit crunch for less financially liquid firms (Berger and Udell, 1998; Guariglia and Mateut, 2006; Nilsen, 2002; among others), while also reinforcing supplier–customer relations (Cunat, 2007). Nilsen (2002) shows that not only do small firms increase their use of trade credit during periods of monetary contraction, but large firms without bond ratings and access to external capital markets also increase their reliance on trade credit finance. Evidence on redistribution during periods of financial crises is mixed, as banking systems are not properly functioning as in ‘normal’ times (Boissay and Gropp, 2007; Love et al., 2007). Love et al. (2007) find empirical evidence of the redistribution effect for a sample of large listed firms in emerging markets; however, they find that this effect contracts during a financial crisis. Consequently, there is a shift in how firms are able to redistribute credit over time, while Choi and Kim (2005) find that when banks refrain from lending to smaller firms during a monetary contraction, large US firms then withdraw financial help to small firms too.

Theoretical support for the substitution effect is evident (Biais and Gollier, 1997; Burkart and Ellingsen, 2004; Cunat, 2007; Meltzer, 1960; Wilner, 2000), and empirically in studies by (Danielson and Scott, 2004; Fishman and Love, 2003; Huang et al., 2011 and Petersen and Rajan, 1997). But following banking crises, other studies demonstrate that small illiquid firms pass on liquidity shocks to their suppliers along the supply chain thus, propagating reduced trade credit and ultimately default in many cases (Love and Zaidi, 2010). Furthermore, the possible substitution between trade credit and bank credit is likely to be determined by operating context (Demirguc-Kunt and Maksimovic, 2001; Hernández-Cánovas and Koeter-Kant, 2011). Love et al. (2007) and Love and Zaidi (2010) examine the role of trade credit during the financial crises in emerging economies of Thailand, Philippines, Indonesia and Korea in the late 1990s and find evidence against the premise that trade credit can act as a substitute to bank credit during financial crises. Instead, they argue that liquidity shocks are passed along the supply chain exacerbating the financial shocks from reductions in official credit lines.

The majority of trade credit studies are based on samples of large US listed firms (Calomiris et al., 1995; Choi and Kim, 2005; Garcia-Appendini and Montoriol-Garrig, 2013; among others) or publicly listed small and large firms from emerging market economies (Coulibaly et al., 2013; Love et al., 2007; Love and Zaidi, 2010). This may reflect the lack of comprehensive data on SMEs and their financing. In a recent study, Garcia-Appendini and Montoriol-Garrig (2013) find that large US firms extended credit to financially weaker large firms throughout the crisis. There is little evidence regarding the role of trade credit financing for SMEs and whether it has acted as substitute for bank financing over the crisis, particularly in the case of unlisted SMEs which are the focus of this study. Carbó-Valverde et al. (2012) examine the role of trade credit for a sample of Spanish SMEs over the period of 2004 to the onset of the financial crisis in 2008 finding that financially constrained SMEs increased their dependence on trade credit to finance capital expenditures at the onset of the crisis. Such findings imply a significant role for trade credit in investment among SMEs by modelling firms constrained in their access to bank finance. This article extends the work of Carbó-Valverde et al. (2012) by examining the effective role of trade credit in financing SMEs over the period of the financial crisis and beyond and focuses on the role of trade credit as a substitute for bank finance for financially constrained SMEs.

Given that our sample consists of unlisted SMEs, we expect the dramatic reduction in bank credit extended in Ireland to have a significant impact on the level of trade credit extended to the SME sector. Bernanke (1983) argued that the disruptions in the banking sector following the Great Depression reduced the efficiency of credit allocation and consequently aggregate demand and output. Thus the secondary effects of increased lending restrictions and credit availability to firms in the years after the onset of the crisis will reduce the overall supply of credit and allocation of credit in the economy, therefore reducing the level of trade credit extended and received too. The the first hypothesis we test is for the overall reduction in trade credit after the crisis:

Hypothesis 1: Net trade credit reduces over the financial crisis.

Overall, the expectation is that net credit redistributed by financially stronger firms rises immediately after the crisis and falls thereafter, such that the impact of the crisis would be seen in year one of the crisis and fall in subsequent years:

Hypothesis 2A: Financially liquid firms extend more trade credit following a financial crisis.

Hypothesis 2B: Financially illiquid firms receive more trade credit following a financial crisis.

According to the redistribution view of trade credit, financially stronger firms have the ability to pass on credit to financially constrained and vulnerable firms via their accounts receivable. Trade credit, therefore, acts as an important source of financing when credit from financial institutions is not available. This leads to our third hypothesis:

Hypothesis 3: Trade credit acts as a substitute for bank finance for SMEs in a financial crisis.

In addition, if trade credit acts as a substitute for bank finance for credit constrained or financially illiquid firms that would have received bank financing prior to the crisis, then it follows that trade credit use increased for this group of firms in the period after the financial crisis of 2008. This leads to Hypothesis 4:

Hypothesis 4: Trade credit received will be strongest for the period immediately after a financial crisis, particularly in terms of volume and length of credit time.

Next, we assess the role of collateral and use of trade credit finance. Tangible assets are a good indicator of collateral, which is previously found to alleviate the problems of information asymmetries for SMEs accessing bank debt (Mac an Bhaird and Lucey, 2010). High growth and innovative SMEs that invest in intangible activity are often restricted in accessing bank finance due to their lack of tangible assets (Vanacker and Manigart, 2010). We assess if there is a relation between the level of intangible assets in balance sheets and the likelihood of receiving trade credit financing. We suspect that the firms with the highest intangible-to-total asset ratios are most likely to be adversely affected by the banking crisis, where increased capital requirements by banks restricted the allocation of credit:

Hypothesis 5: Innovative firms will depend more on trade credit finance in the crisis period as opposed to pre-crisis years.

Petersen and Rajan (1997) show that firm age and financing behaviour are often characterised by a non-linear relationship. Reputation and credit worthiness are much more important in the early years of a firm’s life, and these assets take time to acquire. In addition, older firms are in a better position to accumulate assets and earnings. To account for this non-linear effect of age, we include the square of the age variable alongside the age variable itself. Alternative variables such as retained profits and the size of firm assets are likely to capture the effect of SME age on the level of trade credit granted:

Hypothesis 6: Older SMEs are more likely to extend trade credit than young SMEs.

Data, variables and methodology

The data consist of Irish SME financial statement data obtained from the Amadeus database supplied by Bureau Van Dijk. The Amadeus data are derived from accounts filed at the official Irish government’s Companies Registration Office (CRO). In total, there were 158,666 private limited companies registered in Ireland in 2012 (CRO, 2013) representing 79.3% of the estimated 200,000 total number of enterprises in the economy (Eurostat, 2013, The Structural Business Statistics Database). The sample obtained in this study includes over 7600 SMEs with balance sheet and profit and loss account information over the period 2003–2011.² While the sample only represents a small proportion of the total number of limited companies registered in Ireland, it is much more representative of surviving companies. The figure of 158,666 companies masks the fact that each year an average of 9.1% of the total are new entrants and 8.9% are exits. The sample population is also significantly affected by the provisions of the Companies (Amendment) Act 1986,³ whereby many SMEs in Ireland are exempt from filing complete financial accounts. Thus, of the 15,964 companies in our sample with employee data, only 7600 have financial data for at least two/three of the years of sample period (2003–2011).

SMEs are defined according to the standard European Commission criteria,⁴ which includes firms that employ fewer than 250 workers in a given year and have either an annual turnover of less than €50m or a balance sheet total of less than €43m. We also include micro enterprises in our sample. In this study, we scale by the number of employees and the balance sheet totals of each SME in each year. All financial variables are winsorised at the 1% and 99% levels.⁵ The criteria for our sample are as follows:

Active firms employing less than 250 employees.

Firms with balance sheet total of greater than €43,000,000 or annual turnover greater than €50,000,000 in any of three consecutive years of the sample years are excluded.

Firms that are reported to be listed or delisted are excluded.

All financial and insurance companies, in line with existing empirical studies, are excluded.

Public utilities such as public transport and postal services are excluded.

Firms whose records for creditor days, debtor days, accounts receivables and accounts payables are greater than zero in any given year.

In total, the final sample contains 7618 Irish SMEs and 68,562 firm year observations over the period 2003–2011, all of which are active over the sample period. Based on sales turnover of the last 3 years of the sample and scaling turnover according to European Commission (2008), the panel contains approximately 6002 micro enterprises (78% of total sample),⁶ 864 small enterprises (11.5% of total sample) and 723 medium-sized enterprises (9.5% of the total sample). Firms employing 10 persons or fewer in a given year are classified as micro enterprises, while those that employ between 10 and 49 workers are labelled small and, finally, enterprises employing between 50 and 249 employees are classified as medium-sized enterprises.

