Sunday, 3 April 2011

Irish QIS5 results - par for the course?

Our good friends at the Central Bank did a lot more than scaring the bejeezus out of the senior bondholders of the remaining banks (although as an ardent follower of the 'number 1', I will resist any jokes about severe haircuts) - they also released the Irish-specific QIS5 results.

On the face of it it seemed par for the course compared to the EU-wide exercise - however, they actually fared better on participation (80% ish against 70%). That's where the good news ends, as there were more non-SCR compliant (20% against 15%) and the same for non-MCR compliant (both 5%).

I noted the following on the way through as items of particular interest;
  • Life companies in general saw surplus capital rise, but overall coverage fall (life co participation was 87% in isolation). This is because generally Own Funds rise, but so does required capital (the denominator), and the dynamic in that change determines the extent of benefit/downside.
  • Over 30% of life cos were over 4x covered! 17% are in the zone of "tight or under" SCR
  • Reinsurers and capitives still getting the thin end of the wedge
  • Complaints about complexity in the counterparty default and non-life cat modules, non-life underwriting risk calibrations, as well as the risk margin element of the technical provisions calculation
  • Contract boundary and negative technical provisions feature extensively throughout, and I will look specifically at this (again, to be an actuary!).
  •  A straw poll on data reliability was taken, and the areas of least comfort were in the calcs for Risk Margin, USPs and non-life underwriting risk
  • In the preparedness stakes, 35% of respondents had a problem with SCR calc data, and another 35% with the SCR methodology
  • Technical provisions were down in general (prudent assumptions, discounting for non-life business and negative provisions mooted as reasons, with Risk Margin somewhat offsetting).
  • Contract boundaries determined at zero for unit linked contracts (causing SP and RP inconsistency)
  • "Overwhelming" view that QIS5 definition was out of line with IFRS, uneconomic, unrealistic and not reflective of the risk profile of the contract [very strong words, which sounds like the SAI gave it some thought!]
  • Illiquidity premium approved of in theory, but poorly executed in QIS5
  • Most companies used one of the simplifications for the Risk Margin, as the "sums were too hard" (paraphrasing slightly!)
  • Expected Profits in Future Premiums causing big changes to Own Funds
The rest of the doc is heavily internal model influenced, and while I like the look of it from a scenario challenge perspective, I'll save it for another day.

No comments:

Post a Comment