Characteristics of the recovery process for small financial intermediaries: the case of Italian cooperative banks.

Author:Gibilaro, Lucia


The perspective of the Basel 2 implementation has encouraged banks to conduct a number of studies on the impact that new rules will have on credit management, according to the specific characteristics of the lender (Carretta & Gibilaro, 2005). With regard to Italian Cooperative Banks (hereinafter BCCs), studies tend to highlight the opportunities related to the use of complex credit management tools for internal management purposes (Cesarini & Trillo, 2004), as well as the limitations related to the assessment of the value of the relationship between customer and bank (Comana, 2003).

Available studies on credit management for BCCs focus on the characteristics of this type of bank, such as size, type of customer base and field of activity, showing how its specificity can affect the riskiness of its loan portfolio (Ferri & Di Salvo, 1994). In detail, the evidence provided attempts to prove the greater capacity of these banks to select and monitor their borrowers, compared to the banking system as a whole.

This paper focuses on the phase subsequent to borrower default, with the aim of assessing whether the special nature of these banks is also reflected in the effectiveness of the debt recovery process. Therefore, the analysis is aimed at assessing the significance of the characteristics of the BCCs, with respect to another driver of the expected credit loss, with a view to highlighting if the differences proposed in the relevant literature--in terms of the probability of insolvency--are consistent with the differences reported in terms of the effectiveness of the recovery process, or if the two risk factors offset each other.

By examining the Italian market, the paper underscores the considerable differences featured by these banks, in respect of loss given default (hereinafter LGD), compared to the national average, and shortlists several common characteristics shared by the BCCs, which, in the time horizon considered, implement more effective recovery processes.

The first section of the paper features a review of the relevant literature, aimed at highlighting the factors affecting the effectiveness of the recovery process (paragraph 2.1), and the distinctive characteristics of BCCs capable of influencing insolvency management (paragraph 2.2). Section 2 is dedicated to an empirical analysis of the Italian market in order to identify the differences between the recovery processes put into place by the BCCs and the average of the banking system as a whole (paragraph 3.1) and to highlight the characteristics of the banks determining their greater or lesser success in the recovery process itself (paragraph 3.2). The last paragraph features some brief conclusions (paragraph 4)


Variables Relevant For Assessing The Effectiveness Of The Recovery Process

The study of the debt recovery processes resulting from borrower default, even in the more developed financial markets, usually highlights a success rate below the amount of initial exposure and, therefore, the role played by the LGD in determining the overall credit risk should not be ignored (Covitz & Han, 2004).

The portfolio characteristics capable of affecting the LGD can be either specific, with respect to the characteristics of the individual credit exposures, or general (Resti & Sironi, 2005). The former primarily comprise factors such as the characteristics of the borrower, various aspects of the loan relationship and the distinctive features of the loan contract (Grunert & Weber, 2005), while the latter comprise the (actual or financial) macroeconomic variables approximating the economic cycle (Trauck, Harpainter & Rachev, 2005).

The significance of these factors with respect to credit default and, consequently, recovery rates, ultimately depends on the characteristics of the lender implementing the debt recovery process (Salas & Saurina, 2002).

The following sub-paragraphs present a detailed overview of the characteristics of the loan relationship (paragraph 2.1.1) and the systemic and semi-specific factors (paragraph 2.1.2) that may influence the success of the recovery process, besides the key empirical evidence supporting the assumption that the lender's characteristics can affect the value of the LGD (paragraph 2.1.3)

The Characteristics of The Loan Relationship

The LGD is affected by the characteristics of the borrower and the most significant aspects affecting overall recovery are the legal form of the company concerned, as well as its size, the type and the residence of customers supplied.