Figure 1 shows the changes in both the number of mean and median debtor and creditor collection days and the efficiency in working capital in SMEs over the sample period. The change in micro debtor days appears quite significant. Average micro debtor collection days have slightly increased from 82 pre-crisis to 91 during the crisis, while payment days have reduced from an average of 60 to 59 days for micro enterprises resulting in an increase in the Working Capital Requirement (WCR) from 22 days pre-crisis to 32 days in the crisis. In addition, we also present the median numbers of days over the two periods. However, the median figures show micro debtor days to have risen only from 43 to 45 days over the two periods, suggesting the figures are skewed to the right as we would expect. This is a strong indication that the ‘bad’ debtor days have deteriorated over the two periods for some firms, hence pushing the average figure up. Based on median figures, this represents an increase in WCR days from 16 to 20 days for micro enterprise. The change of nine days in the mean debtor days for micro enterprises represents an approximate move of six standard deviations of the mean of the micro sample indicating that the average has, over the two periods has risen.

Figure 1.

The average (mean) number of debtor and creditor days and the working capital requirement (WCR) of SMEs measured in days.

From Figure 1, we also observe a steep reduction in small firm debtor collection days over the two periods from 59 to 47 days. Small enterprise debtor collection days have reduced from an average of 59 days prior to the crisis to 47 days during the crisis, while small firm payment days have remained the same.⁷ In this case, the median change is actually from 51 to 33 days. This represents an interesting change over the two periods. The reduction of 12 average debtor collection days also represents an approximate movement of six standard deviations of the mean of the small firm sample. These changes are a strong indication that the differences we observe over the two periods are not by chance. Medium-sized enterprise collection days (while still high) have reduced from an average of 59 days in the pre-crisis period to 53 days in the crisis/post-crisis period. In terms of WCR, micro enterprises have been placed under the most pressure from suppliers over the two periods, while small firms appear to be the most efficient in adjusting their WCR policy in the post-crisis period.

When we examine debtor and creditor days across industries, we can see that the average level of trade credit received has increased for food processing, wholesale and business service sectors. Trade credit levels decrease for real estate and community services/residential care services. We find changes within the two periods, debtor days generally reducing in all sectors with the exception of construction and retail (see Table 1).⁸ The largest proportion of SMEs in our sample are in the sectors of real estate services, hospitality and tourism and community services, with the lowest proportion in construction, retail and publishing (see Table 2). The figures suggest that cash in certain sectors is being collected more quickly since the onset of the crisis; however, some sectors have seen an increase in the levels of trade credit financing received over the crisis period. These include food processing, wholesale and business services (Table 3). As pointed out by Love et al. (2007), the redistribution of credit from financially stronger to financially weaker firms during a banking crisis is based on the assumption that firms with better access to external finance will redistribute credit via trade credit to financially weaker firms. In the event of a credit crunch as experienced in the Irish context where bank lending effectively stopped, redistribution requires a transfer of credit from cash-rich firms to those firms that are constrained in access bank finance.

Table 1.

Average debtor and creditor collection days across selection of industry sectors for SMEs over period 2003–2011.

	All	Food/manufacturing	Construction	Wholesale	Retail	Hospitality and recreation	Broadcasting/publishing	Business services
Average debtor days across sectors
Pre-crisis	45	58	47	75	16	25	60	61
Standard deviation	(78)	(64)	(84)	(55)	(54)	(76)	(46)	(82)
Skewness	2.28	2.37	2.2	2.23	2.81	3.18	0.94	2.36
Crisis	45	51	49	65	17	23	51	55
Standard deviation	(110)	(62)	(75)	(75)	(56)	(51)	(68)	(83)
Skewness	2.19	3.21	2.62	2.89	4.56	3.24	3.23	2.44
Average creditor days across sectors
Pre-crisis	26	27	33	25	37	24	28	19
Standard deviation	(80)	(51)	(60)	(42)	(50)	(65)	(88)	(76)
Skewness	2.90	4.49	2.47	3.33	2.79	3.36	2.86	2.97
Crisis	25	32	39	29	34	23	29	24
Standard deviation	(84)	(74)	(93)	(66)	(52)	(75)	(71)	(99)
Skewness	2.99	3.84	2.75	3.89	4.09	3.2	3.26	2.71

SME: small- and medium-sized enterprise.

The pre-crisis represents the average values over the period 2003–2007, and crisis represents the average values from 2008 to 2011. Standard deviations are shown in parentheses. The figures represented illustrate the average number of days calculated using the median.

Table 2.

Summary statistics for the final year of the sample 2011.

Industry sectors	No. of SMEs	Sample (%)	Employment	Firm age	Sales	Cashta
Agriculture, mining and forestry	4	.05	7 (4.6)	24 (12)	2,134,365 (3,521,288)	.40 (.43)
Food processing/manufacturing	261	3.4	51 (51)	24 (13)	9,200,000 (1,000,000)	0.32 (.03)
Construction	204	2.7	14 (22)	16 (8)	2,100,000 (5200,000)	0.37 (.34)
Real estate, management services and related activities	2180	28.6	5 (5.8)	16 (7)	101,180 (546538.2)	0.56 (.32)
Wholesale	346	4.5	25 (31)	22 (12)	7,800,000 (9,400,000)	.27 (.26)
Retail trade	230	3.0	42 (51)	21 (10)	7,500,000 (9,500,000)	0.3 (0.3)
Hospitality, tourism and recreation	1230	16.2	12 (21)	20 (10)	780,000 (2,900,000)	0.7 (.33)
Broadcasting, publishing	70	0.9	27 (45)	15 (7)	4,300,000 (7,800,000)	0.38 (.33)
Business services	603	8.0	14 (24)	17 (7)	2,000,000 (5,200,000)	0.49 (.34)
Community work activities, residential care	2281	29.9	16 (24)	17 (8)	500,000 (1,700,000)	.80 (.27)
Other/missing	209	2.8	12 (3.5)	21 (8.9)	995,027 (1,035,429)	.44 (.05)
Total	7618	100

SME: small- and medium-sized enterprise.

Figures represent mean levels of employment, firm age, sales and cash at bank and in hand of company scaled by firm assets (Cashta). Figures in parentheses denote standard deviation.

Table 3.

Average (mean) levels of trade credit received by sector.

Industry sector	2003	2004	2005	2006	2007	2008	2009	2010	2011	Average	β
Food processing/manufacturing	0.08	0.1	0.08	0.11	0.13	0.14	0.18	0.19	0.2	0.13	.0163***
Construction	0.11	0.14	0.11	0.21	0.18	0.15	0.26	0.31	0.25	0.19	.021***
Real estate	0.28	0.27	0.31	0.28	0.28	0.25	0.24	0.24	0.24	0.27	−.007**
Wholesale	0.11	0.11	0.13	0.15	0.17	0.17	0.17	0.17	0.16	0.15	.008***
Retail trade	0.19	0.13	0.14	0.14	0.14	0.14	0.14	0.14	0.15	0.15	−.002
Hospitality and tourism	0.15	0.11	0.18	0.16	0.15	0.15	0.17	0.15	0.15	0.15	.001
Business services	0.19	0.15	0.21	0.2	0.18	0.2	0.2	0.21	0.25	0.20	.006*
Community services	0.16	0.14	0.15	0.14	0.13	0.13	0.13	0.11	0.1	0.13	−.006***
Average across years	0.16	0.14	0.16	0.17	0.17	0.17	0.19	0.19	0.19	0.17

Trade credit received is calculated as trade payables divided by the total sales of the firm in each year. β is the estimated coefficient of regression TC = α + β₁ (year) + ε₁, indicating how much trade credit increases for each sector for each year of the sample from a simple pooled regression with no control variables. Positive values of β indicate an increase in average levels of trade credit received in a given sector over the sample period. ***, ** and * represent statistical significance at the 1%, 5% and 10% levels, respectively.

