Mr P. G. Scott, F.I.A. (Chairman): I would like to invite Mr Kemp & Mr O'Malley to introduce the paper.

Mr M.H.D. Kemp, F.I.A. and Mr P. M. O'Malley, F.I.A. introduced the paper. The full text of their introduction is given at the Edinburgh discussion on 21 January 2013.

Mr A. J. M. Chamberlain, F.I.A. (opening the discussion): In paragraph 4, the authors describe their objectives and the parameters that they set themselves. While the designers of Basel III and Solvency II might dispute it, I think the authors are justified in focusing on Pillar 1 minimum capital requirements as they are the main influences on so many actions by both banks and insurers.

Taking the paper as a whole, I wonder if the authors have erred in insufficient differentiation of non-life insurers from life insurers. I believe some of their analysis would be simpler had that difference been given greater prominence.

Paragraph 12 considers liquidity risk to insurers. While I agree with the thrust of this paragraph, I feel the underlying issue with lapse risks has been misunderstood, as it is not the availability of funds but rather the loss of profit that is more often the concern in mass lapse scenarios.

Accounting issues are covered in section 2(b). The values of insurance liabilities under Solvency II bear only superficial links to arms-length transaction values. Indeed, the extent to which some valuations, as presently intended, relate to such values arise only from the power a Pillar 1 basis can have in driving transaction prices – a circularity. The value of annuity liabilities is a case in point.

As an aside, the IMF published a technical note on EIOPA (European Insurance and Occupational Pensions Authority) in March 2013, which appears blind to the consequences of strict market consistency which can, under a 0.5% stress condition, lead to phantom insolvencies.

In paragraph 23, the authors point out a material difference in the valuation of assets. In the second bullet of that paragraph, the irony that an insurer who is unlikely to have to realise assets subject to credit spread reports greater volatility from that source than a bank will not have been lost on those present.

The exposition of the key features of Basel III and Solvency II in the paper is very welcome and useful. I sympathise with the authors who, whenever they had finished writing, would have found Solvency II had made some change. Paragraph 44 most clearly illustrates their problem by the timetable it quotes. Nevertheless, the paper is valuable in the consideration of the fundamentals of the two systems. The contrast between paragraphs 79 and 82 on capital risks is a matter on which I hope others joining in the discussion will comment.

The cost of capital in section 6a, does, in places, implicitly assume that relative change is not intended. There is not a level capital playing field today between banks and insurers, so change may, in some cases, be a rectification not an unbalancing of the situation.

I feel the authors also, perhaps, misjudge reinsurance, which is a mechanism critical to insurance markets in both risk mitigation and risk transference. I do, however, think the section on cost of capital is particularly important and I hope others will comment.

One point that the authors do not make is that ultimately excessive capital requirements will be paid for by consumers. Paragraph 119 looks at one side of the issue: the issue of transference of risk back to consumers. But, where risk is retained by institutions and capital charges are very high, the price to the consumer must rise.

In paragraph 125, the Solvency II balance sheet is described as realistic. I would challenge this, at least as Solvency II is being described today. Political moves since the initial concept have moved the balance sheet some way from reality, in places, into a very conservative world.

Dr L. M. Pryor, F.I.A.: The parallel structure of the discussion of banking and insurance is especially valuable, and the level of detail will undoubtedly make the paper an essential reference.

However, I am not sure that the authors have placed enough emphasis on what the differing objectives should be.

As Mr Kemp discussed in his introduction, the insurance and banking industries play different roles in society. The paper mentions the issue in paragraphs 124 and 125 and especially in footnote 74. It says that the nature of capital needed by banks and insurers is naturally different because of their different business models, and notes the possibility of a belief by governments that they are less likely to incur material costs in bailing out insurers.

I think this point deserves more prominence. The banking system and insurance industries interact with other parts of society in very different ways and, in particular, banking and insurance failures play out very differently and over different timescales. The network effects are different, too. It seems that contagion is rather more likely in banking than in insurance.

