D7: Earnings Volatility and Canadian GAAP D7 : La volatilité des gains et les principes comptables généralement reconnus au Canada

Transcription

1 Earnings Volatility and Canadian GAAP (D7) 1 D7: Earnings Volatility and Canadian GAAP D7 : La volatilité des gains et les principes comptables généralement reconnus au Canada November 17 Novembre 2004 Moderator/Modérateur: Panelists/Conférenciers; Byron Corner Allan Brender Arshil Jamal Robert Lefrançois, Partner, Deloitte & Touche LLP, Montreal, QC [Editor s Note: The recorded transcript joined this session already in progress.] Moderator Byron Corner: [The Canadian Asset Liability Method (CALM) is] one of the only significant ones with a prospective only approach dealing with the total contract where all assumptions are open to review and unlocking at any time. It s principle-based which means that there s lots of room for judgment and it tries to balance the solvency and balance sheet/income statement concerns to serve both of those worlds. That obviously is quite different than the US situation. CALM is a complex methodology to value very complex financial instruments. Many of these instruments are long term, which of course means that there can be a lot of leverage to very small changes in assumptions or data. There can also be a lot of volatility associated with actual experience and to value of unhedged, embedded options. So the possible reactions to these realities of our Canadian Generally Accepted Accounting Principles (GAAP) for life insurers, include: you got to live with it that s the way it is it s the nature of the beast, and understand the root causes of your volatility, explain it effectively, and manage your business through decisions about what you sell and how you hedge it to reduce the volatility to an acceptable level. I suppose another possibility would be to try to reduce or manage the volatility within the limits that are allowed by standards. There may be and we may hear discussions about the room that perhaps exist for reducing what some refer to as unnecessary or excessive volatility. And I guess, finally, you could decide to choose a different GAAP. If you don t like this one, you may choose to manage your business to a different standard that perhaps is less volatile. There may be a combination of perhaps all of three reactions. Those are my opening remarks. I am going to ask each of the panelists to provide some opening remarks as well to get things started. The comment in our notes here, external users, in particular analysts, like GAAP models that are understandable and produce stable base results, but periodic readily identifiable capitalization of event impacts. Some believe that Canadian GAAP for insurance fails this test. It is complex, unstable and opaque to users. So, one of the questions that perhaps our panel can answer or comment on in their opening remarks, or address in setting the stage for their position, is Do you agree with the comments in our notes? Is it a problem? What are the issues, if there are any issues? I m going to start with Arshil.

2 La volatilité des gains et les principes comptables généralement reconnus au Canada (D7) 2 Panelist Arshil Jamal: I will start by reiterating that we do have an accounting model that we ve developed and applied to the life insurance industry over a number of years [for other than our treatment of deferred acquisition costs (DAC) and how we model that] is fully prospective. I certainly don t have a lot of sympathy for people who say that the end result is particularly volatile. I mean, it was a foreseeable consequence of the move to a fully prospective method that requires you to use the values at a balance sheet date to project forward. So if we didn t like this when we were doing it, then that s what we should revisit. I would have some sympathy for revisiting the model because we have a model that s very different. US GAAP for life insurers is very different from what banks use for the bulk of their traditional banking business. It s different than what is used to account for pension plans, and pension expense is different from the mutual fund industry. We compete with all of those industries for access to capital and if we don t have stable earnings or equivalently stable earnings, our cost of capital will be that much higher. I think that once we are in this situation with a fully prospective model ignoring values on a particular balance sheet date because you didn t like the value or because markets were closed on that particular day is just a recipe for undermining the credibility of actuaries in the actuarial profession. We have a model, and though you may hear some comments later that international accounting might move in this direction, we re on a leading edge and being on the leading edge means that we are suffering from many of the problems by adopting new methodologies before other industries and before the same industry in other countries have adopted it. But, I have less sympathy for volatility concerns because it was all a foreseeable consequence. Moderator Corner: Allan? Panelist Allan Brender: First of all, I m going to say that I m talking for myself and you re not going to believe that because everybody works for us. Somehow it is assumed that I am telling you what the government has in mind, but I m talking for myself today but with some experience of what is going on back in the office. We re users of the financial statement and we re looking at it from the point of view of users. My real perspective is that there are many ways to tell a financial story of the company s progress, and that accounting is basically a set of rules for how you tell the story. My real personal concern is that once we agree on the rules then let s stick to them, otherwise we haven t a clue as to what we re getting and how to understand it. I also appreciate that, particularly when it comes to volatility, people don t like it. People don t like change. They like predictability. They like everything to be, in some sense or other, known. This, I think, flies in the face of what I understand a lot of the accounting theory to be and, as I ll explain later, gives us all kinds of problems figuring out how to judge the company s progress. Two other things I want to point out. Maybe it s because my head is very much into all this discussion of international stuff but I m very taken with not so much the accounting details today as what I hope they ll be in