(PDF) Ownership, diversification and cost advantages: Evidence from the Italian leasing industry - DOKUMEN.TIPS (2024)

Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Contents lists available at SciVerse ScienceDirect

Journal of International FinancialMarkets, Institutions & Money

journal homepage: www.elsevier.com/locate/ intf in

Ownership, diversification and cost advantages:Evidence from the Italian leasing industry

Marta Degl’Innocenti a,∗, Claudia Girardoneb,1

a Department of Management, University of Bologna, Via Capo di Lucca, 34, 40126 Bologna, Italyb Essex Business School, University of Essex, Wivenhoe Park, Colchester, CO4 3SQ, United Kingdom

a r t i c l e i n f o

Article history:Received 20 September 2011Received in revised form 23 February 2012Accepted 10 May 2012

Available online 19 May 2012

JEL classification:G23D24

Keywords:EfficiencyLeasing industryRandom parameter frontier modelDiversification strategiesOwnership structures

a b s t r a c t

Using Greene’s (2005) random parameter cost frontier model anda unique hand-collected dataset, this paper provides novel evi-dence on the cost advantages of Italian leasing companies over2002–2008, focusing on ownership structures and diversificationstrategies. Results suggest that cost efficiency and economies ofscale have decreased significantly over the period analyzed. Bank-related, independent and domestic leasing companies are moreable to control costs than their captive and foreign counterparts.Diversification strategies can be crucial in determining the costeffectiveness of leasing firms. Nonetheless, smaller, independentand less diversified leasing firms appear to benefit from highereconomies of scale and greater technological advancements.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Over the last few decades the asset-based lending sector has increasingly become a key financialresource in most developed countries, particularly for small and medium businesses. In the US, priorto the 2007–2009 crisis, about a third of the capital equipment used by companies was leased (Chem-manur et al., 2010). In Europe, the leasing penetration rate (measured as the amount of overall newbusinesses divided by investment defined as gross fixed capital formation) rose by nearly 90% over

∗ Corresponding author. Tel.: +39 051 2098085; fax: +39 051 6390612.E-mail address: [emailprotected] (M. Degl’Innocenti).

1 Tel.: +44 1206 874156; fax +44 1206 873429.

1042-4431/$ – see front matter © 2012 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.intfin.2012.05.002

880 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Trend i n Yearly Contract Vol ume s (E UR mil) in the Italian Lea sing Market

(2002-2009)

10,000

20,000

30,000

40,000

50,000

60,000

20092008200720062005200420032002

New

production

Fig. 1. Trend in Yearly Contract Volumes (EUR mil) in the Italian Leasing Market (2002–2009). Source: Assilea, Annual Report,various years. The new production refers to the annual volume of leasing contracts.

the period 2000–2007 (Leaseurope-KPMG, 2008; Leaseurope, 2009) and then dropped significantlyduring the financial turmoil.2

However, the decline in the leasing business as a result of the crisis was partially alleviated by thefact that bank borrowing has generally become harder to obtain. As a result many small and mediumbusinesses resorted to leasing and other forms of specialised credit as alternatives to banking products.Berger and Udell (2006) observe that leasing contracts can be used to provide funds to transparentand opaque firms because the underwriting decision is primarily based on the value of the asset beingleased, a process that helps overcome potential adverse selection problems. Aside from fiscal benefits,another related advantage associated with leasing contracts is that the financial lease includes thepossibility to finance the use of the asset’s economic life with the option to purchase the asset at aprice that is likely to be lower than the fair value at the time the option is exercisable.

Compared to banking, asset-based lending involves specific technologies and processes. Thesefinancial firms specialized in leasing do not collect the funds to finance their business through deposits.Therefore, controlling the production and administrative costs in an efficient way represents a keystrategic driver for ensuring their competitive viability. In this respect, this paper offers novel evi-dence on the way leasing firms manage their own internal resources (financial, human, technology,etc.) and marketing structure (distribution channels).

Specifically, this paper provides a detailed analysis of the cost advantages of leasing companiesoperating in Italy over the period 2002–2008. There are several reasons for the focus on a single coun-try. First, the Italian leasing sector represents the third largest market in Europe with nearly 13% of newproduction, and the second largest market in terms of portfolio of leased assets (outstandings) withapproximately 19% of the entire European market share (Leaseurope, 2010a). Second, nearly one-fifthof the top 50 leasing companies in Europe in terms of new businesses are Italian (Leaseurope, 2010b).Third, Italy’s yearly contract volumes grew by 50% in the first half of the 2000s (Fig. 1) and contributedsignificantly to the growth of the European market.3 Although the financial crisis interrupted thistrend of rapid growth, the importance of the Italian leasing sector in Europe is still high.

For this research we hand-collected a unique dataset covering approximately 90% of companiesoffering leasing services that were members of Assilea, the Italian Leasing Association, during theperiod 2002–2008. The methodology used for the cost efficiency estimation is the random parameterstochastic frontier model (Greene, 2005) which allows us to take into account the latent heterogeneityamong specialized financial intermediaries. In order to calculate the degree of scale economies and

2 A recent survey carried out by the European Central Bank reports that the use of leasing, hire purchase and factoring coveredalmost 34% of the total of external sources of funding for European Small and Medium Enterprises in 2010 (ECB, 2011).

3 The Italian leasing sector benefited from fiscal advantages (the deductibility of lease payments) and from the effects ofthe 1994-95 and 2001-02 so-called “Tremonti Laws” designed to give investment incentives to industrial firms. The ‘TremontiLaws’ were named after the then Finance Minister Giulio Tremonti. See, for more details, Rapporto Assilea (various years).

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 881

technical change with their confidence intervals, we employed the Delta method. In trying to explainthe cost differences in the leasing sector, this study focuses on the role of ownership structure anddiversification. In Italy – as in most European countries – it is common to distinguish between threetypes of specialised leasing companies, namely: bank-related, captive and independent (Leaseurope-KPMG, 2008).4 Since the Italian leasing industry has recently opened up to foreign competition, wealso distinguish between companies that are domestic and foreign-owned.5

On the other hand, diversification strategies are also expected to affect costs and business in generalbecause, for example, many specialised financial intermediaries have strategically chosen to offercollateral services, such as factoring and consumer credit, in addition to the leasing activities. Othersspecialised in one single leasing asset instead of diversifying the portfolio of leased assets. To take thesedifferences into account, we cluster the leasing firms in our sample using two alternative diversificationindexes: the entropy index and the loan diversification index.

Further, we investigate the cost advantages of the leasing firms’ distribution channels follow-ing the Italian Leasing Association’s classifications. Specifically, we distinguish between: ‘suppliers’,‘banking’, ‘direct’, and “mixed” sales channels. In Italy most leasing firms (over 27%) use suppliersas their main sales channels, followed by banking distribution channels (23.4%) and direct channels(19.5%). The remaining 30% uses at least two of the above channels and thus this cluster is typi-cally labelled as ‘mixed’6 (Assilea, 2008). Finally, we consider the leasing companies’ relative size inorder to identify the most cost effective scale of production in this industry. To achieve this aim, thisstudy examines how and to what extent technical changes can affect leasing companies’ economiesof scale and thus may favour the consolidation process. Such investigation has direct policy impli-cations given the considerable growth in M&A activities that have characterized the Italian leasingsector over the last fifteen years or so. Therefore, this study is expected to offer useful insights onhow the growth of the sector, weakened by the recent crisis, may be sustained over the next fewyears.

Results show that independent and bank-related companies are more cost efficient than captives.Regarding firm characteristics, diversification strategies and distribution channels seem to play animportant role in the achievement of cost advantages. Furthermore, the rate of technical changehas decreased the minimum efficient size, despite the industry’s trend of consolidation that mainlyinvolved the bank-related companies during the years 2002–2008. As explained in the next section,there are no studies, as far as we are aware, on the efficiency of specialised financial intermedi-aries such as leasing firms. In addition, a consensus on the implications of diversification strategiesis still lacking. This paper reveals that leasing companies can benefit from leveraging managerialskills and abilities across products for the enhancement of cost efficiency, and across geographicregions for improving the degree of technological progress. Our evidence also suggests that, in linewith the mainstream literature (e.g. Hunter and Timme, 1986), the benefits of technological progressonly exist up to a certain company’s size, after which the degree of economies of scale appearsto decrease.