Multivariate analysis

Debtor and creditor days are one measure of the use of trade credit among SMEs; they indicate the length of time for payment of goods to take place. To avoid potential biases or misleading inferences from our results, it is necessary to use several methods of estimating trade credit use. Most noticeably, it is important to take account of differences among SMEs in terms of their financial vulnerability to the crisis. Similar to Love et al. (2007), we examine the use of trade credit prior to the period of financial crisis and during the financial crisis using panel data. The advantages of panel data are significant. Panel data allow for the study of changes in trade credit financing over a period of time and ultimately gives the researcher more information, variability, greater degrees of freedom and efficient estimates (Baltagi, 2008). Most importantly, in terms of the estimates, panel data allow for the control of unobservable and individual heterogeneity (Askildsen et al., 2003) which often leads to biased results with other forms of data. In addition, panel data are most appropriate in studying the dynamics of adjustment (Rodriguez-Rodriguez, 2006), particularly important since our data cover a period of significant change in financing behaviour. Fixed effects estimation is also employed to capture the net effect of the financial crisis on trade credit use. Fixed effects estimation allows for the controlling of time invariant and unobservable firm-specific characteristics influencing trade credit use (Love et al., 2007). This is particularly important given the dynamic behaviour and diverse characteristics of the SME sector (Berger and Udell, 1998; Jordan et al., 1998). We scale our variables for trade credit by firm sales for account receivables and payables in Tables 4, 5 and 10 and by firm assets in Tables 7 and 8.⁹ In Tables 6 and 7, we estimate the length of credit days using the natural logarithm of creditor and debtor days and by the difference between debtor and creditor days for each firm, in total, our analysis includes nine different measures of trade credit including the three measures of creditor and debtor days.

Table 4.

Trade credit and short-term debt.

	1	1A	2A	2B	3A	3B
	Tradecreditorst	Tradecreditorst	Tradedebtors	Tradedebtors	Netcredit	Netcredit
Crisis	−.024***	−.009*	−.007	−.018***	.003	−.007
	(.00)	(.00)	(.00)	(.01)	(.00)	(.00)
Post 1	−.011**	.003	.012**	−.010	.008	−.012
	(.00)	(.00)	(.00)	(.01)	(.00)	(.00)
Post 2	−.012**	.010	.032***	.0001	.014**	−.015*
	(.00)	(.01)	(.00)	(.00)	(.00)	(.01)
Post 3	−.102***	.011	.046***	.005	.035***	−.002
	(.00)	(.01)	(.00)	(.01)	(.00)	(.01)
Crisis*loansta-1	.005***	.006***	−.001	−.001	−.006***	−.007**
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 1*loansta-1	.007***	.009***	.007**	.009**	−.007***	−.004
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 2*loansta-1	.009***	.010***	.004	.007**	−.005*	−.005*
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 3*loansta-1	.009***	.008***	.006**	.002	−.010***	−.009***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
lag Size		−.0002		−.001**		−.0001
		(.00)		(.00)		(.00)
lag salesgrowth		−.0001**		.001		.0001**
		(.00)		(.00)		(.00)
Age		−.004**		.011***		.009***
		(.00)		(.00)		(.00)
Age²		−.0001		−.0001		−.003
		(.00)		(.00)		(.00)
Fixed effects	YES	YES	YES	YES	YES	YES
Constant	.195***	.269***	.257***	.110**	.078***	−.057
	(.00)	(.00)	(.00)	(.04)	(.00)	(.04)
Observations	23,502	16,194	24,651	17,835	28,067	22,850
Number of groups	5247	4336	5488	4642	6281	5824
Observations per group (average)	4.5	3.7	4.5	3.8	4.4	3.9
Adjusted R²	.65	.72	.71	.75	.56	.60

SME: small- and medium-sized enterprise.

The dependent variables are ‘Tradecreditorst’ calculated as accounts payable scaled by sales, ‘Tradedebtors’ calculated as accounts receivable scaled by sales and ‘Netcredit’ as accounts receivable minus payables scaled by sales. Independent variables include the following: ‘Crisis’ represents a year dummy variable for the year of financial crisis impact in Ireland (2008), while ‘Post 1’, ‘Post 2’ and ‘Post 3’ are time dummy variables for the years 2009, 2010 and 2011, respectively. Crisis*loansta-l represents the SME level of short-term bank loans-to-asset ratio 1 year prior to the crisis year interacted with the crisis year. The interactions with ‘loansta-1’ show the effects of ‘loansta-1’ during the crisis and for the 3 years following the onset of the crisis. The models are estimated with fixed effects and include the independent variables of ‘Size’ represented by the natural logarithm of firm assets, a measure of sales growth ‘salesgrowth’ and the age of the firm (Age) and the squared age of the firm ‘Age²’. Standard errors are represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. Industry sector dummies make no difference to results as they are excluded automatically with fixed effects.

Table 5.

Trade credit and cash.

	1	1A	2A	2B	3A	3B
	Tradecreditorst	Tradecreditorst	Tradedebtors	Tradedebtors	Netcredit	Netcredit
Crisis	−.011	−.001	−.001	−.019**	−.019*	−.038***
	(.01)	(.01)	(.00)	(.01)	(.01)	(.00)
Post 1	.004	.008	.015*	−.014	−.022**	−.044***
	(.01)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 2	−.002	.015	.033***	−.007	−.025**	−.059***
	(.01)	(.00)	(.00)	(.00)	(.01)	(.00)
Post 3	−.013	.004	.037**	−.017	.001	−.043**
	(.01)	(.01)	(.00)	(.00)	(.01)	(.01)
Crisis*Cashta-1	−.013	−.008	.0003	.007	.043**	.054***
	(.01)	(.01)	(.00)	(.00)	(.00)	(.00)
Post 1*Cashta-1	−.026**	−.009	.006	.012	.063***	.061***
	(.01)	(.00)	(.01)	(.00)	(.00)	(.01)
Post 2*Cashta-1	−.014	−.009	.009	.010	.078***	.076***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.01)
Post 3*Cashta-1	.006	.011	.033**	.035**	.062***	.060***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.01)
lag Size		−.001		−.0001		−.001
		(.00)		(.00)		(.00)
lag salesgrowth		−.002***		−.0001		.0001***
		(.00)		(.00)		(.00)
Age		−.004**		.013***		.012***
		(.00)		(.00)		(.00)
Age²		−.001		−.0001		−.0001
		(.00)		(.00)		(.00)
Fixed effects	YES	YES	YES	YES	YES	YES
Constant	.189***	.259***	.245***	.054	.069***	−.123**
	(.00)	(.00)	(.00)	(.04)	(.00)	(.00)
Observations	21,953	15,195	22,637	16,452	25,939	21,143
Number of groups	4851	4051	5034	4283	5797	5383
Observations per group (average)	4.5	3.8	4.5	3.8	4.6	3.9
Adjusted R²	.64	.71	.71	.75	.56	.60

SME: small- and medium-sized enterprise.

The dependent variables are ‘Tradecreditorst’ calculated as accounts payable scaled by sales, ‘Tradedebtors’ calculated as accounts receivable scaled by sales and ‘Netcredit’ as accounts receivable minus payables scaled by sales. Independent variables include the following: ‘Crisis’ represents a year dummy variable for the year of financial crisis impact in Ireland (2008), while ‘Post 1’, Post 2’ and ‘Post 3’ are time dummy variables for the years 2009, 2010 and 2011, respectively. Therefore, Crisis* Cashta-1represents the SME level of cash-to-asset ratio 1 year prior to the crisis year interacted with the crisis year dummy. The interactions with ‘Cashta-l’ show the effects of ‘Cashta-l’ during the crisis and the 2 years following the onset of the crisis. The models arc estimated with fixed effects and include the independent variables of ‘Size’ represented by the natural logarithm of firm assets, a measure of sales growth ‘salesgrowth’, firm age and the age of the firm squared ‘Age²’. Standard errors are represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. Industry sector dummies make no difference to results as they are excluded automatically with fixed effects.

Table 6.

Descriptive statistics of the major variables of the study.