another two, three years and where I think we re going. One of the things that I think is important is that we re moving from a regime where we account by industry to a regime where we account for things. We ll have rules for accounting for various things and if those things live in different industries, in theory they are going to get the same treatment no matter where they live. So, it s a matter now of accounting for insurance contracts, not insurance companies. A second thing: because the Office of the Superintendent of Financial Institutions (OSFI) is an integrated regulator, we see things from several different points of view and one of the ways that things get contracted is as banking or insurance. Now in the coming world that I hope we will eventually see those things will be much more on a par, but right now we get all kinds of comments from the banking side as compared to the insurance side, as to who can do what, who has the discretion to smooth, who doesn t have discretion to smooth, who s forced to use market values for certain things, and who is not and so on. And that definitely influences my desire to see a certain amount of harmony, and to also look at the fact that insurance companies have been getting increasingly into market related products and conclude that some of the volatility comes with their market exposure. To some extent, if you don t like the volatility you shouldn t be in the business. Moderator Corner: Robert? Délibérations de l Institut canadien des actuaires, Vol. XXXVI, n o 1, novembre 2004

3 Earnings Volatility and Canadian GAAP (D7) 3 Panelist Robert Lefrançois: Thank you. I will agree that, in general, I share your view that probably people don t like the volatility, but then I want to point out that members of the Accounting Standard Board are probably not people because we do like volatility. And so, we are not ordinary people. I think volatility exists in this world, and we have to see accounting as a way to reflect things. It s like journalism. When I give you a report on what is occurring today or what has occurred during the year, I m just reporting what happened, and whether I report it or not, things occurred. So, when markets change, the values change. The life cycles of products are shorter than they were in the past and in your business you re probably selling products today that were not possible to sell ten or fifteen years ago; we have a situation where your business has evolved. The risk has evolved, as well, and today we are able to manage risk a different way than we were doing it in the past. I think when you look back to the early 1980s, there were just a few players around the world that we are taking swaps or any form of derivatives other than pure futures. So where we are now, 25 years later, is surprising to me. We believe that these things have been there forever, but that s not the case, those of you who were here in the market years ago know that. My point is that since things happen, the real question is whether you will inform the market of these changes, whether you will inform the investors, and whether you will inform the readers of your financial statement that these things have occurred. Moderator Corner: Thank you. I guess we re going to come back to you, Robert, and have you spend a little bit of time telling us about what accounting theory and principles and Canadian GAAP Standards say with respect to volatility of earnings. There s a question certainly about whether life insurance is a service contract or a financial instrument and whether it should be treated as one or the other, and that has certain implications in terms of its accounting and volatility. Are the differences in our business significant enough to warrant special treatment or consideration for life insurance products? I ll allow you to expand on any of those questions or points. Panelist Lefrançois: The first point I would like to make is that accounting maybe 20 years ago was pretty much based on matching principles. What we were trying to do was match expenses and revenue in the same year if possible, and even if it was even more possible in the same quarter. In doing that we have faced many challenges, and we probably also face many abuses in that model. We have developed a general approach for I m not speaking exclusively for life insurance but globally here all the people in the industry, in all industries. Many people use that as a way to manage earnings and to always achieve what you want to do. For example, they take something that is at cost and if they need the profit at year end, they look at their portfolio and they say, Okay, this one has a nice profit and I ll just recognize this one, so I ll make a nice profit and everybody will be happy. When you look back for the past eight, nine, ten years at the Board, all of our standards have been really issued based on the balance sheet approach. The balance sheet approach really forgets about this matching principle and tries to tell the true story about what kind of assets and liabilities you have in your balance sheet, with more focus on the picture of what is the right balance sheet today, and maybe less focus on matching involves too much judgment anyway and too much freedom. That s a fundamental change. What we have today then is a model that is similar to when we started with income taxes if you remember, and even pensions of all things. Every standard was issued with a more swinging balance sheet approach, trying to get a better balance sheet. Recently, we have issued an exposure draft on financial instruments and we are going forward with a fair value model for most financial instruments except some forms of liabilities., I believe that when you look at everything we do assets, retirement obligations everything is based on the balance sheet approach. We are using a lot of valuation techniques, a lot of discounting factors, a lot of discounting techniques as well. This creates a lot of more work for actuaries and other people and other valuators in the field, but I believe this balance sheet approach is providing us with a way to say, Here s the real pictures, the real economic picture of the company at a specific date. We still have a long road to travel and that the problem we have right now, or most of the problem we have right now, is because it s a mixed model. We still have some fixed assets at cost; we still have a lot of things that are not based on fair value. But, we are really heading towards a fair value model. I believe, from every meeting we have, reviewing what s going on at the international board and from comments that I receive every month, that the international board is going to order a fair value model for life insurance companies. I don t know yet how it will be done exactly. I don t want to take your point out of view, Arshil, but I think you re right they re always three years down the road. Sorry if I get your punch for you, but there always seems to be good reason to delay that project. You have to understand there s good reason right now

