The VA challenge

Variable annuities have revitalised the US life industry. But are they suitable for the European market? Aaron Woolner reports

From the smoothing of with-profits payouts to the hidden reserves within balance sheets, the European life industry is famously opaque to outsiders. But faced with sceptical consumers, together with government pressure to bolster private pension provision in a transparent way, the industry is now beginning to embrace a US success story.

In early 2006, multinational life and pensions provider Axa says that it will launch its first variable annuity (VA) product in Germany. Meanwhile, there are strong indications that heavyweight competitors to Axa are planning to bring their own versions onto the market soon after. What makes the products stand out is the explicit presence of guarantees - which have long been a ubiquitous and popular feature of the European life savings product landscape.

Despite this common feature, VAs are in a sense the antithesis of the European with-profits tradition. Foregoing the risk-sharing and mutuality concepts that underpin this tradition (and provide an excuse for its opacity), they are structured financial products. Instead of giving policyholders the expectation of bonuses from a single underlying fund, VAs offer payouts from transparent underlyings such as equity indices or mutual funds.

They combine explicitly priced financial risk and life insurance elements in a building block format designed to be endlessly customised for individuals. For providers, this building block quality means that marginal economic risk capital is identifiable at individual customer level, and is more easily shifted into wholesale markets.

Although the first VA product was launched over half a century ago by a teacher's retirement fund in the US in 1952, the product has only taken off in recent years. According to figures by the US National Association of Variable Annuities (NAVA), by the end of Q3 2005 there were over $1.2 trillion of net assets invested in VA products. And with total sales of the product currently running at $120 billion per year, demand shows no sign of abating.

And the success in the US has been matched by its popularity in Japan - the world's second largest insurance market but also one that is famed for its conservatism - it appears that the safety net of guarantees this product offers has struck a chord with the country's risk averse investors. VAs were only launched onto the Japanese market in 1999, but in this short amount of time the assets it controls, according to a report by JP Morgan, are already worth $70 billion.

Mathew Andre, chief of investor relations at Axa, is quick to point out that while comparable, the markets in North America, Japan and Europe are by no means identical: "Each market is different but you can have a similar chassis for the product which you then adjust to the individual requirements. Just because it is successful in the US does not mean it will automatically be successful over here, but it is still worth looking at it to see if you can adapt it."

But while differences exist, particularly in tax treatments, the underlying demographics each region faces are similar - a rapidly aging population that, due to improvements in mortality, is facing up to a long-term retirement with doubts over its financial security amid declining occupational-defined benefit and state pension provision.

These demographic pressures were also the catalyst for VAs' move from its status as a niche product to the mainstream at the start of the 1990s. At this stage in its development it was purely a unit-linked style investment product, which offered exposure to equity markets and the benefit of tax-deferred status that the US grants annuity products.

Initially, all investment risk was passed on to consumers - but this situation soon changed, says Craig Raymond, chief risk officer for US-based insurers the Hartford, as increasing competition led to companies fighting for market share by the introduction of the cheerily titled "death benefits" followed by other innovations. These innovations sought to address a perceived drawback of traditional annuities: that while the products were fairly priced from an actuarial perspective, customers who died soon after signing their contracts lost all their capital.

"In the early 1990s, companies started to enhance the death benefits on the contracts - primarily guaranteeing the account value or that your premium would be returned if you died prior to annuitisation on the contract. Then these were enhanced by guarantees that meant after five or seven years of issue, the death benefit would be reset equal to the account value at that point." The number and style of these guarantees multiplied and the introduction of "living benefits" - that guaranteed the value of policyholders if they were alive to see their policies mature - resulted in consumer interest rocketing. (For a full explanation of types of guarantees available see box).

It would be easy to dismiss these guarantees as more marketing devices rather than something of concrete value to customers - a point acknowledged by Raymond. "The guarantee that you can get your money back - without interest - after 15 years is not a particularly strong one, but once you add in the death and living benefits it gives you peace of mind that if bad things happen then there is a floor to your investment."

And it appears that this peace of mind and transparent approach to pricing is a significant pull for consumers - the strength of which was a major factor in Axa's decision to introduce the VA concept to Europe after selling it in the US for over a decade. According to Andre, the existence of these guarantees kept consumer demand high in times of economic uncertainty. "The success of the VA products in the US at the time of the market downturn in 2003 was quite striking. When we researched this we discovered the reason the sales held up was the faith consumers placed in the guarantees, and we decided then to look into launching them onto the European market."

