Transcript of a Press Conference on the Analytic Chapters of the Global Financial Stability Report

April 11, 2012

Washington, D.C.
Wednesday, April 11, 2012

José Viñals, Financial Counsellor and Director of the Monetary and Capital Markets Department, IMF
Laura Kodres, Assistant Director, Monetary and Capital Markets Department, IMF
Erik Oppers, Deputy Chief and lead author of Chapter 4, Global Financial Stability Division, IMF
Silvia Iorgova, Lead author on Chapter 3, Global Financial Stability Division, IMF
Andreas Adriano, External Relations Department

MR. ADRIANO: Good morning and thank you all for coming for this press conference for the launch of the Analytic Chapters of the Global Financial Stability Report. I'm Andreas Adriano with the IMF External Relations Department.

Today with us we have the director of the Monetary and Capital Markets, José Viñals, the head of the Global Financial Stability Division, Laura Kodres; then, Silvia Iorgova, the lead author of Chapter 3 “Safe Assets, Financial System Cornerstone?”; and the lead author of Chapter 4, Erik Oppers. The chapter is titled, “Financial Impact of Longevity Risk.”

I would like to also salute those watching us online. Mr. Viñals will have some opening remarks and then Laura will go over the main points of both chapters. Unfortunately, Mr. Viñals will not be able to stay until the end of the press conference, but I would like to remind you he will have a press conference next Wednesday, a week from today, April 18th, to present the main chapters of the Global Financial Stability Report. So Mr. Viñals, please.

MR. VIÑALS: Thank you very much. Good morning to all of you. Today as has been our tradition for the past few years, we publish the analytic chapters of our Global Financial Stability Report a week or so ahead of the publication of the conjunctural chapters, which as was said will be discussed next Wednesday.

This spring we have chosen two very topical and important issues. The first has to do with safe assets, and the second has to do with the financial impact of longevity risk. The first chapter, the one that talks about safe assets, is a very immediate concern. The first question we need to ask is what constitutes a safe asset in the current financial environment? And this is something which is not as straightforward as we used to think a few years ago. For example, during this present crisis we have found out that a traditional safe asset, which were government bonds, in many cases they have lost some of the safety characteristics that used to be believed by people. This is something which we look at in the chapter.

At the same time, we document the rising demand for safe assets and the forces which are pushing towards a declining supply, suggesting that the price of safety will likely increase. And the key will be to ensure that this process takes place in an orderly manner without excessive volatility, and the chapter outlines a number of policy measures to smooth demand-side pressures on safe assets, and at the same time to replace the lost supply of safe assets.

The other chapter which deals with longevity risks points out that beyond the large upcoming costs of aging, which is a well-known issue, there are additional financial and fiscal costs that come from people living longer than it is now expected. And it turns out that as we document, these costs are very large.

The root of the problem is that how long people live has been consistently underestimated in the past, and although we would all agree that living longer is a good thing, it carries a financial risk. We can run out of money in retirement. And what the chapter does is to look at how much our longer lives will likely cost us as individuals, but also how much it will cost to corporations which fund private pension schemes and to governments which run public pension schemes, Social Security systems, and so on. And it argues that we need to worry now about these future longevity risks and we need to start adjusting now so that the costs do not overwhelm us later.

These are the main themes that we developed in these two analytical chapters. And now, I am going to give the floor to my colleague, Laura Kodres, who is going to give you more on the specifics of each of them.

MS. KODRES: Thank you very much, José. Allow me to highlight just a few of the key messages from each of the chapters to give you an idea of where we're headed with these. In the chapter on safe assets we document the different sources of demand and supply of safe assets. In essence, what we aim to do here is connect the dots to form a broad-brushed picture of the imbalances, something that most discussions of this topic don't do.

To accomplish this effectively, it is necessary to identify what constitutes a safe asset in the first place. In principle, a safe asset can mean something very different to specific types of investors due to their different objectives. To most investors, a safe asset is an asset with low credit and market risks. Hence, this is why government debt such as U.S. Treasuries, German bunds, Japanese government bonds have traditionally played a role as the safest of assets.

