Abstract
Linking the theoretical predictions of the research on lending relationships with those of the literature on managerial incentives, we investigate whether the duration of credit relationships impacts on SMEs’ technical efficiency. Our hypothesis is that the balance between costs and benefits of enduring banking relationships might have heterogeneous effects on managers’ incentives depending on the level of firms’ indebtedness. Using a large sample of European SMEs, observed in the period 2001–2008, and adopting both parametric and non-parametric measures of efficiency, we find that the positive impact of longer lending relationships on efficiency decreases as indebtedness increases, suggesting that moral hazard problems may endanger firms’ technical efficiency.
Similar content being viewed by others
Notes
Focused initially on the issue of the separation between ownership and control in large companies, this theory has found important applications also in the SMEs’ literature (e.g.: Ang 1992; McMahon 2004). In this context, the primary agency relationship is not between owners and managers, as these roles often overlap, but “between insiders and outside suppliers of funds” (McMahon 2004, p.123). For this reason, in the following, when using the term managers we mean owner-managers.
Nonetheless, the multiple borrowing practice does not preclude Italian firms from having strong ties with an individual bank, especially in the case of SMEs (e.g.: Carmignani and Omiccioli 2007).
Vice versa, efficiency may also be defined as the ability to minimise the amount of inputs required to produce a given output level. An output orientation is commonly adopted when it is fair to assume that firms seek to maximize output for given input combinations (as in the manufacturing case; see for instance Milana et al. 2013). By contrast, when producers have a statutory obligation to meet demand, and they also have to guarantee certain quality levels, it is proper to assume that firms attempt to minimise input costs for given output levels.
Schmidt and Wan (2002) recall that the two-step SFA approach is biased if the efficiency determinants and the inputs are correlated. Furthermore, the assumption that the inefficiency term is independently and identically distributed in the first step is at odds with the second step, where the efficiency terms are assumed to be normally distributed and dependent on some explanatory variables.
Banker and Natarajan (2008) show that the OLS estimator in the second-stage regression is statistically consistent under some circumstances, providing theoretical justification for its use. Yet, Simar and Wilson (2011) highlight that this consistency depends crucially on quite restrictive assumptions on the production process, that should not be expected to hold in general (on this issue, see also Badin et al. 2014).
To corroborate our expectation, we run a nonparametric output-based test of returns to scale (Simar and Wilson 2002). In each sector/year, the null hypothesis that the technology is globally CRS (versus VRS) is always rejected, with rare exceptions.
The output is approximated by the sales, deflated by producer prices. Analogously, capital and raw materials volumes are approximated by their respective costs, deflated by inputs prices. Finally, labour is measured by the number of employees. In order to interpret the deflated values as real aggregates, since we lack data on inputs and outputs prices at the firm level, as described in the data section, we employ deflators that are as close as possible to the prices of the group of goods we employ.
Different stochastic frontier models have been proposed for panel data (we refer to Greene, 2008, and Kumbhakar et al. 2014, for some reviews). The model most commonly adopted is Battese and Coelli (1995), which allows for a time-varying inefficiency term. However, it does not capture firm-specific latent heterogeneity, thus firm effects are mixed with inefficiency. The Greene (2005) “true” fixed effects (TFE) or “true” random effects (TRE) models have the merit of disentangling these components. In the TFE model, the producer-specific effects can be correlated with the regressors. However, we discard this model, as it is plagued by an incidental parameters problem when the time span of the panel is limited compared to the number of units, as it is the case in our analysis. Indeed, Belotti et al. (2013) recall that the maximum-likelihood dummy variable (MLDV) approach does not appear appropriate when the length of the panel is T ≤ 10.
Within the TRE model, by parameterizing the variance, one not only allows the inefficiency term uit to be heteroskedastic, but also can investigate the determinants of the average inefficiency. Indeed, the mean of the inefficiency term ui is in direct proportion to its standard deviation, for all commonly adopted distributions of the inefficiency term. Hence, a factor influencing the variance of uit is bound to influence its mean in the same direction.
To preserve the anonymity of the firms surveyed, the EFIGE dataset provides information on industrial sectors in the form of a randomized identifier ranking from 1 to 11.
The preceding question is “What % of your firm’s total bank debt is held at your main bank?”
Incidentally, the EFIGE survey does not provide the identity of a firm’s main bank, and information concerning other lending relationships’ characteristics—such as the percentage of the firm’s total bank debt held by the main bank, and the number of lending banks—is available for 2009 only.
