The usual disclaimer applies. The opinions are those of the discussant only and in no way involve the responsibility of the Bank of Italy.

Transcription

1 Business Models in Banking: Is There a Best Practice? Conference Centre for Applied Research in Finance Università Bocconi September 21, 2009, Milan Tests of Ex Ante versus Ex Post Theories of Collateral using Private and Public Information by Allen N, Berger, W. Scott Frame and Vasso Ioannidou discussion by Francesco Columba The usual disclaimer applies. The opinions are those of the discussant only and in no way involve the responsibility of the Bank of Italy.

2 Summary I What is the economic function of collateral? Two different sets of theories : ex-ante theories: adverse selection problems due information solved thanks to collateral through signaling to asymmetric ex-post theories: frictions motivate collateral in an optimal debt contract to address moral hazard, contract enforcement, costly state verification Empirically i unsettled the relative relevance of the two sets of theories as well as the effect of relationship lending on collateral The paper tests the empirical predictions of the two classes of models discussion by Francesco Columba 2

3 Summary II Empirical test 1 : identification of the effect of the 2 sets of theories studying variation in the incidence of collateral pledges at loan origination Empirical i testt 2: which h one is the empirically ii dominant effect? The ex-ante private information or the ex-post incentive conflict? Results Test 1: consistent with both theories, but ex-ante private information models valid only for borrowers unknown to lenders (short relationships) Test 2: consistent twith testt 1; ex post theories empirically i dominate for firms with longer relationships (where private information is less important) Policy implications: to reduce use of collateral developing credit registries discussion by Francesco Columba 3

4 Relevance of collateral Financial crisis: change in the value of pledged assets correlated across borrowers and amplifies business cycle through pro-cyclical changes in access to credit Key for literature on role of financial frictions leading to procyclicality of financial sector: financial accelerator Effects on loan demand, loan supply, funding and market liquidity, stability of financial intermediaries (Bernanke et al., 1999, Brunnermeier, 2008, Adrian and Shin, 2008, Iacoviello, 2005, Gerali et al. 2009) Originate-to-distribute model; securitisation, retained securitisation; pledgeability in operations with Central Banks. discussion by Francesco Columba 4

5 data and empirical model Bolivian credit registry design: information on outstanding loans, on lenders. alternative measures of unobserved risk (assumed to be not indirectly observed from other sources)? test 1 hinges on assumption that past repayment problems predict future delinquencies (shown in Appendix): how is endogeneity addressed in assessing this relation? discussion by Francesco Columba 5

6 test 1 test 1: observed risk positively associated with incidence of collateral consistently tl with ex-post theories unobserved risk only if interacted with relationship length provides support to ex-ante theories for short relations, when private info is more important both theories hold true or sub-samplingsampling along firm size and loan contract is in favour of ex-post theories (as indicated in robustness checks)? discussion by Francesco Columba 6

7 test 2: test 2 examines relationship between collateral and ex post loan performance effect of collateral on probability of ex post performance is ambiguous with ex post models the positive relationship with collateral could be compensated by the mitigation effect provided by collateral itself with ex ante theories the effect of collateral is negative so if the effect is negative ambiguity on which of the two theories prevails results: collateral is positively associated with ex post defaults consistent with ex post theories, but when interacted with relationship length results change. other ways of disentangling the effect of the two theories effectsof personal guarantees on ex-post loan performance? how is debt restructuring accounted for? discussion by Francesco Columba 7

8 Developments and policy implications Contribution to the literature: not so clear the divide between the two classes of theories Implications for access to credit in terms of quantity and interest rates? Characteristics of credit registries, one size fits all countries? Policy implication: to foster development of credit registries to eliminate the need for costly collateral Welfare analysis? If collateral l is inefficient i why it isused? Are banks lazy? discussion by Francesco Columba 8