The View on Flu

Should life companies worry about bird flu? Swiss Re says yes, but Munich Re and a host of primary insurers claim the risks can be overstated. Richard Irving reports

It is 25 times more deadly than the killer "Spanish flu" virus of 1918; it could ultimately claim the lives of hundreds of millions of people; and there is no cure. Yet the insurance market scarcely batted an eyelid last month when avian influenza, or bird flu, finally surfaced in Britain.

So far, the virus, known as H5N1, has yet to pass from human to human. But the bug, which has proved fatal in 61 of the 118 cases in which it has jumped from bird to man, belongs to a family of influenzas that constantly evolve and scientists believe that it is only a matter of time before the strain "merges" with a common type of flu that circulates among humans every winter. So convinced is the World Health Organisation that the virus is close to making that final leap, that it is warning governments around the world to expect a death toll of 150 million people or more.

How would such a leap in death rates affect the capital position of the life insurance industry? If Swiss Re's actions are anything to go by, the risks could be substantial.

Earlier this year, the Zurich-based reinsurance giant dusted off a specialist bond programme that offloads unwanted risks into the capital markets. In an exquisitely timed deal, Swiss Re shunted more than $362 million of exposure to a pandemic outbreak into a catastrophe bond - the second time in as many years that it has devised deals specifically targeting its exposure to a killer flu bug - and sold it on to fixed income investors.

Implicit in the reinsurer's pandemic bond issuance programme, is the conviction that the market is ill-equipped to weather a killer flu outbreak. Bob Howe, a senior risk manager in Swiss Re's Life & Health unit is reluctant to be drawn: "We concluded that a pandemic is very difficult to predict. There is enough uncertainty out there to warrant a mechanism to hedge the risk."

But the Swiss group's determination to seek alternative ways to lay off its exposure to a spike in death rates suggests deep underlying concerns that a pandemic might trigger a systemic collapse in the reinsurance market. The group, whose US portfolio alone accounts for nearly one-third of life protection covered in the reinsurance market, is especially worried at the credit implications of a pandemic. In the case of an extreme event, it is likely that other reinsurers could face liquidity constraints. "Would everyone be around to pay us? It is very hard to say," Mr Howe says. "Maybe we are a lot more cautious than other people."

His comments come amid a testing time for the markets. Standard & Poor's, the US credit agency, recently put the ratings of 16 insurers and reinsurers on review for possible downgrade after a survey by the US Insurance Services Office put the official estimate of insured losses from Hurricane Katrina, which ripped through the Gulf Coast in September, at $34.4 billion. Analysts decline to put a number on the potential cost of a pandemic outbreak although the US Centers for Disease Control and Prevention have conservatively estimated that a killer flu outbreak could cost upwards of $166 billion in hospitalisation costs alone. Mr Howe will say only that Swiss Re's decision to hedge its exposure in the capital markets is based on detailed analysis of past pandemics, and that its business model is structured in such a way to ensure that the group is in a position to write new business in the immediate aftermath of an outbreak.

Swiss Re's two pandemic bonds are structured in much the same way as other so-called "catastrophe bonds". A special purpose vehicle, Vita Capital, controls the proceeds of the bond offering, diverting capital to Swiss Re if mortality rates in an independently devised index rise above pre-determined levels. Under the first offering, a $400 million bond launched in 2003, investors lose part of their principal if mortality rates jump by more than 30% above normal expectations and all of it if death rates exceed 50%. In return, they receive an interest rate equivalent to 135 basis points over the US dollar Interbank Offered Rate.

When the reinsurer came back to the programme earlier this year, it made two important tweaks: Under the second tranche of bonds, Swiss Re receives payments if mortality rates jump by 10, 15 and 20%, and the trigger period is averaged over a two-year time frame in order to capture a pandemic that straddles the European winter.

Although other reinsurers are understood to be looking at the mechanism, none have so far tried to replicate Swiss Re's pandemic bonds. For one thing, the hedge is not perfect - the bonds are pegged to mortality rates in the general population rather than to the group's specific book. Critics argue that this mismatch creates a basis risk so acute that the resulting hedge may not be worth the cost. And cost, too is an important factor. Quite apart from the fees associated with setting up and running a special purpose vehicle, the borrowing charges embedded in the structure far exceed the costs of reinsuring exposure to a pandemic in the retrocession pool.

