Islamic Banks, Deposit Insurance Reform, and Market Discipline: Evidence from a Natural Framework

1. Islamic Banks, Deposit Insurance Reform, and Market Discipline: Evidence from a Natural Framework Ahmet F. Aysan – Mustafa Disli* – Meryem Duygun – Huseyin Ozturk Developmental Central Banking: Issues, Prospects and Challenges 25-26 April 2016

2. Market discipline? • Depositors actively reward or punish banks in function of banks’ riskiness • Two components of market discipline • Market monitoring • Market influence Depositrate ↑ if bank risk ↑ Deposit base↓ if bank risk ↑

3. Depositsdemand and supplyshifts If with bank risk… i S S’ Market discipline:  i*↑& d*↓ d D

4. Depositsdemand and supplyshifts If with bank risk… i S S’ No market discipline:  i* ↓ & d* ↑ d D

5. Depositsdemand and supplyshifts If with bank risk… i S Inconclusive:  i* ↑ & d* ↑ d D D’

6. Depositsdemand and supplyshifts If with bank risk… i S Inconclusive:  i* ↓ & d* ↓ d D D’

7. Motivation • For conventional banks, the process of market discipline is well-documented: e.g., Martinez Peria and Schmukler (2001, JF), Sironi (2003, JMCB), Demirgüç-Kunt and Huizinga (2004, JME), Nier and Baumann (2006, JFI), etc. • For Islamic banks,e.g., Errico and Farahbaksh (1998, IMF), El-Hawary et al. (2004, WB), Beck et al. (2013, JBF) : • Depositors are quasi-shareholders => market discipline ↑ => promote financial stability. • Surprisingly: no empirical analysis that reveals market discipline on Islamic banks.

8. Problems • Very little known about the mechanisms through which the disciplining occurs in a profit and loss sharing arrangement. • An Islamic deposit contract contains neither debt nor equity compensation • Agency conflicts • Not only from the separation of ownership and control for shareholders, but also from the separation of cash flow and control rights for depositors. • Depositors do not have control rights that shareholders possess and their cash flow rights do not coincide with the rights to control investments.

9. Corporategovernance: ‘exit’ as a form of ‘voice’ • Governance trough direct intervention (‘voice’) • Large blockholders have more incentives than atomistic shareholders to exercise disciplining sanctions and pressure management (Shleifer and Vishny 1986). • Governance through ‘exit’ strategies • In the absence of intervention power, dispersed blockholders can still govern firms (Edmans 2009, Edmans and Manso 2011). • Exit strategy upon negative information, thus leading to a stock price that closely reflects the firm’s fundamental value. • Ex post: Pstock↓ → punishment of management; • Ex ante: threat of exit increases managerial efficiency

10. Corporategovernance: ‘exit’ as a form of ‘voice’ • Governance through ‘exit’ strategies • Deposits typically contain a withdrawal option embedded with each account, which licenses the depositor to sell the deposit to the bank at will. • In a risk-sharing framework: Calomiris and Kahn (1991) explain that liquid deposits substantially help to align the bank’s portfolio choice with depositors’ preferences • Off-loading deposits => deposit price ↓ => yield ↑ = pro-rata profit-sharing should increase.

11. Remark 1: influence of religiouscommitmentondepositor’ssensitivity • Religious individuals show more risk-averse characteristics than non-religious individuals. For example, Miller and Hoffmann (1995) report a negative association at the individual level between religiosity and attitudes towards risk. • Abedifar et al. (2013) point that Islamic depositors may have a strong sense of loyalty toward their banks, thus numbing the sensitivity to bank riskiness.

12. Remark 2: Influence of depositinsuranceonmarket discipline • Literature is silent regarding the impact of deposit coverage on market discipline. • Existence of different modes of Islamic deposit insurance makes clear-cut interpretations even more difficult. • Two forms of Islamic deposit insurance: • Protection through the conventional insurance system • Protection through a separate Islamic deposit insurance scheme.

13. Remark 2: Influence of depositinsuranceonmarket discipline • As in conventional banks, deposit insurance will reduce incentives to exert market discipline • However, disciplining intensity may vary with the deposit insurance mode: • A dual deposit insurance scheme: • Shariah-conformity of the reimbursed funds in case of bankruptcy. This may undercut market discipline. • Conventional insurance scheme • May face difficulties in convincing depositors about the reliability of reimbursed funds when a bank failure occurs.

14. Banks in Turkey • After the collapse of Ihlas Finance in 2001, institution of an Islamic deposit insurance scheme providing insurance up to 50,000 TL for each deposit ownership in each bank. • Islamic deposit insurance scheme was administered by the ‘Union of Private Finance Houses’. • 2005/Dec: dual deposit insurance system was revised and the management of the Islamic deposit insurance fund was transferred to the Savings Deposit Insurance Fund (SDIF).