The paper is structured as follows. Section 2 presents a brief review of the literature. Section 3describes the methodology and data. Section 4 presents and discusses the main findings and Section5 provides some final considerations.

2. Literature review and main research hypotheses

A wide range of literature exists that focuses on the technical and/or cost efficiencies of bankinginstitutions as well as cost savings through scale and scope economies both in the US and in Europe

4 A bank-related leasing company either has a banking license of its own (as in the case of Agrileasing Bank) or it is part of abanking group (e.g. UBI Leasing). A captive is a subsidiary leasing arm of a manufacturing group (e.g. BMW Financial ServicesItalia); while an independent leasing company offers leases directly to consumers and businesses and is generally not affiliatedwith a particular manufacturer (e.g. GE Capital Servizi Finanziari).

5 The market share of foreign companies grew rapidly over the period analysed (from 21.2% to 26.9% on total leasing volumefor the years 2007–2008). In 2009, the foreign shareholders contributed to almost 30% of the sector’s new production.

6 It also includes brokers and other financial operators.

882 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

(for extensive reviews see, for example, Berger and Humphrey, 1997; Goddard et al., 2001; Berger,2007; Hughes and Mester, 2009). Despite the recent spurt in M&A activities and growth in volumes,the empirical literature on specialized financial intermediaries is still in its infancy, mainly owing tolimited data availability and disclosure. In this section we review the relevant prior literature anddevelop our hypotheses.

To the best of our knowledge no recent work has been published specifically on the efficiency of theleasing industry. However, two recent studies (Fiordelisi and Molyneux, 2004; Fiordelisi and Monferrà,2009) focus on the performance of the asset-based lending sector (mainly factoring). Fiordelisi andMolyneux (2004) investigate the efficiency and productivity of the Italian factoring industry between1993 and 1997 using a non-parametric frontier method (Data Envelopment Analysis) and the TotalFactor Productivity (TFP) Malmquist Index. Their results suggest that there are relatively high costinefficiency levels in the sector, ranging between 14% and 22% and, contrary to ownership structure,firm size does not seem to explain differences in efficiency across firms in the industry. The TFP showsa slightly positive productivity change over 1993–1996, followed by a substantial increase in thefollowing two years mainly driven by technological improvements and scale efficiencies. In a morerecent study, Fiordelisi and Monferrà (2009) analyze the shareholder value created by a sample of 13leasing companies and 14 factoring companies operating in Italy over the period 1994–2004. Theirfindings show that both leasing and factoring firms are characterised by a relatively low weightedaverage cost of capital, coherent with their low risk profile. Moreover, the authors find that leasingcompanies in particular display high profitability and Economic Value Added (EVA) levels over theperiod analyzed. For both leasing and factoring companies the highest profitability and shareholdervalue levels were achieved in 2004.

The current paper provides an in-depth analysis of the cost advantages of the leasing sector inItaly in the 2000s and focuses on various aspects that are expected to significantly impact on thecost structure of leasing firms, particularly on ownership and diversification. The reference litera-ture in this context is represented by the many published banking studies in the area. A numberof recent studies have concentrated on the differences between public (state-owned) and privateinstitutions (see, for comprehensive reviews, Goddard et al., 2008, and Wilson et al., 2010). Oneparticular stream of research has studied whether bank affiliation is a source of cost advantages(e.g. Chen and Chen, 1998 and, more recently, Bos and Kolari, 2005). Bank affiliation is particu-larly relevant for the leasing industry, as bank-related financial intermediaries in Italy cover thelargest market share in terms of contract volumes. It seems reasonable to expect that their strongcompetitive position contributes to greater cost efficiency and technological advancements thantheir captive and independent competitors. Therefore our first research hypothesis H1.A statesthat:

H1.A. Bank-related leasing companies benefit from higher cost advantages than independent andcaptive leasing companies.

The period studied was characterised by dramatic changes such as increased prudential regulationin the wake of the Basel Accords that affected banking groups, the introduction of new internationalaccounting criteria and new, tougher, tax laws. In addition, greater competitive pressure resulted,on one hand, in an active consolidation process particularly in terms of M&A activities; and, on theother hand, in new financial intermediaries entering the market–both domestic and foreign. In the-ory, foreign institutions can benefit from compensating advantages that enable them to competewith domestic ones on equal terms despite their initial informational disadvantage. These includetechnological and managerial expertise, marketing know-how, and innovative product capabilities.In addition, the strengthening of the position of foreign financial intermediaries in local markets canprompt efficiency improvements, since domestic firms “can assimilate any superior banking tech-niques and practices of foreign entrants” (Claessens et al., 2001: 906). Thus our second hypothesisH2.A can be formulated as follows:

H2.A. Domestic leasing companies benefit from higher cost advantages than foreign ones.

With respect to the achievement of cost advantages and competitive viability, diversification strate-gies can play a crucial role. Despite the large number of studies, there appears to be no consensus on

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 883

the effects of diversification on banks’ economic profile (see the recent analysis by Berger et al., 2010).Some studies, however, have found evidence that financial institutions should focus on a single line ofbusiness to benefit from management expertise for different reasons: to reduce agency costs (Acharyaet al., 2006; Laeven and Levine, 2007); to reduce the sources of risk (Fauver et al., 2004; Deng andElyasiani, 2008); or to maintain profitability (Stiroh, 2004).

Another stream of literature has shown that diversified financial institutions can acquire man-agerial skills and abilities through diversification across products and geographic regions (e.g.Iskandar-Datta and McLaughlin, 2007; Drucker and Puri, 2009). However, the diversification motivescan also be associated with various other factors including: higher debt capacity (Lewellen, 1971);reduction of the adverse selection problems (Hadlock et al., 2001); benefits through economies ofscale and scope and decrease in idiosyncratic risk (Santomero and Eckles, 2000); and a greater abilityto internalize market failures (Khanna and Palepu, 2000).7 Drawing from these arguments, the thirdhypothesis H3.A, along with the competing hypothesis H3.B, can be stated as:

H3.A (H3.B). Diversification strategies are associated with higher (lower) cost advantages.

In this paper we also compute the level of economies of scale for the leasing firms in our sample,to check whether there are any advantages deriving from size. This is particularly important giventhe numerous M&As which have characterized the Italian leasing sector over the studied period. Suchchanges may have affected the technology rate and the level of cost efficiency of the leasing production,with possible consequences for the long-term growth of the industry. The empirical banking literature,however, does not share the same opinion on the existence of cost advantages for a specific bank size.Some studies found empirical support for the existence of economies of scale for large banks (Tadesse,2006). Others found economies of scale for all bank sizes (Altunbas and Molyneux, 1996; Cavallo andRossi, 2001). In yet other studies, economies of scale are exhausted for larger banks (e.g. Girardoneet al., 2004). Thus our fourth hypothesis H4.A, along with the competing hypothesis H4.B can be statedas:

H4.A (H4.B). The benefits derived from economies of scale increase (decrease) along with leasingcompanies’ size.

Hunter and Timme (1986) argue that the existence or the degree of economies of scale can beaffected by the characteristics of the technical change in the industry. Especially the non-neutraltechnical change is expected to affect the efficient size of the firm.8 This means that increasing the scaleof technical progress entails a similar enlargement of the minimum efficient size at which the averagecost is at the minimum level (see, for a survey on the banking sector, Hughes and Mester, 2009). Thisevidence has important policy implications since it suggests that regulation should favour the increaseof a firm’s relative size in the industry. In the leasing sector, economies of scale and technical change canprovide incentives for industry consolidation since the achievement of cost savings through M&As canmake it more difficult for smaller financial intermediaries to compete with their larger counterparts.It follows that the fifth and final hypothesis we verify in this study is:

H5.A. Technological advancement is associated with greater economies of scale.

