Is the delay in full implementation of Solvency II a boon or a bugbear? What are the implications for the regulators, as well as the regulated? CROs and Solvency II experts debated these and other issues last month in an InsuranceERM roundtable, sponsored by SAS.
Greg Gould, Chief Risk Officer, RSA UK
David Gulland, Chief Risk Officer, MGM Advantage
Simon Kirby, Consultant, SAS UK
Andy Sharpe, Financial Risk Director and Lead on pillar 2 for Solvency II, Prudential
Penny Shaw, Head of Risk Management and Capital Planning and Solvency II Lead, ACE European Group
Dave Smith, Head of Actuarial Development, LV=
Peter Field, Partner, InsuranceERM (chairman)
Note: what follows is the first part of the discussion; read part two (Solvency II roundtable: embedding processes and live testing are benefits of "extra" year)
Peter Field: Is there really an extra year to prepare for Solvency II?
Dave Smith: It is not really an extra year in terms of the IMAP [internal model approval process] and actually getting to full compliance. We have an extra six months in a sense, because the IMAP date was April and now it is October. But we believe we would have been ready.
We are not relieved as such that we have an extra six months, because I think we did a lot of the heavy lifting last year.
We are going to be using the extra year, or whatever it is, to do more embedding. I think that is the bit that is not going to be fully ready for October: to actually demonstrate full embedding across the whole business from top to bottom.
Simon Kirby: Have all of you been given your date now from the FSA for your entry into the IMAP process?
David Gulland: We are slightly different because we are going down the standard formula route. We had planned to go with an internal model but we did a cost-benefit analysis, and given the time and resources available and the nature of our organisation, we changed tack during last year. For us the extra time helps us to actually increase the training that is needed across the company and at the board level as well.
Andy Sharpe: There is the extra year out there but it certainly has not meant an extra year for us. Our IMAP went back six months, so the FSA gave us some time, but you could not really have expected them to give everybody that full extra year because their review time cannot just be crunched.
Probably it gives you a longer period to engage with the FSA between now and actually submitting your IMAP, but also, during that IMAP process, while they are looking at it and coming back to you.
Penny Shaw: We were one of the first to get the PAQC [pre-application qualifying criteria] back in April last year.
We were expecting a work plan within six to eight weeks; that took until October. All that time was actually spent trying to set out what a submission looked like from the contents of application.
At that point, Lloyd's really drove the whole process forward, putting a very clear deliverable that is very numbers-focused so you really knew what it actually meant from a numbers standpoint. That was much more engaging in getting the board and executive management involved.
I think the FSA are catching up now. They are talking about what the final submission is going to look like with indications that this might take on a bit more of a Lloyd's-type final pack, where you do a self-assessment against the six tests and show what you are doing to get compliance against them.
Simon Kirby: The problem was that the FSA had no basis in law to give you formal approval if they had been able to do all the IMAP reviews before the October deadline, which is the date that was set originally.
They would have had to give you a letter which said, "We are of a mind to approve your internal model, as long as you make no changes between now and the go-live date of Solvency II, at which point we will write to you again and give you a formal approval".
What the new timetable does is give them the ability, and the basis in law, to conduct the approval, because the go-live date does not change for them; it only changes for insurance companies.
So actually bifurcation is really more beneficial for the regulator than it is for the insurance companies.
Greg Gould: I cannot imagine many insurance companies baulking at the idea of receiving a letter from the FSA saying, "We are minded." I think we would all be quite comfortable with that.
Does it give us another year? For us it has been pushed back four months. In the last week we have been informed again that the FSA is under further pressure and they expect it to move back even further, which just condenses the period for agreeing with the FSA any nuances or any changes they need us to make post-IMAP, or as part of that process, and actually putting it live.
As soon as we go through IMAP, we have to put the model in and use it. On 1 January 2014 we are expecting all our boards and all our subsidiary companies, not just the single legal entity, to use it. They will have to get very used to dealing with the output of our internal model and using it in earnest to make business decisions. Condensing that process will cause problems.
Penny Shaw: It does have the effect of creating another year of cost. The work fills the space of time that you have. I think that is one of the downsides.
