Abstract

Long-term care (LTC) insurance is a salesman's dream. Millions of well-heeled boomers, anxious to protect their estates from the random expropriation of institutional dependency – what a market! But for Manulife, bleeding $1.5 million a day in LTC claims through subsidiary John Hancock, LTC is a nightmare. Company spokesmen blame unexpected increases in life expectancy. But management's fundamental error was insuring correlated risks. Risk pooling works only when individual risks are uncorrelated. Increases in life expectancy affect all contracts together. Manulife made the same mistake selling equity-linked annuities with guaranteed floors – essentially insuring against stock market declines. Results for shareholders have been catastrophic. Top management, meanwhile, have been honoured and richly rewarded.

The Manufacturers' Life Insurance Company, now Manulife Financial, is almost as old as Canada. Founded in 1887, with Sir John A. Macdonald as first president, it has always been one of the country's blue-chip corporations, a part of every sound conservative investment portfolio. In the present century, however, an aggressive management generated particularly impressive returns for investors. Trading at about $20 in 2000, Manulife's shares doubled to $40 in early 2008. The rather anaemic quarterly dividend of $0.05 per share had quadrupled to $0.26. Manulife was a star.

Then, something happened.

Riches to Rags: Manulife Financial in the 21st Century

Once a darling of the Canadian stock market, [Manulife] has fallen on hard times. … Over the past two years, it has halved its dividend, diluted existing shareholders by doubling the number of outstanding shares and reported bafflingly inconsistent earnings from quarter to quarter. These are the sort of moves that can put a stock on the DL for a long time. (Berman 2011)

In March 2009, Manulife shares were trading under $10. The capitalized value of the company had fallen by three-quarters. The August dividend was cut to $0.13, a level not seen since February 2005. The share price did then recover with the overall market, but only to the mid-$20 range, and by August 2010 was back near $11. The company was haemorrhaging money. An investment in this blue-chip a decade ago would, at time of writing (June 2011), have lost about 20% of its value.1

What went wrong? Two things. Both involve violation of elementary insurance principles in pursuit of market growth, but one in particular demonstrates the fundamental weaknesses of private markets for health insurance (hence this column). The whole episode raises serious questions as to the quality of corporate governance and the incentives faced by top management.

On the correlation of risks

Insurance companies pool individual risks. Whether a particular person will die, or a house burn down, or a car/driver be involved in an accident, or the occurrence of any other adverse event, is inherently unpredictable. But the number of such adverse events that will occur among a large number of similar persons, houses, etc. is much more predictable, as is the corresponding loss – so long as the risks faced by each member of the pool are not correlated with the others. The probability of your death (in the next year, say) is not correlated with the probability of mine. Whether or not you die next year does not influence my chances.

If, by contrast, adverse events were perfectly correlated – either everyone's house burns down next year, or no one's does – then pooling the risks does not reduce them, and the insurer faces exactly the same risk as each of the insured individuals. In such circumstances, there would be no benefit from insurance, and no rational agent would buy or sell it. That is why insurers traditionally excluded from coverage "Acts of God or the King's enemies." Losses associated with war or natural disaster tend to be highly correlated. (Insurers are now able to insure themselves against "local" disasters that might otherwise break the company by re-insuring with giant international companies operating on a global scale, expanding the risk pool to find uncorrelated risks.)

In the simplest terms, Manulife got into trouble – and remains in trouble – by insuring risks that were obviously correlated.

Annuities, variable and otherwise

The core functions of a life insurer are collecting premiums and paying claims. The insurer sells a policy to the insured, a contract specifying the amount the policy holder will pay at periodic intervals to the insurer, and the amount that the insurer will pay to the named beneficiary at the death of the policy holder. Apart from administrative and marketing functions, this activity requires the insurer to calculate the mortality risk, the probability of payouts, for a specific group of insured lives. This is the actuary's job. Premiums can then be set at a level high enough to cover the estimated payouts for the insured pool of lives, plus the various overhead expenses, and provide a profit for the insurer (if commercial).2

With those same skills, however, the insurer can design and offer contracts written in reverse. In return for an initial payment, the insurer contracts to make periodic payments of a specified amount – an annuity – until the death of the purchaser (the annuitant). As with life insurance, the insurer is making a bet on the length of life of the insured (or rather, the average longevity of the pool of policy holders/annuitants). The mathematics are the same; the only difference is that the seller of life insurance policies gains if the insured live a long time, while the seller of annuities wins when the customers die early.

