Considerations for LIBOR transition and U.S. Variable Annuity Guarantee Valuations
This paper discusses risk-free curve selection and setting of the discounting spread (over the risk-free rate) for variable annuity fair valuation.
Oct. 9, 2008
Using a hedging strategy that emphasizes execution and simplicity, Milliman's FRM practice is helping insurers achieve predictable results and maintain confidence.
We asked Ken Mungan, Kamilla Svajgl, and Tamara Burden for perspective.
Q: How have guarantees held up to the current financial crisis?
Mungan: Most life insurers look to protect themselves against directional movements in major equity indices such as the S&P 500, the Russell 2000, or the Nasdaq, and also to protect themselves from movements in interest rates. On those two dimensions—declines in equity markets and declines in interest rates—life insurance companies have had hedge assets in place on their balance sheets, and the value of those assets has been increasing dramatically to offset their increased claims-payment obligations on variable annuities containing guarantees. Some companies also supplement their hedge programs through the use of put options; that's for volatility hedging (also called "vega") and is a longer-term hedge to protect against sustained market turbulence.
In general terms, everyone who was hedging is delighted that they were hedging. For the companies that were not hedging, obviously, the market has already gone down.
Burden: European companies have had a similar experience. The markets there are very volatile. And their hedging programs have performed well, though there are two things to note about their programs. First, most companies are not hedging volatility, and although current IFRS standards don’t require a full mark-to-market update of the volatilities, the recent market conditions have led to an increased volatility assumption for many companies. So even if their hedging programs have performed as expected over the last few weeks, the overall value of their reserves versus their assets has resulted in losses for them because the volatility assumption has had to change. U.S. companies as well, depending on their approaches to updating volatility assumptions, would be seeing that kind of loss without vega hedging. Second, many companies have seen very poor bond performance, primarily because of the credit market. As credit spreads widen, you see falls in the value of bonds currently on the books. Hopefully, credit spreads will tighten again and the bonds will regain much of the value they’ve lost.
Q: Have you discovered new risks that could have been hedged?
Burden: Many companies are not hedging volatility for long-term exposures. That's a known unhedged risk factor. So losses due to increasing the volatility assumption used to value the liabilities are an expected loss. Most companies are aware that they're not hedging credit risk, although the degree of credit spread widening seen in the past few weeks—even among highly rated corporates—is a lot higher than most anyone expected.
Mungan: The credit-risk exposure on bond funds in retirement savings accounts is a secondary or tertiary risk exposure for long-term guarantees. This fact highlights one of the basic principles that we've applied throughout all of our hedging programs since we began doing this in 2002: Complexity is its own source of risk. The kinds of retirement-savings products that invest in "plain vanilla" securities—stocks, bonds, treasury bonds, corporate bonds—have behaved in a fairly predictable fashion. We emphasize simplicity and transparency and avoid complexity. And this crisis, if anything, has validated that this is absolutely the right way to go. We use the simplest and most liquid instruments: the futures contracts on the major equity indices, plain vanilla put options on the major equity indices, and with the interest-rate exposures we tend to use treasury futures or plain vanilla interest-rate swaps. Unlike hedge funds or investment banks, which have experienced a solvency crisis because they were investing in complex securities that were impossible to value, our clients are using the simplest possible securities, where the value is immediately obvious and where those values have moved as expected.
Q: So hedging volatility is difficult—can volatility be reduced?
Mungan: Retirement-savings products and retirement-savings guarantees have a multidecade time horizon. There is no way to completely eliminate risk over such a long time horizon. That is simply not possible. So we work with insurers to minimize those risks.
There are really four basic dimensions of risk. There is the risk that the stock market will drop, the risk that interest rates will drop, the risk that the market will be volatile or turbulent, and that volatility will persist for a multiyear time horizon. And there are demographic risks—the risk, for example, that people with annuity products will live longer than expected. There are derivatives available in the capital markets to hedge the first two risks, and companies can reduce the risk associated with market volatility. Using traditional actuarial methods, insurers can spread the risk associated with mortality/longevity. Companies should recognize that they're reducing risk, not eliminating it. And then they need to measure their net exposure and constrain it within a band that they can handle given their capitalization. That kind of risk-management process is one that's in place at all of the major life insurance companies today and it’s worked really well throughout the crisis.
