PRA - Q&A from December's Industry Briefing

Following on from the PRA's Industry Briefing on Solvency II prior to Christmas, they have released a set of Questions and Answers, which appear to have prompted by questions asked on the day as opposed to being more generic. A useful document to push around your programmes in the UK in particular, even if you are not part of ICAS+/IMAP.

I was particularly interested to see if they had rowed back on any topics where they have previously expressed an opinion, as well as any new insights which aren't necessarily made public as a matter of course. The Q&A is broken out into topics, so I have grouped my thoughts accordingly.

Proportionality

• For ICAS+, the PRA are "...focusing in on those areas which move the dial the most", referring specifically to the ICA calculations.
• They will therefore apply "a relatively higher view" in some areas than others (later referred to in terms of their intensity of focus).
• They use their IMAP approach as an example of this, where they have grouped the 300+ requirements into 15 categories.
• My common-sense read of this is, to be proportionate in the above context, the first 7 consecutive categories on the IMAP template would receive less time and attention than the following 8, all of which have the potential to numerically "move the dial".
• They are already looking prospectively at firm's risk profiles during planning sessions, in advance of ORSA, so that they work and plan proportionally, but with the future in mind.
• They suggest that firms provide their own opinions on materiality if they wish for the favourable application of the proportionality principle (though they don't guarantee it!).

Documentation-specific proportionality

• Confirm that "...there is definitely a need for an improvement in the quality of documentation".
• Seem to specifically align the calibration requirements of Solvency II with the need for additional documentation, therefore a good area to be elaborate in your work.
• On expert judgement, they suggest that documenting why a judgement was made, and what were the alternatives will help firms currently falling short.
• "...no aspiration" to revisit documentation review work already conducted and concluded.

PRA's Supervisory Statement

• They confirm that letters confirming whether a firm is covered by the reporting guidelines will be sent this month. They suggest categories 1-3 (p20) will be the bulk of participants
• There are national-specific reporting templates to come, currently under consideration. While they are not certain on submission format for those, they reiterate that xBRL is required for QRT submission.
• QRTs can be submitted incomplete, but they reserve the right to intervene if you take too many liberties!
• Acknowledge that, due to most firms' ORSA Processes operating around year-end balance sheets, that they expect to receive a flood of ORSA Supervisory Reports, and that this will delay feedback (presumably worse in 2015 than this year?).
• A question regarding 'who is the ORSA for' was answered rather dismissively. However, bearing in mind the content of their "good and bad documentation" letter (which constantly refers to "the reader", which for me meant "the PRA"), it felt like a reasonable question. Clearly if the PRA do not like your style, it will add to your review time.
• They seem to indicate that the amount of time one's Board should spend on reviewing/using their ORSA is not subject to any periodic expectations. That means if you are planning for annual intensive interaction, or something more regular and fluid, you should get a fair hearing from the PRA.

Standard Formula

• Without much subtlety, they seem to be encouraging firms to look at the appropriateness of Standard Formula asap, encouraging further by stressing "...that internal models do not need to be overly complex". This feels like early positioning for their planned SF work with the industry, where no doubt they will find a few firms ripe for USPs/PIMs during the next 18 months. 

Early Warning Indicators (EWIs)

• Following on from a post in October, they are evidently still stuggling to calibrate them!
• They confirm that the purpose of automatic capital add-ons (in the event of a future EWI breach) is to freeze the modelled capital figure to prevent distribution

Trilogue

• Question 29 is a good example of the PRA showing where their hands are now tied, and where EIOPA will be the provider of guidance.
• Interesting confirmation that the Commission has apparently asked that as little as possible of the Draft Implementing Measures be changed. Those of us working on non-mathematical areas can perhaps take some comfort from that!

PRA and Early Warning Indicators - Knowing me, knowing EWI...

Couple of interesting bits have popped up in the last week on the PRA's Early Warning Indicator (EWI) initiative, which I had touched on earlier in the year. At that time it was a little hard to establish the lie of the land, other than the PRA approach to preventing potential 'gaming' of internal model capital requirements was markedly different to that of their European NCA counterparts, and that EIOPA fancied their chances of doing the job on the NCA's collective behalf.