It should also be recalled that the EFIGE dataset neglects firms with less than ten employees, thus implying that our results may not be generalised to the smallest of firms. Furthermore, accounting information refers to firms that are surveyed in 2010, thus defaulted entities are excluded. Hence, our findings are conditional on survival (as in other works based on the same source of data, such as Barba Navaretti et al. 2014; Agostino and Trivieri 2018).
For instance, if the quantile on which the firm is efficient (in the output orientation) is 0.95, this means that there are 5% (1−0.95 = 0.05) of the firms in the comparison set (firms employing at most the same level of inputs) which outperform the considered firm by producing more output. Thus, order-α partial frontiers allow for super-efficient units, i.e. in the output-oriented (input-oriented) case scores may fall short of (exceed) the value of 1.
Following Simar (2003), if we considered as potential outliers all DMUs with super-efficiency higher than 10%, we would have eliminated about 15% of observations.
The choice of α = 99% has been made after applying the following procedure: for each sector in each year, we run series of order-α efficiency analyses using increasing values for α. Subsequently, we plotted the share of super-efficient firms as a function of increasing values of α. In the large majority of sectors/years, we noticed a smooth decrease of the share, which is considered as a symptom of absence of outliers. We also considered three rules for detecting points of discontinuity (two based on either rough or smoothed series of differences in differences, the other based on the BIC detection rule). Yet, the share of outliers based on the most conservative of these rules was too frequently higher than any reasonable upper bound.
The marginal effect of DURAT, and the relative standard error are computed conditional on the level of LEV, accounting for the non-linear nature of the model. We omit the others graphs for the sake of conciseness, making them available upon request.
Given monitoring costs and regulatory constraints, banks may prefer to minimise the counterparty risk by diversifying their exposures with a higher number of firms (Carletti, 2004). Also, banks may tend to act as free riders: believing that other banks will bear the cost of monitoring the firm, if information is a public good no one has incentive to do costly monitoring activities. On the other hand, firms may be inclined to establish multiple relations with banks to hide their effective financial situation avoiding a careful monitoring by banks (Foglia et al.1998).
We do not add the percentage of graduate employees, proxy of human capital, as defined on a limited number of observations (even after imputing).
The Z-score is the sum of return on assets plus the capital asset ratio divided by the standard deviation of return on assets, the latter being computed over 3-year rolling time windows.
We cannot emphasise the results of this test for two reasons: first, we do not consider the other determinants included in the efficiency equation. Moreover, we encounter computational problems in several sectors/years, as the splitting above described reduces the number of observations available. Therefore, we proceed applying the test to each sector, separately for each country, pooling observations over the years.
For the sake of conciseness, we omit the estimates concerning the frontier. Incidentally, in the benchmark model, the estimated output elasticities at the mean of the inputs are 0.02 for capital, 0.26 for labour and 0.58 for raw materials.
Karakaplan and Kutlu (2017) suggest a one-step maximum likelihood-based methodology that allows to account for endogenous variables both in the frontier and the inefficiency model. Further, it provides a test of endogeneity, relying “on ideas similar to the standard Durbin-Wu-Hausman test for endogeneity” (Karakaplan and Kutlu 2017, p. 6).
This result is based on instrumental variables defined in 1936 at regional level: the number of branches per million inhabitants (p.m.i); the number of saving banks (p.m.i).; the number of mutual cooperative banks (p.m.i).; the share of branches owned by cooperative Popolari banks, and the share of branches owned by large banks.
