What a complex web we weave. The unveiling yesterday of the International Association of Insurance Supervisors’ (IAIS) Higher Loss Absorbency (HLA) capital buffer for global systemically important insurers (G-SIIs) was certainly an important moment for that world. But it did throw up a lot of questions, many of which are unanswerable at present, if nothing else because there’s three years to run before it kicks in and there’s a ton of definitional work and testing to do ahead of the effective go-live date.
Hence one of my questions is: with key definitions still outstanding around its own work, not to mention key work still taking place in related areas (such as NBNI SIFIs), did the IAIS jump the gun and go out too early with its supplementary capital framework? Given how fluid the outcomes are at present, I guess a broad framework today can’t do any harm. But then again, in the absence of hard lines, you can certainly question the specific utility of yesterday’s pronouncements from a firm-level strategy and capital planning perspective.
Beyond the insurance specifics, though, what struck me generally in this whole debate was how the inter-connectedness between and cross-over with banking, insurance and asset management both at the stand-alone level but particularly within diversified financial groups has rendered the question of broad financial sector solvency and capital quantum and calibration a rather convoluted and complex one.
How should silo regulators treat the non-silo activities of their primary agents? How should the banking activities carried out by G-SIIs be treated for capital purposes by insurance regulators if those banking activities are covered by banking regulations? Should banking activities conducted by G-SIIs be treated by definition as G-SIB activities? Should insurance companies be treated at all in the same way as banks? This all calls for incredibly tight regulatory co-ordination. I’m not sure we have that at present.
Bank/insurance crossover
Regulators do seem to think that banks and insurers share a lot of common ground; the IAIS consultation was replete with examples of where the G-SII process engaged in G-SIB look-across for guidance, equivalence and for the purposes of figuring out what a level playing field might look like. Pertinent to the insurance question, I wonder whether it’s right for the IAIS to be basing its HLA buffer for G-SIIs on the sorry tale of what happened to AIG following the Lehman crisis.
AIG was subject to a US$182bn bailout NOT because it was writing dodgy insurance business but because it was recklessly writing dodgy credit derivatives business. AIG had crossed the line into banking territory but without the requisite understanding of what they were exposing themselves to, and without any proper due diligence or banking regulatory oversight. That, as they say, is history.
My point is: can insurance – as in peddling insurance premiums – truly be systemically dangerous? Given that the already-existing Basic Capital Requirement (BCR) and the HLA will apply both to the insurance and non-insurance activities of insurance companies, policy makers presumably think yes. But should we expect core insurance activities to be assigned capital adequacy standards that reflect its core nature?
The AIG story certainly underlined the approach taken by policy makers on insurance capital adequacy and solvency for the nine G-SIIs – Allianz, AIG, Assicurazioni Generali, Aviva, Axa, MetLife, Ping An, Prudential Financial and Prudential plc. The G-SIIs are clearly unhappy at being given that discriminatory (from a capital and competition perspective) designation in the absence of clear and transparent metrics and the unlevel playing field to which it gives rise.
MetLife, of course, took the issue of its G-SII designation to federal court at the start of the year; a route Prudential Financial considered at one point. General Electric, of course, opted to flog its financial business to swerve the G-SIB/G-SII thing altogether.
On the question of whether insurance business can ever really be systemically dangerous, eight of the nine G-SIIs, writing In response to the June-August IAIS HLA consultation as the ‘Group of Eight G-SIIs’ (i.e. excluding AIG) said:
“the focus for HLA should be on activities that give rise to systemic risk. Specifically, to achieve the stated IAIS and FSB objectives for HLA, it is important that the application of the capital measure is surgically targeted at activities that pose risk to the global financial system. [My italics]
“We would also note, however, that there is no clear understanding of what systemic activity is,” they continue, “and certain G-SIIs fell that the current NTNI definition is not a good proxy for systemic risk.” (NTNI refers to the non-traditional, non-insurance business undertaken by insurance companies).
On a different point, the Group of Eight reckons that future considerations of systemic risk in the sector should not focus on individual insurance groups but instead on activities or risk attributes, whether they are conducted inside or outside of a GSII.
“We have strong reservations over the rationale to designate entities rather than activities as systemic and the current designation methodology, which is heavily influenced by the size of an insurer, the definition of NTNI, and the desire to somehow rank GSIIs against their peers rather than an objective measure of risk and does not consider the size of activities in context of the financial system as a whole.”
An element of talking their own book maybe, but this strikes me as a valid point. The problem with systemically important status is that it’s black and white. You either are or you aren’t. This roils the competitive landscape for reasons that aren’t always immediately clear, certainly vis-à-vis immediate competitors.
Systemic risk reduction
The IAIS says the primary purpose of the HLA, which will apply from January 2019 as a complement to last October’s BCR, is to “help reduce the probability and impact on the financial system of the distress or failure of a G-SII”. But as no-one yet knows which bits of which insurance company needs to be treated in which way, that comment can only be taken in a conceptual sense at this point.
IAIS executive committee chair Felix Hufeld described the capital and solvency requirements for G-SIIs as a “major milestone in its commitment to address risks to the global financial system”. Next stop: a comprehensive group-wide supervisory and regulatory framework for Internationally Active Insurance Groups (IAIGs) a.k.a ComFrame. Can’t wait.
Total insurance capital will have six components. Existing BCR (let’s call it proto-BCR); the 33% BCR uplift announced yesterday (excluding regulated banking non-insurance activities – where banking sector rules kick in); and the HLA will each be calculated separately for insurance and non-insurance activities.
I must say the uplift – expected to be phased in from 2016 in equal measures over three years – looked like an oddity. The IAIS said it was only introduced to reduce the gap regulators expect between proto-BCR and the risk-based global Insurance Capital Standard that will come into effect in late 2019. In that respect it has elements of making sure that the G-SII calibration going forward matches up with the expected ICS outcome working backwards. Messy.
G-SIIs will be assigned to low, mid or high HLA capital buckets based on a number of specific risk exposures. Just like the highest 3.5% G-SIB supplementary capital bucket is empty as a stark reminder that big isn’t always beautiful from the regulators’ perspective; the IAIS reckons the high insurance bucket will similarly initially remain unpopulated as a disincentive for G-SIIs to increase their systemic importance.
It’s taken two years to get to this point and we’re far from an end-point. The HLA public consultation ran from June until August 2015 but the assessment methodology and policy measures for G-SIIs came out as far back as July 2013. And we’ve still got two more major public consultations ahead of us around the not-insignificant issue of how NTNI activities are defined. Where we end up on these and coming up with workable definitions will be key to making this whole construct a success.
We’re not going to get any quick answers. Field testing as part of the review and refinement process will continue until 2018. And bearing in mind the related work the FSB is conducting on NBNI SIFIs is yet to be concluded, I wonder whether better co-ordination is called for here. “If more appropriate globally comparable regulatory capital requirements are implemented, the IAIS will review them and consider their application to the BCR and HLA”, the IAIS said in its responses to the consultation. I rest my case.
I guess they felt forced into going now because the insurance HLA proposal is supposed to be going in the same packet as the bank TLAC blueprint to Antalya for ratification by the G20 in November.