David Gulland: We are actually taking the delay as a chance to reduce some of the contractual spend, because now that we have a bit more time, we are looking quite closely at where money has been spent and saying, "Well, actually, can we use our internal permanent staff?"
So I am trying to make sure that we do not actually carry on using contractors at the current rate of spend.
Greg Gould: I think this is one of the differences between the internal model and standard formula. With the standard formula I would agree that is the approach to take.
However, once you have had your internal contractual resource helping you build your internal model, you really do want them helping you through the IMAP process.
Penny Shaw: If you go back, a lot of firms were not given much choice about the internal model vs the standard formula.
David Gulland: For us, the difference in going with the standard formula compared with an internal model is the costs.
Peter Field: So it sounds like the FSA is actually trying to catch up as much as, say, some of the European regulators who have apparently been in favour of the delay?
Simon Kirby: Other regulators in Europe took an approach that said, "We recognise the resources that we have to conduct internal model approvals, and we will restrict the number of applications to those firms we know we can approve".
If I understand correctly, BaFin, the German regulator, restricted their internal model approval to half a dozen or so firms, because that is all they could actually resource.
The FSA took a different approach. They said to the industry, "Tell us if you want internal model approval, and we will then try to resource up".
That has been very difficult in terms of getting the people (and I was personally involved in the recruitment process), and then trying to retain them, because they drain the FSA of everything they can learn, and then of course they go out to industry with the FSA on their CV and earn a great deal more.
Penny Shaw: If you look at how the Swiss regulator [FINMA] approves internal models, you essentially have an external reviewer and FINMA set the remit. That then goes into FINMA for approval.
I think the FSA has started that to a degree with the data audit. Now they are going to rely more so on the data audit than conducting their own detailed review, essentially looking for that to give a clean bill of health on internal model readiness of the data in order for them to give approval. I think that is a more effective use of their time. This is more challenging for the firm, provided the audit is conducted properly.
Peter Field: Is FINMA a good example, because have they haven't actually approved any models yet?
Penny Shaw: We have dealt with FINMA on ACE's re-domestication to Switzerland. We did the SST [Swiss Solvency Test], and have some experience of that. I would say it is very transparent what you need to do.
The approach, and how to engage with them, is very clear, and the rules were out there two or three years before they actually came into effect. The focus there is very much on capital and risk transfer instruments, so you really have to get the legal-entity dimension correct, for example, how your intra-group reinsurances, dividend payments and capital fungibility works.
Andy Sharpe: We all talk about the IMAP and say we feel as though 95% of people at the FSA are there to do IMAP. But actually they still have the current regulatory regime to deal with. Our supervisory team is already spending a lot of time, for example, reviewing our ICA from last year and giving us challenges around that. They have to approve all of that, plus they have IMAP.
Penny Shaw: And they have the internal split between the two.
Andy Sharpe: They have their own jobs to think about as well. It must be a difficult place to be at the moment.
Peter Field: Would the extra time give the FSA more chance to go back and ask you to do things again or to revise things you have already done?
Penny Shaw: We were already quite a way through our feedback with them and we have had a number of walkthrough sessions. Their approach has been very detailed. The depth of questioning is quite intensive. That is slightly different to the Lloyd's approach, which has been more a case of getting you to self-evidence, and then they will come in and look at specific things. That is a very efficient way of doing it.
Whilst it was intense and there is a huge amount you need to do, at the end of it you had to lay out very clearly what your gaps were and how you were addressing them.
Peter Field: Are insurance firms tempted to go back and revise some of what they have done already because they have the extra time?
Dave Smith: We are not quite in the position of revision just yet! We are having external validation assurance. We are trying to tease out what may be the elephant in the room that we have missed. Not that we think there are any.
We have the models in place, we can do the numbers; it is just getting that end bit. My big concern is that, at the point of IMAP, we would not be able to say, hand on heart, "We will be 100% compliant, we have been using this stuff for three to six months".
It comes down to whether they will be doing that amazing FSA thing where they pour over everything in minute detail, irrespective of whether it is relevant or not. Principles-based, with random deep dives, as you say Penny, seems a lot more sensible.
