BAIL-IN RULES under repair
EURO CLEARING in London under threat
BASIS RISK and the FRTB
COMMENTARY: Standardisation risk
Criticism is rising over the Fundamental Review of the Trading Book’s rules on offsets – which would penalise all but perfectly offsetting hedges. The result, many fear, will be what could be called standardisation risk – the market loses diversity as higher prices push end-users towards a small set of standardised products, leading to a more fragile monoculture.
Tighter constraints on internal modelling will have the same result – this means when market conditions move outside the envelope within which these models work, everyone will suddenly experience problems simultaneously. The ‘monoculture’ metaphor refers to a similar problem in farming; if everyone grows the same strain of crop, then a problem, such as an infection or an extreme weather event, will affect everyone at once and cause a famine. Diverse crops guarantee that some at least will survive. Or, alternatively, if everyone uses the same risk model, their investment decisions are likely to be similar too – leading to the classic instability of a tightly linked network.
This problem has come up before, of course. In 2009, credit practitioners warned the industry’s convergence on a single credit risk model could be catastrophic. Insurers in 2011 warned Solvency II regulations – which they said imposed too much conformity in models – could lead to systemic risk across the industry, effectively driving herding behaviour and exacerbating a future market crash. And there were elements of crowding risk in the lead-up to the crisis, as many investors relied on the opinions of the three major credit rating agencies on mortgage-backed structured products.
But the recent argument over the introduction of the standardised measurement approach (SMA) for operational risk exemplifies the problems on both sides. The SMA has come under sustained criticism – not because of crowding risk, but because industry practitioners complain the SMA approach is too insensitive. However, the advanced measurement approach (AMA), which it will replace, had flaws too. Allowing banks to customise their own internal operational risk capital models may have provided more sensitivity, but at the cost of comparability.
Use of a common op risk model won’t cause crowding in the investment sphere, but it could mean a change in the risk environment will go undetected by everyone at once – and when the losses start to come in, every institution will find itself undercapitalised simultaneously, with obvious systemic consequences.
Allow too much latitude and you lose comparability, not to mention opening the way for banks to game the rules; allow too little and you risk inaccurate models and herding risk. The FRTB dispute is only the latest instance of what could fairly be called the central problem of financial risk modelling – one that remains, at present, frustratingly insoluble.
STAT OF THE WEEK
With four weeks left before the start of mandatory variation margining for non-cleared derivatives, banks say they have completed the required paperwork with between 2% and 10% of their counterparties.
QUOTE OF THE WEEK
“My view is that CCPs disclose a lot already. We publish all our margin methodologies on the website, including our external validations and our stress-test methodologies. I have a little trouble seeing how much more information we can put out there which will bring value” – Frederik Ekström, Nasdaq Clearing
The week on Risk.net, December 2–8, 2017Receive this by email