Autori

Juan Sergio Lopez e Maria Carmela Mazzilis (Federcasse- Economic analysis and statistical reporting department)
Manfredi Aliberti (University of Cambridge – MPhil in Economic Research Candidate)
Stefano Grimaldi (EGS-Economia, Geografia, Statistica)

Abstract

The paper follows that part of the banking literature stating that banking diversity is an asset towards achieving more resilient and functional banking systems examining the case of the Italian Cooperative Credit Banks (CCBs). CCBs are local banks highly focused on strengthening the social capital of the local communities where they operate. They are characterized by a better knowledge of local economic agents than larger banks. The study investigates the role and performance of local cooperative banks by identifying how much of their behavior and performance can be explained by their local nature and how much by their mutualistic rules. In particular, it has been tested the significance of some mutualism indices (measured in terms of intensity of the member-CCB credit relationship) in explaining the dynamics of the NPL-Asset ratio occurred in the 2007-2017 period.

Introduction

CCBs have a strong expertise in the traditional originate-to-hold intermediation model: banks provide loans to firms and individuals and hold them in their balance sheet until maturity, bearing the related credit risk. This model differs from the originate-to-distribute one, widespread in large banks, where banks distribute their loans to other market participants through securitizations, thus transferring the credit risk to other parties.

In addition, Italian CCBs are the only banks characterized by “prevailing mutualism”, consisting in a variety of regulatory constraints including that a minimum of 51% of risk activities must be carried out with members and that at least 95% of lending must be granted in the catchment area.

These peculiar characteristics make the CCBs a unique category in the Italian banking system.

The acquisition over the years and the careful analysis of the so-called “soft” information is the key to the success of this category of banks. CCBs success is not only and not so much linked to the short-term profitability, as it is mainly due to the aim of maintaining a healthy local economic environment, considered as a prerequisite for the stability of the bank and its long-term survival.

It is believed that, by focusing mainly on retail banking and applying the relationship banking model, cooperative banks can support the financial inclusion of marginal customers and reduce the credit rationing of borrowers, particularly households and small and medium-sized enterprises (SME’S). Several studies carried out in recent years have shown that, in times of crisis, CCBs have been able to act asynchronously to large banks, managing to mitigate the effects of a credit crunch through the provision of loans to customers linked to the bank by long, stratified and fruitful relationships, largely CCBs members [1].

One relevant issue in investigating the role and performance of local and cooperative banks is identifying how much of their behavior and performance can be explained by their local nature and how much by their mutualistic rules. This distinction can be quite helpful for the internal management, in order to evaluate strategic choices, and for public authorities when they have to choose among different policy options. In order to bring a contribution to this debate we exploit the mutuality rule that CCBs have to respect (at least 51% of RWA must be exposures to members). Even though 51% is the lower bound, each bank can choose to work more with members. This choice may have pros and cons. For example, members are better known and have a stake in the bank (so they may be considered safer) but they usually enjoy better conditions so the bank may have to forfeit the additional earning coming from making loans to non-members. Moreover, to maintain a high ratio of mutuality a bank may have to increase the number of members, but this may impact on consolidated governance equilibria. Therefore, the intensity of mutuality is a choice of the management that may affect the bank performance.

In this study we use credit quality as a measure of performance. The relationship banking model envisages that small local banks may make better use of soft information than large banks. We test if CCBs with a higher rate of mutuality achieve a better performance, namely a better credit quality. This would suggest that the localism combined with mutualism leads to better performance in term of screening and monitoring of clients.

As measure of credit quality, we use the ratio of bad loans to total loans. The main variable of interest is the intensity of mutualism measured by the ratio of loans to members to total loans.

Beside CCBs data, regional and county-level control variable are used to better represent the essentially local nature of these institutions.

Explanatory and control variables have been employed to account for national macroeconomic trends and social capital local differences. The data have been analyzed within a dynamic panel data model structure to take into account endogeneity issue. The model can thus be estimated through the employment of explanatory variables and of an autoregressive element.