Variables	Observations	Mean	2003–2007				2008–2011
			Standard deviation	0.25	Median	0.75	Observations	Mean	Standard deviation	0.25	Median	0.75
Firm age, Age	38,090	13.2	51	5	9	15	38,090	17	51	4	14	20
Total assets	38,090	71,0000	2,200,000	13036.00	43,636	200,000	31,432	920,000	2,400,000	18,014	60,023	260,000
Number of employees, EMP	5096	19	32	3	7	20	15,203	17	29	3	6	18
Creditor payment days, Creditordays	8194	53	76	10	26	62	14,224	53	79	9	26	61
Debtor collection days, Debtordays	8982	78	95	17	45	97	15,965	85	110	16	45	104
Debtor days minus creditor days, Numdays	5724	21	85	−12	11	52	10,630	30	97	−10	13	56
Accounts receivable/firm assets, Tradedebtorsassets	11,518	0.39	.31	0.11	0.33	0.63	18,632	.37	.31	0.09	0.29	0.6
Accounts payable/firm assets Tradecreditorassets	12,447	0.45	.76	0.06	0.2	0.51	17,553	.44	.81	0.05	0.16	0.46
Accounts receivable minus payables/firm assets, Netcredita	11,518	−0.01	.68	−0.13	0.03	0.31	17,553	−.02	.70	−0.09	0.04	0.29
Total firms short-term debt/firm assets, Loansta	36,520	0.52	2.61	0	0	0.08	30,608	0.66	2.94	0	0	.16
Accounts payable/sales, Tradecreditorst	8536	0.19	0.37	0.02	0.07	0.17	15,209	0.19	0.37	0.02	0.08	0.17
Accounts receivable/sales Tradedebtors	8375	0.26	0.43	0.04	0.12	0.28	16,499	0.28	0.44	0.04	0.12	0.31
Accounts receivable minus payables/sales, Netcredit	8375	0.05	0.36	0	0	0.06	15,209	0.09	0.44	−0.03	0.02	0.15
Credit extended by banks as % of GDP, PcreditGDP	38,090	156	29.0	133	159	179	30,472	222.00	9.55	215	225	229
Gross domestic product per capita growth, GDPpcg	38,090	2.9	0.54				38,090	−2.20	2.69
Inter-money market lending rate, irmoneymkt	38,090	2.99	1.03	2.13	2.4	3.64	30,472	1.35	0.97	0.64	0.97	2.06
Cash and Cash equivalent/firm assets, Cashta	33,809	0.61	0.35	0.27	0.69	0.96	28,823	0.62	0.35	0.3	0.73	0.96
Cash of firm and deposits at bank/firm assets, Cashflowta	9187	0.39	1.51	0.02	0.17	0.45	15,779	0.04	14.36	−0.02	0.1	0.35
Intangible to total assets, Intang	23,971	0.01	0.07	0	0	0.01	19,990	0.01	0.07	0	0	0.01
Firm assets (year1) − firm assets (year 0)/firm assets (year 0), Invest	18,337	0.32	1.73	−0.13	0.01	0.13	18,931	0.04	0.87	−0.2	−0.03	0
Net sales plus operating revenues/firm assets, Opprev	21,204	5.07	12.07	1.01	2.06	4.23	28,440	4.73	11.26	0.84	1.82	4
Firm sales (year 1) − firm sales (year 0)/firm sales (year 0), Salesgrowth	13,370	1.83	29.000	−0.05	0.07	0.25	26,865	0.05	0.55	−0.13	−0.01	0.09

Table 7.

Trade credit and asset intangibility.

	1	1A	2A	2B	3A	3B
	Tradecreditorassets	Tradecreditorassets	Tradedebtorassets	Tradedebtorassets	Netcredita	Netcredita
Crisis	−.028***	.011	−.027***	−.010**	.002	−.006
	(.01)	(.01)	(.00)	(.00)	(.01)	(.01)
Post 1	−.039***	.018	−.041**	−.012**	−.006	.019
	(.01)	(.02)	(.00)	(.00)	(.00)	(.01)
Post 2	−.038***	.014	−.034***	.002	.0001	−.003
	(.01)	(.02)	(.00)	(.01)	(.01)	(.01)
Post 3	−.026**	.011	−.030**	.011	−.007	.005
	(.00)	(.02)	(.00)	(.01)	(.01)	(.02)
Crisis*Constrained-1	−.149	−.189	.041	.055	.087	.131
	(.14)	(.16)	(.04)	(.06)	(.11)	(.13)
Post 1*Constrained-1	.151	−.113	.013	−.019	−.168	−.020
	(.14)	(.16)	(.04)	(.06)	(.12)	(.13)
Post 2*Constrained-1	.247*	.216	.033	.020	−.217*	−.224*
	(.10)	(.15)	(.05)	(.05)	(.11)	(.13)
Post 3*Constrained-1	.260**	.223	.005	.001	−.274**	−.267**
	(.00)	(.16)	(.05)	(.00)	(.11)	(.13)
lag salesgrowth		−.0002		−.0001		−.0003
		(.00)		(.00)		(.00)
Age		−.012*		−.010***		−.001
		(.00)		(.00)		(.00)
Age²		−.001		−.0001		−.0001
		(.00)		(.05)		(.00)
Fixed effects	YES	YES	YES	YES	YES	YES
Constant	.484***	.579***	.353***	.518***	−.093***	.018
	(.00)	(.15)	(.00)	(.05)	(.00)	(.10)
Observations	20,730	11,232	18,469	10,739	23,716	13,812
Number of groups	3749	3089	3632	3025	4265	3669
Observations per group (average)	5.5	3.6	5.1	3.6	5.6	3.8
Adjusted R²	.63	.70	.69	.73	.60	.67

SME: small- and medium-sized enterprise.

The dependent variables are ‘Tradecreditorassets’ calculated as accounts payable scaled by total assets, ‘Tradedebtorsassets’ calculated as accounts receivable scaled by total assets and ‘Netcredita’ as accounts receivable minus payables scaled by total assets. Independent variables include the following: ‘Crisis’ represents a year dummy variable for the year of financial crisis impact in Ireland (2008), while ‘Post 1’, ‘Post 2’ and ‘Post 3’ are time dummy variables for the years 2009, 2010 and 2011, respectively. Crisis*Constrained-l represents the SME level intangible asset-to-total asset ratio 1 year prior to the crisis year. The interactions with ‘Constrained-1’ show the effects of ‘loansta-1’ during the crisis and the 3 years following the onset of the crisis. The models are estimated with fixed effects and include the independent variables of measure of sales growth ‘salesgrowth’ and the age of the firm and the squared age of the firm ‘Age²’. Standard errors are represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. Industry sector dummies make no difference to results as they are excluded automatically with fixed effects.

Table 8.

Trade credit and cash flow.

	1	1A	2A	2B	3A	3B
	Tradecreditorassets	Tradecreditorassets	Tradedebtorassets	Tradedebtorassets	Netcredita	Netcredita
Crisis	−.024**	.015	−.025***	−.009**	−.010	−.014
	(.00)	(.01)	(.00)	(.00)	(.01)	(.01)
Post 1	−.035***	.011	−.040***	−.017***	−.018**	−.012
	(.00)	(.00)	(.00)	(.00)	(.00)	(.01)
Post 2	−.033***	.025	−.032***	−.004	−.013	−.006
	(.00)	(.00)	(.00)	(.01)	(.01)	(.01)
Post3	−.039***	.035	−.028***	.002	−.014	−.009
	(.00)	(.00)	(.00)	(.01)	(.01)	(.01)
Crisis*Cash flow-1	−.007***	−.015***	−.0001	.001	.012***	.026***
	(.00)	(.01)	(.00)	(.00)	(.00)	(.00)
Post 1*Cash flow-1	−.004	−.011***	.003***	.005***	.008***	.015***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 2*Cash flow-1	.001	−.007***	−.0001	.001	.010***	.017***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Post 3*Cash flow-1	.007**	−.009***	.001	.004**	.001	.012***
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
lag Size		−.002***		−.001***		−.001*
		(.00)		(.00)		(.00)
l.cashta		−.081***		−.099**		−.015
		(.00)		(.00)		(.00)
Age		−.015***		−.006***		.003
		(.00)		(.00)		(.00)
Age²		−.0001		−.0001		−.0001
		(.00)		(.00)		(.00)
Fixed effects	YES	YES	YES	YES	YES	YES
Constant	.377***	.646***	.326***	.457***	−.027***	−.002
	(.00)	(.00)	(.00)	(.00)	(.00)	(.00)
Observations	15,117	13,393	13,546	12,730	17,332	16,321
Number of groups	2629	2571	2590	2554	3055	3020
Observations per group (average)	5.8	5.2	5.2	5.0	5.7	5.4
Adjusted R²	.58	.60	.69	.69	.57	.58

SME: small- and medium-sized enterprise.

The dependent variables are ‘Tradecreditorassets’ calculated as accounts payable scaled by total assets, ‘Tradedebtorassets’ calculated as accounts receivable scaled by total assets and ‘Netcredita’ as accounts receivable minus payables scaled by total assets. Independent variables include the following: ‘Crisis’ represents a year dummy variable for the year of financial crisis impact in Ireland (2008), while ‘Post 1’, ‘Post 2’ and ‘Post 3’ are time dummy variables for the years 2009, 2010 and 2011, respectively. Crisis*Cash flow-l represents the SME level of cash flow-to-asset ratio 1 year prior to the crisis year. The interactions with ‘Cash flow-1’ show the effects of ‘Cash flow-1’ during the crisis and the 3 years following the onset of the crisis. The models are estimated with fixed effects and include the independent variables of ‘Size’ represented by the natural logarithm of firm assets, a measure firm cash and bank deposits lagged ‘l.cashta’ and the age of the firm and the squared age of the firm ‘Age²’. Standard errors arc represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. Industry sector dummies make no difference to results as they are excluded automatically with fixed effects.