The public face of the VA story is the offer of rider guarantees that consumers can take or leave as they wish, but the engine room that powers it is the advances in risk management that have occurred over the last 10 to 15 years.

The positive impression of the impact that guarantees have on consumer confidence is uniform among those with an interest in the VA industry, but views on how to manage the risk they throw up are more diverse. According to Gary Finkelstein, UK practice leader of the financial risk management team at actuarial consultants Milliman, new techniques in risk management of VAs are "a key factor for the product's continued success. Modern risk management allows the guarantees to be managed in a capital efficient way, and therefore they can be priced at a level that is attractive to both consumers and insurers".

"In the early 1990s, these guarantees were generally not risk managed in the way they are today. The insurer may have reinsured, remained naked, or perhaps followed very crude hedging that merely involved taking an offsetting position in its balance sheet. While it is possible for the insurance company to partially hedge the market risk this way, they are still exposed to changes in the offsetting positions required (delta and gamma risks), volatility (vega) and interest rate (rho) risks, which can now be managed by trading in the markets or with banks." The path towards risk management self-sufficiency has already been trodden by the Hartford, and Raymond is confident that the system the Connecticut-based insurance company has put in place is more than capable of mastering "three Greek hedging".

"There are (variable annuity) companies that just hedge delta - the sensitivity to changes in the market - but this is a risky strategy. We feel strongly that given the size of our portfolio (currently standing at $24 billion) we needed to robustly hedge a wide range of Greeks, vega and rho included." While companies like Milliman provide software platforms for managing VA risk, Raymond believes that the commercial packages do not go far enough.

The Hartford has built up a bespoke internal hedging model that is made up of a combination of in-house developed software and commercially available programs that have had to be adapted. "I don't think that there is software available to do this that meets our standards," says Raymond. The internal hedge programs are run on an internal supercomputer made up of a series of servers linking around 500 high-powered PCs.

The result of this complex system is a sophisticated internal hedging structure that is managed on a daily basis and contains a portfolio of options and futures that essentially contains a mirror image of the benefits that the company has sold the policyholder. Raymond says: "We have offset the risks we have in the withdrawal benefit guarantee with a hedge portfolio that has similar but opposite risks."

"The easiest way to think of this is that the benefit that we have sold the policyholder looks like a series of options," he says. "We have given them an option to put their contract back to us at a fixed price, at a point in the future. So what we do is buy a series of options that look like those options, and every night we run a detailed stochastic analysis model of the benefit guarantees within our portfolio that determines all the sensitivities to market conditions of the liability portfolio."

The aim of this - aside from the need to ensure that all liabilities are met - is to create a system that is not just the envy of their peers in the insurance business but to rival companies that are entirely focussed on investment management. "We want to build a hedging capability that is comparable to a Wall Street company or an investment bank."

The re-insurance alternative

While the Hartford and Axa are implementing a sophisticated internal hedging programme, one of their main rivals in the US VA market, MetLife, is still managing some of the risk from its guarantees via reinsurance. At the end of January, the New York-based insurer transferred the whole of its VA guaranteed minimum death benefit exposure to Catalyst Re - a 100% wholly owned subsidiary of French investment bank Societe Generale.

MetLife trumpeted this deal as an innovation for the industry - as it is the first such transaction to go to a non-traditional bank-owned vehicle - and argued it will protect the firm against adverse equity market scenarios, but it appears at first glance to be a retrograde step. However, Lisa Kuklinski, vice president and actuary for the company, describes it as a pragmatic response to the prevailing regulatory regime.

While MetLife has had an internal hedging system in place since 2004, this is only used to risk-manage its living benefits. Under US General Accounting Procedures (GAP), the death and living benefits are treated differently - the settlements contained within living benefits are categorised as derivatives, which are marked-to-market in the US. But the insurance component of death benefits means it is not marked-to-market.

"We are in a position to be able to hedge the death benefits as well," says Kuklinski. "But under GAAP accounting rules it would create a mis-match between assets and liabilities, meaning there will be more volatility within the income statement than if we use reinsurance."

MetLife's action in reinsuring death benefits raises question marks over Raymond's assertions that a narrowing of liquidity in the reinsurance market was a prime driver in the Hartford's move to internal hedging, but He is now looking to the future. "The latest talk is all about cross-Greeks - for example, the sensitivities of volatility when the market changes. These interrelationships are very important and understanding them can minimise the amount of trading that we do - and trading costs money. There are significant diversification benefits among investment accounts, and in the future we will be looking at some significant diversification benefits across products."

