Regulatory Lessons from the Crisis Jeff Carmichael CEO Promontory Financial Group Australasia Date Bond Project University Python 12 December Symposium 2007 Apr 2010 Bond University Symposium 9 April 2010 1
Outline Themes: What were the main regulatory lessons from the crisis? What have been the responses to date? Are they appropriate? Two areas: Architecture Prudential tools and methods Closing Thoughts 2
1.A Architecture The Lessons All regulatory failures are ultimately failures of implementation but architecture can support or inhibit sound implementation The fragmented, industry-based, and competitive regulatory architecture in the US has been heavily discredited: Gaps: regulatory system must be comprehensive and eliminate gaps in coverage (Geithner) ibanks / State insurance Jurisdiction shopping: must end practice of allowing financial companies to choose their regulator simply by changing their charters (Geithner) Arbitrage: Risk held at the point of lowest regulatory cost (holding of sub-prime in SIVs, CDS in insurance and hedge funds, etc) Consolidated supervision: Fed, SEC, OTS all weak Too much focus on trees and not enough on forest 3
Architecture (cont.) UK FSA also criticised but failure of model or of implementation? Northern Rock highlighted the need to keep a proper balance between prudential and conduct regulation Bank failures highlighted the risks that attach to blurring prudential objectives with market development objectives was FSA a captive of the industry? The home/host regulator model also failed to deliver Iceland put paid to the idea of unfettered cross-border branch banking International coordination was poor and well below what had been expected countries closed ranks and sought to protect their own Cross border failure resolution highlighted (BCCI) 4
1.B Architecture - Response US response Eliminate OTS Consolidated supervision (by the Fed) of Tier 1 FIs New Consumer Financial Protection Agency New Oversight Council UK response If Conservatives get in they will undo the FSA International response Colleges of regulators Retreat into silos Cross-border resolution on the radar of FSB 5
1.C Architecture - Reflections Response: totally inadequate yet highly predictable The problem is that financial conglomerates span industry sectors, domestic boundaries and international boundaries Unless agency authorities align with industry structures the regulators are massively handicapped in attempting to regulate the group who regulated AIG Financial Products Unit? AIG perpetuated a massive capital arbitrage because the US banking and insurance regulators were in silos We know what is needed. but politically it is too hard Let s not pretend the domestic problem has been solved let alone the international problem 6
2.A Tools and Methods - Lessons Prudential regulators have much about which to be modest Regulatory capital was inadequate No international consistency Market risk inadequate (3 x 10 day VaR) Looking at the wrong part of the distribution Liquidity was inadequate Risk concentration need to look beneath the surface Consolidated supervision was inadequate Pro-cyclicality in regulatory and accounting rules Regulate smarter skepticism beats a rulebook any day 7
2.B Tools Response Focus on BCBS response to FSB and G20 proposals Proposals are in 3 documents and include: Liquidity (consistent approach): Liquidity Coverage Ratio (LCR) for 30-days stress scenario Net Stable Funding ratio (NSFR) funding to exceed needs Capital Basel III (consistent approach): Remove existing discretions Tighten rules for deductions CAR for counterparties, market risk, and securitisation to increase Simple leverage ratio Countercyclical capital buffer Capital penalty for too big to fail 8
2.C Tools - Reflections Some are unobjectionable but some are of questionable logic Combined impact is likely to be of a magnitude not previously experienced consequences could be major Focus has been too much on the tools and rules and not enough on methodology the agenda has been captured by those who were the biggest failures 9
2.C Liquidity some fuzzy logic? LCR based on a 30-day survival scenario no basis in reality: Liquidity runs are not addressed by selling assets into the market Failure is a matter of 5 days not 30 Standard must be relevant not calibrated on crisis Composition is even more unrealistic approach is rule based (Government securities plus, maybe, corporate and covered bonds) In fiscally responsible countries (Australia) the entire stock is less than the requirement! Problems with the proposal is its prescriptive, one-size-fits-all foundation unlike capital, liquidity is a local concept Liquidity should be set at a realistic level and must be defined against central bank LLR facility with local conditions prevailing 10
Liquidity more fuzzy logic? NSFR is not a new concept but regulatory rule is also fuzzy Intention of promoting more stable funding is fine but devil is in the detail and again the level of prescription Required funding = all illiquid assets (loans, mortgages, etc) plus securities Available funding = a weighted menu (term deposits 100%, stable retail deposits 85%, less stable 70% etc) Ratio (Available/Required) > 1 Where is the science in 1 with such arbitrary weights? Banks are in the business of liquidity transformation natural position is for liquid deposits to fund illiquid loans If banks don t provide liquidity transformation then who will? 11
Liquidity Beware the Impact Regardless of logic for liquidity there is a matter of impact Australian banks currently hold approx $200 b of liquid assets New requirement (if not amended) would require $500 b Requires an adjustment in either: Lending Australian deposits (low price elasticity?), or Foreign wholesale deposits (no, no, no) While impact on GDP and employment is hard to estimate it is likely to be non-trivial (multiplier of.3?) transition? Again this impact varies from country to country 12
2.C Capital Some Quibbles Hard to disagree with the proposition that more capital is better Australia was a good lesson to others Hard to disagree with focus on Fundamental Tier 1 real capital Some examples where more debate is needed: NTA of an unconsolidated subsidiary used to be deducted from T2, then 50/50 T1 and T2, now all T1 1,250% risk weight fine when CAR is 8% but when it is higher, the bank would hold more capital then the size of the investment Unweighted leverage ratio rolls back 30 years of risk focus didn t stop US banks lesson on regulating smarter Countercyclical capital buffers concept sound but implementation tough Same with capital penalty for too big to fail 13
Capital Who will pay the Bill? The real question is how much capital is enough and who will bear the cost? Historically, banks have offered an ROE of between 15% and 20% ROE is generated by earning a spread between borrowing and lending rates after allowing for risk Rule changes that increase capital, increase holdings of low-yield assets (Govt. securities), increase ratio of retail deposits, etc will increase the level of interest rates and the spread Start with impact on profitability Studies by JP Morgan, UBS and others range of assumptions but similar results: The proposed changes will, ceteris paribus, reduce bank profitability by between 40% and 60% (ROE on global banks from 13.3% to 5.4%) - varies by country 14
Capital from Profits to Rates to GDP To attract new capital, banks must restore an acceptable ROE Unfortunately, there is scant statistical modelling to guide us Only one at this stage that models the flow through effects but largely back of the envelope in style They allow for banks to bid up retail deposits, for tighter underwriting standards, and for non-banks to offset the reduction in bank lending For the US they estimate that bank lending growth drops from 7.3% to 3.8% after 4 years The cumulative impact on GDP is around 2% over 4-5 years Impact in Australia is likely to be much greater the liquidity impact in the US is negligible the non-bank substitution offset is very high in the US, and The US reliance on wholesale funding is much lower (LCR closer to 1) 15
Closing Comments There have been many important lessons from the crisis It is still early days and we have yet to see which proposals are translated into practice However, the early indications are that: Global regulators are addressing the wrong lessons they are attacking the easy problems not the important ones They are focusing too much on adding rules and prescription rather than focusing on how to regulate smarter Focus is on regulation - when it should be on supervision The changes are coming thick and fast and are being pushed by political agendas We need to assess the costs and weigh the combined impact of the proposals otherwise the recovery may be dead 16