Table 9.

Financial strength and length of credit days.

	1	1A	2A	2B	3A	3B
	logcreditordays	logcreditordays	logdebtordays	logdebtordays	lognumdays	lognumdays
Crisis	−.088***	.024	−.087***	−.162***	−.107***	−.155***
	(.03)	(.04)	(.02)	(.04)	(.04)	(.05)
Post 1	−.079**	.054	−.066**	−.159***	.002	−.078
	(.03)	(.00)	(.00)	(.03)	(.04)	(.06)
Post 2	−.085***	.056	−.089***	−.218***	.027	−.080
	(.00)	(.00)	(.03)	(.04)	(.04)	(.06)
Post 3	−.156***	.003	−.088***	−.263***	−.031	−.166***
	(.00)	(.00)	(.03)	(.04)	(.04)	(.07)
Crisis*Cashta-1	−.046	−.119*	.100***	.126**	.213**	.191
	(.05)	(.06)	(.04)	(.06)	(.09)	(.11)
Post 1*Cashta-1	−.107**	−.154**	.171***	.198***	.198**	.178
	(.05)	(.00)	(.00)	(.05)	(.09)	(.07)
Post 2*Cashta-1	−.067	−.097	.340***	.364***	.275***	.204*
	(.05)	(.06)	(.04)	(.05)	(.09)	(.11)
Post 3*Cashta-1	.059	.037	.434***	.465***	.556***	.484***
	(.01)	(.06)	(.05)	(.05)	(.09)	(.11)
lag Size		−.0001**		−.0001		−.0001
		(.00)		(.00)		(.00)
lag salesgrowth		.0001		.0001		−.0001
		(.00)		(.00)		(.00)
Age		−.032***		.034***		.025***
		(.00)		(.00)		(.01)
Age²		.00001		−.0001		−.001
		(.00)		(.00)		(.00)
Fixed effects	YES	YES	YES	YES	YES	YES
Constant	3.19***	3.61***	3.56***	3.04***	3.47***	2.91***
	(.01)	(.00)	(.00)	(.00)	(.00)	(.35)
Observations	21,072	14,583	22,727	15,956	9334	6679
Number of groups	4759	3956	4973	3.8	2809	2299
Observations per group (average)	4.4	3.7	4.6	3.8	3.3	3.0
Adjusted R²	.64	.67	.72	.75	.65	.75

SME: small- and medium-sized enterprise.

The dependent variables are ‘logcreditordays’ calculated as the natural logarithm of creditor days, ‘logdebtordays’ calculated as the natural logarithm of debtor days and ‘lognumdays’ calculated as the natural logarithm of the difference between debtor and creditor days. Independent variables include the following: ‘Crisis’ represents a year dummy variable for the year of financial crisis impact in Ireland (2008), while ‘Post 1’, ‘Post 2’ and ‘Post 3’ are time dummy variables for the years 2009, 2010 and 2011, respectively. Cashta-1*Crisis represents the SME level of cash-to-asset ratio 1 year prior to the crisis year. The interactions with ‘Cashta-1’ show the effects of ‘Cashta-1’ during the crisis and the 2 years following the onset of the crisis. The models are estimated with fixed effects and include the independent variables of ‘Size’ represented by the natural logarithm of firm assets, a measure of sales growth ‘salesgrowth’ and the age of the firm ‘Age²’. Standard errors are represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. Industry sector dummies make no difference to results as they are excluded automatically with fixed effects.

The variables for credit received are accounts payable scaled by firm sales (tradecreditorst), accounts payable scaled by firm assets (tradecreditorassets) and the number of creditor collection days. The variables for credit extended are accounts receivable scaled by firm sales (tradedebtors), accounts receivable scaled by firm assets (tradedebtorassets) and the number of debtor collection days. Table 6 illustrates differences in mean levels of the variables for the two periods of the pre- and post-crisis. Given that the sample period (2003–2011) covers a period of economic boom and recession, we split the descriptive statistics into two separate periods (2003–2007) and (2008–2011). All the economic indicators such as firm sales growth, profits, GDP per capita growth and retained profits and credit extended by the banking sector to private enterprises (PcreditGDP) are different from pre-crisis to crisis periods. It is also worth noting that the sample as a whole is made up of mainly mature SMEs with an average age of 17 years. Therefore, this research is based on what might be termed ‘resilient firms’ which survived over the 8-year period in question.

To test the set of hypotheses outlined in this study, it is necessary to distinguish between the characteristics of the SMEs in our sample. We employ panel data fixed effects and control for the financial position or strength of SMEs entering the financial crisis period based on their financial vulnerability to the crisis, as measured by (a) the level of short-term debt financing, (b) the level of ‘Cash’ held by the SME prior to and during the crisis, (c) the level of intangible assets to total assets and (d) the level of cash flow of the firm. The Cash variable is a measure of cash stocks held by the company and deposited in banks. As a first measure, we examine the ratio of short-term debt to assets prior to the crisis. Reliance on short-term debt is used as a proxy for vulnerability to the crisis in several studies (Guariglia and Mateut, 2006; Love et al., 2007). As per Hypotheses 2A and 2B, we expect firms with higher short-term debt to reduce their provision of credit as a result of the financial crisis and increase their use of trade credit financing, relatively more so than those with lower short-term debt ratios (Love et al., 2007) given the difficulty in obtaining financing from banks.

Basic regressions for trade credit take the form of the equation below, where t and i indicate the time period and individual SMEs, respectively; α is the firm fixed effect. X is a vector of firm-specific control variables. To examine the responses of SMEs to the crisis, we use the interactions of the financial position of the firm in the pre-crisis year (2007) with the crisis year (2008) and the post-crisis years (Postcrisis), where FST_i(−1) represents the financial strength of SME (i) measured in the pre-crisis year and this value is fixed. Financial strength or position of the firms is measured using the four factors above in separate regressions. ε_it represents the error term which is composed of unobserved time invariant $(\cap_{i})$ and time variant (V_it) factors.

T C_{i t} = α_{i} + β_{1}^{*} c r i s i s + β_{2}^{*} Postcrisis + β_{3}^{*} F S T_{i (- 1)} * c r i s i s + β_{4}^{*} F S T_{i (- 1)} * Postcrisis + X_{i t} + ε_{i t}

Variants of this approach are applied where $ε_{i t} = \cap_{i} + V_{i t}$ .

Causal factors that are time invariant, including industry effects which influence trade credit are captured by the fixed effects. All other explanatory variables change over time and are predicted to influence the level of trade credit. These include age, growth in sales (salesgrowth), cash reserves (Cashta), size (log of total assets) the level of economic activity indicated by GDP per capita (GDPpcg). Our first table shows the significance of an SME financial position and use/provision of trade credit. A Hausman test was also conducted, and this showed in favour of the fixed effect regression over random effects.

Findings

In Table 4,¹⁰ we examine the use of trade credit since the crisis taking financial stance into account by using both year dummies and interactive dummy variables to capture the relationship between the financial position of SMEs entering the crisis and their use of trade credit during and after the crisis. We capture both the levels of credit received and extended as well as a variable to capture the net change in trade credit (Netcredit). To avoid any potential endogeneity/simultaneity, all explanatory variables are lagged. Age is included as well as the values of age squared. Overall, the results indicate that firms with greater short-term debt-to-asset ratios entering the crisis receive more credit (as measured by the interactive dummies in Columns 1 and 1A), and extended less credit compared to pre-crisis levels (Columns 2 and 2A). While overall net credit extended by firms with greater short-term debt-to-asset ratios entering the crisis extended significantly less trade credit in the years 2008, 2009, 2010 and 2011, as indicated in Column 3B. These findings support H2B. The results are also statistically significant in both Columns 1 and 1A, when we include additional firm-specific control variables of age, size and sales growth. Consistently, firms that were more vulnerable to the crisis in terms of their reliance on short-term bank finance extend less trade credit to their customers over the same period. Older firms and higher growth firms appear to be net providers of credit, and it is consistently shown within each regression format that older firms receive less trade credit financing from their suppliers supporting H6.

In Table 5, we examine whether firms with a better cash position prior to the crisis provide more trade finance to their customers during the crisis period. As expected, when we control for firm size, sales growth and firm age, those with the greatest levels of cash and cash equivalent reserves, on entering the crisis, extended more trade credit finance. This result is shown to be statistically significant particularly for our variable capturing net credit extended and thus, supporting H2A. There was an increased reliance on trade credit financing among firms most financially vulnerable at the time of the banking crisis, that is, firms with the highest levels of short-term debt financing and lowest levels of cash reserves. These firms are more likely to receive trade credit financing after the onset of the financial crisis, when faced with increased difficulty in rolling over short-term bank debt and overdrafts. This is evidence of a substitution effect for firms most financially vulnerable at the time of the crisis supporting H3.