4 La volatilité des gains et les principes comptables généralement reconnus au Canada (D7) 4 for that and that there s valid reason. We have a new International Board for the past three years and they have a lot of things to do. They have resolved (and they still have remaining a lot of) problems with International Accounting Standard (IAS) 39 on derivatives. As soon as they have dealt with that, I think they will have more time to deal with these issues. We can probably have more confidence in their planning process at this time for the future than we had in the past maybe I m just optimistic, but I believe it will be so. So that s where we are heading. When I joined the Board, eight years ago, I was not really very keen on fair value and I don t believe I would have been able to stay there all that time if they had not convinced me, because it s very frustrating process if you re not in agreement with the whole fundamental concept of financial reporting. So I think I ve bought the concept that we have to put less emphasis on the matching principle and move toward the balance sheet approach. Now that has many implications for you. To answer your question at last, Is it a financial instrument or form of service contract? I ll be honest: I don t know. For example, I believe that should you have some perpetual debt which could never be repaid in the future determining at what point in time somebody will die is even more complex because you know that this debt is there for 100 or maybe 200 years, then we don t knowif it is a financial instrument. I believe that it has many characteristics of a financial instrument, and I believe that with your techniques and with your expertise we can have a good idea of when it will be paid. So, yes, it looks very much like financial instrument, but we can also make the case that it s a service that you are rendering and that for all kinds of services we will wait and recognize this instrument as the service is provided along the road, but I believe that the most important thing for your industry is not necessarily how we ll deal with contracts on life insurance, but its rather what we will do with the rest of your balance sheet. There s a lot of work to be done there. Moderator Corner: Arshil? Panelist Jamal: I suppose that I might reiterate what Robert said. I think a big driver of the accounting problem is this mixed state where a lot of industries are on the matching principle, and other industries on other parts of the standards are using the balance sheet approach, and they get compared. Even though we are on a balance sheet approach as you know, CALM is a Canadian Asset Liability Method we re getting compared period to period with industries that use the matching principle approach. I would happily live for the day where everyone was on the balance sheet approach. For example, an oil and gas company taking their proven reserves, at the beginning of the quarter multiplying it by the spot price of oil, at the end of the quarter multiplying it by the then spot price of oil, and then taking that difference through their profit and loss (P&L) statement. If that was happening, then analysts wouldn t be bugging actuaries about what reserve basis changes you made in the quarter because the oil and gas company s figure is an order of magnitude bigger number, and I think we d have the same issues around inventories and all sorts of other things. I think this is really something that I ll say probably three or four times over the course of the hour. It s a problem we visited on ourselves, right? I mean that we adopted this model of having assets and liabilities linked in a fully prospective valuation and at a time when no other industry was doing it. Perhaps the international accounting standards in three years time will put everybody on the same page, but right now there s this difference between where we are and where everyone else is and that creates enormous pressure on actuaries who are trying to apply their judgment to come up with financial statements that are meaningful and then react to the commentary that we get from the analyst community and the people who ultimately own the company s stock. It really undermines the credibility of the profession that we ve gone out on this limb and not a lot of people are ready to follow us yet. Moderator Corner: Heading off in a slightly different direction for a minute. Allan, why is the regulator concerned about GAAP financial reporting? What are the regulator s concerns with respect to volatility producing mechanisms? Maybe you could just remind us of what the key disclosure requirements are, the things that OSFI needs to receive and understand with respect to valuation of policy levels. Panelist Brender: If I could provide a little historical perspective, there was a time when Canadian insurance accounting was like what you find in US statutory accounting: dictated assumptions, dictated methods, and no real appointed actuary/valuation actuary any of that stuff. That changed in 1978, which sounds like a long time ago to some of you but it s not all that long ago. I remember reading insurance accounting books at the time, and coming across a quotation from the then Superintendent of Insurance who agreed that we should have an appointed valuation actuary at the time, and that the valuation actuary should be given the discretion to choose appropriate assumptions appropriate was a new word in the vocabulary in that sense and, importantly, that there should only be one set of reserves. Whatever was in the statutory statement with respect to policy Délibérations de l Institut canadien des actuaires, Vol. XXXVI, n o 1, novembre 2004