The next section details the main methodological issues and the data sources used for the empiricalanalysis.

7 Aside from diversification strategies, the development of marketing strategies can also provide unique intangible resourceswhich contribute to the organizational success of a firm through a more efficient interaction with the environment (McDanieland Kolari, 1987).

8 Non-neutrality in technology can be driven by input biases or biases with respect to the scale characteristics of the produc-tion technology. The latter implies a modification of the range of outputs over which a given degree of scale economies can beachieved (e.g. Tadesse, 2006).

884 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

3. Methodology and data

3.1. The stochastic cost frontier

The stochastic frontier was first introduced by Aigner et al. (1977), Meeusen and Van den Broeck(1977), Battese and Corra (1977) and recently extended by Kumbhakar and Lovell (2000). There havebeen several models for stochastic frontiers developed for panel data settings. These may be broadlydivided in time-varying inefficiency models and time invariant inefficiency models.9 In this paper,we employ a time-variant random parameters stochastic frontier model which embodies both therandom and fixed effect models’ characteristics (Greene, 2005). The random coefficient model hasthe advantage that it allows us to take into account the inter-individual heterogeneity through thevariations in the parameters across sectional units, without assuming any trend in the cost efficiencyover time. Therefore, we consider Greene’s model as more ideally suited for our research aims com-pared with other models because, by assuming a time-variant inefficiency component, it takes betteraccount of the impact of regulatory (fiscal and prudential), structural changes (e.g. M&As) and theinter-company heterogeneity in the Italian leasing industry over the studied period. However, themodel presupposes that the variance of the noise term is constant and free of heteroscedasticity.

The functional form used in this study is a translog cost function with three inputs and two outputsthat can be written as follows:

Ln (TCit) = ˇ0 +∑2

m=1ˇm ln Yimt +

∑2

j=1ˇj ln Wijt

+ 12

(∑2

k=1

∑2

j=1ˇkj ln Wikt ln Wijt +

∑2

m=1

∑2

n=1ˇmn ln Yimt ln Yint

)

+∑2

j=1

∑2

m=1ˇjm ln Yimt ln Wijt + t1Tit + 1

2t2T2

it +∑2

j=1ˇjT ln WijtTit

+∑2

m=1ˇmT ln YimtTit + ϕIASit + uit + vit for k /= j and m /= n (1)

where TC is total costs; Y1 is total lease receivables; Y2 is other receivables and loans; P1 is the ratioof personnel expenses to number of employees; P2 is the ratio of other administrative expenses totangible assets for own use; and P3 is the ratio of interest expenses to financial debts. In order toinclude the financial intermediaries specialized only in leasing activities, we add 1 to the calculationof natural logarithm of Y1 and Y2. The inputs prices and the total costs are normalized by W3, the priceof financial funds. To impose linear hom*ogeneity in input prices, the variables were normalized in(1) with respect to P3 so that: TCit = ln(TCit/P3t), Wjt = ln(Pjt/P3t), j = 1, 2. Symmetry restrictions werealso imposed on the translog coefficients. The trend T is included in equation (1) to measure technicalchange. Since the accounting representations of leasing operations may follow two alternative criteria(the local GAAP and the International Accounting Standard) we control for these differences by addinga dummy IAS directly in the cost function. The inefficiency term, uit must be positive, uit =

∣∣Uit

∣∣ where

Uit∼N[0, �2

u

], and can be interpreted as the percentage deviation of observed best performance from

a firm’s own frontier performance. The random error term (vit) accounts for random external factors,such as luck or unexpected disturbances, measurement errors and other factors unspecified in thecost function and is assumed to be an iid normal random variable: vit∼N[0, �2

v ].The cost efficiency estimates are obtained by calculating CEi = [exp(-ui)] − 1. This measure takes on

a value between 0 and 1 where 1 (or 100%) refers to the highest attainable cost efficiency level andthe distance from it denotes cost inefficiency. The parameter �, that is calculated as the ratio between�u and �v, measures the amount of variation of inefficiency as relative to noise for the whole sample.

9 The second category is used to capture heterogeneity in panel data through a time invariant, firm-specific constant term.For example, Schmidt and Sickles’ (1984) fixed effects formulation or Pitt and Lee’s (1981) random effects model treat theinefficiency term as time invariant.

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 885

The parameters of ln Y1 and ln Y2 are randomly distributed with constant means. Changes in regula-tory burden should be hom*ogeneously reflected in the financial reports of leasing firms. However, thedifferences in the distribution structure and external costs associated with the leasing contracts mayinstead heterogeneously alter the variable total costs. This form of heterogeneity is captured throughthe random coefficient of ln Y1. By doing so, we control for latent heterogeneity. Moreover, since ln Y2refers to multiple financial contracts, such as factoring, consumer credit, and other, the assumptionof its hom*ogeneous elasticity with TC appears inappropriate. These arguments give further supportto our choice of Greene’s (2005) model over other models. In the model we impose an unrestrictedcovariance 2 × 2 matrix Var [ln Y1, ln Y2] = � � ′, using the Cholesky factorisation. We choose 250 Hal-ton draws as a number of random replications. On this point, it is worth noting that several authors(e.g., Bath, 2001; Train, 2009) demonstrated that when using a Halton sequence with a relatively smallnumber of replications, it is often the most effective method.

3.2. Economies of scale and technical change

Economies of scale (ES) occur when a firm is able to reduce costs per unit of output as it gets bigger.An increase in ES implies a decline in the (long-run) average cost curve as a consequence of a marginalreduction in costs as the output increases; all other factors (such as input prices) being constant. ES ismeasured by the following cost elasticity, differentiating from the translog functional form in equation(1):

ES = 1(∑2m=1∂ ln Ct (W, Y)

)/∂ ln Ym

(2)

ES > 1 implies economies of scale, while ES < 1 diseconomies of scale. Furthermore we carry out ananalysis of the technical change (TCC) that allows us to estimate whether a financial firm experiencesa reduction of the costs over time. Given the cost function in Eq. (1), the rate of technical change canbe measured as:

−(

∂ ln Ct (W, Y)∂T

)= −

(t1 + t2T +

∑2

j=1ˇjT ln Wijt +

∑2

m=1ˇmT ln Yimt

)(3)

If TCC > 0, then for a given level of input prices, the firm can produce a greater output at lower costsover time. The technical change can be biased with respect both to inputs (IBTC) and the scale charac-teristics of the production function (OBTC). As concerns IBTC, holding the output fixed, the technicalchange is Hicks-neutral if the slope of production isoquant is independent of technical change. In thecase of cost-neutral technical change, the Hicks-neutral concept is related to technical changes thatdo not modify the cost shares of inputs. The input-biased technical change is calculated by:

IBTC = ∂2 ln Ct(W, Y)∂ ln Wi∂T

(4)

IBTC = 0 indicates a neutral technical change, while IBTC > 0 and IBTC < 0 implies ith factor-usingand ith factor-saving technical advancements, respectively.

We also measure the response of scale economies changes in technology that is given by:

OBTC = ∂(1/(ES))∂T

(5)

If OBTC is <0, then changes in technology increase scale economies, whilst in the case of OBTC > 0scale economies are decreased. Finally t1+t2T refers to the pure technical change (PTCC). It accountsfor the reductions in the totality of costs attainable with a constant efficient scale of production forany specific mix of outputs, and with constant shares of each of the inputs in total cost (Altunbas et al.,1999). In order to verify the significance of OBTC and IBTC and pure technical change we employ theWald test.