Taking this notion of safety as a starting point, we examine how the demand for safe assets has risen and what pressures would propel it to rise further. One factor increasing demand is general uncertainty about the economic and financial environment and the actions of central banks in advanced economies to provide liquidity to the system and to keep interest rates low, direct results of the ongoing crisis.

Another factor is related to new regulatory reforms. For instance, those requiring more safe assets to be held by banks, and those requiring safe collateral to be held in central counterparties to reduce the risks in over-the-counter derivatives markets. While these reforms are needed to build safer institutions over time, they will inevitably drive up demand for safe assets as they are implemented.

At the same time that demand is rising, the supply of traditional and not so traditional safe assets is falling. The creditworthiness of sovereign debt in a number of advanced economies has suffered from fiscal strains, impairing some of their safety features. Other private sources of safe assets that were produced in large numbers before the crisis, such as securitized products, turned out not to be so safe after all.

Both the increased demand and falling supplies suggest a higher price of safe assets going forward. This is not an undesirable outcome, in part because the price of safety was arguably too low before the crisis. Nonetheless, it's important to make sure that the adjustment process proceeds smoothly.

On the demand side, a smooth transition to higher safe asset prices can be supported by regulations that do not label assets as either safe or unsafe, but rather differentiate the assets by their safety features along a continuum. In addition, regulations should be implemented so that institutions' holdings of safe assets build up slowly and steadily to avoid destabilizing price effects that may result if the timing of changes is too abrupt.

On the supply side, we advocate that those advanced economy governments whose debt is no longer perceived as safe, and those in danger of it being so, commit to strategies of reducing debt levels and strengthening debt management. Obviously, the primary reason is to lower their financing costs and rollover risks. However, by doing so they will also help replace lost safe asset supply. The private sector can also contribute. It should not be impeded from issuing safe assets as long as those safe assets that are issued are based on transparent and well-conceived methods so that investors can price those assets appropriately.

Turning to the chapter on longevity risk, as José noted longevity risk is the risk associated with people living longer than expected. Let me say right up front that living a longer life than expected is generally a good thing. So, calling this a risk does seem a bit odd. But if we consider the issue from the financial planning point of view, there is a clear risk associated with living longer than expected. You might run out of money in retirement. And anticipating that you live longer than expected is not unreasonable. Nearly all countries have continuously underestimated how long people will live by an average of about three years.

Unexpected longevity risk is not just a risk for individuals, but also for those that provide guaranteed retirement income, defined benefit pension plans, insurance companies that provide annuities and governments if they provide universal old age pensions. And it is not just for advanced economies either. There is a number of emerging market economies facing this risk as well.

If you think about the global extent of longevity risk, it is really our own individual retirement finance problem multiplied a few billion times. This gets very large. The chapter presents estimates that the size of longevity risk amounts to roughly about 50 percent of 2010 GDP in advanced economies, and 25 percent of GDP in emerging market economies. In dollars, this runs into the trillions.

Given those estimates, the chapter suggests various solutions. One effective way to offset the risk is to encourage people to work longer. This can be achieved by linking the retirement age to expected developments in longevity. This would serve to keep the expected number of years in retirement more or less constant, keeping retirement affordable. We can also use financial markets to transfer those risks. Through various types of strategies and instruments it may be possible to move the risk and the potential costs from those weighed down by it to some who might be better able to bear it.

Let me conclude by stressing that longevity risk is not some way off in the future type of risk. True, it is not on the front page like many of the headline financial stability issues of today, but it should not be relegated to the back page.

The reason is that longevity risk is like other pension issues. The longer you ignore it, the more difficult it becomes to resolve. Think about your own pension planning process. We are all told to start saving for retirement in our twenties. Those that wait until they are in their forties or their fifties to start saving find it very difficult to make that up. The same holds for longevity risk. Relatively modest measures can be taken now to start dealing with it, but the longer we wait the more dramatic and potentially disruptive measures have to be taken in the future. We argue in the chapter that the time to act is now. Thank you. I'll take some questions.

QUESTION: I think in the report you mention that by 2016, 9 trillion safest assets might be removed from the supply. I just wondered, could you elaborate more about the time period? How long we might expect the price may go up and when it might be reversed? Thank you.

