Prof. dr. Enrico Perotti

1
Prof. dr. Enrico Perotti
Regulating Systemic Risk Creation

• Duisenberg school of finance
• University of Amsterdam
• Oxford University

2
Systemic risk creation as an externality

• We need to discourage systemic risk creation at
  the level of individual bank
• Define systemic risk as vulnerability to extreme
  correlation
• Associated with disruptive jumps in risk premia
• Strategies which increase systemic risk create a
  negative externality on other banks
• A Pigovian pollution tax can correct such an
  externality, improve incentives
• So what is the externality ?

3
Lessons from the crisis

• Regulatory arbitrage reduced bank capital and
  created opacity by slicing up assets
• Securitization encouraged riskier lending
• Once the first losses occurred, runs accelerated
  due to extreme maturity mismatch
• long term opaque assets funded with short term
  wholesale borrowing

4
All systemic crises involve runs

• In a crisis, whatever the source, investors
  escape risk, seek cash at all cost
• As prices no longer adjust supply, access to
  credit, not its cost, becomes central
• Maturity mismatch compounds any shock, spreads
  runs
• Rapid withdrawals led to fire sales, rapid price
  drop, as repricing speed is high
• More losses lead to margin calls and more fire
  sales
• So runs are a cause of extreme correlation, not
  (only) their consequence
• Crisis which do not spread to general credit (eg
  Internet bubble) do not qualify

5
How to measure systemic risk creation ?

• Maturity mismatch produces rollover risk
• Most of the time, rollover risk is just price
  risk, can be dealth with capital requirements in
  extreme events, it forces sales at any price
• Then reliance on short term funding increases the
  risk of crisis propagation
• As maturity mismatch amplifies co-movements in
  assets with little economic correlation, it may
  be taken as a proxy measure for marginal
  contribution to systemic volatility

6
Liquidity Charge

• We propose a liquidity charge indexed to maturity
  mismatch, paid continuously to a supervisory
  institution able to supply liquidity (and some
  capital insurance) upon massive runs
• Only activated by systemic runs (no single bank)
• Insured deposits to be exempted (sluggish)
• May be seen as insurance premia providers of
  liquidity by those who draw it
• Aim is to affect incentives, discouraging
  rollover risk by flattening yield curve at short
  maturities

7
Insurance charges or taxes ?

• Charges meant to correct an externality
• Short term borrowing is too cheap as it bears
  little risk, does not price systemic rollover
  risk it causes
• Charges best designed as a tax, to allow the
  central bank ex post discretion on volume and
  allocation of support
• Main aim is to shape incentives by flattening the
  short term borrowing curve
• Powerful ex ante prudential tool

8
Combine with capital requirements

• Geneva report call for capital requirements
• indexed to maturity mismatch, asset growth and
  leverage
• No need for high capital in good times, only
  funding contingent on a crisis
• May trigger specific intervention tools (limits
  on bank strategies)
• More political support, as prepayment of
  intervention
• Liquidity insurance is complementary to capital
  requirements, targets extreme rollover risk
  (which is not an asset price risk !)

9
Relying only on capital ratios causes incentive
problems

• Capital levels to prevent a systemic crisis would
  need to be huge during normal times
• Because of tax, information and agency issues,
  bank capital is expensive.
• Or short term debt is deceivingly cheap.
• Bank capital is legally owned by shareholders.
• As an asset rather than a provision, it creates
  legal issues in rescues of bankrupt banks
• Banks with excess capital will lever it up
  through riskier loans and strategies.

10
Combine with capital insurance

• Capital insurance offers contingent capital,
  ideal for a rare event (Kashyap Rajan Stein 2008)
  
• Private insurance no longer seem realistic after
  September
• Contingent liquidity support can support a larger
  fraction of the balance sheet
• Capital injections should be part of the scheme,
  perhaps event for contractual amount
• Full capital insurance needed to stop runs would
  be large, best to combine with liqudiity
  insurance
• In the fall, government purchases of bank equity
  worked at first but were clearly insufficient

11
How is it different ?

• Why banks should pay a liquidity charge and still
  have to provide securities to ''obtain''
  liquidity ?
• Right now this service is free of charge
• Can be seen as payment for LLR services
• Can it be justified in price terms ?
• Yes, the market value of liquidity in a crisis is
  huge, and collateral is currently generously
  refinanced
• Gives legitimacy to massive support on a crisis
• Does it really add to crisis management tools ?
• It adds more to ex ante tools.
• But the trigger of insurance payment allows
  capital contributions (with no not state funds),
  associated with prudential restrictions

12
Powerful ex ante supervisory tool

• Insurance charge to be based on asset weighted
  maturity mismatch. Should be large enough to
  discourage banks to grow simply by reusing cheap
  wholesale funding to buy market assets.
• Potentially powerful tool to control risk.
• Example, to slow down mismatch (or rapid
  leveraging), CB can increase charges for higher
  mismatch (eg very short term funding). An extreme
  level 1 day annual rate 1 year rate
• Alternatively, insurance may have larger marginal
  than average rates.

13
Unintended consequences

• Could this shift rollover risk to shadow banking
  ?
• Not if charges apply to exposure to unregulated
  intermediaries with maturity mismatch
• Critical to distinguish between intermediaries,
  firms
• Regulating the boundary indispensable for any
  policy
• How to treat credit lines ? Any contingent
  exposure should be fully charged as it would be
  drawn for sure in a run
• If the pricing of bank contingent liabilities
  fully reflects exposure to mismatched
  intermediaries, it would ensure better pricing of
  bank liquidity to shadow sector
• The IFRS demanded full recognition of
  bank-supported SIV on bank balance sheets.

14
Summary

• A new regulatory tool to tax system risk creation
• All systemic crises involve liquidity runs, force
  support by monetary authorities
• Public liquidity charges realign incentives by
  taxing marginal liquidity risk creation as
  proxied by maturity mismatch
• New prudential tool for monetary authorities,
  potentially enables to slow down asset growth
• Requires an internationally coordinated effort to
  avoid regulatory arbitrage

15
Conclusions

• Public liquidity insurance, combined with some
  capital insurance, offers better incentives than
  high capital requirements
• no need for high capital in good times,
• better intervention tools upon a crisis
• may be better supported politically, as it may be
  seen as a prepayment of potential future support.
• As prepayment of intervention costs, it would
  legitimize central bank support in a crisis

16
Prof. dr. Enrico Perotti

• Duisenberg school of finance
• University of Amsterdam
• Oxford University