Axa's launch of VAs onto the European market will not just be limited to Germany. Discussions at the Paris-based company are already at an advanced stage over when to follow in the Belgian and French markets. However, no decision has yet been reached over the UK, the largest market in Europe. With several other firms active in the North American market rumoured to be considering a similar move, it appears that it will not be long before the question of whether VAs will enjoy the same success here as in North America and Japan will be answered.

There is, however, a finite source of customers for long-term savings products and Milliman's Finkelstein is sure that with-profits funds will be an early casualty when VAs hit the European market. Its transparent menu of clearly priced guarantees is an obvious draw to consumers who - especially in the UK - have become cautious of the opacity and the divergence between investment returns and bonuses paid to policyholders.

"With-profits was by many measures a successful product - it lasted over 50 years (in its equity investing form). But products structured like VAs, with transparent menus of options that customers can select whether or not to purchase, are clearly an improvement from the point of view of customer choice and fair treatment. We expect the next generation of products to come out of the VA mould - products that include attractive benefits chosen to meet customer needs, but not necessarily fitting the VAs' original specification exactly."

This is already happening - Hartford Gold, launched by the eponymous company in the UK in 2004, contains some of the living and death benefits of VAs, within a unit-linked framework. But John Enos, managing director of the company's UK arm, is not convinced that with-profits providers are in any shape to compete with VAs and their offspring.

"In the past, with-profits would have been the main competition for these kinds of savings products. If you go back four or five years, there were around £20 billion of with-profits bonds sold - last year it was around £2 billion. It has killed itself." Instead, his company is targeting the large surpluses of cash sat in low-yielding bank or building society accounts. "We think there is around £1 trillion sitting around in people's accounts, doing nothing. A lot of it has gone into ISAs (a UK tax exempt savings product) and other tax advantaged vehicles, but there is a limit to how much you can put into these. There is a lot of cash out there, and we want to do something with it."

A VARIABLE ANNUITY BACKLASH?

Despite the much-trumpeted success of VA products in the US, there are signs of a consumer backlash against them, reminiscent of mis-selling scandals in the UK. As in the UK, the arguments stem from the way sales people and advisers are incentivised to sell the products, and the fees involved.

In July 2005, Citizens Financial Group, a subsidiary of Royal Bank of Scotland, was fined $3 million by Massachusetts securities regulators and was ordered to reimburse elderly VA policyholders who were mis-sold the products. Meanwhile, the retail brokerage arms of Morgan Stanley and Merrill Lynch are facing class action lawsuits in California over allegations that lucrative VA commission-sharing arrangements with insurers were not fully disclosed to investors.

Most controversially of all, certain large New York-based hedge funds are thought to have exploited the favourable US tax regime to make substantial investments in VA products under false pretences, in order to trade the mutual funds underlying the products. These allegations, which are being investigated by New York attorney general Eliot Spitzer as part of an inquiry into mutual fund abuses, lend support to with-profits traditionalists who argue that full transparency is not always in policyholders' interests.

VARIABLE ANNUITIES - RIDER BENEFITS

The first wave of rider benefits in the early 1990s came in the form of Guaranteed Minimum Death Benefits (GMDB), and while there are a diversity of options available now, they all share two characteristics - they are optional, and have a clearly expressed price.

Death benefits

These guaranteed the account value or that your premium would be returned if you died prior to annuitisation on the contract. These benefits were later enhanced so that after a certain date - usually five or seven years after the policy was issued - the death benefit would be reset equal to the account value at that point, locking in any gains if you died after that point and the market went down.

Living benefits

These death benefits were tweaked further, but the next major stage was the introduction of living benefits - in other words guaranteeing your investment even if you survived to see your policy mature. The initial stage of this was the Guaranteed Minimum Income Benefit (GMIB), which guarantees the policyholder that their monthly annuity payments will be above a certain level.

As with the death benefits, these guarantees multiplied in line with competition in the market, with the first being a Guaranteed Minimum Accumulation Benefit (GMAB) - which is much like the GMIB, but instead guarantees a minimum size of the pot that will be annuitised.

The latest stage in product development is centred on a Guaranteed Minimum Withdrawal Benefit (GMWB) that allows policyholders to withdraw a set amount of their investment - typically around 7% - annually, while still guaranteeing that they will receive at least their initial investment when the policy matures.