Solvency II assumptions and realities

Gideon Benari editor Solvency II Wire 2

All the talk this year has been about reporting. Surveys, press releases, events and tweets galore fill the Solvency-II-sphere, like birds and bees in early spring. In truth, not much is new here. Most of the problems for asset reporting were identified and articulated in some detail two or three years ago, around the time the first EIOPA consultations on the Quantitative Reporting Templates (QRTs) surfaced.

Many firms say they are well ahead in their Pillar I work, with Pillar II not too far behind. Although – with the exception of the Dutch and some of the Nordic markets – it is not clear how many ORSAs have been done and how many ORSA reports regulators have seen. Pillar III, the “forgotten pillar” as I have called it before, is now having its day in the sun. Projections of firms’ (and again regulators’) readiness for the 1 January 2016 implementation date are assured.

But there is, to quote Colombo, “just one more thing”. Much of this confidence is based on assumptions and projections of the amount of work required to get the Pillar III asset data. Both may be very (very) far off the mark.

Firms have been running some trials of SCR and internal model calculations, of ORSAs and variations of ORSA reports (and in some cases even quantitative reports both public and private). These are based in part on actual data, but by and large on estimations and proxies. The overarching assumption is that the “real” data can be easily substituted to do the “real thing” when the time comes. This is a problem – a very (very) big problem.

I have been roaming the Solvency-II-sphere for over three years now. Solvency II Wire has reported on look-through and its challenges in great detail since 2011, but I have yet to meet an insurer, or anyone who knows of an insurer, that has completed their full look-through reporting exercise and can replicate it at the frequency and timescales required by Solvency II.

All the people I have spoken to in the various associations on both the insurance and asset management side of the Pillar III road haven’t been able to tell me about a single insurer they know that has completed their look-through to anything close to what Solvency II will require either. Not even in Denmark or the UK. One even burst out laughing when I asked him … it seemed more out of despair than anything else.

The problem is that getting the ‘real data’ is proving a lot more difficult than anyone anticipated. Those who have been actively working on providing look-through solutions are fast discovering the enormity of the task – possibly greater than what was anticipated and agitated by the worst doomsayers (regardless of their motives).
What takes all of this to the next level of complexity (or panic?) is that Solvency II requires data consistency across all three pillars. In other words, the data used in Pillar III must be the same data used for the capital calculations and the ORSA. There should be only “one truth” when it comes to the numbers.

Listening to the media buzz, you could be bamboozled into thinking this is the only big challenge running up to implementation. Two things stack up against this view.

First, narrative reporting is potentially no less of a challenge. Talk to insurers who have completed an ORSA report and they will tell you how tricky it can be to describe what you found in the data or to define those unquantifiable risks (such as reputational risk). A challenge amplified by the requirement to make much of this information public. The difference between the two is that QRT cells will only accept one value that must be reconciled and validated while text affords you a lot more flexibility.

Second, as supervisors are quick to point out, quantitative data reporting is as much about data governance as it is about getting the data. Putting the right number in the right box is important, but the supervisor will also want to know where that number came from and may well ask you to demonstrate that you understand it and what assumptions you used to choose it over another: something no formula or XBRL tool can do for you.

Thought should also be given to insurers’ state of mind. I get a strong sense that something happened after the slowdown at the end of 2012 when political uncertainty re-emerged. Many firms shed external consultants and carried on with Solvency II work internally, while projects were re-evaluated on a business case basis. The experience may have been sobering. Now they look to take on the reporting workload and apply the same criteria to any new projects while the memory is fresh in their minds. If you consider the need for governance and understanding the data, a better picture of what is going on may emerge.

Given all this, assessing the state of readiness in the Solvency-II-sphere could be tricky. My guess is that at some point, probably towards the end of the year, the chasm between the assumptions and reality will become apparent, derailing workplans based on rosy assumptions. Meanwhile if you’re an insurer, or know of an insurer that has completed the full look-through exercise – call me.


Gideon Benari is the editor of Solvency II Wire.