5 Earnings Volatility and Canadian GAAP (D7) 5 liabilities or reserves, had to be the same number that was in the GAAP statement or the published financial statement, though I m not so sure that it was really GAAP at that time. An interesting point. I went to the Department of Insurance in 1983, and the Superintendent who had made that decision, and written that letter, was at that time retired but was still working as a consultant. So, I had a really great opportunity to ask him, Why were you willing to do that? To give up your control of the liabilities? I think his answer was very illuminating because it tells you why, later on, we went the further step of using a totally GAAP financial statement with no statutory statement at all. Whatever the accounting profession says is subject to CIA rules with respect to actuarial standards. Why are they willing to do that? And the answer was because a realistic measure of income is a leading indicator of future solvency, because income is a change in surplus. If you can measure and you have some confidence in the measure how the surplus is trending, then you have some comfort from the regulator s point of view as to what s going on. Basically the decision was made that we re going to trust in GAAP. So the next question then was; Can you believe it? So we come up against the fact that the industry has changed in many ways, and there s a lot more volatility. Now, if volatility is perceived in some people s minds as a disease, then what s the cure? The cure is a word called smoothing. So what s the problem with smoothing? Well, we don t think it s what the rules say. We think it obscures what we see, what we think is happening, and (generally speaking as a user) we would have a problem because we don t really know any more what we have. The problem is it s not hard to smooth when life is getting better, you just tame it down a bit so no one minds. But what happens when things are going the other way? The analogy I ll give you is this: par business, with profits business in the United Kingdom, was sold basically for the individual to participate in the equity market without suffering the vagaries of, or the daily changes in, the market. The company will smooth your experience, so you don t have to worry about reading the paper everyday to look at the index. Generally speaking, once every six months is good enough. The problem is, of course, what then happened in the UK: the insurance industry invested heavily in equities, and then the market dropped like a rock. The problem is that the funds they were holding lost a lot of value, and there weren t enough of them to smooth on the downside. And our whole problem with smoothing with respect to income is precisely that: if you have smoothing, how do you know when they re getting into trouble and how do you know that they re going to be able to bring it to a soft landing? And the answer is we don t think that we do know. Therefore we have concerns about smoothing. The other part of things is that let s face it everyone believes in actuarial judgment, my best actuarial opinion and so on. There s a heck of a lot of discretion buried in that, and there s probably more ability to do smoothing in insurance because of the actuarial judgment than there is in most other industries. We therefore have a few problems. One of the things we discovered lately is that there are a number of companies that have very significant amounts of money tucked away, called bulk reserves, and I don t think we re very comfortable with this. We recognize it, and as an appointed actuary, I had a bulk reserve once upon a time but it was there for a specific purpose. [Laughter.] It was there for a specific purpose for a limited amount of time one year because we were changing valuation systems, didn t know what was going to happen and were afraid that reserves were going to go up. It didn t happen, so we released it, but the point is that occasionally things do happen there are certainly plenty of cases where people say we re doing mergers, we haven t changed, we haven t really made a full adjustment in the integration of systems and companies and so on, so we made some provision this year so, take the bulk reserve out next year. That I can understand, but in general we re not comfortable with the idea. We get to this whole issue of disclosure from that point of view, and if you look at the Memorandum of the Appointed Actuary this year, you ll find that last year, for the first time, we said tell us about your reserves. This year we re saying, not only do we want a list of them and how much they are, but tell us why they exist, how they are built up, how they are going to be run down and when. We want to know what s going on, and they re not so secret messages discretionary reserves for cookie jar purposes shouldn t be there. Now in terms of volatility and products that are subject to the market, people are proposing various things and saying various things, but I ve got to tell you that with the issue in particular is emerging with respect to segregated (seg) funds. If you start thinking about seg funds today you have to conclude that most of the seg fund guarantees that are outstanding today they re

6 La volatilité des gains et les principes comptables généralement reconnus au Canada (D7) 6 really a major issue with respect to liabilities and surplus are probably guarantees that have on average five, six, seven years to run. They re essentially like put options. The point is that a bank might sell an option running five, six years, and the bank has to value that option at market, and my banking colleagues and accounting policy colleagues who do keep on reminding me of that. So that becomes a really great issue, and we have a big internal problem about that in terms of consistency of treatment and so on. Now, of course, if I were to talk about today s accounting, I would get all kinds of rebuttal about things that the banks do, that the insurance companies get away with and that the insurance companies don t get away with et cetera, which is why I m putting my hopes on the International Accounting Standards [Laughter.] defining one set of rules that s going to apply uniformly to everybody by product and not by industry. Moderator Corner: Okay. We re going to get back to the other two panellists with the same basic questions. Why are smoother earnings good and rougher earnings bad? What drives this sense of smoother being good? Is volatility a communications challenge or a real issue to certain audiences and stakeholders? Panelist Jamal: My view would certainly be that volatility is a real issue. You know, I think