Simon Kirby: When I was a supervisor at the FSA, we saw people who would literally just sit down once a year or once every two years and create their ICA. And then that was it; the process was over. Then they carried on running the business and one year or two years later, they did it again.
They would literally pull the individuals from claims, actuarial, etc, put them in a virtual team for two or three months, do all of the modelling, produce all the documentation, then run it into the FSA and pray it was going to be all right.
There is this concern about embedding that means you actually have to embed right the way down through the deepest structures of the organisation. I don't think the FSA's view is that at all. It is more that you create a process that actually works throughout the year. As long as it is used at the highest echelons of the business, regularly, that is probably all right.
Penny Shaw: It is a good point, because it has meant that our company is actually more aligned with what we were doing for the syndicate, which was to look at capital before going into the next year.
David Gulland: We are going to try to make the capital work more than just an annual planning routine. We definitely see it as a living thing that is going to get tried and tested, and we are building another dashboard for the board.
Andy Sharpe: That sounds like one of the real benefits of the delay. If you take our current solvency metrics, during end of 2008 we were doing an approximate daily evaluation. It was not absolutely correct, but it was near enough to tell us whether we actually had to open up the can and have a better look.
When you think of how familiar we are with the old metrics, we can do that. We can pump a few bond yields and equity values in and get near enough to tell you the appropriate solvency position. You probably cannot do that sort of thing yet with the new metrics.
Penny Shaw: With pillar 2 I think the value is when you actually look at everything you have in play already. You bring that together, put a more executive focus on that and try to show what things are aligned to the model but above appetite. That is the value of Solvency II.
Moving to a discounted best-estimate technical provisions and running a big model to determine the SCR is not that different to the ICA.
At ACE we built a global capital model that we can now use at the European entities level, and that is really good for ACE because we get all the intra-group reinsurance modelled and that promotes wide model usage, as well as consistency across the group. But it is a big model. When it comes down to day-to-day business decisions you have to also build models which are connected to the global capital model but that can be run real-time and have more granularity as required to support certain business uses.
Peter Field: Are you going to run ICA and SCR calculations in parallel in 2013?
Dave Smith: We are still under Solvency I right now, so we still have to do ICAs and all that horrible peak one/peak two stuff. Even though there are going to be some technical differences between the two, we still intend to do a full ICA calculation and a full Solvency II balance sheet. We have to do the full Solvency II for IMAP. It is a question of trying to align the timelines.
Penny Shaw: Do you think there is value in doing both?
Dave Smith: We are not in our submission year, but the ICA is an ongoing continuous requirement. I assume not doing an ICA of some sort would be frowned upon by the FSA.
Penny Shaw: Lloyd's has said -- though they have got to agree this with the FSA - you will not need to do an ICA and you can run the SCR as long as you have the proxy. The model we have built can do both, but a huge amount of time is needed to load it up with ICA information.
Dave Smith: We are not one of the top 10 companies that the FSA is focused on and have no information back from them yet saying that kind of thing. Is the FSA advice that they are happy with an ICA-light approach this year?
David Gulland: I think it depends where you are. Julian Adams [director of the FSA's insurance division] has been very clear that he expects people like us on the standard formula to be concentrating on the ICA. This fits in well with what we are planning anyway, because we regard that as being a core part of our pillar 2 ORSA [own risk and solvency assessment] in the future.
Penny Shaw: So you intend to keep the ICA model running, even though you are on the standard formula?
David Gulland: Yes, absolutely. That will merge into the ORSA.
Peter Field: Could the Euro crisis derail Solvency II?
Andy Sharpe: One of the interesting things for parallel running is going to be what happens in the financial markets over the next six to eight months, because we are still actually run by Solvency I rules, not Solvency II.
If you have turbulent and falling financial markets, then, quite frankly, that is when you will need to do an absolutely full-works parallel run because you will want to manage the company on metrics you are familiar with, and try to keep an eye on what is going to happen with the other ones.
Penny Shaw: This is certainly in the present scenario that we are running on the Euro. Only so much can be done in the European parliament at any one time. Solvency II has a question mark: does it delay it further?