Optimal Credit Risk Transfer, Monitored Finance, and Banks

Transcription

1 Optimal Credit Risk Transfer, Monitored Finance, and Banks Journal of Financial Intermediation 17 (2008) Gabriella Chiesa Department of Economics University of Bologna

2 Loans held by banks until maturity/default risk-management tool: construction of diversified portfolios No discrepancy between real life and banking paradigm (Diamond 84): Loans retention and diversification: debt-financed bks retain monitoring incentives: perform delegated monitoring

3 bks keep being debt-financed extensively engage in credit risk transfer (CRT): - first traded in 96, CRT vol. $4.5 trill.; - CRT of unrated firms (bk m.) is steadily increasing; -- portfolio products: loan portfolio securitization: 26% of CRT new pattern of financ. Intermediation: Originate-To-Distribute (OTD) model

4 Concern: mixed feelings about CRT Buffett: CRT stability: makes banks relinquinsh monit./screen Greenspan: CRT has insulated bks and f.mkt from corporate failures (2000 US recession): merits of CRT as risk-management tool

5 Current credit mkt turmoil seems to support Buffett s view and raise doubts about the provision of incentives underlying OTD model Questions: Is the OTD model necessarily harmful? or are the CRT instruments used distortionary? If so why are they used? What s the role for prudential regulation?

6 Previous literature CRT monit./screen. incentives.: undermines the premise of financial stability This paper revisits the issue

7 This paper Optimal CRT monitoring: Maximizes incentive-based lending, for any given bk capital financial intermediation and real invest activity by contrast to previous literature

8 Optimal CRT instruments are based on loan portfolio and have credit-enhancement guarantees pretty much as bks do in practice, but:

9 Credit enhancement needs to be precisely delimited: it must be within a defined interval. Outside that interval, monitoring incentives are undermined -- the quality of assets that back CRT deteriorates

10 Delicate issue. while insufficient credit-enhanc. is never profitable (dilution of mon. inc. factored into the price of insuff. credit-enhanced securities) Excessive credit-enhanc. is profitable: wealth is transferred from bk s financiers (tax payers via bail out) to the bk properly designed risk-based cap. requirements are shown to restore efficiency making it profitable for bk to engage in optimal CRT

11 This paper s message CRT as dynamite: very useful tool, if used properly Prudential regulation part of the conditions for proper use

12 w.r.t. current credit mkt turmoil: The problem is not that credit risk has been transferred (which indeed can be efficiency enhancing), but rather that it has been retained via the (excessive) guarantees provided by the bks to the SIVs

13 Intuition Bks raise outside finance via debt With debt, the better the outcome the greater the bk s income; high outcomes may result from good luck rather than from monitoring: Debt rewards the bk for good luck rather than for monitoring

14 We allow for aggregate risk: loans are s.t. idiosyncratic risks and common risk factor (in line with the evidence: correlation of defaults is driven by the bus. cycle) monitoring improves loan expected return, most valuable in downturns

15 Project return distribution Bank s action m cost F Upturn: θ (p) R Downturn: θ (1-p) R prob. α Ø R R prob. α

16 monit.rev.outcomes are not high outcomes. High outcomes result from good luck rather than m: Debt is suboptimal: it rewards good luck rather than m a CRT arrang that reallocates bk s income from lucky states to the monit.revealing states improves incentives

17 Capital per unit of lending lowers: Incentive-based lending expands: Bk raises more funds, lends more, and still monitors

18 Incentive-Based Capacity L K / c * L * o Ø m Ø m K/ c D L o K o K

19 CRT is welfare improving but bk must have the incentive to engage in optimal CRT Time inconsistency/commitment problem of hedging: after having borrowed funds and made loans, unregulated bk has the incentive to retain loan risk role for prudential regulation: riskbased cap. req.