it s pretty basic in finance that smooth earnings indicate less risk. Less risk means that you get a better price earnings multiple which means your cost of capital is lower. If you re perceived as an industry that has particularly high volatility then your cost of capital will be higher. If we re on a accounting model that s different from what the bulk of the banks are on and what the mutual fund industry is on, then we have a problem. Their cost of capital is going to be lower than ours and they re going to capture that margin on the next sale, as opposed to us where we have a similar product. There are a number of investment analysts who follow the companies. Certainly the sell side analyst will spend a fair bit of time trying to understand what each of the companies is doing. They like their forecast to be right, and smoother earnings means its more likely that their forecast of your earnings is going to be correct, so there s certainly a bias from that group to get stable earnings. And then on the buy side, you ve got an analyst working at a pension plan covering broadly speaking a universe of 500 to 600 stocks, 100 of which he s following pretty closely, 30 or 40 of which he might buy or currently hold in his portfolio, but with another 50 or 60 that he d like to buy, and he s covering them every quarter when earnings come out. That s a huge number of companies to follow and he s not spending more than five or ten minutes looking through your earnings. When that analyst is an making that allocation of capital decision, he is going to want smooth earnings, something that s predictable, and something that he can understand in five or ten minutes. We can talk about, doing a better job of disclosure and can we have 50 page press releases with 60 page decks of slides and two hour analyst conference calls or whatever, but none of that s going to get into the buy side analyst s world. They re looking for concise information, something they can act on in the next five or ten minutes or put away to look at again in a quarter s time from now. Moderator Corner: Robert? Panelist Lefrançois: Yes, I think I ll have to complain to Peter Martin, the Director of the Accounting Standards Board of the Canadian Institute of Chartered Accountants [CICA] because he told me this panel was an opportunity to sell the volatility side to a lot of people, but I fear a certain absence of people on the other side here so if someone in the room wants to stand up? I think that we are effectively in a world of transparency today. I think you all know that everything that has happened in the market during the past three, four, five years has led the accounting profession, and many other people as well, to a more transparent world and it has some merit. Just as an example, for years we ve been exchanging comments with the security commissions in the US and in Canada. Normally, when a company received a letter from the security commission, we dealt with them and tried to provide more information and to help the security commission understand what the company was doing. But now a letter from the Securities and Exchange Commission (SEC) and the comments received are the information answers provided by companies are published on the SEC website. So, when they ask you for further information, that s public information available to everyone. So when we are speaking about smoothing or non-smoothing, what we say is that since volatility exists we want you to be more transparent in disclosing that volatility to the market preferably in the balance sheet and not in the remote note 26 in your financial statements, for example. It s really about putting that directly in the balance sheet. What we say is that we want you to be really transparent and show everything to the market. When we had this debate at the Board, at some point I said to the Délibérations de l Institut canadien des actuaires, Vol. XXXVI, n o 1, novembre 2004

7 Earnings Volatility and Canadian GAAP (D7) 7 Chair at that time who was Trish O Malley, she s now at the International Board I said I have the feeling that, with all that you request from companies, it s rather like asking people to pull their pants down. And she said, Yes, that s exactly what we re doing. Many times we were trying to put some of these things in, what we call in the US, other comprehensive income (OCI). So, we were trying to bury some of these fluctuations in this account which is a parking lot where we can park gains and losses until the time we deem it appropriate to put them back in the P&L. What happened is that I said we had another project, an international project which is Reporting Financial Performance, that essentially says though, you ll get used to OCI because it s coming to Canada, soon this international project will kick in and just kill the OCI, sending everything through to the P&L. I said not only you ask people to pull their pants down, but you force them not to cover themselves, like this. [Laughter.] So it s not fair. At least allow them to put their hands in front of themselves, but they said, No, we want you to be transparent. All joking aside, I believe that we are in a different world, and I m not saying that smoothing is bad. I m saying that we should have a greater emphasis on the balance sheet approach and we should really give that information to the market. I recognize it s difficult for everyone because we are in a world where people look to make money in the next two weeks or sometimes in the next two hours. We are in a different world than we were in the past, but people will need to learn how to deal with this volatility. Just because we don t speak about it doesn t mean that volatility does not exist, and we really believe it should be disclosed. Moderator Corner: Allan? Panelist Brender: Just on that point I, too, really do believe. If people are having trouble with volatility now, then what are we going to do when we have a full blown fair value system? And if we know that people are having trouble now and the accounting community knows, isn t it a little bit insane to be continuing to march down this road? And what we really are going to need is a whole psychological transformation of people in business, to understand volatility and learn to live with it, and love it. Quite honestly, I really think that s where we are going to have to head; as long as it is a problem then we re not going to be at rest. I don t mean a rest like not moving, I