Both the economies of scale and the technical changes are non-linear functions of the estimatedparameters, and therefore the exact standard errors cannot be calculated. The standard errors of these

886 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

statistics are computed though the Delta Method as the standard deviations of the first order Taylorseries approximations of the respective statistics around the parameter estimates.10

3.3. Diversification indexes

In order to assess the level of diversification of the leasing firms, two different indexes are cal-culated: a modified entropy index and a loan diversification index. EI is a measure of corporatediversification that takes into account the diversification level both within and across industry groups.In this study EI can be written as follows:

EI =∑M

j=1V∗

j ln

(1

V∗j

)(6)

where V∗j

is the proportion of the jth volume, Vj in a specific category’s asset for the proportion of thecontracts’ numbers, Nj, in the relative category:

V∗j =

(Vj

(Vj +∑n

i=1Vi)

)∗(

Nj

(Nj +∑n

i=1Ni)

)for i /= j (7)

We consider four categories of leasing assets: real estate, equipment, automotive and other assets(ship and aircraft). If EI equals 0 the firm is specialised: the higher EI, the higher the level of firmdiversification. Additionally, we cluster the companies according to whether the index assumes avalue equal to 0 (specialized), a value lower than the mean value of the numbers but higher than 0(low diversification), or one greater than that same mean value (highly diversified portfolio of assets).

Since the management of financial portfolios is often a marginal activity for leasing firms, we alsoemploy a modified version of the Loan Diversification Index (LDI) proposed by Leaven and Levine(2007). It focuses specifically on lease receivables rather than on a bank’s earning assets. It can bewritten as follows:

LDI = (1 − |(lease receivables − other receivables and loans)/net loans)| (8)

In equation (8) the item ‘other receivables and loans’ refers to all those activities that do not fallunder leasing assets. The sum of lease receivables and other receivables and loans is equal to net loans(i.e. net of amortisation or depreciation). LDI can take on any value from 0 to 1. We classify the firmsas poorly diversified, if LDI assumes a value in the range of 0–33%; highly diversified if LDI is above66%. In addition, a firm is considered to have a medium level of diversification if the index falls inthe range of 33%–66%. We also distinguish across companies based on the geographical location ofleasing contract draws. In particular, following the scheme proposed by Assilea, we classify as highlydiversified those companies that have at least 70% of the yearly contract volumes realized in morethan two regions. The rest of the companies are considered to be geographically diversified on a lowlevel.

As concerns firms’ size, we follow Tadesse (2006) and identify three main groups: “Large” “Medium”and “Small”. “Large” companies are those with a value of the inputs (Wj) and outputs (Ym) greater thanthe relative average values. We classify as “Medium” those companies whose inputs and outputs arehigher than the minimum value of the companies classified as “Medium” by Assilea (i.e. ranked 12–30in terms of yearly contract volumes). The remaining firms are classified as “Small”.

We verify the significance of the differences in mean efficiency levels across the clusters describedabove (ownership structures, domestic versus foreign-owned firms, size, diversification strategies,distribution channels and geographical diversification) using a standard mean comparison two-tail

10 Specifically, given a vector of statistics (G) that are functions of the vector of estimated parameters, the approximatevariance-covariance matrix of the statistics is calculated by dGVdGt where dG is the matrix of the gradients of G with respectto the parameters evaluated for each observation and V, is the variance-covariance matrix. In this paper the standard errorsare calculated for each observation. On this point we are grateful to Professor William Greene of Stern School of Business (NewYork University) for helpful comments and practical suggestions.

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 887

Table 1Sample description by ownership.

Ownership type Number of leasing firms Observations over 2002–2008 Average asset size*

Number Percentage

Bank-Related 45 293 72.2 2,722,709Captive 10 55 13.5 1,386,672Independents 9 58 14.3 312,174Foreign 19 121 29.8 1,200,642Domestic 45 285 70.2 2,620,524

Total 64 406 100 2,197,357

* Note: Values are in thousands of euros. The average size is calculated in terms of total assets for each category.

test (t-test).11 As concerns the distribution channels, they can produce cost differences among financialintermediaries. However, we do not consider them to be relevant in determining variations in termsof economies of scale and technical change.12

3.4. Data sources and input–output approach

The dataset used for the empirical investigation is hand-collected and spans the period 2002–2008.It is drawn from three different sources: Assilea, OSSFIN13 and the financial reports available in thedataset of the Italian Chamber of Commerce. The final dataset is made up of 64 financial firms mainlyspecialized in financial lease and under the supervision of the Bank of Italy over seven years. The samplecoverage is approximately 90% of the yearly volume’s contracts provided in 2008 by the leasing firmsassociated to Assilea. Table 1 reports the details of the sample by ownership type: bank-related com-panies, independents and captives. It shows that the majority of firms are bank-related and domesticand that their average size is relatively high compared to their counterparts.

We exclude from our sample the multi-product banks for whom leasing is a marginal activity aswell as foreign firms operating in Italy through subsidiaries.14 The panel data is unbalanced to accountfor the many M&A operations and cases of new firms entering the market that occurred in the periodunder study.

In terms of description of the production process of the leasing firms, among the alternativeapproaches proposed in the literature, the asset approach is chosen for this study. It posits that theoutputs are strictly defined by assets and mainly by the production of loans, in which some firms canhave competitive advantages over others (Favero and Papi, 1995). This approach seems to be appro-priate for non-bank financial institutions since it does not take into account most of the products andservices provided by banks. The descriptive statistics of the variables used in this study are reportedin Table 2.15

11 The ownership structure and diversification strategies variables were not included directly in the distribution of cost effi-ciency, in order to provide a better comparative analysis of the impact of these two variables on cost efficiency, economies ofscale and technical change. The inclusion of these variables in the distribution of the inefficiency term would make any directcomparisons with other sources of cost advantages more difficult, especially where non-linear functions are concerned, as inthe case of economies of scale and technical change, and the heterogeneity regards the heteroscedasticity in ui (inefficiency) orvi (noise) and it is not limited to the mean of the inefficiency term.

12 A specific distribution channel can potentially contribute to a technical change in the cases in which, for example, a leasingcompany is able to reduce the costs over time for the effect of an increase in knowledge of the supplier’s reliability and asset’squality. The information that we have on distribution channels is substantially invariant over time and not detailed enough toinfer an effect on the technical change.

13 OSSFIN is the “Osservatorio sugli intermediari finanziari non bancari” that belongs to the Scuola di Direzione Aziendale(SDA) of Bocconi University (Milan).

14 For these four cases the accounting data relative to the unconsolidated Italian subsidiaries are not available with theexception of Toyota Financial Services (UK) PLC for four years.

15 To calculate the input prices for the empirical analysis we used balance sheet items that in some cases were classified differ-ently using Local GAAP and IAS. For the purposes of this study, we have made every effort to ensure year-on-year consistency.In the case of the Local GAAP, activities in leasing include the expired credits plus the fixed assets given in lease (Legislative

888 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Table 2Descriptive statistics of total costs, input and outputs.

Variables Mean Std Dev Median Min Max

TC (+) 83,483.51 136,134 34,675 350 1,110,019Y1 (+) 1,659,883 2,558,916 707,356 7,225 18,318,255Y2 (+) 409,284 1,153,632 27,582 1 9,374,614P1 (+) 64.689 16.789 64.169 17.599 140.644P2 32.358 171.765 7.126 .064 3,011.802P3 .0419 .084 .034 .002 1.687

Note: Total number of observations: 406. (TC) = total costs are calculated as the sum of other administrative expenses, personnelexpenses and interest expenses; (Y1) Total lease receivables; (Y2) Other receivables and loans (includes the assets generatedby factoring, consumer credit and other financial activities). (P1) is the ratio of personnel expenses to number of employees;(P2) is the ratio of the other administrative expenses to tangible assets for own use; and (P3) is the ratio of interest expenses tofinancial debts. (+) The values are expressed in thousands of euros.

Table 3Random parameters stochastic frontier.