Q: Who benefits from the relative success of this approach to risk management?
Mungan: We're giving middle-class people who are the consumers of retirement-savings products access to basic financial risk-management techniques; that's our value proposition. Most people saving for retirement have no capability to understand or execute basic hedging strategies. We can do that on behalf of life insurance companies who then realize a tremendous economy of scale and pass on not only the service to these consumers but the efficiency that they get from doing this for a million people as part of an aggregate group. The people saving for retirement in the United States really need the service. They need simple, straightforward, and transparent products. They don’t need any of this complex stuff that was done by the investment banks or the hedge funds.
Q: How has the experience for policyholders with guarantees differed for those invested exclusively in the stock market?
Svajgl: Based on the September data, compared to where we were a month ago, account values have dropped for policyholders. But they're protected because they have guarantees. From the retirement gains point of view, they're still protected and the losses are not increasing. If policyholders were just invested in the stock market, they would have seen declines in their accounts and they would have seen losses. In these crazy few days that we've had, with the markets diving 700 points and more, we are seeing gains on hedge assets. Financial risk-management hedging is beneficial to policyholders because it gives them confidence to stay invested during turbulent times.
Mungan: Think about the typical advice a financial planner gives to somebody saving for retirement: "Stay the course." That advice becomes very difficult to follow if you have no guarantees and the market drops as it has. People get nervous. They want to feel like they're in control. Having a guarantee gives these people the peace of mind to know that they can stay the course. Even if the down market deepens and lengthens, the insurance companies will make good on the withdrawals that the person is entitled to.
Burden: This is even more obvious in markets with these kinds of wild swings, where it will go down 4% one day and back up 4% the next day. When the market goes up by 4% the day after a downturn, those who pulled out don’t get that gain. Even if they go back into the market the next day, they're out that 4%. They've lost it permanently.
Q: What was Monday, Sept. 29, like inside your practice? The market went through a record 777-point drop. Was it business as usual?
Svajgl: It was business as usual in terms of operating the way we normally would. We just had to watch the market more carefully. And the trades may have been more frequent than usual. Throughout the crisis we've had to communicate more regularly with our clients, to ensure they're aware of what’s going on and make sure that we're hedged and everything is up to date. But in terms of doing anything extraordinary, that didn't happen.
Mungan: We come into the office and prepare for massive turbulence in the stock market every single day. Ninety-nine times out of 100 that doesn't happen, but we follow the same process on days when the market is turbulent as when the market is calm. We've got everything worked out in advance, so that when the turbulence comes we’re prepared for it. That's really the only way to do it. You cannot have an event like a 777-point drop happen and then be running around trying to figure out what to do. Everything has to happen immediately and we're training for that every single day so people know exactly what to do: The processes have been worked out in advance, the technology has been developed over a decade, and then it's just a matter of executing.
Burden: In some cases the same processes are going on within the client organization—not even on the Milliman trading floor. The experience for these clients last week was very good. They’re using their own personnel on our systems and still getting good results.
Q: What's the difference between your approach to hedging and a hedge fund?
Mungan: In many ways the term "hedge funds" is a misnomer. Hedge funds pursue complex investment strategies for high-net-worth, very sophisticated investors. What we do is work with major life insurance companies to create retirement-savings products that give ordinary consumers access to basic hedging. That's one key difference. The second key point is from the life insurance company’s perspective: The retirement-savings guarantees that they offer have no liquidity on the guarantee whatsoever. There is no cash value associated with the guarantee. And because of that critically important contractual element, life insurance companies are not exposed to forced liquidation, while hedge funds are. Hedge funds have a prime brokerage relationship where they are making their investments with borrowed money; the investment bank that’s standing behind them is financing their trading activity. And as the markets go down, as the markets become turbulent and their mark-to-market positions deteriorate, then they are closed out with no control over that process. So that makes hedge funds fundamentally more risky than what the life insurance companies are doing, and we’ve seen that come through the results for the life insurance companies. As a general point, life insurance companies have not had solvency problems stemming from variable-annuity (VA) guarantees and the associated hedging. For example, several major insurers have recently held investor presentations and conducted conference calls on their third-quarter results, and they have noted that their hedging programs have performed very well.