Julian Adams gave a speech at the Insurance Institute of London last week where he commenced with a rather familiar lament about how, at the outset of the FSA/BoE splice, "...we knew that there was some concern in the insurance industry that the Bank didn’t understand insurance". After collecting the award for 'understatement of the year', he then waxed lyrical about the five lessons learned by the BoE from the financial crisis, and how these are being applied currently to insurance supervision in the UK;

1. "...the need for a prudential supervisor to focus on the risks that might materialise for firms in the future"
2. "...not enough to focus on static point-in-time assessments of firms’ solvency positions"
3. "...for some issues it is not enough to just look at the risks within individual firms and address them at an individual firm level. Instead, we need to supplement this analysis with work across the system as a whole"
4. "...the need for clarity about the objectives of prudential regulation"
5. "...the need to have multiple reference points when assessing firms’ capital positions. Simple crude measures are not sufficient in themselves, but neither are complex models"

The last one is of course our context setter for comments later in the speech justifying the EWI approach, which for me reads as a laundry list of why existing model validation activity, assumption setting processes and expert judgements must be considered ineffective by default, rather than after PRA review.

One brush - two industries?

Most significantly from a strategic perspective is the confirmation that the PRA "...wouldn’t expect firms who fall beneath [the EWI limits] to release capital until we, and they, have reviewed the appropriateness of the modelled calibration".

Alongside this speech, Chris Finney has published a scathing article on the PRA's approach to EWIs, focusing on its legality at both national and EU level, its unwillingness to consult on the matter, and ultimately how the PRA's experiences with banks a few years back (well. the FSA/BoE, but let's not split hairs) have seemingly allowed the insurance industry to be smeared with the same tarry brush.

With the PRA seemingly using their "lessons learned" rationale alongside the equally informal "interest" of the IAIS and EIOPA in non-modelled capital backstops, is that really enough to fend off (current) UK statute and (pending) EU law if the industry fights back?

More from EIOPA and PRA on internal models - anything you can do...

Certainly looks like there is a bit more mileage in this back-and-forth around internal models and the assessing of their appropriateness. As I posted recently, the UK's PRA have come out fighting on the subject of 'Early Warning Indicators' (EWIs) to highlight where a company's internal model, for fair reasons or foul, may be measuring the Solvency Capital Requirement in a way which does not meet the required calibration standard of the 1-year VaR at 99.5th percentile.

Early Warnings - time well spent?

EIOPA, through the words of their Chairman, have already made a pitch for assessing internal models on behalf of our overworked national supervisors, at least at some point in the medium-term future. It would now appear they have also piggy-backed the PRA's work around developing and monitoring EWIs as well!

That article is on Risk.net, so for non-subscribers, EIOPAs work is specifically looking at (in the author's words, not EIOPA's) "...threats to a firms solvency that are not picked up by the company's internal model". These will include qualitative and quantitative measures, and are due to be discussed (by their internal models committee) in the Autumn -  you know, that relatively quiet point in the Solvency II implementation calendar...

The EWI article was published at the same time as an interview with Sr. Bernadino and Insurance Risk magazine was released, where the tone of the interview is particularly aggressive around EIOPA's inability to compel national supervisors to "comply" with their recently-issued preparatory guidance. Whilst it is hard not to admire his optimism ("we are moving closer to the date of implementation"), he is quite fixated with "some countries" who, without EIOPA's guiding hand, he feels would end up divergent from the rest of the market.

I certainly am not familiar enough with continental Europe's positions as it stands, but it certainly reads like a not-too-subtle finger point at Royaume-Uni at the very least. He also goes on to mention that, while 'full Board support' was received at EIOPA to issue the preparatory guidance, not all participants were 'completely happy'. One would certainly feel aggrieved at the smaller end of insurance supervisors to comply-or-explain on preparations for a new regime which doesn't have a kick-off date, while still having to administrate the existing one!

An interesting side-issue around potential discord amongst the states is this recent request (which I have procured through the magic of Google!) from most of the Member States for a second "quick-fix" Directive, based on the fact that some of the implementation countdown dates that feature in the first one are next month (good spot Gideon)! According to the letter, the Commission already appear to have singularly ruled out the need for another quick-fix,

Four countries didn't make it onto the list of senders (Ireland, Netherlands, Malta and Lithuania) - not concerned, or not consulted?