Variable Mean p-value Variable Mean p-value Variable Mean p-value

˛ −3.555*** .000 Y2W1 −.021** .002 Mean random parametersW1 1.220*** .000 Y2W2 −.006*** .000 Y1 .541*** .000W2 .197*** .004 T .202 *** .000 Y2 .291 *** .000W1W1 −.014 .466 T2 −.015*** .001 Diagonal elements of Cholesky matrixW2W2 .021*** .000 T Y1 .017*** .000 Y1 .444*** .000W1W2 −.047*** .000 T Y2 −.008*** .000 Y2 .070*** .000Y1Y1 .031*** .000 T W1 −.048*** .000 Below diagonal elements of

Cholesky matrixY2Y2 .017*** .000 T W2 .011*** .000 Y1 Y2 −.277*** .000Y1Y2 −.013*** .000 IAS −.156*** .015Y1W1 −.018** .030 .Y1W2 .007** .033˙ .205*** .000� 1.731*** .000Log L −198.260 p-value < 01***, .05**, .10*

Note: TC (Total Costs) is the dependent variable; Y1 = Total lease receivables; Y2 = Other receivables and loans (including theassets generated by factoring, consumer credit and other financial activities); P1 is the ratio of personnel expenses to number ofemployees; P2 is the ratio of the other administrative expenses to tangible assets for own use; P3 is the ratio of interest expensesto financial debts. T indicates the time trend. IAS is a dummy variable which accounts for changes in the accounting rules (it is1 in the case of IAS and 0 in the case of Local GAAP). The number of Halton draws is 250. The values are expressed in logarithm.

4. Empirical findings

4.1. Cost advantages and technical change in the Italian leasing industry

In this study we use the random parameter frontier model (Greene, 2005) with 250 Halton drawswhere the two outputs Y1 (total lease receivables) and Y2 (other receivables and loans) are used asrandom parameters.16 Table 3 reports the parameter estimates of the cost function (equation (1)) andtheir p-values.

The estimated coefficients support the parametric specification of the monotonicity restrictionssince both inputs and outputs have non-negative shadow prices.17 The input and output coefficients

Decree D. Lgs. No. 87 /1992 and Bank of Italy regulatory requirement No. 216/1996) minus depreciations. Under IAS, a financialintermediary recognizes the asset held under a financial lease in the balance sheet as receivables in an amount equal to thenet investments in the lease. In the case of operating lease the financial intermediary essentially takes the asset risk, which isreflected in the IAS accounting by treating the leased asset as a fixed asset (Leaseurope-KPMG, 2008).

16 In this estimation, the log likelihood of the model with only the first order of outputs random is higher than that of themodel with a higher number of outputs as a random parameter (in this case the log likelihood is equal to −244). Results areavailable with the authors upon request.

17 ı ln TCt (W, Y) /ıW1 is equal to .249 and ı ln TCt (W, Y) /ıW2 has a mean equal to .023.

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 889

Table 4Cost efficiencies estimates by year.

Variables Obs Mean Std. Err. Confidence interval 95% Group Diff

2002 55 .879 .008 .864 .896 Eff (03-02) .0062003 57 .885 .007 .872 .899 Eff (04-03) .017*

2004 57 .903 .008 .888 .918 Eff (05-04) −.0182005 60 .884 .008 .867 .901 Eff (06-05) −.022*

2006 61 .862 .007 .847 .877 Eff (07-06) −.0142007 60 .847 .009 .829 .866 Eff (08-07) −.035*

2008 56 .813 .015 .781 .844

Note: The column “diff” reports the results of the t-test.* p-value < .10.

**p-value < .05.***p-value < .01.

are all significant and positively related to TC. The parameters � (the sum of �u and �v) and � (the ratiobetween �u and �v that measures the degree of variation of inefficiency relative to noise) are equalto .205 and 1.731 respectively and are both significant. This suggests that the inefficiency componentsignificantly contributes to determine the deviations from the frontier. Both coefficients of the randomparameters Y1 and Y2 are significant. As concerns the variable explaining the accounting changes (IAS),it is significantly and negatively related to total cost (it is useful to recall that IAS is equal to 1 in caseof adoption of the IAS, otherwise it is 0).

Average cost inefficiencies for our sample of leasing firms is approximately 14% over the periodanalyzed. These results are broadly in line with several international banking studies (for example,Berger and Humphrey, 1997, Hughes and Mester, 2009) and are also comparable to the study onefficiency in the factoring industry by Fiordelisi and Molyneux (2004). Concerning the efficiency trend,overall it appears to have decreased considerably over the period (from 87.9% to 81.3%). However,as reported in Table 4, some fluctuations occur in the general trend of this increase. In 2004 thecost efficiency of Italian leasing firms improved significantly. This increase occurred despite the dropin contract volumes in 2003 due to the expiration of the so-called “Tremonti Law” that providedinvestment incentives to industrial firms.

The performance in 2006 may have been affected in part by the introduction of new capital require-ments for banks and some specialized financial intermediaries in accordance with the Basel 2 Accord.As expected, the biggest decline in efficiency is reported in 2008 as leasing firms have been affectedby the global financial crisis.

Table 5 presents the estimates of the economies of scale measured using equation (2). Overall, theindustry seems to be characterized by relatively high and statistically significant economies of scaleover the years 2002–2008. The average value of scale economies for the entire sector is equal to 1.178and statistically significant. Looking at the changes over time, there is some evidence of deteriorationafter 2006. Finally, the industry has been characterized by an overall positive but insignificant technicalchange (the average value is .005 with at t-statistic of .80).

4.2. Ownership structure and shareholder type

In support of our first hypothesis H1.A, Table 6 suggests that bank-related and independentleasing companies are significantly more cost efficient than their captive competitors. One possiblereason for this is that bank-related firms can count on the financial support of the parent/holdingcompanies, while the independents can economically benefit from their specialization in a particularmarket or clientele. Indeed this fact can also explain our evidence on the level of scale economies.As shown in Table 7, the firms that are most able to exploit economies of scale are by far theindependents.

Moreover, as shown in Table 8, the findings for the technical change suggest that bank-relatedleasing companies have been characterized by a lower technological advancement than their

890 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Table 5Economies of scale and technical change by year.

Variables Obs Mean Std. Err. Confidence interval 95% Group Diff

Economies of scale2002 55 1.215*** .009 1.194 1.237 Eff(03-02) −.0052003 57 1.209*** .009 1.192 1.228 Eff(04-03) −.0122004 57 1.198*** .009 1.179 1.216 Eff(05-04) −.0062005 60 1.192*** .007 1.176 1.208 Eff(06-05) −.021*

2006 61 1.170*** .008 1.154 1.186 Eff(07-06) −.030***

2007 60 1.139*** .009 1.121 1.158 Eff(08-07) −.018*

2008 56 1.122*** .011 1.101 1.143

Technical change2002 55 −.049*** .013 −.075 −.024 Eff(03-02) .023***

2003 57 −.026*** .009 −.045 −.008 Eff(04-03) .013*

2004 57 −.013* .007 −.027 .001 Eff(05-04) .033***

2005 60 .019*** .006 .007 .032 Eff(06-05) .0062006 61 .025*** .007 .011 .040 Eff(07-06) .0042007 60 .029*** .011 .008 .051 Eff(08-07) .014**

2008 56 .042*** .015 .014 .072

Note: The column “diff” reports the results of the t-test. The Std. Err. and the confidence interval are calculated through theDelta method and are computed by average over sample observations.

* p-value < .10.** p-value < .05.

*** p-value < .01.

counterparts. This result further confirms the idea that specialization in a particular clientele andmarket can favour technological progress.

Fig. 2 illustrates the trend in cost efficiency by ownership types over 2002–2008. Noticeably, theoverall trend is declining over the period and a sharp reduction seems particularly evident for captiveleasing firms. Although the whole industry appears to experience a drop in cost efficiency from 2007(in connection with the crisis), on average after 2005 the independent leasing firms show healthiercost efficiency levels than the bank-related ones.

Apart from the impact of the 2007–2008 global financial turmoil, among the possible reasonsfor these results are the consolidation process and changes in the regulatory burden. During the early2000s, numerous M&As have characterised the industry particularly in the case of bank-related leasing

Cost Effic iencies in the Italia n Lea sing In dust ry by O wnership Struc ture

0.690

0.740

0.790

0.840

0.890

0.940

2008200720062005200420032002

BR

I

C

FO

NZ

Fig. 2. Cost Efficiencies in the Italian Leasing Industry by Ownership Structure Note: BR refers to bank-related compa-nies, C to captive companies, and I to independent companies, FO to foreign-owned companies and NZ to national-ownedcompanies.