Impact of firm characteristics

Large firms (as measured by the log of assets) extend more and receive less in the form of trade credit, supporting the proposition of Berger and Udell (1998) that trade credit financing is more important for financing small firms. This result supports the premise that larger and older firms can access funds from institutions because of their relatively greater stocks of collateral, larger cash reserves and well-established banking relationships. The implementation of fixed effects isolates the specific individual effects of the crisis on the level of trade credit extended within the SME sector. There are a number of benefits of fixed effects in this scenario. It allows us to control for unobserved heterogeneous factors as well as time invariant factors that influence the level of trade credit extended. With the inclusion of fixed effects, we can say with more reliability that firms with greater levels of cash extended more credit in the times of crisis, holding other unobservable and industry factors constant. As in Petersen and Rajan (1994, 1997), we use both firm age and size as a proxy for credit worthiness; older firms are thought to be more creditworthy.

Tables 4 and 5 estimate the relation between the financial position of the firms and the level of credit extended and received in terms of quantity. Table 9 tests whether the results hold for the length of time in which credit is extended and received.

We examine the relationship between the financial strength of the firm (measured by cash reserves) and the length of credit extended measured by the number of days in which payment is received by customers and the number of days in which payments are made to suppliers. Results are consistent with expectations, in that financially stronger firms receive less credit in terms of time and extend more over the crisis period. On average, firms with greater cash reserves extend between 12 and 46 percent more time to their customers to repay over the crisis period (Column 2B), holding all other firm characteristics constant supporting H2A and H4. This could reflect customer refusal to repay on time or the willingness on the part of financially stronger firms to allow flexibility in repayments to financially constrained business partners.

The results for credit received show that financially stronger firms are offered less time for repayments (1A) in comparison to pre-crisis periods. However, these results are not as statistically strong as for credit extension. Columns 3A and 3B also include a variable that captures net extension of credit measured in terms of time (lognumdays). This variable confirms the finding that financially stronger firms allowed a net extension of time for repayments greater than pre-crisis periods; this remains statistically significant when we include additional control variables.

The impact of asset intangibility

The final two measures of the financial position of SMEs are derived from a firm’s ratio of intangible to total assets and the levels of cash flow. One of the benefits of asset tangibility, other than reducing asymmetric information, is that tangible assets can be used as collateral in times of bankruptcy and protecting creditor rights (Berger and Udell, 1998; Michaelas et al., 1999). Consequently, we expect firms with a higher ratio of intangible to total assets on their balance sheet to be more likely to be financially constrained over the crisis. Therefore, we expect this group of firms to access trade credit over the crisis period, and on an involuntary basis. We expect firms with greater liquidity measured by their cash flow to extend more trade credit to their customers (Love et al., 2007).

Tables 7 and 8 provide another robustness check to the hypothesis that financially vulnerable/constrained firms were net receivers of credit from informal sources over the crisis period and that financially stronger firms played a significant role as financial intermediaries when bank lending was absent. The results from Table 7 show that firms with less asset tangibility at the time of the crisis received significantly more trade credit (Column 1) and extended significantly less over the subsequent years (Columns 3A and 3B), supporting H5. We suspect, given the reasons outlined above, that this could be evidence of the involuntary granting of credit. Similarly, the results are consistent for cash flow, that is, firms with greater cash flow entering crisis extended more trade credit and received less over the post-crisis period (2008–2011). Again, our regressions find that older firms tend to receive less credit and extend more supporting H6.

Impact of macroeconomic factors

Finally, in Table 10 we test the relationship between trade credit and macroeconomic factors, including the percentage of credit extended by the banking sector within Ireland and a proxy for the interest rate that is charged to SMEs on bank loans. We use fixed effects estimation and use two different specifications for trade credit finance received (tradecreditorst) which is a proxy for the quantity of credit received and logcreditordays which captures the length of time in which creditors are repaid. The final variable (Netcredit) captures the net credit extended by firms. The results show that firms with higher levels of cash-to-asset ratio receive less credit, both in terms of quantity and length of time, controlling for firm-specific characteristics and time. In Column 1, we observe an inverse relation between dependence on short-term bank finance and trade credit. We would assume this to be the case given that trade credit is generally viewed as a short-term means of finance. This is further support of substitution between trade credit and bank credit and, indeed, the counter cyclical nature of trade credit as found in Huang et al. (2011). Furthermore, it highlights the financial vulnerability of firms entering the financial crisis with high dependence on short-term bank finance. As shown in Table 4, firms more vulnerable to the crisis required increased trade credit financing from their suppliers.

Table 10.

Trade credit and bank credit.

	Tradecreditorst	Credit days	Netcredit	Netcredit
l.Turasset	−.001	−.005***	−.001	−.001
	(.00)	(.00)	(.00)	(.00)
l.loansta	−.003*	.005	.0001	.001
	(.00)	(.00)	(.00)	(.00)
l.irmoneymkt	−.002**	.001	−.004***	−.003**
	(.00)	(.01)	(.00)	(.00)
l.Cashta	−.030***	−.183***	−.058***	−.058***
	(.00)	(.00)	(.00)	(.00)
PcreditGDP				.000** (.00)
Time dummies	YES	YES	YES	NO
Fixed effects	YES	YES	YES	YES
Constant	197***	3.179***	.126***	.042***
	(.00)	(.02)	(.01)	(.01)
Observations	18,368	17,643	25,753	25,753
Number of groups	4970	4695	6103	6103
Adjusted R²	.69	.70	.57	.58
R² overall	.77	.75	.67	.68

GDP: gross domestic product.

The dependent variables are ‘Tradecreditorst’ calculated as accounts payable scaled by sales, ‘Credit days’ calculated as the natural logarithm of the number of creditor days and ‘Netcredit’ as accounts receivable minus payables scaled by sales. ‘l.loansta’ is the lag of the ratio of short-term loans from financial institutions to firm current liabilities, ‘l.Turasset’ is the 1 year lag of the ratio of firm profits to assets. ‘l.Cashta’ is the 1 year lag of firm cash and deposits scaled by firm assets. ‘l.irmoneymkt’ is a proxy for the cost of bank funds, calculated as the 1 year lag of the money market rate. Standard errors are represented in parentheses, while ***, ** and * represent coefficients significant at the 1%, 5% and 10% levels, respectively. No time dummies are included in Column 4 due to correlation with PcreditGDP.

The net interest margins (ratio of net interest income to average interest earning assets) of Irish banks declined steadily over the period 1997–2012 and in particular, over the period of the financial crisis (Central Bank of Ireland, 2013). The majority of Irish bank operating income is sourced from net interest income. While the inter-money market bank rate¹¹ is a blunt estimation of the rates at which bank credit is extended to SMEs within the Irish economy. The variable ‘l.moneymkt’ can be considered as a proxy for the cost of bank financing. The higher the cost of money on the international markets, the more banks charge on the money they lend to SMEs. Therefore, we would expect that the higher the cost for external financing, the more we would expect firms to seek trade credit financing. We examine the effect of the lag of the money market rate (the rate at which banks borrow for funding purposes) and the effect of the money market rate on trade credit use. The negative coefficient of inter-bank lending rates and the amount of trade credit extended and received by SMEs is interesting. In all three columns, with the exception of ‘credit days’, the results show that the higher the money market rate, the lower the amount of credit extended and received in the economy. However, we are unable to detect a significant association between the percentage of credit extended by the banking sector as a proportion of GDP (PcreditGDP) and the level of trade credit due to this being a one-country study.

The reduction in net credit extended by the banking sector in Ireland over the period 2007–2011 coincides with a reduction in the level of trade credit extended and received within the SME sector (H1). However, as we have seen in the analysis, this is not the case for all sectors and for all firms. Therefore, we cannot conclusively support the hypothesis that liquidity shocks are propagated along the supply chain (Boissay and Gropp, 2007; Love and Zaidi, 2010) during a systematic shock leading to a reduction in credit to all firms. Nevertheless, given the severity of the crisis, and the fact that firms that would not have difficulty receiving credit from banks in normal times experienced significant difficulties in obtaining funds during the crisis, we find strong evidence supporting a substitution effect. Firms that can obtain trade credit do; however, the instance of the substitutability may be limited to a period of time. Interestingly, our results do show that larger firms with greater cash reserves and liquidity at the time of the crisis extended significantly more trade finance to less financially liquid firms for a period of time post the onset of the financial crisis. This shows that there is evidence of an adjustment process in financing for some SMEs. On this basis, when we model trade credit using the panel regressions, we show that profitable firms are more likely to voluntarily extend credit, although the period of extension may be limited.