References
Agostino M, Gagliardi F, Trivieri F (2012) Bank competition, lending relationships and firms’ default risk. An investigation on Italian SMEs. Int Small Bus J 30(8):907–943
Agostino M, Trivieri F (2017) Collateral in lending relationships. A study on European SMEs microdata. Int Rev Appl Econ 31(3):399–356
Agostino M, Trivieri F (2018) Who benefits from longer lending relationships? An analysis on European SMEs. J Small Bus Manag 56(2):274–293
Alessandrini P, Presbitero AF, Zazzaro A (2009) Geographical organization of banking systems and innovation diffusion. In: Alessandrini P, Fratianni M, Zazzaro A (eds) The changing geography of banking and finance. Springer, New York, p 75–108
Alvarez R, Crespi G (2003) Determinants of technical efficiency in small firms. Small Bus Econ 20:233–244
Ang JS (1992) On the theory of finance for privately held firms. J Small Bus Financ 1(3):185–203
Aw B, Chung S, Roberts MJ (2000) Productivity and turnover in the export market: Micro-level evidence from the Republic of Korea and Taiwan. World Bank Econ Rev 14(1):65–90
Badin L, Daraio C, Simar L (2014) Explaining inefficiency in nonparametric production models: the state of the art. Ann Oper Res 214(1):5–30
Banker RD, Natarajan R (2008) Evaluating contextual variables affecting productivity using data envelopment analysis. Oper Res 56(1):48–58
Barba Navaretti G, Castellani D, Pieri F (2014) Age and firm growth: evidence from three European countries. Small Bus Econ 43:823–837
Battese GE, Coelli TJ (1995) A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empir Econ 20:325–332
Beck T, Demirguc-Kunt A (2006) Small and medium-size enterprises: access to finance as a growth constraint. J Bank Financ 30:2931–2943
Belotti F, Daidone S, Ilardi G, Atella V (2013) Stochastic frontier analysis using Stata. Stata J 13(4):719–758
Berger AN, Udell GF (1995) Relationship lending and lines of credit in small firm finance. J Bus 68(3):351–381
Bhandari A, Maiti P (2012) Efficiency of the Indian leather firms: some results obtained using the two conventional methods. J Product Anal 37(1):73–93
Bharath S, Dahiya S, Saunders A, Srinivasan A (2011) Lending relationships and loan contract terms. Rev Financ Stud 24:1141–1203
Blackwell D, Winters DB (1997) Banking relationships and the effect of monitoring on loan pricing. J Financ Res 20:275–289
Bolton P, Scharfstein DS (1996) Optimal debt structure and the number of creditors. J Political Econ 104(1):1–25
Boot A (2000) Relationship banking: what do we know? J Financ Inter 9:7–25
Carletti E (2004) The structure of bank relationships, endogenous monitoring, and loan rates. J Financ Inter 13(1):58–86
Carmignani A, Omiccioli M (2007). Costs and benefits of creditor concentration: an empirical approach, Bank of Italy, Working Paper N. 645.
Castelli A, Dwyer GP, Hasan I (2012) Bank relationships and firms’ financial performance: the Italian experience. Eur Financ Manag 18(1):28–67
Castiglione C, Infante D, Zieba M (2017). Technical efficiency in the Italian performing arts companies. Small Business Economics, https://doi.org/10.1007/s11187-017-9931-1
Charoenrat T, Harvie C (2014) The efficiency of SMEs in Thai manufacturing: a stochastic frontier analysis. Econ Model 43:372–393
Daraio C, Simar L (2007) Advanced robust and nonparametric methods in efficiency analysis: methodology and applications. Springer, New York
Daraio C, Simar L (2007) Conditional nonparametric frontier models for convex and nonconvex technologies: a unifying approach. J Product Anal 28(1-2):13–32
Daraio C, Simar L, Wilson PW (2018). Central limit theorems for conditional efficiency measures and tests of the ‘separability’ condition in non-parametric, two-stage models of production. Econometrics Journal, https://doi.org/10.1111/ectj.12103
De Bonis R, Ferri G, Rotondi Z (2015) Do bank-firm relationships influence firm internationalization. Int Econ 142:60–80
Ding S, Guariglia A, Harris R (2016) The determinants of productivity in Chinese large and medium-sized industrial firms, 1998–2007. J Product Anal 45:131–155
Djankov S, Murrell P (2002) Enterprise restructuring in transition: a quantitative survey. J Econ Lit 40(3):739–792
Farsi M, Filippini M, Kuenzle M (2005) Unobserved heterogeneity in stochastic cost frontier models: an application to Swiss nursing homes. Appl Econ 37(18):2127–2141
Fisman R (2001) Trade credit and productive efficiency in developing countries. World Dev 29(2):311–321
Foglia A, Laviola S, Reedtz M (1998) Multiple banking relationships and the fragility of corporate borrowers. J Bank Financ 22:1441–1456
Gambini A, Zazzaro A (2013) Long-lasting bank relationships and growth of firms. Small Bus Econ 40:977–1007
Geroski PA (1995) What do we know about entry? Int J Ind Organ 13:421–440
Giannetti C (2012) Relationship lending and firm innovativeness. J Empir Financ 19:762–781