The literature on the influence of a company's legal form on the debt recovery rates is rather limited at present; in the case of limited partnerships or listed companies, the lender is able to assert rights only on the firm by releasing a judicial trial: this specificity is very important if one assumes that the recovery rate of an individual exposure is a function of the aggregate recovery rate of the company as a whole (Carey & Gordy, 2004). Moreover, the higher the leverage, the lower the probability that, in the event of the company filing for bankruptcy, its available assets will be sufficient to ensure the recovery for all creditors: the recovery rates for subordinate credit are lower than those for privileged credit (Carty et al., 1998), and the difference increases proportionally to the debt cushion (Van de Castle & Keisman, 1999) and the degree of complexity of the borrower's financial organization (Hamilton & Carthy, 1999).

Furthermore, larger companies may become marginally riskier--ex post--because banks usually prefer not to immediately undertake recovery proceedings and tend to grant extensions and/or offer to renegotiate the loan (Asarnow & Edwards, 1995): apparently, however, these conclusions are not confirmed if we consider the amount of exposure instead of the size of the company, also with respect to smaller enterprises (Davydenko & Franks, 2008).

There is a negative relationship between the LGD and, (i) the degree of interrelation between bank and borrower, and (ii) the length of their relationship. The greater the economic importance and the longer the relationship, in fact, the higher the likelihood that the borrower will honour its commitments, because otherwise it would be very difficult to find other lenders on the market willing to offer credit at the same conditions (Berger & Undell, 1995). Customer loyalty, therefore, is a factor of mitigation of the LGD, because the greater the availability of information, the lower the risk of misjudging the borrower (Longhofer & Santos, 1999); this aspect is particularly important in the case of small firms, due to less information available, which shortcoming, however, can be remedied by relationship lending, in cases when the usual hard information is unavailable (Allen, DeLong & Saunders, 2004). Further information on the implications of relationship lending on the credit risk and, ultimately, the recovery rates can be obtained through the analysis of the institutional relationship between lender and borrower. When the borrower controls the lender its interest is to direct its cash flows from the bank to the enterprise, even financing very risky projects, for as long as the profits made by the enterprise are higher than those made by the bank. This aspect entails the interpretation of relationship lending according to the looting view, as well as the information view: the predominance of the former--which can lead to delays in classifying the borrower's default - determines 30% lower recovery rates, on average, with respect to the credits receivable from the party controlling the bank, compared to other credits (La Porta et al., 2003).

The recovery rate also depends on the characteristics of the defaulting relationship, such as the amount of exposure, the manner of repayment and any guarantees provided. Setting aside the institutional and operating characteristics of the bank, the more recent literature highlights a positive relationship between the amount of exposure and the recovery rate, due to the greater focus on the assessment of the borrower's solvency, in connection with both the decision to grant the loan and the subsequent monitoring phase (Grunert & Weber, 2005), and the greater effectiveness and efficiency of the recovery process (Couwenberg & De Jong, 2007). The proper definition of the terms and conditions of contract, regardless of the degree of complexity of the customer's borrowing structure, may limit the bank's exposure and/or speed up the recovery process (Singh, 2003). Instead, the debt repayment procedure is the key factor that can affect the EAD, with respect to a loan transaction, and non-progressive repayment arrangements are those that feature the highest likelihood of low recovery rates. Lastly, the success and length of the recovery process is affected by the availability of guarantees (Altman & Krishore, 1996) and by the possibility of identifying and enforcing the insolvent borrower's guarantees, in accordance with the hierarchy of the various creditors (Eberhart et al., 1990).

The Systemic And Semi-Specific Factors

Recent developments in the literature about the LGD (Frye, 2000) and the capital adequacy regulations (Basel Committee on Banking Supervision, 2005), have placed the focus on the systemic properties of the recovery risk. The recovered amounts, in fact, depend on the winding up of the defaulting borrower, on the sale of any guarantees or security provided and on the enforcement of any bonds (Querci, 2007). During a period of economic recession, increased insolvency, above the average long-term levels, entails an excess supply on the markets of the assets disposed of by the banks to reduce the LGD, which causes prices to drop and reduces the overall value of the recoveries. Moreover, an excessive number of insolvencies may also affect the length of the recovery process, due to the increased activity of both...

To continue reading