mean a rest like not comfortable. I just want to mention an example from the other side. Over the last couple of weeks the public companies reported third quarter earnings. So, you start hearing from the analysts and so on. Now, there s at least one analyst s letter that is a lot longer than usual, and was quite significant because basically the analyst said: The company didn t quite make what we thought they were going to make and we thought they could smooth their way out of anything. [Laughter.] That s really what he said, and therefore he may have thought: Well, OSFI is bringing out source of earnings, and we think this is OSFI saying we don t want you smooth. Well, as a matter of fact, quite honestly, we never thought about the smoothing issue in connection with source of earnings, but we are nevertheless concerned about the smoothing issue. So, he s sort of right. But basically, what did he say? He said that they can t quite smooth as well any more because they have pressure on them, so we re going to have to rethink all our models for all the life companies and we are going to have to downgrade them from outperform to perform. Whatever the rating was, he was downgrading and he s saying, We re rethinking our whole model. I m not making this up this is reality and the fact is that companies will get penalized because this is the analyst s psychology which supposedly reflects the investor s psychology. If we re heading for an accounting system that is supposed to show it all, then the only way that s going to work is for the psychology to change. And I m not so sure that s going to be that easy to achieve, but I really think there s some strange stuff going on in terms of how we are going to adapt or why we are adapting with these rules. Moderator Corner: It s often mentioned that the life insurance side of our business is extremely long term and certainly anybody that s been involved in valuing the business using CALM has encountered all the complexities that go with it. It begs the question whether the long term nature of some of our contracts the insurance side of the business justifies some element of volatility reduction, and so my question is, Do our current standards allow for such techniques? Are there possible ways of reducing what some would refer to as unnecessary or excessive volatility? Panelist Jamal: I can certainly point to two examples that come to mind that do tend to smooth volatility in the prescribed interest rate scenarios that are in the standards. We have the three to ten percent range for short term rates and the five to twelve percent range for long term rates. So irregardless of where you are at the balance sheet start date over a reasonable period of time, all of your interest rates are going to remain that range whether you are at the top of it, in the middle of it, at the bottom of it then that s built right into the standards. I think people didn t recognize that. I was surprised that the standards were adopted

8 La volatilité des gains et les principes comptables généralement reconnus au Canada (D7) 8 at the same time that we also agreed to capitalize the mismatch risk and put that through the P&L because that s an exposure, not a future cost, we may or may not lose money on mismatch risk, and you can address that through capital requirement, but you certainly don t need to hold a reserve against that exposure in advance and put changes in that through the P&L. So, I guess what we gave with one hand and we then took away with the other within the standards. I guess that more recent changes to move to DAC approaches for segregated fund products without guarantees to try to put us on a comparable footing with the mutual fund industry is another example. Certainly, within the profession at that time, I think there was resistance to doing that, but that s the one part of the reserve that period to period unless you get into a DAC write-off situation is very stable. So if you want an accounting model that is a stable model, and produces predictable earnings, this is it. Absent dramatic moves in the market, I can tell you what we are going to earn in the seg fund line in a number of places to a very high degree of accuracy. One other comment is that even when we adopt approaches that are consistent with other industries and give us a more stable result like our DAC treatment We pile on top of it. Requiring you to test DAC recoverability using padded assumptions, for example. There s no other industry other than the life insurance industry that would take an asset and then test for impairment using more conservative assumptions than your best estimate assumptions. I don t think that the conservative nature of actuaries always leads us to the best conclusion, but it s probably heresy saying that in front of a group of Canadian actuaries. From an income perspective, US GAAP does have a lot of advantages. They ve really adopted their accounting models with a clear focus on what it would do to the income statement first, and the balance sheet is second. We ll see what happens in three or four years, but I d certainly really feel uncomfortable being the only person in the room with my pants down. [Laughter.] I mean if we re all at the gym and we were all getting changed in the locker room, that s one thing and analysts will have to accept the volatility. If we re the only one, however, we d simply be drawing undue attention to ourselves, and I think that s where we are right now. Moderator Corner: The imagery. [Laughing.] Panelist Jamal: I apologize for any negative mental image that people might have formed. [Laughter.] Moderator Corner: The other aspect of our standards which come to mind, which certainly needs to be well considered by anybody valuing business, is your level of aggregation as the standards call for review, consideration, and possibly unlocking of any and all assumptions. So assumptions are open to updating at each valuation and there s the possibility, in some instances, that one might try to make a case for reversion to the mean or smooth in recent new experience and things like that. Hence, I think there are aspects within our standards that allow or tolerate a certain amount of volatility reduction to the points that other people have made. Disclosure is critical. Consistency from one period to the next, a common set of criteria and practice standards is absolutely critical if you re unlocking assumptions. Panelist Brender: I just