MS. KODRES: In terms of this estimate for the possibility of the removal of safe assets, let me note that what we accomplished there was to try to look at various countries who were perhaps at risk or had already lost some of their safe asset status, use our world economic outlook forecasts for the debt that they might be needing in the future, gross debt, and then calculated what it would mean if those continued to be in the status that they are now. So, there are a couple of different measures in the chart, which is 3.13, if you're interested in looking at that.

I would like to stress that those are really estimates based on a current conjuncture in terms of what we expect to happen going forward. If things improve, obviously things will look slightly different than that. The point that we're trying to make is that the supply of safe assets in fact has degraded, and there will be some time before that supply will come back onto the market.

Do you want to add something, Sylvia?

MS. IORGOVA: Sure. The notion is really that it's not inconceivable for certain countries to lose their safety, and we actually wanted to get a stylized view and see how large the problem may be if that were to happen.

QUESTION: The study of the U.S. system I think ended in 2007, the U.S. Department of Labor figures, and you said a majority of those were using 25-year-old data or longer for life expectancy. Can you just give us an update of where the U.S. is on actuarial -- those mortality tables? Are U.S. pension plans still using very old data?

Secondly, you know, a large part of the political debate over cutting U.S. deficit and debt levels is obviously over Social Security costs, retirement costs. Does this study mean that they're really arguing over the wrong numbers? If you're saying that you could add another 7 trillion dollars in costs to entitlements? Isn't it a lot bigger, the entitlements a lot bigger? And finally, on Table 4.2 you look at the longevity risk and fiscal challenges in selected countries. Is my understanding correct that you are saying that Japan could be facing costs if their 3-year is rule is correct of 3-1/2 times 2010 GDP?

MS. KODRES: Thank you. A number of questions. Let me just begin by starting to frame the first question so that people that are listening can understand. We do a study of the U.S. system of pensions using a form called 5500. That form is collected by the Department of Labor, and the most recent data that the Department of Labor produces is in 2007. So, we took as recent data as we could possibly take for the purposes of our study. I am going to let Erik answer the questions, I just wanted to frame that one so that people understand what we're talking about there.

MR. OPPERS: On the mortality assumptions we have a table in the report, Table 4.3, that indicates in detail what mortality tables are used over time by the U.S. pension plans. And the trend that you see in that is that over time pension plans are using more updated tables. That is the trend that you see out of this table. It is true that until recently a large number of pension plans were using the 1983 GAM table, which is a bit outdated, but over time you see pension plans updating their mortality assumptions.

QUESTION: So just to clarify, are you extrapolating then, that in the five years since then, U.S. public and private planners are still likely using old mortality data? That seems to be the implication you're using by implying the three-year rule.

MR. OPPERS: No, I think the trend that we see in the data until 2007 will probably have been extended. So since 2007, pension plans have further updated their mortality assumptions.

MS. KODRES: That said, in the study and in the table there's an "other" category which tells us the proportion of pension funds that don't declare what table they use, and there is a growing number of pension plans that don't declare which table that they're using.

Because that number is so large, we thought maybe our results might be biased in case they're using a more recent table. So we assumed that that 57 percent of pension plans in the table labeled “other” used the most recent actuarial table available, and we find our results just don't change very much. So as it turns out, it really does mean that there is about a 3 percent increase in total liabilities each year or – said another way, in the study we show that there's a 9 percent increase, I believe, in the typical liabilities of U.S. pensions due to a 3-year longevity shock. So we tried to factor in the fact that we don't know all the tables that the pensions are currently using. Erik?

MR. OPPERS: Yes. On the assumptions that are being used in the debate on Social Security, I think it is reasonable to use a baseline number in that debate. What we point out in the report is that beyond the baseline numbers there is that longevity risk that is a contingent risk, a contingent liability, that should be taken into account in the discussions but it is still true that taking a baseline assumption for this discussion I think is reasonable.

On your question on Japan, what we show in Table 4.2 is some more detail on the size of longevity risk. There is some detail on different countries. I think in general the table shows that longevity risk is a big issue for most countries in the world. It is a bigger risk for countries that have more aging populations, and that does include Japan.