Limitations and suggestions for further research

Our analysis ends in 2011, and it would be interesting to assess the extent to which profitable firms continued to extend trade credit in subsequent years, when bank financing to SMEs was still very much restricted and aggregate demand remained weak, both in Ireland and the EU. A key question relates to how the financial crisis will impact on trade credit use, survivorship, economic growth and recovery in the long term. Our analysis has focused on the substitution and redistribution of financing in surviving SMEs, as at present, our data do not provide adequate coverage of failed firms. Extending the analysis to failed firms is an important avenue for future research. While we used an unbalanced panel in this study, we do not believe our results are influenced by an attrition bias. Subject to data availability, it would be interesting to study the effects on supplier and customer relations to assess how trade credit use influenced future lines of business relations and growth. Furthermore, quarterly data as opposed to annual data would also improve our understanding in terms of firms’ immediate behaviour in the aftermath of shocks to the financial system.

In addition, data limitations prevent us from reporting on the growing use of business credit cards issued by banks and other financial institutions as a direct attempt to get firms to reduce reliance on overdraft facilities. This is an important development which in part reflects rising risk but also the high capital costs of providing overdraft facilities under Basel regulations.¹² National Small Business Association (NSBA, 2008) found that 44% of small businesses in the United States used credit cards at the beginning of the financial crisis to finance capital needs, further reflecting a trend of increased business credit card use since the 1990s. Fraser (2009) states that while credit cards are the most widely used financial product among UK small businesses, the study found that in both periods 2001–2004 and 2005–2008, the level of usage appears to remain stable at 55.3% and 54.4%, respectively. Unfortunately, our data do not give us detailed information on the use of business credit cards for our sample. Amadeus reports business credit cards, overdrafts and short-term loans of less than one year in aggregate form as short-term liabilities. Nevertheless, our aggregate figures indicate that any increase in the use of credit cards over the crisis period did not mitigate the increased role of trade credit. This point is reinforced by findings for the United Kingdom (Cosh et al., 2009). In addition, our summary statistics show that the sample is not biased to sectors where uptake over time of such cards is more commonplace (e.g. B2B and personal services) and the industry composition of our sample is also stable over time. Finally, further research on the role of market power and SME trade credit contract terms would be an important extension to the research.

Conclusion

This study shows that unlisted financially vulnerable Irish SMEs entering the financial crisis received more trade credit from suppliers and extended less trade credit to their customers during the crisis and thereafter. This article is one of the first, to our knowledge, that shows empirical evidence of a substitution effect in the context of a panel data sample of unlisted SMEs post the 2008 financial crisis. The contextual setting for this research makes an interesting case. During this period, there was a boom in bank lending and a sudden and very dramatic shock to the economy and the SME sector. The timeframe covers the period of economic and financial expansion as well as a financial crisis and finds evidence of an adjustment process and substitution effect in the financing of SMEs. While aggregate levels of trade credit declined over the crisis, the data unequivocally show that trade credit financing played an important role in the financing of SMEs throughout the banking crisis. We find strong support of a substitution effect between trade credit and bank credit over the recent financial crisis period for financially vulnerable SMEs. We suggest that both redistribution and substitution effects are best specified in terms of the financial position and financial strength of firms at the time of the crisis rather than the age or size of the firm. The policy implications of this article are important in light of the recent financial crisis. If during a financial crisis, larger, financially stronger and liquid firms have the ability to redistribute credit to financially constrained SMEs, this provides a source of finance to firms that otherwise would not be available to them. Therefore, any policy that restricts the profitability, cash reserves and access to finance for larger/more financially liquid firms has adverse effects for SMEs by restricting their ability to receive trade credit in place of bank finance when bank lending is restricted. Given the importance of SMEs in terms of national output and employment, this issue has potential significance for economic recovery and the avoidance of compounding the growth crisis. For SMEs in Ireland, survey data shows that the majority of demand requests for credit are for working capital and cash flow purposes (Mazars, 2011). The late payment for goods is particularly important for the working capital financing of micro and financially weaker SMEs. The European Commission (2008) “ Think Small First” a “ Small Business Act” for Europe makes specific reference to trade credit in its 10-point plan and highlights that on average, SMEs wait between 20 and 100 days for the payment of goods, with one in four insolvencies due to late payment. Therefore, this is clearly an issue for further consideration and importance for other countries within the EU. Our results show that while trade credit is used for transaction purposes within the economy in non-crisis periods, there appears to be some degree of substitutability between the cost of bank credit and the use of trade credit as measured by our proxy for the cost of inter-bank lending. Finally, while the results of this study suggest that some involuntary use of trade credit is evident in our data, the findings robustly show that financially strong firms are more likely to extend finance, even though the period of extension may be limited.

Footnotes

Appendix 1

Table 13.

Average (mean) levels of net credit extended by sector.

Industry sector	2004	2005	2006	2007	2008	2009	2010	2011	Average	β
Food processing/manufacturing	0.09	0.1	0.07	0.07	0.04	0.001	−0.0006	0.024	0.05	−0.014**
Construction	0.04	0.06	0.02	0.08	0.06	0.007	0.005	−0.03	0.03	−.0009*
Real estate	0.08	0.09	0.1	0.14	0.15	0.18	0.19	0.22	0.14	020***
Wholesale	0.19	0.11	0.14	0.12	0.08	0.06	0.07	0.06	0.10	−.016***
Retail trade	−0.02	−0.03	−0.03	−0.05	−0.06	−0.06	−0.04	−0.06	−0.04	−.005**
Hospitality and tourism	0.002	0.005	0.001	−0.001	−0.006	−0.011	−0.0002	−0.002	0.00	−.006*
Business services	0.05	0.07	0.05	0.08	0.05	0.06	0.06	0.066	0.06	.001
Community services	0.001	0.005	0.006	0.004	0.001	0.004	0.001	0.008	0.00	.0002
Average across years	0.05	0.05	0.04	0.06	0.04	0.03	0.04	0.04	0.04

Net credit is calculated as trade receivables minus payables divided by the total sales of the firm in each year. β is the estimated coefficient of the regression TC = α + β₁ (year) + ϵ₁, indicating how much trade credit extended has reduced/increased for each sector for each year of the sample from a simple pooled regression with no control variables. Positive values of β indicate an increase in average levels of trade credit extended in a given sector over the sample period, while negative values of β indicate how much they have reduced. ***, ** and * represent statistical significance at the 1%, 5% and 10% levels, respectively.

Funding

This research was undertaken by Gerard McGuinness while a candidate for a PhD in Finance at Dublin City University. The research was funded under the DCU Business School Doctoral Scholarship Programme.

Notes

Author biographies

Gerard McGuinness is a PhD candidate in Finance at Dublin City University Business School. Email: gerard.mcguinness@dcu.ie

Teresa Hogan is a Director of Doctoral Studies and a senior lecturer in Finance and Entrepreneurship at Dublin City University Business School. Email teresa.hogan@dcu.ie

References

Agostino

Gagliardi

Trivieri

(2012) Bank competition, lending relationships and firm default risk: An investigation of Italian SMEs. International Small Business Journal 30: 907–943.

Aktas

De Bodt

Lobez

. (2012) The information content of trade credit. Journal of Banking & Finance 36: 1402–1413.

Askildsen

Baltagi

Holmas

(2003) Wage policy in the health care sector: A panel data analysis of nurses’ labour supply. Health Economics 12: 705–719.

Baltagi

(2008) Econometric Analysis of Panel Data (4th edn). Chichester: John Wiley & Sons.

Banos-Caballero

García-Teruel

Martínez-Solano

(2012) How does working capital management affect the profitability of Spanish SMEs? Small Business Economics 39: 517–529.

Bastos

Pindado

(2013) Trade credit during a financial crisis: A panel data analysis. Journal of Business Research 66: 614–620.

Beck

Demirgüç-Kunt

Maksimovic

(2008) Financing patterns around the world: Are small firms different? Journal of Financial Economics 89: 467–487.

Berger

Udell

(1998) The economics of small business, finance, the role of private equity and debt markets in the financial growth cycle. Journal of Banking & Finance 22: 613–673.

Bernanke

B S

(1983) Non-monetary effects of the financial crisis in the propagation of the Great Depression. Working Paper 1054, January. Cambridge, MA: National Bureau of Economic Research.

10.