Charlie Shamieh, head of integrated risk management at Munich Re, comments: "You shouldn't decide to launch a catastrophe bond programme just because it's sexy. One should not let the glamour associated with transferring risks out to the capital markets detract from the need for integrating it into the capital management process." Shamieh, adds: "We're not so sure that there are any free lunches out there. It's a useful instrument to have in the toolkit, but it has to be integrated into how you think about capital."

The German reinsurance giant is also skeptical of efforts to strip out exposure to a calamitous event such as a pandemic: "When we assess capital, we take a long-term view. That way, we can see the trend risks that start to dominate."

In other words, Munich Re believes it is sufficiently well reserved to withstand a short-term shock such as a pandemic. What really concerns its risk managers, are the long-term health and mortality trends that can decimate profits at what is already a low-margin business. "Trends might be a major risk that do not easily diversify and you should really focus risk management activities on these kinds of risks at the same time as focusing on one-off shocks like influenza. One-off shock type risks like influenza can diversify with other shock- like events in a global reinsurance portfolio."

The London-based risk manager at one rival is even more forthright in his opinions: "We have looked at the costs of issuing a pandemic bond and we have looked at our exposure to the retrocession pool. As far as we are concerned, the retrocession pool is adequately capitalised to withstand a killer outbreak. There is an extremely low probability that bird flu will trigger a pandemic and an even smaller risk that the resulting loss of life, however tragic, will bring down the retrocession pool."

A similar complacency pervades the life insurance industry at large. According to research by Life & Pensions, insurers hedge out less than 0.5% of the $140 billion annual bill for death claims in the event of a pandemic. Even in the world's three largest markets - Japan, the UK and the US, life companies have hedged less than 0.8% of their exposure to a pandemic.

In response to public concerns, the Association of British Insurers, recently asked a committee of health specialists to review the impact of a pandemic on the industry. The committee declined to recommend any official guidance to members. Meanwhile, Life & Pensions has been unable to identify a single life insurance company that has tested its exposure to a killer flu bug with specific computer simulations.

The industry's complacency appears to be born largely of the conviction that life insurers benefit from a natural hedge. Providers in the UK offer two main types of life cover: term assurance, which is aimed at middle-aged homeowners; and guaranteed annuity cover, which, in return for a lump sum, provides pensioners with an annual income until death. Provided that a pandemic were to indiscriminately kill men and women, young and old, one might expect any losses arising from a surge in payouts to holders of term cover to be offset by "gains" in the annuity book, where providers will be released early from open-ended commitments to pensioners. Further, life insurers typically pass on between 70-90% of the mortality risk, embedded in their term assurance book to reinsurance specialists such as Swiss Re.

In practice, therefore, the industry sees itself as almost entirely protected against the "downside" risk associated with a spike in death rates among holders of term assurance while retaining any "upside" benefit pegged to its annuities book.

According to Jonathan French of the ABI, providers believe their thinking to be in line with that of contingency planners. "The latest guidance from the Department of Health, is that a pandemic is expected to target the very young and the very old. The insurance industry takes that view very much into account."

However, scientists believe that H5N1 might pack a very nasty surprise. Because the strain has never before appeared in man, there is no natural immunity to the bug. When it strikes, the body's response is severe, causing the lungs to disintegrate and multiple organs to fail in what physicians call a cytokine storm. The fitter the victim, the greater the body's response and the more likely the patient is to die. This could make younger men more prone to the disease.

Analysis of mortality rates during the 1918 Spanish flu pandemic appears to bear the theory out. The virus, which originated in China at the start of the year but mutated into a killer bug in Boston in the early summer, hit 20-40 year old Caucasian men hardest. According to official figures, seven times as many young men died from the flu than were killed in action during the whole of the First World War. By October 1918 - just months after the virus had established itself, the projected average lifespan of a 20-40 year old man had dropped by 12 years; and by June 1919, the virus had left more than 100 million people dead.

Insurers claim that they construct their books carefully to eliminate any risk of an age, gender or geographical bias. John Gill, finance director of Standard Life's UK life and pensions business, one of the largest life insurance providers in the UK, explains: "While it is true that mortality rates are becoming increasingly volatile, the main worry for us is our exposure to market and credit risk.