Also worth highlighting the recent spectacularly ill-timed London press strafing of the internal modelling world - an strange development, particularly at the same time that the UK Independence Party has been lifting up its dress for London's financial sector to have a gander underneath. With some of the UK's bigger beasts such as L&G and Pru already openly hostile to Solvency II (and Resolution being an early faller in the IMAP stakes), is there potential for the rinsed-out UK insurance sector to throw its weight behind a political force such as UKIP, who would happily make an EU Directive bonfire, using Solvency II for kindling?

EIOPA and PRA - model appropriateness, expanding remits and early warnings

Another short flurry of activity recently on the Solvency II front ahead of the day (well, 'month') of reckoning for the Long Term Guarantees assessment. The industry media is certainly very chipper around the prospects of a deal being done off the back of the LTGA (here and here), highlighting that some German products may be the beneficiaries of this new found camaraderie (here).

German products - carved out from 

LTG requirements?

EIOPA themselves have been vocal on power extension this week, with Sr Bernadino making a few waves in his submission to a public hearing on Financial Supervision in the EU. Apart from his Partridge-esque diatribe about 'evolution, not revolution', he also highlighted what he considers enhancements to the existing supervisory structure in the EU, namely:

1. Strengthening EIOPA's operational independence - effectively through a change in its funding arrangements, going as far a potentially levying the industry direct, and also by having more money in any case (referred to cheekily as 'budgetary flexibility'!).
2. Reinforcing its existing 'independent challenge' role - by securing access to national supervisors' QRT data and allowing EIOPA to conduct EU-wide investigations of conduct-related issues (effectively an FCA for Europe!)
3. Enhancing both its mandate and powers - perhaps most controversially, fishing for centralised oversight of internal models, as well as powers to ban or restrict activities in member states.

One might say to EIOPA 'don't walk before you can toddle', but I guess if we are serious about operating a single market, the UK's consumers shouldn't need to rely on generally being in the vanguard on these matters (both producing nefarious financial products, then banning them and recouping the profits for compensation!)

Back to the UK's national regulator, the PRA dropped a few pearls of administrative agony for the insurance industry this week, with a couple of 'dear CEO' letters which were part-briefing and part-data request, ultimately driven by the UK's aggressive take on assessing internal model adequacy (i.e the same activity which EIOPA wishes to expropriate from national hands!). 

The purpose of the letters is nicely summarised by Chris Finney here, so I only need to highlight a couple of elements for my own interest;

Solvency II update letter

• "Unlikely to be any certainty" on timetable before autumn
• "Just under half" of IMAP candidates have applied to participate in ICAS+
• They are hoping to share learnings from ICAS+, "particularly developments used by firms in their modelling techniques" - danger here of the early birds determining what's hot and what's not in the world of assumptions/calibration/expert judgement/documentation for those not participating in ICAS+?
• Highlight pension risk as one area where the standard formula is potentially not suitable
• Next industry briefing forecast for late November (post-Omnibus II ratification?) - that's what I call 'scaling back' on costs!

Early Warning Indicators letter - remember here that the PRA's plans are potentially at odds with EIOPA's, to the extent that the PRA are already braced for some kind of legal challenge

• EWI's aim to test calibrations of internal models as well as "monitor any downward drift in capital" - presumably just quantity for the latter?
• Ratios being monitored are of pre-corridor MCRs (as illustrated in firm's LTGA submissions earlier this year) against current Individual Capital Guidance - once we go live, this is likely to be replaced by modelled SCR
• Special treatment for With-profits business to account for the fineries of that sector (cost of guarantees, level of free assets and the proportion of non-profits written in the book)
• Ratios are deliberately set so that 10% of affected firms will fall below
• Information required to conduct this work is covered by the data collection exercise below

Data collection exercise letter

• Fishing for data from all internal model applicants using YE2012 balance sheets (unless you can excuse using earlier data) which covers standard formula SCR, Internal Model SCR and ICAS by end of July.
• Also asking Life firms for key percentiles from distributions for 'risk variables' - one assumes this means each one of the risk drivers in one's SCR calculation.
• Conducting what seems to be peer review work around credit stresses and stochastic simulation files (for anyone using them).

With EIOPA and the PRA both seemingly interested in being top dog in the world of assessing model appropriateness, it looks like we might need a walk-off...

Judging model appropriateness - EIOPA or PRA?