Greene W (2005) Reconsidering heterogeneity in panel data estimators of the stochastic frontier model. J Econom 126(2):269–303
Greene WH (2008). The econometric approach to efficiency analysis. In: Fried HO, Lovell CAK, Schmidt SS (eds), The measurement of productive efficiency and productivity change. Oxford UniversityPress, 92–250
Guiso L, Sapienza P, Zingales L (2004) Does local financial development matter? Q J Econ 119:929–969
Houston JF, Lin C, Lin P, Ma Y (2010) Creditor rights, information sharing, and bank risk taking. J Financ Econ 96:485–512
Jensen MC, Meckling W (1976) Theory of the firm: managerial behavior, agency costs, and capital structure. J Financ Econ 3:305–360
Jin JY, Kanagaretnam K, Lobo GJ, Mathieu R (2013) Impact of FDICIA internal controls on bank risk taking. J Bank Financ 37:614–624
Johnson AL, Kuosmanen T (2012) One-stage and two-stage DEA estimation of the effects of contextual variables. Eur J Oper Res 220:559–570
Karakaplan MU, Kutlu L (2017) Handling endogeneity in stochastic frontier analysis. Econ Bull 37(2):889–901
Kittelsen SAC, Persson BA, Anthun KS, Goude F, Hope Ø, Häkkinen U, Kalseth B, Kilsmark J, Medin E, Rehnberg C, Rättö H (2015) Decomposing the productivity differences between hospitals in the Nordic countries. J Product Anal 43(3):281–293
Kneip A, Simar L, Wilson PW (2016) Testing hypotheses in nonparametric models of production. J Bus Econ Stat 34(3):435–456
Krishnan K, Nandy D, Puri M (2015) Does financing spur small business productivity? Evidence from a natural experiment. Rev Financ Stud 28:1768–1809
Kumbhakar SC, Lien G, Hardaker JB (2014) Technical efficiency in competing panel data models: a study of Norwegian grain farming. J Product Anal 41:321–337
Laeven L, Levine R (2009) Bank governance, regulation and risk taking. J Financ Econ 93:259–275
Latruffe L, Davidov S, Balcombe K (2008) Application of a double bootstrap to investigation of determinants of technical efficiency of farms in Central Europe. J Product Anal 29:183–191
Margaritis D, Psillaki M (2007) Capital structure and firm efficiency. J Bus Financ Account 34(9/10):1447–1469
McMahon RGP (2004) Equity agency costs amongst manufacturing SMEs. Small Bus Econ 22:121–140
Milana C, Nascia L, Zeli A (2013) Decomposing multifactor productivity in Italy from 1998 to 2004: evidence from large firms and SMEs using DEA. J Product Anal 40:99–109
Montoriol Garriga J (2006). The effect of relationship lending on firm performance. Universitat Autonoma deBarcelona Documents de Treball 5
Mundlak Y (1978) On the pooling of time-series and cross section data. Econometrica 46(1):69–85
Nickell S, Nicolitsas D (1999) How does financial pressure affect firms. Eur Econ Rev 43:1435–1456
Rajan RG (1992) Insiders and outsiders: the choice between informed and arm’s-length debt. J Financ 47(4):1367–1399
Schmidt P, Wan H (2002) One-step and two-step estimation of the effects of exogenous variables on technical efficiency levels. J Product Anal 18:129–144
Simar L (2003) Detecting outliers in frontier models: a simple approach. J Product Anal 20:391–424
Simar L, Wilson PW (2002) Nonparametric tests of return to scale. Eur J Oper Res 139:115–132
Simar L, Wilson PW (2007) Estimation and inference in two-stage, semi-parametric models of production processes. J Econom 136:31–64
Simar L, Wilson PW (2011) Two-stage DEA: caveat emptor. J Product Anal 36(2):205–218
Su D, He X (2012) Ownership structure, corporate governance and productive efficiency in China. J Product Anal 38:303–318
Taymaz E (2005) Are small firms really less productive? Small Bus Econ 25:429–445
Tran TB, Grafton RQ, Kompas T (2008) Firm efficiency in a transitional economy: evidence from Vietnam. Asian Econ J 22(1):47–66
Voordeckers W, Steijvers T (2006) Business collateral and personal commitments in SME lending. J Bank Financ 30:3067–3086
Weill L (2008) Leverage and corporate performance: does institutional environment matter? Small Bus Econ 30:251–265
Williamson OE (1967) Hierarchical control and optimum firm size. J Political Econ 75:123–138
Yang CH, Chen KH (2009) Are small firms less efficient? Small Bus Econ 32:375–395
Yildirim A (2017). The effect of relationship banking on firm efficiency. Available at SSRN: https://ssrn.com/abstract=2849466
Acknowledgements
We are very grateful to three anonymous reviewers, and the Editor for their precious comments and suggestions. We also wish to thank the participants to the RSA-SIE 2017 conference (Università della Calabria, Rende), where an earlier version of the paper was presented. Any remaining errors are solely our responsibility.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Rights and permissions
About this article
Cite this article
Agostino, M., Ruberto, S. & Trivieri, F. Lasting lending relationships and technical efficiency. Evidence on European SMEs. J Prod Anal 50, 25–40 (2018). https://doi.org/10.1007/s11123-018-0532-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11123-018-0532-z