QUESTION: I wonder if you can have some more specific analysis on the emerging market situation regarding the longevity risk for the public account. Particularly I wanted to know if you did not consider the Brazilian situation in this study because we have in Brazil nowadays a very clear acceleration in the longevity and increasing new population. Do you have some more specific information on both the emerging market situation and Brazil also?

MS. KODRES: Let me just note that we do have Figure 4.2 which shows on average how emerging markets are situated in terms of their longevity risk, so we particularly pool together a number of emerging markets and we show that maybe not surprisingly the longevity risk itself is lower in part because the proportion of aging individuals in these countries is smaller than many of the advanced economies. But I think what was particularly impressive to us is that the additional cost of longevity is about the same. It's quite a large proportion of the existing risk. So even though those populations have less aging individuals in them at the moment, that increase of 3 years which we assumed does matter.

In terms of our analysis of individual countries, there is a shortage of data to be able to say a lot about how this would be impacting individual counties. We have two tables in the report that examine individual countries. One is 4.2 that I pointed out, and then we have another toward the end of the document that looks at those countries that have linked their longevity risk to their retirement plans. I think what's striking about that last table is that there are a large number of emerging markets that have started to address this risk. I think particularly there are several that are well along in terms of addressing this risk. Chile is one that has completely linked their payments to longevity risk through a defined contribution plan, so I think that might be a place to look for individual country results. Erik, do you have anything you'd like to add?

MR. OPPERS: Perhaps just some detail on what you said at the beginning which is that the size of longevity risk is pretty much the same in emerging markets when you compare to the cost of aging. Both for advanced economies and for emerging markets, longevity risk adds about half again to the already large cost of aging. And in advanced economies this comes out to be about 50 percent of 2010 GDP, and 25 percent of 2010 GDP for emerging markets.

MR. ADRIANO: I'd like to take a question online exactly on that point, Erik. Could you explain and provide figures if possible on the calculation that 3 years longer life expectancy would increase the cost of aging by 50 percent?

MR. OPPERS: What we did there is we looked at how much resources do people need in retirement and it is a common assumption to say between 60 and 80 percent of pre-retirement income is what people need in retirement. What we did is we said if people live on average 3 years longer, how much extra money would they need, 60 to 80 percent of their pre-retirement income, you add that up over the population and that's how you get our estimates.

MS. KODRES: I'd like to just make a clarification point. I just would like to make sure that people understand that in our safe assets chapter when we recommend that countries address some of their fiscal strains that it's not for the purposes of producing safe assets, that it's not the government's job, if you will, to produce debt in order to have safe assets out there in the economy. The purpose of course is to fund the government and the purpose from the treasury's point of view is to lower the funding costs of the debt and that's the first order reason for issuing debt. It has an ancillary effect that if there are good fiscal fundamentals underpinning the debt, then that debt can serve a purpose of being a safe asset and that safe asset can be used in a number of roles.

I would point to another table that we have in that chapter where we identify all of the different demand and supply roles that safe assets play and attempt to give a feel for whether they're going to go up or down. So these safe assets serve a number of roles. They serve as collateral, they serve in the purposes for prudential regulations, can be stored by banks and by insurance companies, they can be used by pension funds, they are used by reserve managers and so there is a large number of roles for safe assets but we want to make clear that the reason for our recommendation on the fiscal side is not to produce safe assets per se, but to produce really good fiscal fundamentals that will underpin good debt characteristics.

Let me at the same time focus a little bit on the production of safe asset by the private sector. Again we've focused on sovereign debt because traditionally that has been considered the safest of assets, but there is no reason that the private sector can't produce safe assets also to help the supply of safe assets that can be traded. I would note that high-grade corporate debt is now considered a safer asset in some countries relative to the sovereign debt.

Again there is possibility of production of safe assets by the corporate sector that could be used in terms of collateral, for example. And the private sector can produce safe assets in the sense of securitization and covered bonds as well. Securitization would need to be restarted on a sounder footing than we saw pre-crisis where we saw a number of issues having to do with securitization which made those assets less safe than many people had anticipated. We have a previous chapter that looks at how to restart securitization in a safe manner in a previous GFSR. But we should not forget that the private sector can also produce safe assets as well.