[1] See for ex. Demma C. (2015), Stefani M.L. e Vacca V. (coordinatori) (2016), Barone G., de Blasio G. e Mocetti S. (2016)

The determinants of non-performing loans: a review of the literature

The recent academic literature that have analyzed the determinants of NPLs focused on the relevance of two main drivers: bank-specific and macroeconomic factors.
Jimenez and Saurina (2006) examine the Spanish banking sector. They provide precise and robust evidence of a positive, although quite lagged, relationship between rapid credit growth and future nonperforming loans of banks. Moreover, they also find a direct relation between the phase of the lending cycle and the quality and standards of the loans granted. During lending booms, riskier borrowers obtain funds, and collateral requirements are significantly decreased. Lower credit standards and a substantial lag between decisions made on loan portfolios and the final appearance of loan losses point toward credit risk significantly increasing during good times. Therefore, credit risk increases in boom periods, although it only pops up as loan losses during bad times.
Louzis et al (2010) use dynamic panel data methods to examine the determinants of non-performing loans (NPLs) in the Greek banking sector, separately for each loan category (consumer loans, business loans and

mortgages). The study is motivated by the hypothesis that both macroeconomic and bank-specific variables have an effect on loan quality and that these effects vary between different loan categories. The results show that, for all loan categories, NPLs in the Greek banking system can be explained mainly by macroeconomic variables (GDP, unemployment, interest rates, public debt) and management quality. Differences in the quantitative impact of macroeconomic factors among loan categories are evident, with non-performing mortgages being the least responsive to changes in the macroeconomic conditions.
Nkusu (2011) tests the nexus between NPLs and macroeconomic variables of 26 advanced economies over the period 1998 to 2009. In his panel analysis, he found that unemployment rate and policy rate have positive impact on NPLs, whereas house prices and equity prices have negative impact on the same.
Bofondi and Ropele (2011) investigate the macroeconomic determinants of the quality of loans to
households and firms in Italy in the period 1990–2010. To measure the quality of loans they used the ratio of new bad loans to the stock of performing loans at the end of previous period (NBL ratio). The NBL ratio seems to be well explained by just a few macroeconomic variables bearing on general economic conditions, the cost of borrowing and the burden of debt. In particular, the NBL ratio for lending to households varies inversely with the growth rates of real gross domestic product and house prices while it varies directly with the unemployment rate and the short‐term nominal interest rate. As for firms, the NBL ratio increases with the unemployment rate and the ratio of net interest expenses to gross operating profits, while it diminishes as the consumption of durables increases. The above‐mentioned macroeconomic determinants affect the evolution of the NBL ratios with different time lags.
Angelini, Bofondi and Zingales (2017) try to identify what fraction of NPLs was “unavoidable”, due to the economic recession, and which one was caused by poor ex-ante lending decisions by banks. The analysis suggests that macroeconomic conditions explain almost 90% of the NPLs flows observed in the period 2008-2016. A micro analysis focusing on the non-financial corporate sector (where most of NPLs originated) relying on a simple counterfactual exercise suggests that at least 50% of defaults were “unavoidable”. A mix of a weak corporate sector at the beginning of the recession and a weak ability of banks to select borrowers even played a role.
Cerulli et al (2017) show that In Euro countries (2006-2016) higher levels of judicial inefficiency are related to greater levels of NPLs in the following year: a reduction of 30 days in the average time period required to enforce a contract corresponds, ceteris paribus, to a mean decline of the NPL ratio by 0.24 percentage points. Similarly, greater NPL levels are preceded by higher benchmark rates suggesting that higher interest rates make it difficult for companies to repay their debts and this results in a higher stock of NPLs. A reduction of the 100 basis points of the benchmark rates corresponds, ceteris paribus, to a mean decline of the NPL ratio by 0.51 percentage points. The evidence is even stronger combining economic stagnation and judicial inefficiency in a single indicator: a worse economic framework precedes NPLs.
Mohanty et al (2018) investigate the determinants of NPLs of the Indian banking system for the
period 2000-2016. The major findings of the study are as follows. Among macroeconomic variables, economic growth, stock market index and market capitalization ratio have negative impact on the Gross NPLs ratio, whereas, expansionary fiscal policy escalates the Gross NPL ratio; with regard to corporate specific variables, net sales growth and net profit margin have statistically negative impact on the Gross NPL ratio; finally, with respect to bank specific variables, higher credit deposit ratio, growth in bank branches, higher return on equity and higher Capital to Risk Weighted Asset Ratio will lower Gross NPL ratio. Higher operating expense ratio has significant positive impact on the NPLs, which is indicative of inefficiency of the banks.
Lee and Roseknratz (2019) analyse the determinants of NPLs in some Asian countries in the period spanning from 1997 to 2017. The data employed have been divided into macroeconomic data and lagged bank-level data to better distinguish between local-management issues and national or international trends in determining the trend of NPLs. The findings suggest that NPLs’ trend is mainly influenced by macroeconomic trends, with economic growth exerting downward pressure on NPLs and the exchange rate against the dollar, the VIX global risk aversion index and the annual variation of unemployment rate causing increments in the NPL ratio. Lagged bank-specific variables also explain the trend of NPLs but to a lesser extent, as the associated coefficients are somewhat relatively smaller. The regressions give that equity-to-assets ratio and return on equity both tend to reduce the NPL ratio, while variations in loans-to-deposits ratio and loans growth rate are associated with increments in NPLs.

