NSFR problems, LEI delays and how to fix op risk

The week on Risk.net, 25 November - 1 December, 2017

7 days in 60 seconds
Infopro montage

Basel concession strengthens US opposition to NSFR

Lobbyists say change to gross derivatives liabilities measure shows whole ratio is flawed

Esma downplays fears of LEI logjam

Applications have risen sharply in recent months, but concerns remain Asia is lagging behind other jurisdictions

How to save op risk modelling

Drop loss categories and correlations and adopt simple loss distribution, advises AMA expert


COMMENTARY: A different classification

The advanced measurement approach (AMA) for calculating operational risk capital is approaching its sunset, but the planned replacement, the standardised measurement approach (SMA), which will also supersede the two basic approaches to op risk capital, is still far from universally accepted. Critics argue with some justification that it is over-simplified and too backward-looking; this week Risk.net published two suggestions for how to improve matters.

Writing in the Journal of Operational Risk, Fabio Piacenza and Claudia Belloni propose a way to solve one of the SMA’s most egregious flaws, its insensitivity to insurance coverage, using the same operational capital-at-risk model as the SMA itself. Meanwhile, Ruben Cohen argues that a heavily revised AMA still looks like a superior alternative to the SMA. In particular, he favours one that abandons the AMA’s dependence on the seven categories of operational risk laid down in the Basel II rules, and classifies losses simply as conduct or non-conduct, with different approaches to modelling each.

Cohen (as he acknowledges) is far from the first to suggest ditching the Basel event categories. In fact, JP Morgan’s chief risk officer, Ashley Bacon, is moving one stage further and sidelining the categories of market, credit and operational risk in favour of a home-grown six-bucket approach.

There are two very good reasons for thinking it’s time to rethink how we categorise op risk. Cohen’s argument is based on modelling: if several categories of event all follow the same distribution, what’s the point in imposing artificial distinctions during the modelling process? In this approach, the event categorisation drops out of the modelling process. Bacon’s approach, on the other hand, is based on root cause; risks resulting from adverse economic conditions all fall into the same bucket, whether they’re losses due to falling asset values or customer defaults on loans.

From a modelling perspective, one is superior; from the point of view of a risk manager trying to reduce or avoid exposure, the other is definitely more practical. Enforcing either one by means of capital calculation regimes looks like a lost cause, but at least losing the AMA will mean op risk modellers are no longer straitjacketed into a system of event categories that is increasingly accepted to be inappropriate.



Despite reducing their exposure since the June 2016 referendum on EU membership, EU countries had a total asset exposure of €1.585 trillion ($1.9 trillion) and a total liability exposure of €1.335 trillion towards the UK in June 2017



“I started my career at the best investment bank in the world – though some might argue with that – and ended up at the most famous investment bank in the world – no-one would argue with that” – Xavier Rolet, LSE, whose banking career started at Goldman Sachs in 1984 and ended at Lehman Brothers in 2008

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here