Cash holdings of listed and unlisted firms: new evidence from the euro area

Abstract

This paper examines the cash holdings behavior of listed and unlisted firms. We argue that unlisted firms, which are smaller, face a greater wedge between the cost of external and internal finance and as a result they need to rely more on the later. Relying on internal funds means that firms have a precautionary motive to hold cash. We test our theory using an unbalanced panel of mainly small medium enterprises within the euro area over the period 2003–2017 paying special attention to the role of financial pressure, financial constraints and the recent financial crisis. Our findings reveal that unlisted firms hold more cash than their listed counterparts due to precautionary motives. In addition, when considering the effect of financial pressure, the results show that the difference in cash holdings between listed and unlisted firms exhibit a ‘U-shaped’ relationship. Finally, unlisted firms have a higher sensitivity to save cash out of cash flow than listed firms. Our results are robust to using different specifications and different financial pressure measures.

Keywords:

• Cash holdings
• listed and unlisted firms
• precautionary motives
• financial pressure

JEL Classifications:

• C23
• C26
• G01
• G30
• G32

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Stylianos Asimakopoulos http://orcid.org/0000-0002-1362-5865

Notes

1 Such as dividend payout, cash flow management, working capital, and investment plans.

2 See Chen, Goldstein, and Jiang (2007) and Edmans, Goldstein, and Jiang (2012) for empirical evidence.

3 The unlisted firm classification includes both private firms and unlisted public firms. The listed firm classification includes all type of listed firms. So, the term ‘listed’ is similar to the term ‘public’ as used in literature. The separation between unlisted and listed firms is crucial and in this paper we follow the work of Mortal and Reisel (2013). For each firm into investigation its Initial Public Offering (IPO) date and delisting date from the stock market is checked. Then firms are reclassified as unlisted or listed based on this information. See more details in Section 3.

4 Chava and Purnanandam (2011) find that bank-dependent firms are more affected during banking crises than firms with access to public debt markets. Carvalho, Ferreira, and Matos (2015) find that borrowers with pre-crisis relationships with less healthy lenders were more affected by the 2007–2009 financial crisis compared to borrowers of healthier lenders.

5 For empirical support of trade-off theory see Opler et al. (1999), Almeida, Campello, and Weisbach (2004), Ozkan and Ozkan (2004), Han and Qiu (2007) and Bates, Kahle, and Stulz (2009), among others.

6 For empirical support of pecking order theory see de Haan and Hinloopen (2003), Dittmar, Mahrt-Smith, and Servaes (2003), Ferreira and Vilela (2004) and Bigelli and Sánchez-Vidal (2012).

7 For empirical support of free cash-flow theory see Harford (1999), Dittmar and Mahrt-Smith (2007) and Kalcheva and Lins (2007).

8 China has a unique environment due to lage share of government ownership of the firms.

9 Furthermore, SMEs have more difficulties in raising finance compared to large firms and normally encounter difficulties in signalling their creditworthiness (De Maeseneire and Claeys 2012).

10 Gao, Harford, and Li (2013) and Farre-Mensa (2014) employ a sample of firms from Capital IQ database. This database only reports information on private and public firms with minimum annual revenue of approximately 5 million euros. Amadeus, which is the database used in this paper, includes information on firms with less than 2 million euros.

11 Austria, Belgium, Finland, France, Germany, Greece, Italy, Luxembourg, Portugal and Spain.

12 There is another strand of literature assessing temporary and permanent cash flow shocks and their impact on cash holdings, e.g. Gryglewicz et al. (2017). However, our paper is deviating from these papers as our variable of interest utilizes a permanent firm status rather than a shock.

13 Denis (2011) claims that firms with costly external financing can undertake valuable investments opportunities only by keeping larger cash reserves.

14 Information asymmetry is considered as an important barrier for unlisted firms, as they are subject to lower levels of disclosure, supervision and external auditing. Additionally, unlisted firms lack a public price as a mechanism to signal their quality to investors, they do not benefit from the presence of analysts and they are subject to less accurate and less efficient monitoring (Mantecon 2008).

15 Asimakopoulos et al. (2017) show that under asymmetric information investors by using an appropriate econometric technique that utilize a time-disaggregated dividend–price ratio could reveal the link between dividend yield and future dividend growth, exploiting on that way good listed-firm prospects which are embedded in the stock price while dividends are sticky or smoothed. Also, Brav (2009) points out that public firms are reluctant to alter their dividend policy in response to changes in their performance, contrary to what happens in private firms which are more sensitive to their operating performance.

16 This argument is supported by recent studies which explore firm behavior in the Euro-area (see for example Ferrando and Mulier 2013).

17 Riskier firms hold higher levels of cash as a precaution since they have higher levels of debt relative to their cash flows. Low-risk firms also hold higher levels of cash due to a pecking order issue.

18 Our results remain unchanged when we use the propensity score matching approach for the unlisted firms so as to match the listed firms sample to deal with the fact that the majority of our sample consists of unlisted firms.

19 To overcome the limited data availability of R&D expenses we follow Brown and Petersen (2011) and Guney, Karpuz, and Ozkan (2017) and we use the cost of employees as a proxy for R&D defined as the firms' total personnel. In addition, as in Bigelli and Sánchez-Vidal (2012) and Mulier, Schoors, and Merlevede (2016), data on dividends are unavailable in Amadeus and as a result we are not able to use them in our analysis. See Table A1 in the Appendix for the definition of the variables in the data-set.

20 See Section 3.1 for details.

21 It should be noted that the latter test can be relatively weak for large samples. Blundell, Bond, and Windmeijer (2001) shows using Monte Carlo experiments that this test tends to over-reject the null hypothesis of valid instruments for the system GMM, especially for large samples. Chen and Guariglia (2013) confirm this result using a large panel of Chinese firms.

22 It should be noted that Ireland does not report information for public firms. Also Netherlands does not have enough information on listed/unlisted firms. As a result these countries are dropped from the database. We do not incude any non-Eurozone countries in our analysis so as to eliminate the impact from currency changes.

23 It should be noted however, that Amadeus database only provides a contemporaneous information rather than a historical information.

24 For example, if a firm had an IPO in 2009 and it also has accounting information from 2003 to 2011, Amadeus database classifies the firm as public throughout the sample period. Thus, in this situation the firm is reclassified as unlisted from 2003 to 2008 and as listed from 2009 to 2011. The same methodology is employed for the delisting case.

25 See Tables A2 and A3 for the structure of the panel in the appendix.

26 According to the European Comission (EU), in 2012 Italy introduced fiscal incentives for the issuing of minibonds by unlisted firms. Similarly, Spain created an Alternative Fixed-Income Market (‘Mercado Alternativo de Renta Fija-MARF’) which was focused on trading bond of SMEs. Finally, Portugal simplified its legislative framework, making it easier to issue commercial paper for SMEs.

27 We have also examined classifying firms regarding their level of total assets at a different percentile (i.e. 80th, 60th percentile etc.). We do not show the results here to save space, however, we would like to mention that the coefficient of the unlisted dummy becomes positive again at the 66th percentile.

28 We need to use the nearest neighbor matching approach for total assets for a given country and industry because it is impossible to find a listed and an unlisted firm with the exact same level of assets for a given country and industry.

29 For simplicity we present only the result from the baseline estimation. However, the results of the other estimations remain valid and are availiable upon request.