want to remind you if we re talking about standards that standards do say in a couple of places but should not smooth. Certainly the old provision for adverse devision [PfAD] paper does; I m not sure whether the words found their way back into the Consolidated Standards of Practice (CSOP). But the old PfAD paper certainly said the reason for the margins is a shift in the mean and stuff like that. It specifically said it s not to cover fluctuations in experience; didn t it? I see that Simon s nodding. Unidentified Speaker: I ve read CSOP back and forth many times. Byron and I have talked about this. The CSOP is surprising silent on this. There may be a gap in the standards. Panelist Brender: But wasn t it in the old PAD paper? Unidentified Speaker: We seem to have lost it in the CSOP. Panelist Brender: I always thought it said don t account for experience and therefore implicitly said don t smooth. Panelist Jamal: I don t want to steal my Committee on Life Insurance Financial Reporting (CLIFR) colleagues thunder, but that Délibérations de l Institut canadien des actuaires, Vol. XXXVI, n o 1, novembre 2004

9 Earnings Volatility and Canadian GAAP (D7) 9 is a change in CSOP that s floating around in one of the drafts that we are now working on and should be exposed to membership. I don t want to give a date. I will let Micheline Dionne do that. But I think the principle certainly exists not just in actuarial standards in the past maybe not particularly in CSOP but certainly in all the accounting literature that period to period consistency is very important. And we are making changes from one period to the other, changes in estimates, and how they get disclosed. I think you could look to accounting literature and what s required there, and not just to the actuarial literature. Moderator Corner: Robert, banks have marked to market derivatives in their trading book. Our seg fund guarantees are similar to derivatives but why should the segregated fund guarantees be treated differently? What s likely to happen when fairer value accounting comes into effect? Panelist Lefrançois: I don t, in fact, know the answer why segregated fund guarantees should be treated differently. I believe that they should probably be treated in the same way. That s the way I understand how it works. Again, I believe that not all bank derivatives are marked to market at this time. These are the one that you mentioned that are in their trading book. You have others that are still part of some hedging relationship that can be established, because they have large portfolio and that the hedging rules despite their complexity and their difficulties still allow a number of hedging relationships. Sometimes you can have a business macro hedging relationship. You ll need to get the different designation for accounting purposes but you still get there. That means that you can still have on paper a designation that will not reflect your true objective, but at the end you get the same result. I believe that in the future we ll get to a world where all derivatives will be at fair value first with the new exposure draft. Some of the changes due to fair value will be put in the OCI parking lot for a while, until we get this new reporting financial performance that will essentially kill the OCI and present income statement as a form of matrix with all the changes in the estimates in one column. That will make it easier for everyone to focus on these changes. I would be happy to hear why someone thinks that guarantees should not be treated this way. Panelist Jamal: I guess the point that I d make on seg fund derivatives and those sorts of things, and the behaviour of policy holders, is do they really follow rational market views on how things might evolve? The example I d use is on the banking side. I m pretty sure in the banking book they don t mark their derivatives to market. The last time I renewed my mortgage by putting in place a mortgage commitment before I bought a new house, the bank was more than willing to give me a 90 day or 120 day mortgage commitment and fix my rate. When that option dipped into the money, I didn t rush out that day and say I m not going to spend half a million dollars, and buy a new house, because my option is in the money. So I m sure their models didn t put the market value of their option through their P&L until I decided to let the option expire worthless after 90 days. So I think you have to be really careful with market value and how people interact as things aren t often symmetrical. They re often not symmetrical regarding what s in the best interest of the policyholder and what s detrimental to the company. I m thinking of minimum interest rate guarantees on universal life products, for example, where you might expect that people would be happy to lock into a three percent rate on their account value because three per cent is twenty five basis points more than what we could earn in the outside world. Maybe three percent was too high, but if that policy holder needs eight percent to get their policy to perform the way that they want the policy to perform, then they re still likely to invest a fair chunk of their proceeds in equity funds. Market values of traded instruments are not always the best reliable indicator of market values on guarantees associated with insurance products and other banking products. I don t think that there s a panacea there. Panelist Lefrançois: I can almost hear somebody from the banking industry telling me that it s not fair because they have a relief only on loan commitments when you had the relief on everything because of I really believe that there are a lot of things that are difficult to value, and probably, yes, the standard is not perfect we have some scope exceptions that may be questionable. Let s say you re producing aluminum, or any form of commodities that are, traded in the market, and you always sell at a fixed price thirty days or in sixty days in the future. Is that a derivative? The answer is yes, it s a derivative. However, you have a scope exception in the standards to say although it s a derivative, because it s in your normal course of business you ll get an exception. While it s difficult to determine if it was the real reason for the treatment of loan commitments, I do believe with this exception or a few other exceptions (it s not the only one) that most of the derivatives are at market except for those which are in a hedging relationship.