20 π π nm πcrt nm π m K L D L* L

21 Model Bk funds lending out of internal funds (capital K) and outside finance. The supply of funds is perfectly elastic at a gross rate of return normalized to 1 (zero risk-free int.rate). Lending consists in project financing. Project requires 1 unit at date 0 and delivers a return X {0,R} at date 1. Success prob. depends on: bk s monit/non non-monit monit. Realization of a common risk factor

22 Sequence of events θ Return realization Lending m/ø θ Return realization

23 Project return distribution Bank s action Upturn: θ (p) Downturn: θ (1-p) m (cost F) R R prob. α Ø R R prob. α [p+(1-p) α] R < 1; [p+(1-p) α]r >1+F

24 Portfolio Outcomes For a diversified portfolio, solvency rate s = 1, α, α Highest outcome s=1 good luck (upturn) Monit.rev.outcome: s = α

25 Optimal Contract Maximizes bk s profits: Maximizes outside finance, and hence bk lending s.t. monit.inc.constraint and final investors part.constraint makes use of the information conveyed by loan portf.outcome and rewards the bk as much as possible for the outcomes that signal monitoring

26 Contract (W 1,W α,w α ) Up-turn: θ Down-turn: θ m 1 α W 1 W α Ø 1 W 1 α W α (=0) Bk penalized for nm

27 Optimal contract (W 1,W α ) Max [p W 1 +(1-p) W α - FL-K ] s.t. p W 1 +(1-p) W α - FL-K > pw 1 -K (IC) p(rl-w 1 )+ (1-p)(α RL- W α ) = L- K (PC) W 1 > W α (MC) Rearranging, the max. problem:

28 L < [(1-p)/F] W α Max L s.t. (IC) p(rl-w 1 )+ (1-p)(α RL- W α ) = L- K (PC) W 1 > W α (MC) maximizing profits amounts to max L. From (IC) this amounts to maximizing the reward for monit. W α, because of (PC) this requires minimizing the reward for good luck W 1. At the optimum W 1 = W α

29 Prop.1 At the optimum, bk s lending capacity is: p F p p c c K L + + = = 1 ] ) (1 [ 1 ; * * * α Bank s income: [ ] { } K R p p L W W + + = = 1 ) 1( * * 1 * α α

30 Debt Financing Bk s payoff sched. is a portf.outcome conting. sched. with W 1 > W α Monitoring is under-rewarded: incent.-based lending capacity * * * 1 ) 1 ( c p F R c c K L c K L D D D > + < = α

31 Incentive-Based Capacity L K / c * L * o Ø m Ø m K/ c D L o K o K

32 debt is suboptimal: rewards good luck CRT is the tool for addressing the incentive distortion of debt financing. Consider the arrangament: (assuming Bk can commit to CRT)

33 CRT 1. The bank raises deposits L*-K and finances the optimal loan portfolio L* 2. It forms an SPV and securitizes/sells the loan portfolio for a total price P 0 3. It credit-enhances the deal by giving investors the option to sell their claims back to the bank for a total price P. To back this guarantee, the bank injects P as cash collateral.

34 Prop 3 Any CRT mechanism (P* 0,P*) with ( ) [ ] R L P L p F R P P P P R L p R pl P * * * * * * 0 ; 1, 1 α α α + + implements the optimal contract.

35 The bk s income in securitization is the same as in the optimal contract: W 1 = W α =L* {[p+(1-p) α]r-1}+k ; W α =0 Lending capacity is maximized Depositors and investors break-even All loans are monitored

36 Bk s incentive for CRT Prudential Regulation Will the bk engage in optimal CRT, once the funds have been raised? An unregulated bk profits by creditenhancing the CRT deal excessively, to such an extent that effectively amounts to retaining the entire credit risk

37 π π nm πcrt nm π m K L D L* L

38 Prop. 4 A CRT mechanism backed by excessive creditenhancement undermines the bk s monit.inc. and entails an ex post wealth transfer from depositors to the bk. A capital req. on loans conditioned on the extent of retained risk c P / L * P = c + p α R L Prevents such a wealth transfer, restoring efficiency

39 Results Risk Management: monitoring incentives: more outside finance more lending and real investment Diversification: risk-management tool for idiosyncratic risk; Optimal CRT: risk-management tool for common risk

40 Results Optimal CRT based on loan portfolio and backed by a precise extent of creditenhancement. Excessive cred-enhanc undermines monit (loan quality ) = ex post wealth transfer from depositors to the bk Risk-based capital req. provide bks with the incentive to engage in optimal CRT