The Italian Cooperative Credit Banks: features and importance

CCBs play a very specific role within the Italian banking industry with regard to governance and
organizational structure.
With reference to governance, Italian CCBs are the only banks characterized by “prevailing
mutualism”, which consists in the following legal features:
• adoption of the “one-head one-vote” rule;
• members may hold shares of up to €50,000;
• members must have their domicile and/or continuative business within the area where the bank operates;
• at least 51% of risk activities must be related to members;
• at least 95% of the loans must be granted in the catchment area;
• at least 70% of profits must be allocated to the legal reserve (with 3% dedicated to Mutual Funds for the promotion and development of the cooperation);
• reserves cannot be distributed to members;
• derivatives may only be used to reduce the risk of loss (hedging).
As for the organisational structure, from the outset CCBs have traditionally been independent banks that joined together to become a national horizontal network. The network was structured into two parts which fulfilled different functions: a political side (an association structured in regional federations and a national association) and a corporate side for business initiatives. The latter was designed to ensure integration, synergy and economies of scale to benefit the single banks of the network and hinged on three Central Institutions (Gruppo Bancario Iccrea, Cassa Centrale Banca di Trento and Cassa Centrale Raiffeisen dell’Alto Adige).
In the first part of 2019, a very relevant process of reform (started with a new law approved in 2016) came to an end. ICCREA and Cassa Centrale Banca became the parent companies of two banking groups, to which 141 and 83 CCB’s respectively belong . The 39 banks of the Bolzano province did not join one of the two banking groups; instead they opted for the creation of an institutional protection scheme (IPS). The establishment of an IPS ensure a mutual capital and liquidity support among the participating banks and allows the recognition of certain regulatory benefits.
The 2019 reform changed the organisational structure of CCB’s but it did not change the nature of the BCCs, which remain independent local banks with their own boards of directors, focused on providing loans to firms and individuals settled in the catchment area.
Today CCB’s represent the greatest majority of local banks in Italy: by mid-2019 there were 263 CCBs (53% of the total number of banks) with over 1.3 million members and 4,231 branches (17% of total branches) located in 2,638 municipalities. CCBs are the only bank service providers in 630 municipalities, mostly characterized by having less than 5,000 permanent residents. CCBs loans were 7.2% of total loans and the market share of CCBs rises to more than 20% in elective sectors such as agriculture (20.9%), tourism services (21.5%) and artisan firms (23.5%) as well as in loans to minor and micro enterprises (24.2% and 19.3% respectively).
With respect to credit quality, the share of bad loans over total loans is 6.9% (5.1% for the whole banking system ); CCBs share of bad loans is significantly lower than the average of the banking industry in important sectors: households, micro enterprises and non-profit institutions.
A further interesting figure regarding Italian BCCs, is the Common Equity Tier 1 ratio (CET1) that, in 2018, averages 16.4% of total risk-weighted assets (RWA) (13.3% the average for the banking industry);
In the period analysed the credit supply of the Italian banking industry has experienced some stress due to funding problems, need for recapitalisation and exponential growth in non-performing loans. The Bank Lending Survey, the periodic survey conducted by the ECB and the Bank of Italy on the state of credit supply and demand in the countries of the Eurozone clearly shows this phenomenon: from mid-2007 it is clear that the supply criteria are progressively more restrictive until they reach their peak value in mid-2009. Only from 2014 onwards the supply conditions seem to be more relaxed.
In this scenario, CCBs are characterised by a significantly higher growth in lending than other banks. In the period 2008-2018 CCBs increased their loans by more than €11 billion, corresponding to an increase of 9.1%, significantly higher than that for the other banks (+4.3%).

Source: Bank of Italy

The strong growth of CCBs loans, despite the deep economic crisis, has definitely helped the local economy of the areas of reference, but has physiologically determined a significant growth in non-performing loans as in the banking industry.
CCBs gross non performing loans were €17.6 billion at the end of 2018, more than doubled over the decade; a similar trend was recorded for the other banks, despite a lower increase of loans.
In the period 2008-2018 CCBs Gross NPL ratio has increased from 6.9% to 13.7% while other banks Gross NPL ratio has passed from 5.5% to 10%. In the same years the Gross bad loans ratio has passed from 2.8% to 7.4% for CCBs and from 2.6% to 5.6% for other banks.