15. Data, empiricalstrategy Erdogan-era 2002/Q4 2005/Q4 2012/Q4

16. Data, empiricalstrategy Conventional deposit insurance scheme 2002/Q4 2005/Q4 2012/Q4 Islamic deposit insurance scheme

17. Data, empiricalstrategy Conventional deposit insurance scheme Conventional deposit insurance scheme 2002/Q4 2005/Q4 2012/Q4 Islamic deposit insurance scheme

18. Data, empiricalstrategy Conventional banks: • 3 state banks • 21 private banks • 19 foreign banks 2002/Q4 2005/Q4 2012/Q4 6 Islamic banks

19. Data, empiricalstrategy Market discipline if α1 > 0 and β1 < 0 with bank capitalization for the control group (i.e., conventional banks) in the pre-treatment period

20. Data, empiricalstrategy α2 and β2 seize the difference in means between the treatment (i.e., Islamic banks) and control groups before the treatment takes place

21. Data, empiricalstrategy α3 and β3 capture for both the treatment and control group the change in depositor sensitivity with the unification of the deposit insurance scheme

22. Data, empiricalstrategy Difference-in-differences coefficients α4 and β4: additional shift in depositor sensitivity of the treatment group after the treatment.

23. Data, empiricalstrategy Y X C Before the reform After the reform

24. Empiricalresults 2002Q4-2012Q4

25. Empiricalresults2002Q4-2012Q4 Conventional banks: evidence of market discipline

26. Depositsdemand and supplyshifts Banks with lower capital ratios… i S S’ Market discipline:  i*↑& d*↓ d D

27. Empiricalresults 2002Q4-2012Q4 Islamic banks: no evidence of market discipline

28. Empiricalresults2002Q4-2012Q4 Islamic banks: evidence of market discipline

29. Findings No market discipline Market discipline Conventional deposit insurance scheme 2002/Q4 2005/Q4 2012/Q4 Islamic deposit insurance scheme

30. Robustness check 1: Before and after 2001 Crisis • Ihlas Finans, the then largest Islamic bank, faced a run on its deposits. • In February 2001, the BRSA revoked the operating license of Ihlas Finans on the grounds that it failed to fulfill its liabilities. • Unlike the conventional deposits, the deposits of Islamic banks did not enjoy insurance coverage with the rationale that profit-and-loss accounts involved no guarantee of return.

31. Robustness check 1: Before and after 2001 Crisis • In a way, the actual creation of the Islamic deposit insurance scheme right after the collapse of Ihlas provides us a second natural experiment: • both banking models were affected by the crisis with the difference being that conventional banks enjoyed deposit insurance, whereas Islamic banks only did so right after the collapse Ihlas.

32. Robustness check 1: Before and after 2001 Crisis Outbreak of the financial crisis occurred concurrently with the introduction of the Islamic deposit insurance scheme. We are not able to disentangle their potential opposing effects on market discipline. However, it might be possible that in the immediate aftermath of the crisis, the wake-up call effects might transcend the numbing effect of the newly instituted deposit insurance scheme.

33. Robustness check 1: Before and after 2001 Crisis

34. Robustness check 1: Before and after 2001 Crisis Conventional banks: evidence of market discipline

35. Robustness check 1: Before and after 2001 Crisis (0.0949) Islamic banks: evidence of market discipline

36. Robustness check 1: Before and after 2001 Crisis Islamic banks: no evidence of market discipline

37. Robustness check 2: Full sample period • We estimate the sensitivity of deposit growth to bank risk factors by looking at the differences in sensitivities across the two depositor groups and three distinct periods: before the 2001Q1 crisis, the 2001Q1 (outbreak of the crisis) and 2005Q3 (before deposit insurance reform) period and the post-reform period.

38. Robustness check 2: Full sample period

39. Robustness check 2: Full sample period

40. Discussion & concludingremarks • Islamic bank depositors behave differently than their conventional counterparts • Presence of both insurance schemes: • Conventional bank depositors were sensitive to bank capitalization • Islamic banks were not disciplined at all. • Deposit insurance reform • produced more sensitive Islamic depositors to bank risk.

41. Discussion & concludingremarks • Pre-reform period • Conventional bank depositors were sensitive to bank capitalization • Islamic banks were not disciplined at all. • Deposit insurance reform • produced more sensitive Islamic depositors to bank risk. Specific design of the Islamic deposit insurance scheme: • consisting of only a handful of Islamic banks • empowered to act as the resolution authority for non-viable banks, and was charged to promote sound risk management practices over its members • mandated to detect early warning signals so that it could intervene timely in the resolution of troubled banks

42. Discussion & concludingremarks • Deposit insurance reform • produced more sensitive Islamic depositors to bank risk. • Depositors may have started to hesitate about the Shariah-compliance of the deposit insurance under SDIF protection • Viable argument since the purpose of a separate Islamic deposit insurance scheme was to signal a radical break from the conventional banking system in order to soothe the sensitivities of the religiously inspired depositors.

43. Finalnote • Risk aversion of depositors is contingent on the Shariah-compliance of Islamic banking. • If depositors become suspicious about the operating environment of banks, they become more vigilant and pay greater attention to bank risk. • This intuition suggests that the degree of risk aversion depends on the religious commitment of Islamic banks. In environments where the commitment is high, loyalty may win over risk aversion (and vice versa)