"Obviously, there are some low-probability extreme events that can impact death rates but they are easy to diversify away in the reinsurance market - it's not an issue that keeps us awake at night."

Actuaries at the group routinely assess volatility in mortality rates alongside a number of other variables such as changing interest rates, fluctuating stock market performance and surging inflation rates, but the group does not model a pandemic outbreak as such.

Like Munich Re's analysts, Standard Life worries most about longer-term trends such as the ever-lengthening lifespan of pensioners, rather than on killer flu bugs.

New research recently released by the UK's actuarial profession (see Life & Pensions, October 2005) shows the average lifespan of a 65-year old male jumping from 17-18 years up to 20-21 years - a rise of nearly 20%. Gill comments: "A jump like that can easily wipe out your profit margin in what is already a highly competitive environment."

Against such a background, the threat of a pandemic outbreak appears more benign: "We think we have taken enough steps to diversify our risk away from short-term blips in mortality rates. Historically, mortality risk has been much more predictable than market or credit risk."

Tony Jupp, chief underwriter at Norwich Union Life is similarly unconcerned: "It's on our radar screen but you have to ask, what - as an insurer - can we reasonably do? The fact is that we have taken on liabilities to those who buy life insurance from us and we make sufficient reserves to ensure that we can meet those liabilities. In this industry, you have to decide the degree of business risk you are happy to take - if you tried to hedge out every exposure you would be squeezed out of the market very quickly."

Like its competitors, Norwich Union believes that, through a mixture of reserving and reinsurance, it is more than appropriately hedged. "Regulators require us to hold enough reserves to cope with the type of event that strikes once every 200 years. Given that we have had three pandemic outbreaks in the last century, it is reasonable to assume that our internal assumptions estimate that threat accordingly," Mr Jupp says.

Norwich Union places "well in excess of 80%" of its mortality risk with reinsurers: "The market is currently as competitive as ever, which would suggest that everyone is taking very similar views when it comes to a potential pandemic."

THE MARKET IMPACT.

Life companies face something of a double whammy in the event of a pandemic outbreak: unhedged exposure to a sharp spike in mortality rates might stretch reserves, but free falling financial markets might push them to breaking point.

As one actuary, who declined to go on the record, conceded: "I suppose the doomsday scenario would include a flu pandemic that targeted 45-55 year old executives with the biggest life cover. If the outbreak prompted a stock market collapse, plunging property values and a slump in bond yields, then, yes, I guess we would be in crisis."

While risk-averse asset allocation might cushion the industry from the worst, there are still concerns that losses from a pandemic outbreak together with a collapse in financial markets might fuel a crisis.

The Association of British Insurers recently commissioned Deloitte & Touche to examine the possible relationship between a spike in death rates and stock market volatility. As a rule, market risk and mortality risk are not considered to be correlated although the consultant concluded that a catastrophic loss of life might have a significant impact on the capital investments in an insurer's portfolio.

But the unpredictability of the 1918 Spanish flu outbreak again casts doubts on received wisdom. In the UK, where the pandemic killed more than 250,000 people, the gross annual return on equities rose 5% in the year of the outbreak (according to Barclays Capital). Elsewhere, stock markets appeared similarly immune to the crisis. Between January 1918 and December 1919, for example, the Buenos Aires All-Stock Index rose in all but one quarter.

Economists argue that the world was a different place at the start of the Twentieth Century: Sherry Cooper, a senior analyst at BMO Nesbitt Burns, explains: "Spanish flu did not come to a world characterised by economic interdependence or at a time when international trade as a percentage of world GDP was at an all-time high." In other words, the very globalisation of the world economy that has fuelled financial and commodities markets could be their undoing. "The bottom line is that a pandemic - even one significantly less virulent than the 1918 outbreak - would have a disruptive effect comparable the Great Depression of the 1930s."

But these are not the fantasies of frustrated bears run wild. The SARS outbreak of 2003 provides some valuable clues. In pandemic terms, the virus was benign, killing 775 of the 8,097 that were infected. Yet the economic damage was vast: According to the WHO, GDP in Asia during the first two quarters of 2003 plunged 5 percentage points. In Canada, where the virus surfaced in Toronto, GDP plunged 4.2 percentage points in a single quarter. The Province of Ontario estimates that the outbreak, which prompted Air Canada to file for bankruptcy protection, cost more than 28,000 jobs and C$2 billion.