In the decade analysed, CCBs loans provided to members have increased more than CCBs total loans: at the end of 2018 CCBs members gross loans were more than €62 billion, up by 13.9% compared to 2008.

With a more intense growth in member loans, throughout the period considered there has been a gross bad loans ratio for the members constantly better than that for all CCBs customers.

A clear difference in the quality of the loans provided to members is evident in all geographical areas and is consistent in the period analyzed (in the following figure it is possible to see, for example, the evidence relating to 2018).

It can be assumed that a greater incidence of credit provided to members may have partially mitigated the worsening of credit quality that has characterized recent years and facilitated the gradual recovery of the most recent period.

Aggregate analyses show that where the intensity of mutualism is higher CCBs show, on average, a better credit quality.

These preliminary results suggest the opportunity to further analyse, with an econometric model, this issue.

The model

The aim of the analysis is to test the hypothesis that a higher level of mutualistic intensity, measured by the percentage of loans allocated to CCBs’ members, might determine a better quality of loans provided by CCBs.
Our goal is to check whether, taking into account the economic and banking environments where each CCB operates we can confirm that a significant and negative relation between a measure of credit quality (the ratio of bad loans to total loans) and the mutualistic intensity exist.
To this end, we take the single CCBs as statistical units and we build a longitudinal dataset which covers the period 2008-2017 to include both the recessions of the recent past.
In line with the literature (Klein 2013, Lee and Resenkranz 2019), in our model we use three types of explanatory variables in order to investigate the determinants of CCBs bad loans:

macroeconomic variables to account for the dynamic of the local economy;
local social variables to account for the propensity to abide to social standards;
bank specific variables to account for the choices made by the bank management, including the propensity to work with members
As macroeconomic variable we use the variations of GDP at regional level. We also add the average default rate at county level in order to take into consideration the riskiness of the area where each CCB operates.
As social variable we select the micro criminality rate at county level.
As bank specific variables we use:
Tier1 capital; less capitalized banks can choose to issue riskier loans (moral hazard hypothesis, Berger and De Young 1997); on the other hand the lower capital ratio may constrain the growth of loans and reduce the probability of higher NPLs (capital constraint model, Keeton 1999);
Loans growth: higher credit growth may be achieved reducing credit standards (Keeton 1999)
Productivity (loans per employee) and cost efficiency (cost to income ratio): the impact of this variables is not straightforward: according to Berger and Young (1997) a high level of cost may signal bad management that may result in the accumulation of higher bad loans; on the other side high cost may reflect a policy devoting more resource to screening and monitoring that would result in lower bad loans;
Size: larger banks may diversify their portfolio and use more sophisticated approaches in screening their clients; conversely, larger banks may experience increasing difficulties in apply the relationship banking model;
Composition of loans (share of loans to enterprises): we expect a positive relationship between the share of loans to enterprises and the level of bad loans since loans to enterprises are usually riskier than loans to households.
Finally, we include a variable that measure the strength of the mutualistic relationship measured by the share of loans to members
The following table summarises and explains the variables used:

As in much of the literature on NPLs, we use the logit transformation of the NPL ratio as the logit specification ensures that the dependent variable spans over the interval (-∞,+∞) and hence is distributed symmetrically and not restricted to take on values in the (0,1) interval.

As the dataset comprises between 421 CCBs (as in 2008) and 262 CCBs (as in 2017), panel data is not balanced. To account for this feature of data, following the literature on this topic, we have decided to include an autoregressive element, in the form of the lagged dependent variable. The following is the model employed:

Where the dependent variable y_(i,t) denotes the logit transformation of the NPL ratio for bank i at year t . The regressors are classified into four groups:
X_t is the vector including the macroeconomic variable ∆GDP and Decfnz on a regional level; S_t is the variable which describes the social environment (Crim); B_t is the vector of bank-specific variables and M_t denotes the variable accounting for mutualistic intensity. The term ε(i,t) is the composite error term consisting of bank fixed effects u_i and the idiosyncratic term e(i,t).
The Arellano-Bond estimation technique has been employed, a technique in which additional instruments can be obtained exploiting the orthogonality between the lagged dependent variable and the error term e_(i,t).
The results are presented in the following table:

The results show that the NPL ratio is negatively affected by the annual variation of GDP and positively by the average riskiness of the area. The social environment variable have a positive coefficient. The level of capitalisation shows significant negative coefficients, as expected.
Significative and positive is the coefficient of the logarithmic transformation of total assets, which evidences that the size of a CCB directly influences the quality of loans; also, as expected, a higher share of loans to enterprises affect positively the bad loans ratio.