QUESTION: Can you give us some sort of timeline on the longevity issue? You recommend a mandatory rise in retirement age for governments facing these longevity risk issues. What sort of timeline are we talking about there? Then secondly, you raise the possibility of disturbing fiscal sustainability and financial markets. Also, what sort of timeline are we talking about there? Are we talking about in the next couple of years or are we talking about in the next couple of decades?

MR. OPPERS: I think the answer to both is the same and it is really what Laura said in her opening remarks. The timeline here is how you choose it, but the earlier you start with tackling longevity risk the easier it is. It's really like the issue of your retirement savings. When you start that, what is the timeline? The timeline is you should start in your twenties and if you start in your twenties it's much easier, it's not disruptive, you can save for your retirement. If you start in your forties or fifties, you're too late, and there is some stretch there, but the earlier you start the better. So that's really the only timeline that we offer.

QUESTION: I understand that and I understand that it's a safe sort of, please forgive me, vague sort of answer. I understand the nature of it. I understand the rationality of it. But I think surely you can give some -- if we don't say in the U.S. put forward a plan to address this issue in the next 2 years then are debt markets going to look at the U.S. debt and say maybe we should reassess their credit rating again because of this issue. Or are you saying we have 10 years where we can really start to build this issue without causing any sort of financial or fiscal turmoil?

MR. OPPERS: We think that the chance of this causing disruption in markets now is fairly low. We don't see that happening. But again the earlier you start the better. Again, the chance that this will lead to market disruptions in the next few years is really very low.

MS. KODRES: Let me add that while it is low in terms of some sort of disruptive, discontinuous kind of price response, it is a growing vulnerability and as we can note in the pre-crisis period, we had a number of those vulnerabilities building and I think that behooves us to attack this earlier rather than later. Let me also note that a World Economic Outlook (WEO) chapter quite a while ago, I'm not sure exactly when, noted that many structural reforms to economies tended to happen during crisis periods, that there tended to be more movement on some of these very difficult structural issues in part because of the shakeup of some sort of crisis in the country. In that sense, while we're not at what I would call it an optimal time in many respects, but nonetheless there is a lot more attention to structural policies now than perhaps there would be if things were going very smoothly.

MR. ADRIANO: Thank you, Laura. I'll take one question online then we'll go back to the room. Will the reduction of safer assets create a bubble and drive up prices?

MS. IORGOVA: In principle it could and this is likely to happen if the process actually is left to go too far. But the immediate concern truly is much more about ensuring a smoother transition to a higher price. Necessarily we don't take a view that a bubble is likely to happen per se. We are much more concerned about introducing the right policies to actually move to a new price of safety that is justified by the underlying risks associated with safety.

MS. KODRES: Let me make an additional point to that which is that if safe assets are not so discrete, safe and unsafe, then you will also be less likely to develop bubbles in particular assets. So if an asset is deemed to be the only safe asset and people need to buy that one, you can see that indeed the price would be bid up and its yield would come down and that could encourage this kind of bubble-like behavior. So the notion that we've recommended that there be more of a continuum of safe assets in terms of prudential regulation will help alleviate some of these types of issues about the development of bubbles or the development of situations which become vulnerable later on to abrupt changes.

QUESTION: I was wondering how long should countries raise their retirement age, countries like Spain, for example, the decision recently to raise the age until 67 years old. And I wonder does it sound reasonable? Is that a good age to be able to face the increasing cost of retirement?

MR. OPPERS: Our recommendation is that countries let the retirement age rise with developments in longevity so that could be a good guide. If longevity goes up by 1 year, it would be a good thing to raise the retirement by 1 year as well. Let me also make the point that if you do this once, if you say we'll raise it from 65 to 67 and then we're done, that's not how we recommend it. This is a dynamic process, people keep living longer and longer and the retirement age should keep pace with that. So this is not a matter of raising the retirement age once, but to have it keep pace. Perhaps have an automatic adjustment of the retirement age with adjustments in longevity.

QUESTION: Do you know in the case of Spain how much it's going to increase?