10 La volatilité des gains et les principes comptables généralement reconnus au Canada (D7) 10 Panelist Jamal: We could go back and forth on this one. We started with an oil and gas example, and I could point to mining companies. I could write a derivative on an oil and gas contract and promise to deliver oil at some future point. That s a derivative but all the oil that they have buried in the ground, that s a proven reserve, is not even in their balance sheet. Certainly, the change in the value of those proven reserves is not going to be their income statement. Panelist Brender: I had a lot of debate with a client that says the water behind his dam is electricity, and they wanted to fair value the inventory. [Laughter.] Panelist Jamal: I m assuming that that would lead to an upward revision in their inventory values? Moderator Corner: I am going skip ahead here a little bit, because I want to leave some time for questions at the end. I ve heard it said that CALM it has some or all of the following characteristics. It allows the actuary to override balance sheet values provided by other professional standards. It makes the key component of the movements of the reserves driven by matching as opposed to the contingencies which were seen to be our expertise. It s inherently volatile. There s no comparable GAAP model world wide that has similar characteristics. And it s complex, unstable, difficult to describe and define. With that lead in, does it make sense to have CALM as a GAAP model? Do we need a change? Does that put us at a competitive disadvantage with other GAAP standards? Really without change meaning that we ll see the end of CALM as a basis for GAAP, perhaps as a statutory capital measure. Let s start with Allan. Panelist Brender: I ve been hearing this argument a lot lately and maybe I ve come to understand it. CALM is a really neat idea from a strictly actuarial point of view, saying that we re concerned about matching cash flows and so on. It kind of makes a lot of good sense that way, though I can see the volatility and what it could do to income statements. My sense is international accounting is going back to what we used to call the Policy Premium Method (PPM), and we re going to be back there anyhow. Even several years ago, we said keep your PPM software. [Laughter.] On the other hand I ll tell you this keep your CALM softwares because CALM software is a great way to measure asset liability management (ALM) mismatched cashflow risks. If you take that measure of mismatch risk out of the financial statement, given that the world has heard this whole phrase about total balance sheet approach, then it s natural that this thing is going to reappear in capital requirements. So save the computer programs. I don t have too much trouble with this. Panelist Jamal: If we re going to go down this international route with a discounted PPM, based on everything that I ve read the two places that where our traditional approaches are most vulnerable is assuming an equity risk premium in advance, and assuming a credit risk premium in advance in the valuation. I think that in both of those areas we re not going to be able to sustain what we now have in the Canadian valuation approach. We can criticize the things we don t like about CALM and its volatility, but those are two things that I think most people like and appreciate, and will likely lose. I would really emphasize that if international accounting standards don t deliver a consistent platform for the Canadian life insurance industry, then I think people will move away from CALM methods, and methods that result in undue volatility, or more volatility than alternate methods. I wouldn t be surprised if things on the international side were deferred a little bit should a major Canadian company decide to report using US GAAP as their performance measure, and thereby lead analysts down that path in terms of valuing their company. Moderator Corner: Robert? Panelist Lefrançois: I think that if we were not considering harmonization with the US somehow in the future, and harmonization with international GAAP, we d probably have moved ahead with a project on life insurance accounting before now. We decided to wait because we didn t want to push Canadian life insurers to adopt something in advance of the others maybe you would have to change it back given some form of harmonized standard in the future. But I think in my view, and certainly in view of some other of my colleagues at the board, we believe this is probably not an optimal method. Maybe it s good for other purposes, but not for financial reporting. Let me tell you a short story. I am not an industry expert. Years ago, I spent several weeks discussing an issue with a life insurance company and we debated and debated. It was a very hard process. It was a good fight. After some time, my client agreed Délibérations de l Institut canadien des actuaires, Vol. XXXVI, n o 1, novembre 2004