Finally, the share of loans provided to members appears to be significant and with a negative sign showing that where the share of loans to members is higher, the ratio of bad loans to total loans is lower.
In order to evaluate the robustness of the results obtained, the dependent variable was replaced by the NPL on risk-weighted assets (RWA). Also, alternative variables and specifications were tried. The results obtained substantially confirm the results obtained.

Conclusion

The work confirms that diversity of ownership, size, organization and governance represents an element of strength of the banking industry, because of the diversification of behaviours and operational approaches. More specifically, it emerges that the specific nature of cooperatives and mutual societies, in addition to the local bank feature, plays a significant role in determining the performance of CCBs.
Despite the pressure to standardise organisational and governance models, empirical evidence confirms the validity of the specific model of a small local and mutual bank for financing of the local economy.
This model can be further strengthened with a coherent implementation of the reform that led to the creation of the cooperative banking groups.

References

Angelini,P., Bofondi, M., Zingales, L., P. Angelini, M., 2017 ,“The origins of Italian NPLs”, mimeo, Banca d’Italia and University of Chicago;
Barone G., de Blasio G. e Mocetti S., 2016, “The real effects of credit crunch in the Great Recession: evidence from Italian provinces”; Temi di Discussione Banca d’Italia n.1057;
Berger, Allen N., and Robert DeYoung, 1997. “Problem Loans and Cost Efficiency in Commercial Banks.” Journal of Banking and Finance 21 (6): 849–70.
Bofondi, M. and T. Ropele, 2011, “Macroeconomic determinants of Bad Loans: Evidence from Italian debt”. Banca d’Italia, Occasional papers n. 89;
Cerulli, G., D’Apice, V., Fiordelisi, F., Masala, F., 2017, “ Non-Performing Loans in Europe: The Role of Systematic and Idiosyncratic Factors”, Banking Review Policy Journal, 3(5), 34-59;
Demma C., 2015, “Localismo bancario e crisi finanziaria”; Questioni di Economia e Finanza Banca d’Italia n. 264;
Dimitrios P. L., Vouldis, A.T., Metaxas, V.L., ,2012. Macroeconomic and bank-specific determinants of non-performing loans in Greece: A comparative study of mortgage, business and consumer loan portfolios, Journal of Banking & Finance n.36;
Jimenez, G., Saurina, J., 2006, “Credit Cycles, Credit Risk, and Financial Regulation.”,International Journal of Central Banking, vol. 2, pp 65‐98.
Keeton, W.,1999, “Does Faster Loan Growth Lead to Higher Loan Losses? Economic Review, Federal reserve Bank of Kansas City, second quarter”.
Keeton, W., Morris, C., 1987. “Why Do Banks’ Loan Losses Differ?” Federal Reserve Bank of Kansas City. Economic Review (May): 3–21.
Klein, Nir. 2013. “Non-Performing Loans in CESEE: Determinants and Impact on MacroeconomicPerformance.” IMF Working Paper No. WP/13/72.
Lee, J., Roseknratz, P., 2019, “Non performing loans in Asia: determinants and macrofinancial linkages”, ADB Economics Working Paper Series, n. 574;
Louzis D. P., Vouldis, A. T., Metaxas, V. L., ,2010, “Macroeconomic and bank‐specific determinants of non‐performing loans in Greece: a comparative study of mortgage, business and consumer loan portfolios.” Bank of Greece, Working paper no. 118.
Mohanty, A. R., Das, B. R., Kumar, S., 2018, “Determinants of non-performing loans in India: A system GMM panel approach, Prajnan, 47(1), 37-56.
Nkusu (2011) Nkusu, M., 2011, “Nonperforming Loans and Macrofinancial Vulnerabilities in Advanced Economies”, International Monetary Fund Working Paper 11/161.
Stefani M.L. e Vacca V. (coordinatori) ,2016, “Le banche locali e il finanziamento dei territori: evidenze per l’Italia (2007-2014)”; Questioni di Economia e Finanza Banca d’Italia n. 324;

Le BCC a sostegno della popolazione ucraina

Attenzione alle truffe perpetrate a nome di “BCC Federcassa Spa”

Resta aggiornato sul mondo del Credito Cooperativo