Since their introduction in the 1980s, variable annuities (VAs) have evolved significantly. In their basic form they resemble classical mutual funds, but with the added advantages of periodic life-long payments and tax deferrals on investments and gains.

Today, most VAs include a death-benefit rider that lets beneficiaries receive a benefit if the policyholder dies before payments are complete. Many also offer sophisticated living-benefit (wealth preservation) riders, such as a guaranteed minimum withdrawal benefit. Other features include a guaranteed lifetime withdrawal or a guaranteed minimum income benefit.

The development of such a diversity of attractive guarantees catering to individual concerns is one reason why variable annuities (VAs) have boomed in the North American insurance market in the last decade. There they have overtaken traditional fixed annuities to become the primary form of protected investment. Similarly, in Japan the VA market has grown from $1.3 billion in 2001, up to more than $140 billion in March 2008.

Now, with VAs making inroads into European markets, the current financial climate has revealed design weaknesses. Unfortunately, the severe decline in equity values means most guarantees embedded in VAs have become “in-the-money.” That is, the guaranteed benefits have greater value than the assets accumulated in the policyholder’s account balances.

According to a recent working paper from risklab, an investment and risk advisory boutique,* the market downturn will lead to greater demand for variable annuity products, even as it has exposed profitability and risk management issues.

Dr. Bernhard Brunner, senior vice president at risklab, says “the general market downturn has underlined the customer perception of the value of guaranteed products.” Providers now face challenges in terms of profitability. Written by Dr. Brunner and Mikhail Krayzler, The Variable Annuities Dilemma paper points out that the insurer’s fee is based on actual account values – and these have dropped considerably as equity prices have fallen. One consequence is that increased equity volatility is driving the guarantee costs demanded by VA providers higher.

As high product fees are unattractive to purchasers, the paper argues, insurance companies will need to revise product and rider designs. Modifications are likely to include increased restrictions on the risk of the underlying funds and the avoidance of expensive (to the provider) riders, the feature most attractive to purchasers.

However, the most important implications relate to risk management and hedging. “As guarantees become more valuable and the possibility increases that guarantees will end up in-the-money, insurance companies will be forced to increase risk-based capital requirements,” explains Dr. Brunner. “This means sophisticated hedging programs to counter increased liabilities or reinsurance arrangements to transfer risk will become more critical.”

While the paper concludes that existing programs have been largely effective in mitigating the increase in liabilities, a relative comparison reveals considerable differences in terms of hedging performance. Most programs apply dynamic hedging strategies in order to dynamically manage a portfolio of derivatives with sensitivities corresponding to VA liabilities. This, in industry parlance, is known as “Greeks hedging.”

Depending on the type of sensitivities or risks to be hedged, a variety of dynamic hedging approaches are common. Delta hedging insures against movements in the underlying fund price, rho hedging that protects against negative movements in risk free interest rates, and vega hedging that takes changes in the volatility into account. The so called “higher-order Greeks” or “cross-Greeks” are often used in more comprehensive hedging programs.

According to Dr. Brunner, high market volatility has generated substantial losses to programs that failed to include vega hedging. “It is often argued that these losses are not realized cash losses, yet the impact on the required risk-based capital can be significant,” says Dr. Brunner.

Recent experiences could result in risk management being broadened to include pricing and product design phases. Potential risks may be eliminated by the ability to change guarantee costs and by placing restrictions on fund investments that reduce volatility and diversify the portfolio. Placing caps on benefit levels, as well as different callable features (such as the possibility for the VA provider to cancel the guarantee), would help to reduce the costs of expensive guarantee riders.

In terms of investment funds underlying the VA contracts, the authors observe a trend to use funds that employ risk mitigation strategies. These funds include volatility target funds and funds with a built-in downside protection. This implies the shifting of guarantee costs from the VA level to the fund level. In order to avoid additional guarantee costs in the VA contracts it is essential to model these funds properly.

* risklab is owned by Allianz Global Investors

Published by PROJECT M in September 2009