Biais

Gollier

(1997) Trade credit and credit rationing. Review of Financial Studies 10: 903–937.

11.

Boissay

Gropp

(2007) Trade credit defaults and liquidity provision by firms. ECB Working Paper no. 753. Available at: http://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp753.pdf

12.

Burkart

Ellingsen

(2004) In-kind finance: A theory of trade credit. The American Economic Review 94: 569–590.

13.

Calomiris

Himmelberg

Wachtel

(1995) Commercial paper, corporate finance and the business cycle, a microeconomic perspective. Carnegie-Rochester Conference Series on Public Policy 42: 230–250.

14.

Carbó-Valverde

Rodriquez-Fernandez

Udell

(2012) Trade credit, the financial crisis and firm access to finance. Working paper. Available at: http://www.centralbank.ie/stability/Documents/SME%20Conference/Session%202/paper.pdf (accessed 25 September 2013).

15.

Central Bank of Ireland (2012) SMEs in Ireland: Stylised facts from the real economy and credit market. Available at: http://www.centralbank.ie/publications/Documents/SMEs%20in%20Ireland%20Stylised%20facts%20from%20the%20real%20economy%20and%20credit%20market.pdf (accessed 25 September 2013).

16.

Central Bank of Ireland (2013) The impact of the financial crisis on banks’ net interest margins, vol. 2013, issue 1. Available at: http://www.centralbank.ie/stability/Documents/Economic%20Letter%20-%20Vol%202013,%20No.%201.pdf (accessed 25 September 2013).

17.

Cosh

Hughes

Bullock

. (2009) SME Finance and Innovation in the Current Economic Crisis. Centre for Business Research University, Cambridge: University of Cambridge. Available at: http://www.cbr.cam.ac.uk/pdf/CrCr_EconCrisis.pdf (accessed 20 August 2014).

18.

Choi

Kim

(2005) Trade credit and the effect of macro financial shocks: Evidence from US panel data. Journal of Financial and Quantitative Analysis 40: 897–925.

19.

Company Registration Office (CRO) Report 2012 (2013) Company Registration Office. Available at: https://www.cro.ie/Portals/0/Corporate%20Publications/Companies%20Registration%20Office/2012%20English%20Version%20CRO%20annual%20report.pdf [accessed 14 August 2014]

20.

Coulibaly

Sapriza

Zlate

(2013) Financial frictions, trade credit and the 2008-2009 global financial crisis. International Review of Economics & Finance 26: 25–38.

21.

Cowling

Liu

Ledger

(2012) Small business financing in the UK before and during the current financial crisis. International Small Business Journal 30: 778–800.

22.

Cunat

(2007) Trade credit: Suppliers as debt collectors and insurance providers. The Review of Financial Studies 20: 491–527.

23.

Danielson

Scott

(2004) Bank loan availability and trade credit demand. The Financial Review 39: 579–600.

24.

Demirguc-Kunt

Maksimovic

(2001) Firms as financial intermediaries: Evidence from trade credit data. World Bank Policy Research Working Paper no. 2696. Available at: http://www-wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2001/11/28/000094946_0111090402236/Rendered/PDF/multi0page.pdf [accessed 14 November 2014]

25.

Drakos

(2013) Bank loan terms and conditions for Eurozone SMEs. Small Business Economics 41: 717–732.

26.

European Commission (2008) Communication from the commission to the council, the European Parliament, the European economic and social committee and the committee of the regions: ‘Think small first’ a ‘Small Business Act’ for Europe. Available at: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2008:0394:FIN:EN:PDF (accessed 30 November 2013).

27.

Eurostat (2013) The Structural Business Database, Business Demography. Available at: http://epp.eurostat.ec.europa.eu/portal/page/portal/european_business/data/database [accessed 10 August 2014]

28.

Fishman

Love

(2003) Trade credit, financial intermediary development and industry growth. Journal of Finance 58: 353–374.

29.

Fraser

(2009) Small firms in the credit crisis: Evidence from the UK survey of SME finances. Economic & Social Research Council. Available at: http://www2.warwick.ac.uk/fac/soc/wbs/research/ei/latest/small_firms_in_the_credit_crisis_v3test.pdf (accessed 25 May 2014).

30.

Garcia-Appendini

Montoriol-Garrig

(2013) Firms as liquidity providers: Evidence from the 2007-2008 financial crisis. Journal of Financial Economics 109: 272–291.

31.

Qiu

(2007) Financial development, bank discrimination and trade credit. Journal of Banking & Finance 31: 513–530.

32.

Giannetti

Burkart

Ellingsen

(2011) What you sell is what you lend? Explaining trade credit contracts. Review of Financial Studies 24: 1261–1298.

33.

Guariglia

Mateut

(2006) Credit Channel, trade credit channel, and inventory investment: Evidence from a panel of UK firms. Journal of Banking and Finance 30: 2835–2856.

34.

Hernández-Cánovas

Koeter-Kant

(2011) SME financing in Europe: Cross-country determinants of bank loan maturity. International Small Business Journal 29: 489–507.

35.

Huang

Shi

Zhang

(2011) Counter-cyclical substitution between trade credit and bank credit. Journal of Banking & Finance 35: 1859–1878.

36.

Jordan

Lowe

Taylor

(1998) Strategy and financial policy in UK small firms. Journal of Business, Finance & Accounting 25: 1–27.

37.

Lane

McQuade

(2013) Growth and international capital flows. Scandinavian Journal of Economics 1: 218–252.

38.

Love

Zaidi

(2010) Trade credit, bank credit and financial crisis. International Review of Finance 10(1): 125–147.

39.

Love

Preve

Sarria-Allende

(2007) Trade credit and bank credit: Evidence from the recent financial rises. Journal of Financial Economics 83: 453–469.

40.

Mac an Bhaird

Lucey

(2010) Determinants of capital structure in Irish SMEs. Small Business Economics 35: 357–375.

41.

Martínez-Solo

García-Tereul

Martínez-Solano

(2013) Trade credit and SME profitability. Small Business Economics. Epub ahead of print 19 June 2013. DOI: 10.1007/s11187-013-9491-y.

42.

Marotta

(1997) Does trade credit redistribution thwart monetary policy? Evidence from Italy. Applied Economics 29: 1619–1629.

43.

Mazars (2011) SME lending demand study: Prepared on behalf of the department of finance for the period April to September 2011. Available at: http://www.finance.gov.ie/sites/default/files/mazarbanklend_0.pdf [accessed 30 November 2013]

44.

Meltzer

(1960) Mercantile credit, monetary policy, and the size of firms. Review of Economics and Statistics 4: 429–436.

45.

Michaelas

Chittenden

Poutziouris

(1999) Financial policy and capital structure choice in U.K. SMEs: Empirical evidence from company panel data. Small Business Economics 12: 113–130.

46.

National Small Business Association (NSBA) (2008) 2008 Survey of small and mid-sized business. Available at: http://www.nsba.biz/docs/2008bizsurvey.pdf (accessed 09 September 2014).

47.

Kilholm Smith

Smith

(1999) Evidence on the determinants of credit terms used in interfirm trade. Journal of Finance 49: 3–37.

48.

Nilsen

(2002) Trade credit and the bank lending channel. Journal of Money, Credit and Banking 34: 226–253.

49.

Petersen

Rajan

(1994) The benefits of lending relationships: Evidence from small business data. Journal of Finance 49: 3–37.

50.

Petersen

Rajan

(1997) Trade credit: Theories and evidence. Review of Financial Studies 10: 661–691.

51.

Rodriguez-Rodriguez

(2006) Trade credit in small and medium sized firms: An application of the system estimator with panel data. Small Business Economics 27: 103–126.

52.

Stiglitz

Weiss

(1981) Credit rationing in markets with imperfect information. The American Economic Review 71: 393–410.

53.

Tsuruta

(2008) Bank information monopoly and trade credit: Do banks have information of small businesses? Applied Economics 40: 981–996.

54.

Vanacker

Manigart

(2010) Pecking order and debt capacity considerations for high growth companies seeking finance. Small Business Economics 35: 53–69.

55.

Vermoesen

Deloof

Laveren

(2013) Long-term debt maturity and financing constraints of SMEs during the global financial crisis. Small Business Economics 41: 433–448.

56.

Whelan

(2013) Ireland’s economic crisis: The good, the bad, and the ugly. Journal of Macroeconomics 39: 424–440.

57.

Wilner

(2000) The exploration of relationships in financial distress: The case of trade credit. Journal of Finance 55: 153–178.

58.

Rui

(2011) Trade credit, cash holdings and financial deepening: Evidence from a transitional economy. Journal of Banking & Finance 36: 2868–2883.