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 891

Table 6Cost efficiencies estimates by ownership, size, diversification, and distribution channels.

Variables Cost efficiency

Obs Mean Std. Err. Confidence Interval95%

Group Diff

Ownership structureBank-related leasing firms (BR) 293 .875 .003 .868 .882 Eff(BR-I) .001Independents (I) 58 .874 .007 .861 .887 Eff(BR-C) . 053***

Captives (C) 55 .822 .019 .783 .859 Eff(I-C) .053***

Type of shareholderForeign ownership (FO) 121 .857 .008 .841 .872 Eff(N-FO) .015*

National ownership (N) 285 .873 .004 .864 .881

SizeLarge (L) 110 .864 .006 .849 .878 Eff(M-L) .012Medium (M) 78 .876 .007 .862 .889 Eff(L-S) −.002Small (S) 218 .866 .005 .856 .878 Eff(M-S) .009

Loan diversificationHigh (HD) 36 .862 .016 .829 .896 Eff(HD-MD) .009Medium (MD) 85 .853 .009 .835 .871 Eff(LD-MD) .020**

Low (LD) 285 .873 .004 .865 .881 Eff(LD- HD) .011

Entropy indexHigh diversification (EIH) 242 .876 .003 .869 .883 Eff(EIH-EIL) .011Low diversification (EIL) 84 .866 .007 .851 .879 Eff(EIL-SPE) .020Specialization (SPE) 80 .846 .014 .818 .873 Eff(EIH-SPE) .031***

Geographic diversificationHigh (HGD) 264 .875 .005 .864 .884 Eff(HGD-LGD) .011Low (LGD) 142 .864 .005 .854 .874

Distribution channelsa

Direct (DIR) 83 .878 .005 .868 .888 Eff(DIR-BK) .007Bank-related (BK) 113 .871 .006 .859 .883 Eff(DIR-SU) .049***

Supplier (SU) 68 .828 .016 .797 .860 Eff(BK-SU) .042***

Mixed (MX) 132 .877 .005 .867 .886 Eff(MX-BK) .005Eff(DIR-MX) .001Eff(MX-SU) .048***

Note: The column “diff” reports the results of the t-test. (SPE) refers to specialization in a single leased asset; if the leased assetdiversification index assumes a value higher than 0 but lower than the mean value, the leasing firm is considered to be poorlydiversified (EIL); otherwise the firm is classified as highly diversified (EIH). (LD) refers to low loan diversification (modifiedLeaven and Levine index from 0% to 33%); (MD) refers to medium loan diversification (modified Leaven and Levine index from33% to 66%); (HD) refers to high loan diversification (modified Leaven and Levine index from 66% to 1); (LGD) refers to the highgeographic diversification in the case of 70% of contract volumes drawn in either one or two regions, while (HGD) refers to thehigh geographic diversification in the case of 70% of the contract volumes drawn in more than two regions.

a We assume that the main use of distribution channels in 2003 is the same as in 2002 given the relative stasis of theinformation. We combined the information provided by Assilea with that included in the financial reports.

* p-value < .10.** p-value < .05.

*** p-value < .01.

firms.18 While efficiency and performance are often considered among the main motives for M&Aactivity in financial firms (De Young et al., 2009), there may be some delays before the benefits fromconsolidation start to impact significantly on firms’ performance. This is because after a merger oracquisition, it is not infrequent for jobs to overlap throughout the organization’s various departments

18 Among the most important M&A operations carried out in the Italian leasing industry that involved some large bankinggroups are the following: in 2006 the acquisition of Italeasing by Banca Italease; and in 2008 the merger by incorporation of:San Paolo Leasint into Intesa Leasing; UniCredit Global Leasing into Locat (later known as Unicredit Leasing); Locafit into BNPParibas Lease Group; and BPU Esaleasing into SBS Esaleasing.

892 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Table 7Economies of scale estimates by ownership, size and diversification.

Variables Economies of scale

Obs Mean Std. Err. Confidence interval95%

Group Diff

Ownership structureBank-related leasing firms (BR) 293 1.168 .005 1.149 1.186 Eff(BR-I) −.051***

Independents (I) 58 1.219 .009 1.201 1.238 Eff(C-BR) .018**

Captives (C) 55 1.186 .009 1.167 1.206 Eff(I-C) .033**

Type of shareholderForeign ownership (FO) 121 1.185 .008 1.171 1.201 Eff(N-FO) −.011National ownership (N) 285 1.174 .008 1.158 1.191

SizeLarge (L) 110 1.117 .014 1.090 1.145 Eff(M-L) .038***

Medium (M) 78 1.156 .009 1.138 1.174 Eff(L-S) −.097***

Small (S) 218 1.216 .012 1.192 1.239 Eff(S-M) .059***

Loan diversificationHigh (HD) 36 1.147 .010 1.126 1.168 Eff(LD-HD) .041***

Medium (MD) 85 1.158 .009 1.140 1.177 Eff(LD-MD) .029***

Low (LD) 285 1.188 .010 1.167 1.207 Eff(HD- MD) −.012

Entropy indexHigh diversification (EIH) 242 1.171 .009 1.152 1.189 Eff(EIH-EIL) .000Low diversification (EIL) 84 1.171 .007 1.156 1.184 Eff(EIL-SPE) −.036***

Specialization (SPE) 80 1.207 .010 1.186 1.227 Eff(EIH-SPE) −.036***

Geographic diversificationHigh (HGD) 264 1.164 .008 1.149 1.180 Eff(HGD-LGD) −.037***

Low (LGD) 142 1.202 .011 1.179 1.225

Note: The column “diff” reports the results of test on mean comparison, t-test. (SPE) refers to specialization in a single leasedasset; if the leased asset diversification index assumes a value higher than 0 but lower than the mean value, the leasing firmis considered to be poorly diversified (EIL); otherwise the firm is classified as highly diversified (EIH). (LD) refers to low loandiversification (modified Leaven and Levine index from 0% to 33%); (MD) refers to medium loan diversification (modified Leavenand Levine index from 33% to 66%); (HD) refers to high loan diversification (modified Leaven and Levine index from 66% to 1);(LGD) refers to the high geographic diversification in the case of 70% of contract volumes drawn in either one or two regions,while (HGD) refers to the high geographic diversification in the case of 70% of the contract volumes drawn in more than tworegions. The Std. Err. and the confidence interval are calculated through the Delta method and are computed by average oversample observations. The average mean over each category results to be significant at 1%.*p-value < .10.

** p-value < .05.*** p-value < .01.

and it may take a few years before internal restructuring and reorganisations effectively translate intocost savings.

Corroborating our second hypothesis H2.A, the foreign-owned leasing firms appear to be less costefficient on average than their domestic counterparts. This can also be explained by the fact that foreignownership in the leasing sector is a very recent trend. As discussed in the introductory section, themarket share of foreign leasing firms in Italy increased since 2007. It is to note, however, that foreign-owned companies appear to have a higher and more significant technical change than national-ownedones (Table 8), which can however contribute to the creation of higher cost advantages in the future.

4.3. Diversification strategies and firm size

In this section we verify our third hypothesis H3.A that diversification strategies in the leasingmarket are associated with higher cost advantages, against its competing hypothesis H3.B. Overall,our results are mixed: geographic diversification do not seem to provide any significant advantagesin terms of cost efficiency, thereby rejecting our hypothesis. Conversely, a high entropy index (thatmeasures diversification in terms of leased assets) and a low loan diversification index, appear to

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 893

Table 8Technical change estimates by ownership, size and diversification.

