CORRECTED-Regulators to review bank capital rules "labyrinth"

Thu Jan 31, 2013 1:08pm EST

(Corrects RWA reduction figure in para 18 To 55 bln euros from 47 billion euros)

* Basel to test bank risk-weightings again

* Banks could face curbs on use of in-house risk models

* Deutsche Bank defends "optimised" risk weightings

By Huw Jones

LONDON, Jan 31 (Reuters) - Global regulators will review how their rules are used by banks to determine the size of capital buffers after an initial probe showed variations big enough to confuse investors.

The world's biggest banks use their own models to tot up risks, sparking calls from British and U.S. regulators for a simpler system that does not rely on lenders' own arithmetic.

Last November, the Bank of England called on UK regulators to take action, arguing banks' capital positions could be overstated by not adjusting the value of assets properly to take into account risk, making it harder to restore investor confidence in the sector.

The Basel Committee, made up of regulators from nearly 30 countries, on Thursday published its study of how 16 top banks use its rules to add up risks on trading books. It revealed some of the main findings last week.

Basel said it found considerable variations, and the explanations from banks were of insufficient quality for investors to make informed assessments.

The ratio of market risk-weighted assets to total trading assets, seen as an average risk weight, ranged from a low of 10 percent at BNP Paribas to 80 percent at UniCredit.

Faced with such a range and pressure from U.S. and British regulators for simpler rules, the committee said it would now carry out a further test and consider rule changes.

"This will include other, more complex, hypothetical test portfolios, with the aim of helping the committee to deepen its analysis of the variation in risk measurement of trading books across banks," the committee said in a statement.

A sizeable portion of the differences were due to some national supervisors being tougher and limiting modelling options.

Basel said the way banks set in-house models lay behind some of the variations in risk weightings uncovered by its study.

The committee was already testing the use of risk weights in banking books.

Reforms could include curbs on model choices, requiring better public disclosure on models, and more harmonisation of supervision.

Some policymakers want more radical changes, but these could come after Basel's fundamental review of bank trading books is completed over the next year or so.

LABYRINTH

A Bank of Italy paper last year described Basel's risk-weighting system as a labyrinth and offered advice on how not to get lost.

The Bank of England says the rules are complex and opaque, and differences in how banks add up risks have also sparked complaints among the lenders themselves.

JPMorgan Chief Executive Jamie Dimon complained in 2011 that rivals were gaming the system by attaching aggressively low risk weights to assets, in comments thought to refer to European lenders.

Deutsche Bank's chief financial officer Stefan Krause on Thursday mounted a staunch defence of how his bank "optimised" risk weightings, adding that he understood the concern over the lack of transparency regarding the models.

A 55 billion euro ($74.7 billion) reduction in risk-weighted assets in the last quarter helped the bank report a better than expected core tier one capital ratio of 8.0 percent - and put off having to raise expensive capital.

"We will see these tests coming in terms of our RWA comparability and I can tell you that we are very comfortable with the results," Krause told a conference call, adding that only a quarter of Deutsche's reduction was due to model changes.

Basel's study showed Deutsche's market risk-weighted assets at the end of 2011 were mid-range at just over 30 percent

Espirito Santo analyst Andrew Lim said on Thursday the bank's model tweaks would be viewed by markets as "low quality". ($1 = 0.7367 euros) (Additional reporting by Steve Slater and Laura Noonan; Editing by Helen Massy-Beresford)

Comments (0)
This discussion is now closed. We welcome comments on our articles for a limited period after their publication.