It is also likely that the ultimate effects of a banking system failure would be different from the ultimate effects of an insurance system failure, although I think it would be simplistic to say that one would be more severe than the other. The immediate effects are certainly very different in terms of who picks up the tab in the short term, as the paper points out. But failures in the two systems will also have very different knock on effects on industry, and the economy in general.

It is, therefore, entirely appropriate that the capital requirements should be different as the consequences, as well as the causes, of failure are so different. As a result of that, it is also entirely appropriate that the differing capital requirements should encourage migration of risk between the two sectors. It would be lovely to think that all these effects had been rigorously thought through, but I suspect that the title of the paper is right and that the interactions are, on the whole, unintended.

In many cases they might also be unexpected, at least by those drafting the regulations. The paper suggests that empirical investigation of the effects would be a good thing, and I cannot possibly disagree. But it does look as if the real world will be running a gigantic experiment in a chaotic environment.

Mr M. G. White, F.I.A.: I think the elephant in the room, which we have not talked about at all, is the integrity of the assets and liabilities in these institutions’ financial statements.

Also, I wonder where the “cost of capital” concept comes from. Banks are expected to achieve a high rate of return on capital, which, of course, they do by heavily gearing. Is that really the result of an intelligent requirement from the shareholders, or is it something generated within the banking community to justify their rewards?

Mr C. D. Pickup, F.I.A.: What has the IMF done with this paper? Has it changed anything, influenced anything, or do they say “That is quite interesting” and it then goes on the shelf somewhere?

Mr Kemp: I think, as with most IMF working papers of this type, that its main raison d’être is to generate discussion and debate. The IMF does not have as much power as you might otherwise expect, at least when it comes to regulatory frameworks. I am not aware of this paper having explicitly altered some particular regulatory approach in either of these regulatory frameworks, but what it has done is to help frame the debate out of which regulatory refinements have arisen.

Chairman: We should not be surprised by that answer because the systems are being driven by very different regulators. The Basel III system being driven globally with a lot of influence from the financial stability boards, whereas, so far as Solvency II is concerned, it is very much a European event driven very specifically within the European bureaucracy.

I support much of what this paper has analysed. I think that it is the first piece of work that I have seen which is trying to bridge across the two systems which are not being driven by any need or desire to be consistent, but, at their basics, they are both risk-based capital adequacy type systems.

The one thing that I would have preferred to have seen brought out in the paper, from my experience, is that Solvency II is very much mark to market on the asset side with the risk-free rate of interest only on the liability side. That is very different from the banking systems and, amongst other things, of course, you do not have full mark to market. That is not the way the accounting systems require banks to report.

There are real, fundamental differences in the nature of the risks in banks and in insurers, and there are some fundamental differences in how the modelling works to assess what is on the liabilities side and what is on the assets side of the balance sheet.

Mr C. D. O'Brien, F.I.A.: The paper talks about the differential effects on the cost of capital for banks and insurers. My recollection from the Treasury paper on the cost benefit analysis for Solvency II is that they foresaw a reduction in the cost of capital for insurers which was based upon, two factors. Firstly, there would be greater transparency in that there would be more information in the public domain arising from Pillar 3.

Due to the increased security that there was going to be for policyholders and, implicitly, insurance companies, they were less likely to fail from the risk-based capital regime that Solvency II was implementing.

My recollection is that reduction in the cost of capital was an important part of the benefits of Solvency II because you need only a very small percentage reduction in the cost of capital, multiplied by a large amount of capital in the industry, to produce a big figure that is higher than the cost of implementation.

Hence, my first question would be if there are any comments about the cost of capital reducing for insurers?

Secondly, the authors have referred to the way in which the origins of Basel III and Solvency II are quite different, and that Basel III is intended to apply worldwide if countries implement it, whereas Solvency II is driven by the European Union. Where does this leave the IAIS (International Association of Insurance Supervisors)? That, in a sense, is the equivalent body to the instigators of Basel III.

Where is that in terms of the IAIS principles for insurance regulation and minimum capital requirements? Has its role been usurped by the way that the Europeans are implementing Solvency II, which is then leading to some other territories, in particular the US, to look at Solvency II developments to see how their own regulatory regime might need to adapt?