Variables Technical change

Obs Mean Std. Err. Confidenceinterval 95%

Group Diff

Ownership structureBank-related financial intermediaries (BR) 293 −.004 .006 −.017 . 008 Eff (BR-I) −.031***

Independents (I) 58 .027*** .006 .016 .039 Eff (C-BR) .032***

Captives (C) 55 .027*** .007 .015 .042 Eff (I-C) .004

Type of shareholderForeign ownership (FO) 121 .014** .006 .003 .026 Eff(N-FO) −.014***

National ownership (N) 285 .001 .005 −.011 .012

SizeLarge (L) 110 .002 .007 −.013 .018 Eff (M-L) .002Medium (M) 78 .005 .006 −.008 .018 Eff (L-S) −.003Small (S) 218 .006 .005 −.005 .016 Eff (M-S) −.001

Loan diversificationHigh (HD) 36 .029*** .008 .014 .044 Eff (LD-HD) −.035***

Medium (MD) 85 .026*** .007 .013 .039 Eff (LD-MD) −.034***

Low (LD) 285 −.005 .006 −.017 .007 Eff (HD-MD) .002

Entropy indexHigh diversification (EIH) 242 −.005 .006 −.017 .008 Eff (EIH-EIL) −.017***

Low diversification (EIL) 84 .013*** .006 .001 .024 Eff (EIL-SPE) −.012*

Specialization (SPE) 80 .024*** .006 .012 .036 Eff (EIH-SPE) −.028***

Geographic diversificationHigh (HGD) 264 .008 .006 −.004 .021 Eff (HGD-LGD) .011**

Low (LGD) 142 −.003 .005 −.013 .008

Note: The column “diff” reports the results of the t-test. (SPE) refers to specialization in a single leased asset; if the leased assetdiversification index assumes a value higher than 0 but lower than the mean value, the leasing firm is considered to be poorlydiversified (EIL); otherwise the firm is classified as highly diversified (EIH). (LD) refers to low loan diversification (modifiedLeaven and Levine index from 0% to 33%); (MD) refers to medium loan diversification (modified Leaven and Levine index from33% to 66%); (HD) refers to high loan diversification (modified Leaven and Levine index from 66% to 1); (LGD) refers to the highgeographic diversification in the case of 70% of contract volumes drawn in either one or two regions, while (HGD) refers to thehigh geographic diversification in the case of 70% of the contract volumes drawn in more than two regions. The Std. Err. andthe confidence interval are calculated through the Delta method and are computed by average over sample observations.

* p-value < .10.** p-value < .05.

*** p-value < .01.

be associated with higher cost efficiency levels thus giving support respectively to H3.A and H3.B.Turning to leasing firms’ distribution channels, they seem to play an important role in terms of costadvantages. In particular, leasing firms using primarily the direct, bank-related and mixed distributionchannels appear to be more cost effective than the other firms in the sample. This is not surprisingsince the direct distribution channels can be associated with lower costs. Instead, the mixed channelsare usually chosen because they provide a convenient combination of the other distribution channelsand they can improve the coverage of the customers’ demand. Finally, the bank distribution channelscan typically count on higher territory coverage through the parents’ branches.

Regarding the degree of scale economies and technical change, our evidence suggests that thosefinancial firms with a low and medium loan diversification index appear to enjoy greater economiesof scale but a lower technological progress (Tables 7 and 8). On the other hand, a low entropy indexappears associated with high economies of scale and technological progress. These results indicatethat specialized financial intermediaries are better able to achieve higher economies of scale (with adifference of .036 on average with respect to the diversified companies) and technological progresswith respect to their counterparts. Furthermore, geographical diversification seems to be related tosignificantly lower economies of scale but higher technical progress. As discussed in Section 3.3, theinformation we collected on distribution channels is substantially invariant over time and not detailed

894 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

enough to infer an effect on the technical change. Therefore, we do not include them as possibledetermining factors of the differences in economies of scale and technical change.

With respect to the effect of the size variable (H4.A and H4.B), our evidence suggests that medium-sized leasing companies appear on average more cost efficient than both their smaller and largercounterparts. This result however is not statistically significant. This finding is in line with the existingbanking literature that the relationship between size and cost efficiency is not unambiguous. All leas-ing firms appear to benefit from a relatively high and significant degree of scale economies. However,our evidence suggests that the firms that are likely to obtain higher economies of scale are the inde-pendents as well as the smallest leasing firms in our sample. As mentioned above, it is possible thatthe benefits from consolidation are not yet fully captured by our measures of scale economies. This isnot an unusual finding because in the long run above a certain output size firms should be expectedto experience greater scale diseconomies. In addition, it is often found in the literature that smallerfinancial firms enjoy many advantages over operating locally (e.g. Girardone et al., 2004; Coccorese,2005). However, we do not uncover any significant differences in terms of technological regress overtime (Table 8).

4.4. Is technological progress associated with greater economies of scale?

To test the last hypothesis H5.6 that technological change is associated with greater scale economieswe employ equation (5). Keeping the input prices constant, technical change allows a firm to producethe same output at lower costs. This is important because it allows us to identify how the leasingcompanies are able to enhance productivity without a relative increase in its costs. In the case of Hicksinput neutrality, the shadow-input prices and hence cost minimizing input demand ratios and relativecost shares, are independent of t (Chambers and Färe, 1994). Focusing on the pure technical change,calculated as (t1 + t2T), results are positive with a mean of .143, and highly significant (p-value = 005).This implies that the pure technical change has increased the total costs during the period underinvestigation. From equation (4) we verify that there was a decline in the labour cost share (whosecoefficient is negative) while at the same time we find a slight increase in the share of the capital cost(whose coefficient is positive) in the total costs.

An OBTC > 0 implies that the minimum efficient size has decreased over time. This refers to the firmsize at which the long-run cost is at a minimum. For the overall sample, the marginal positive OBTC19

in equation (5) is significantly different from zero, thus indicating that the degree of scale economieshas on average decreased from 2002 to 2008, strengthening the previous findings. Therefore, theoutput level at which the minimum average cost can be achieved, has decreased over time to a lowerthreshold. However the coefficient for TY2 is negative and significant, thus indicating that the efficientscale is increased for that output.

To summarise, there has been a decrease in the efficient size of leasing firms and, as a result, sizedoes not seem to be an advantage for large financial firms. Therefore, our data provides evidence of adecrease of cost advantages over a specific threshold, in a period of heavy consolidation. As numerousM&As imply a reorganization of the parent group, this may potentially conflict with the achievementof an efficient size.

5. Conclusions

This study provides novel evidence on the potential cost advantages of leasing firms with alternativeownership structures, size and diversification strategies. Despite the relevance of the lease as a form offinancing to fund long-term investments, the focus on this sector in the literature was only marginal –we believe – due to its relative recent expansion and to the difficulties to collect the relevant data. Weemploy a hand-collected sample of data representing approximately 90% of all leasing firms operatingin Italy (based on the market volume covered by Assilea, the Italian Leasing Association, in 2008) overthe period f2002–2008, a period characterised by important institutional and environmental changes.

19 Applying the Wald test, the mean .009 has a standard error equal to 0.02 (significant at 1%).

M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896 895

The model used for the estimation of cost efficiency is the random parameter frontier model (Greene,2005), which allows us to take into account of heterogeneity across the sample through randomparameters.

The findings show that although the leasing sector has experienced a decline in terms of cost effi-ciency over the period, it has also benefited from relatively high (albeit decreasing) levels of economiesof scale. Our main results suggest that bank-related, independents and domestic leasing companiesappear to be more cost efficient than their captive competitors. Diversification strategies seem to playan important role for the achievement of cost efficiencies in the sector. Geographical diversificationseems to be associated with higher technological advancements. Nevertheless, small, independentand less diversified leasing firms appear to benefit from higher economies of scale and technologicaladvancements. In addition, the direct distribution channels seem to be the most cost efficient ones ascompared to their ‘bank-related’ and ‘supplier’ counterparts. Finally, technological process seems tohave played an important role over time in the leasing sector, by decreasing the minimum efficientsize, i.e. the scale of operation at which the total costs are minimized.

References

Acharya, V.V., Hasan, I., Saunders, A., 2006. Should banks be diversified? Evidence from individual bank loan portfolios. Journalof Business 79, 1355–1412.