MR. OPPERS: No, I don't. We didn't look into specific cases. But I think the general rule is relatively specific. Again, if longevity goes up by 1 year, raise your retirement age by 1 year. I think that that could be a very good, practical and sometimes hopefully automatic guide.

MS. KODRES: It's useful to note too that in many countries they don't collect enough information to know exactly what the longevity is or what the longevity risk is. So a part of our message in the chapter is for governments to collect more information about that type of risk and that will allow the private sector to be able to price the transfer of that risk more easily as well. So it serves both the purpose of maybe guiding the policy in terms of recommendations about the retirement age, but it also allows the private market to price that kind of risk and transfer it more effectively.

MR. OPPERS: And perhaps a final point. Raising the retirement age could be seen as a politically difficult thing to do, and if you institute a rule that automatically links the retirement age to increases in longevity, that may make the political process easier so that you don't have to revisit increasing the retirement age every time this comes up, but you just institute that rule and have that automatic adjustment then.

MR. ADRIANO: Thank you, Erik and Laura. I would like to take another question online on Chapter 3 this time: At this moment in the markets how will the safety of eurobonds be perceived and how successful could they be?

MS. KODRES: This obviously is a very sensitive issue, but let me give an overview of what our safe assets chapter would imply about the issuance of these kinds of bonds. Generally the issuance of bonds that depend on the ability and the willingness of a group of countries to jointly and severally honor their payment obligations could in fact be a source of safe asset production. What would happen is that the sharing of this credit worthiness would help reduce the chances of a sharp change in the price of that asset based on perhaps the characteristics or the fundamentals of any one country that's involved in the bonds, and so that allows perhaps some smoothing of the price behavior because it's a pooled risk relative to an individual country.

However, the securities are only going to be considered safe to the extent that there is some sort of framework internally to ensure that those bonds would be in fact paid off on the terms that they were issued. As well, the countries involved in that, some will benefit in terms of having a lower cost of funding for those particular securities, and other countries whose credit worthiness is probably higher than the average, if you will, would find that the costs are slightly higher relative to issuing their own debt. But these sorts of pooling types of securities have been done before. In fact, there are lots of cases in which pooling of risk can help on average overall. Another example is the use of these centralized counterparties where all of the clearing members essentially pay into their default funds and each one is severally and jointly responsible in case the clearing facility has a difficulty. Again, that sharing and pooling of risk makes it attractive and makes the system stronger relative to individual risks of the individual clearing members or in this case of the individual countries involved.

MR. ADRIANO: Thank you. We'll take one final question online: Could countries like Brazil, India and China which are seeing wealth increase be facing the possibility of longevity shocks?

MR. OPPERS: Again I think in the report we make the general point that this is not just an issue for advanced economies; this is also very much an issue that is faced by emerging economies. There as well the additional costs of longevity and about one-half again to the cost of aging which also in emerging markets are relatively large.

MS. KODRES: I think it's worth noting that the advances in longevity are likely higher in emerging markets because much of the longevity increases have been based on medical advances. So the extent that medicine is more widely available and health care is improved in those countries, they are likely to see larger increases in longevity. From the point of view of longevity risk that we look at, longevity risk is the costs in retirement of the financial implications of longevity increasing. Because emerging markets start out with say less individuals covered in such retirement plans, that's sort of an offset in terms of the costs, and that's why we see that there is the demographic issue that they have less older people in the pipeline, there is less usage of formal retirement programs, and so again the fiscal and the financial costs are somewhat lower, but the advances are likely higher.

What this really means is that the individuals bear more of the cost of longevity risk in emerging markets because these formal systems are less in place. So the risk doesn't go away. It's just borne by a different party. And one of the main points that we want to make in this chapter is that these costs are high enough that they can't be absorbed by any one sector in and of themselves. They can't be absorbed only by government, only by corporations or only by individuals. And so that risk has to be acknowledged, it has to be assessed and then it has to be shared in a suitable way.

MR. ADRIANO: Thank you, Laura. If there are no more questions in the room, I'd like to wrap up this press conference. Thank you very much for coming and for watching us online. I'd like to remind you of the press conference next week for Chapters 1 and 2 of the Global Financial Stability Report to be presented by Director José Viñals. Thank you very much and have a good day.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6220 Phone: 202-623-7100