Mr P. J. Tuley, F.I.A. (closing the discussion): The paper is necessarily very strong on the comparisons between the two systems in front of us. A very strong comment has been made that these differences entirely flow from the different situations, the different risks that the different industries pose. Unfortunately, in a way, there are also very strong similarities. The paper, and one of the authors in the introduction, raised internal models as the banking view of internal models is entirely gloomy and suspicious.

For an insurance industry that is, in theory, walking towards approval of a number of internal models that is entirely unhelpful and very much an unintended consequence from the insurance industry side of the street.

Cost of capital also received a fair bit of time and consideration, pointing out that this may not be an unintended consequence: it may be a move to a better place. The last contributor pointed out that the Treasury thought this was moving us to a lower cost of capital in the insurance industry because we are having a more transparent and understandable system.

That could easily be undermined by Mr Chamberlain's comment on Solvency II that, although it has on the tin “mark to market” and “economic reality”, the political process has meant that some elements have been changed, and the end result may not be quite the transparent and obvious insight into insurers’ business as one hoped: embedded value may yet live in the future.

So, there may be unintended consequences for the cost of capital, which might be unexpected, but, quite possibly, could be entirely positive.

The last theme I would pick up is mark to market. It is quite easy for an insurer to look down on a bank's accounts, amortising loans that have turned out to be drastically poor and causing continuous provisioning in the accounts of a bank. One is always very conscious that our own mark to market purity is entirely dubious in the sense that it is a mathematical construct which is very rarely testable against the market, and our new system will rely very heavily on the regulation and criteria being set down by Solvency II.

We remain quite a regulatory driven business in terms of how we recognise and present our balance sheet. There are similarities for the banks because of that position.

Mr Kemp (replying): Mr White suggested the integrity of the balance sheet of insurers may be slightly more trustworthy than banks. My own anecdote in this area dates from when I was a fund manager. At that time fund managers tended to believe that they understood what banks did more than they understood what insurers did. Maybe the picture has changed in the last five or ten years. However, I would not underestimate the lack of clarity and lack of transparency that some people still think applies to some aspects of insurance company balance sheets. Perhaps this ties in with Mr O'Brien's comment about a Treasury consultation which noted that greater transparency and greater clarity of what the balance sheet contains might help insurers.

We will certainly take on board comments like those of Dr Pryor about the desire for greater emphasis on the different objectives that ought to apply to banks and insurers within the regulatory frameworks placed upon them.

Mr O'Brien highlighted one other question to do with where Basel III and Solvency II developments might leave the IAIS. I would note that some of the ideas originally introduced in Solvency II are now being considered in a more global framework. For example, the ORSA concept is something that is now increasingly being referred to across the globe and not just within the EU in isolation.

Mr O'Malley: I should like to add something on the point regarding Solvency II and the IAIS that was raised by Mr O'Brien. I think it was the case that Solvency II was gaining impetus and traction globally. However, over the last couple of years, given the delays, and so on, with Solvency II, I think that it has lost a lot of that impetus. Certainly, when we look at the U.S., they will be much stronger nowadays in terms of saying that their own system is preferable to Solvency II than they would have been a couple of years ago.

Chairman: Thank you very much indeed. The final thought for me is that I have no doubt that the Basel III system is significantly increasing the capital requirements on banks and so it is needed. There is no doubt that the banks were woefully undercapitalised going into the financial crisis. I think the concern so far as banking capital is concerned has to be more whether the system is leading to excessive capitalisation of banks. There is nothing more likely to reduce lending in the economy than if the capital requirement is calibrated too high.

So far as Solvency II is concerned, I do not expect it to be adopted outside of the European Union as I do not think we know whether it is likely to lead to higher capital requirements or lower capital requirements because the system is designed to be highly volatile.

If there is any one weakness from my perspective in Solvency II it is the degree of volatility that is built into the system which will take a lot of experience to determine what is the right level of calibration.

Finally, it remains for me to express my thanks and I am sure the thanks of all of us to the authors, the opener, the closer and all those who participated in this evening's discussion.