Aigner, D., Lovell, K., Schmidt, P., 1977. Formulation and estimation of Stochastic Frontier Production Function Models. Journalof Econometrics 6, 21–37.

Altunbas, Y., Molyneux, P., 1996. Economies of scale and scope in European Banking. Applied Financial Economics 6, 367–375.Altunbas, Y., Goddard, J., Molyneux, P., 1999. Technical change in banking. Economics Letters 64, 215–221.Assilea, 2008. Rapporto annuale, Retrieved 23 July, 2011 from: https://www.assilea.it/repository/annual report/Rapporto-

2008.pdf.Bath, C., 2001. Quasi-random maximum Simulated Likelihood estimation of the mixed multinomial logit model. Transportation

Research 35B, 667–693.Battese, G., Corra, G., 1977. Estimation of a Production Frontier Model: With Application to the Pastoral Zone of Eastern Australia.

Australian Journal of Agricultural Economics 21, 167–179.Berger, A.N., Humphrey, D.B., 1997. Efficiency of financial institutions: international survey and directions for further research.

European Journal of Operational Research 98, 175–212.Berger, A.N., Udell, G.F., 2006. A more complete conceptual framework for SME finance. Journal of Banking & Finance 30,

2945–2966.Berger, A.N., 2007. International comparisons of banking efficiency. Financial Markets, Institutions & Instruments 16 (3),

119–144.Berger, A.N., Hasan, I., Zhou, M., 2010. The effects of focus versus diversification on bank performance: evidence from Chinese

banks. Journal of Banking & Finance 34 (7), 1417–1435.Bos, J.W.B., Kolari, J.W., 2005. Large bank efficiency in Europe and the United States: are there economic motivations for

geographic expansion in financial services? Journal of Business 78 (4), 1555–1592.Cavallo, L., Rossi, S.PS., 2001. Scale and scope economies in the European banking system. Journal of Multinational Financial

Management 11, 515–531.Chambers, R.G., Färe, R., 1994. Hicks neutrality and trade biased growth: a taxonomy. Journal of Economic Theory 64, 554–567.Claessens, S., Demirgv̈c-Kunt, A., Huizinga, H., 2001. How does foreign entry affect domestic banking markets. Journal of Banking

and Finance 25, 891–911.Chen, J.T., Chen, C.T., 1998. Holding company affiliation versus branching by independent banks: a cost analysis for interstate

banking. Review of Financial Economies 7 (1), 87–101.Coccorese, P., 2005. Competition in markets with dominant firms: a note on the evidence from the Italian banking industry.

Journal of Banking & Finance 29 (5), 1083–1093.Deng, S., Elyasiani, E., 2008. Geographic diversification and BHC return and risk performance. Journal of Money, Credit and

Banking 40, 1217–1238.De Young, R., Evanoff, D.D., Molyneux, P., 2009. Mergers and acquisitions of financial institutions: a review of the post-2000

literature. Journal of Financial Services Research 36 (2–3), 87–110.Drucker, S., Puri, M., 2009. On loan sales, loan contracting, and lending relationships. Review of Financial Studies 22, 2835–2872.ECB, 2011. Survey on the access to finance of SMEs in the Euro area (September 2010–Febraury 2011). Retrived from

http://www.ecb.europa.eu/stats/money/surveys/sme/html/index.en.html on 15/01/2012.Fauver, L., Houston, J.F., Naranjo, A., 2004. Cross-country evidence on the value of corporate industrial and international

diversification. Journal of Corporate Finance 10, 729–752.Favero, C.A., Papi, L., 1995. Technical efficiency and scale efficiency in the Italian banking sector: a non-parametric approach.

Applied Economics 27 (4), 385–395.Fiordelisi, F., Molyneux, P., 2004. Efficiency in the Italian factoring industry. Applied Economics 90, 947–959.Fiordelisi, F., Monferrà, S., 2009. Measuring Shareholder Value in asset based-lending industries. Managerial Finance 35,

885–903.Girardone, C., Molyneux, P., Gardener, E.P.M., 2004. Analysing the determinants of bank efficiency: the case of Italian banks.

Applied Economics 36 (3), 215–227.Goddard, J.A., Molyneux, P., Wilson, J.O.S., 2001. European Banking: Efficiency, Technology, and Growth. John Wiley, Chichester.

896 M. Degl’Innocenti, C. Girardone / Int. Fin. Markets, Inst. and Money 22 (2012) 879– 896

Goddard, J., McKillop, D., Wilson, J.O.S., 2008. The diversification and financial performance of US credit unions. Journal ofBanking & Finance 32, 1836–1849.

Greene, W., 2005. Reconsidering Heterogeneity in Panel Data Estimators of the Stochastic Frontier Model. Journal of Economet-rics 126, 269–303.

Hadlock, C., Ryngaert, M., Thomas, S., 2001. Corporate structure and equity offerings: are there benefits to diversification?Journal of Business 74, 613–635.

Hughes, J.P., Mester, L.J., 2009. Efficiency in banking: Theory, practice and evidence. In: Berger, A.N., Molyneux, P., Wilson, J.O.S.(Eds.). Oxford Handbook of Banking. Oxford University Press, Oxford.

Hunter, W.C., Timme, S.G., 1986. Technical change, organizational form, and the structure of bank production. Journal of Money,Credit and Banking 18 (2), 152–166.

Iskandar-Datta, M., McLaughlin, R., 2007. Global diversification: new evidence from corporate operating performance. CorporateOwnership and Control 4, 228–250.

Khanna, T., Palepu, K., 2000. Is group affiliation profitable in emerging markets? An analysis of diversified Indian businessgroups. Journal of Finance 55 (2), 867–891.

Kumbhakar, S., Lovell, K., 2000. Stochastic Frontier Analysis. Cambridge University Press, Cambridge.Laeven, L., Levine, R., 2007. Is there a diversification discount in financial conglomerates? Journal of Financial Economics 85,

331–367.Leaseurope, 2009. Leasing Facts & Figures. Retrieved 18 April, 2011 from http://www.leaseurope.org/index.php?page=key-

facts-figures.Leaseurope, 2010a. Annual Statistical Enquiry. Retrieved 10 February, 2012 from http://www.leaseurope.org/uploads/

documents/stats/European%20Leasing%20Market%202010.pdf.Leaseurope, 2010b. Ranking Survey. Retrieved 10 February, 2012 from http://www.leaseurope.org/index.php?page=ranking.Leaseurope-KPMG, 2008. European Leasing. Private document, obtained by request.Lewellen, W.G., 1971. A pure financial rationale for the conglomerate merger. Journal of Finance 26, 521–545.McDaniel, S.W., Kolari, J.W., 1987. Marketing strategy implications of the miles and snow strategic typology. The Journal of

Marketing 51 (4), 19–30.Meeusen, W., Van den Broeck, J., 1977. Efficiency estimation from cobb-douglas production functions with composed error.

International Economic Review 18, 435–444.Pitt, M., Lee, L., 1981. The Measurement and Sources of Technical Inefficiency in the Indonesian Weaving Industry. Journal of

Development Economics 9, 43–64.Santomero, A.M., Eckles, D.L., 2000. The determinants of success in the new financial services environment: now that firms can

do everything, what should they do and why should regulators care? Federal Reserve Bank of New York, Economic PolicyReview, 6, No. 4.

Schmidt, P., Sickles, R., 1984. Production frontiers and panel data. Journal of Business and Economic Statistics 2, 367–374.Stiroh, K.J., 2004. Diversification in banking: Is noninterest income the answer? Journal of Money, Credit, and Banking 36 (5),

853–882.Tadesse, S., 2006. Consolidation, scale economies and technological change in Japanese banking. Journal of International Finan-

cial Markets, Institutions and Money 16 (5), 425–445.Train, K., 2009. Discrete Choice Methods with Simulation. Cambridge University Press, Cambridge.Wilson, J.O.S., Casu, B., Girardone, C., Molyneux, P., 2010. Emerging themes in banking: recent literature and directions for future

research. British Accounting Review 42 (3), 153–169.