Burden: Another way of thinking about it: Insurance companies are not exposed to a run on the bank. A variable annuity with a guarantee has two sources of value: the policyholder’s deposit, which is invested in equity and bond funds selected by the policyholder, and the guarantee. If a VA policyholder goes to the insurance company and says, "I want my money back right now," the cash value of the guarantee is zero, so the only money that they can get is the current value of the deposit. And that money is in a separate account for them. It's theirs. It has not been borrowed or leveraged or invested in something that's illiquid or anything. Effectively their investment is being held like cash in the vault at the insurance company. With a hedge fund, when investors ask for their money back, the hedge fund has to unwind all of the positions they have in order to change it back into cash to hand to the investor. If they can’t do this quickly enough, they collapse.
Mungan: In the crash of 1987, a lot of pension plans were pursuing portfolio insurance strategies. Those strategies did not work. What we're doing with life insurance companies today is completely different from that because the obligations of the insurance companies are spread over 30 years. The hedging objective is completely different with today's strategies.
Q: Does the recent criticism of mark-to-market accounting fit into FRM's focus on simplicity and transparency?
Mungan: The mark-to-market debate highlights one key difference between life insurance companies and investment banks and hedge funds. A hedging program that's operated as part of a strategy for a hedge fund, for example, is constantly exposed to forced liquidation. Because investment banks and hedge funds are highly leveraged, they are essentially investing with borrowed money. They're exposed. When the market goes down, they're exposed to having all of their positions liquidated or closed out on at no notice in a manner that's completely beyond their control. Variable annuities are different. Contractually, VA guarantees have no cash value. That means that, as the market goes down, policyholders cannot line up outside of the life insurance company's door and demand some sort of mark-to-market of their retirement-savings guarantee. The insurance company's actual exposure is to the long-term payment of claims. And so the valuation on the insurance company's balance sheet is centered around a fair value for that ultimate long-term obligation. In that sense, the P&Ls of insurance companies have been more stable than the financial results for hedge funds or investment banks. As a result, the solvency and security of our clients is much stronger today than banks.
Svajgl: I don't see any panic among our clients. They all are performing as expected. Insurance companies that have used hedging aren't seeing the effects that we're seeing with banks or with some of the structured derivatives.
Q: What lessons will the insurance industry take from this experience?
Mungan: Hedging is likely to become more widely used, though perhaps not in the way you'd anticipate. Hedging is already ubiquitous throughout retirement-savings guarantees and the life insurance industry. The top 25 VA writers all have hedge programs. The hedging programs will increase primarily because sales are going to increase.
Burden: If you look back, there was an increase in sales of VA products with guarantees after the dot-com bust. I expect we will see an even greater interest in guarantees now. The dot-com bust affected a lot of people, but there were many others who thought "Well, I wasn't invested in technology so I'm fine." But this crisis is affecting everything in the market, not just a few segments. Diversification isn't the answer to a market like this one; guarantees are. This will drive the interest in VA-style products.
Q: Can the VA model be applied to other retirement products?
Mungan: We're developing something called the retirement guarantee network (RGN) that makes these types of guarantees available to people more broadly. The interest in RGN has increased dramatically in the past month, especially from the major mutual fund companies. We'll soon see guaranteed withdrawal benefits available on 401(k) accounts, IRA accounts, and mutual fund accounts.
Svajgl: When we can build guarantees into 401(k)s it will allow people to more aggressively put more money in equity funds. This allows participation in an upward scenario when the market goes up while allowing protection when the market goes down.