All of which is very old news – so how do you "sex up" the product to give your marketing people some traction? The variable annuity links the periodic payout associated with a given initial purchase to the value of some type of capital asset – for example, the return on a portfolio of particular stocks or bonds. Suppose the initial payment is invested by the insurer in more risky assets, such as equities, rather than in stodgy old government bonds. If stock markets rise, as historically they have, and yield returns that are on average greater than government bonds, as they tend to do, then the annuity payments can be escalated over time. But – greater returns, greater risk – variable annuity payments can also fall when equity markets fall. So the marketing trick was to guarantee the annuity purchaser against a reduction in the initial annuity. Heads you win, tails you do not lose (your initial investment).

This offer was understandably popular, particularly with unsophisticated investors who were suspicious of the stock market. Manulife, in particular, sold a lot of these variable annuities with a guaranteed floor. The product enabled people to feel that they were participating in booming stock markets without risking their retirement savings. The variable annuity also had the potential advantage of providing some protection against inflation. The purchasing power of a fixed annuity declines over time, as prices rise. Insofar as inflation is reflected in rising corporate profits and thus in share prices and dividends, the variable annuity will compensate for this decline.

It is also quite a profitable product to sell. Guarantees come at a price. In effect, Manulife and others were selling insurance against stock market declines. The variable annuity was – had to be – priced so as to cover all the costs and profits in what was, in effect, a whole new insurance market linked to an investment vehicle. It was not, in fact, a very good buy, but it sure was a good sell. Manulife prospered accordingly.

The catch, of course, was that in moving from insuring mortality risks to insuring market returns, the "lifecos" were not only moving from a field where they had a comparative advantage in expertise to one where they did not. They were also moving from insuring uncorrelated risks to insuring correlated ones. If the markets crash, the insurer has to make good on all the guarantees. It would be akin to an extensive and virulent epidemic – a whole lot of life insurance policies all paying out at once. The quondam stodgy old insurance company, now high flyer, had become massively exposed to fluctuations on the equity markets.

And that is why, in March 2009, Manulife's capitalized value had fallen by 75%.

That Was Then; This Is Now: Imperial Overstretch

But there is more, and this is where our story really begins. Capital markets have since recovered. Not all the way to their pre-crash highs, perhaps, but a long way off their bottoms. Canada's banks, for example, came roaring back from their March 2009 lows. Manulife, not so much. Why not?

Remember John Hancock?

In September 2003, the Manufacturers' Life Insurance Co. announced that it was buying the American insurer John Hancock, and in 2004 re-branded the combined company as Manulife Financial. It seemed like a good idea at the time, although Manufacturers' shares fell on the announcement, while Hancock's rose. Eight years later, Hancock is a major drain on Manulife's profits. The problem? Long-term care insurance (Perkins 2011).

Public health insurance coverage in the United States takes many forms, in total reimbursing about 60% of health expenditures in this nominally "private" system. The two principal programs are Medicare, for the elderly, and Medicaid, for the indigent. But Medicare does not cover long-term care, so Medicaid has become, by default, primarily long-term care insurance. To qualify, however, one must be indigent. Those in long-term care must spend down all their own assets before qualifying for public support. The "spend down" requirement makes long-term care a capricious but quite effective destroyer of inheritances – and a market for private insurance.

John Hancock – and thus Manulife – is heavily involved in that market. And they are losing their shirts.

The long-term-care business has been a big part of John Hancock's problems: In fact, it was the main factor cited by Moody's when the rating agency downgraded Manulife in November. (Perkins 2011)

***

For some Hancock clients, the insurance is an important part of their financial plan for their golden years. For Manulife, it represents something else: The policies are a money pit, and also an illustration of how much has gone wrong since it vaulted itself to the top of the Canadian financial services industry in 2003 with the $15-billion acquisition of Hancock. Which is why Ms. Harrison is here in Lansing – to make the case for why the company should be allowed to raise prices on long-term-care insurance as much as 90 per cent. … The company is attempting to do this throughout the U.S., imposing average rate increases of about 40 per cent on long-term care. But first it must persuade regulators in each state. … John Hancock is looking to … make hundreds of thousands of people pay a lot more for something they have already bought. (Perkins 2011)

***

Once viewed as the boldest foreign acquisition in Canadian financial services history, Hancock has become Manulife's albatross, sucking up resources to such an extent that some analysts think it might be time for the company to sell it and flee the U.S. for the promise of Asia. Manulife took a $1-billion writeoff last year because of diminished prospects for its U.S. business…. (Perkins 2011)

Where the hell were the actuaries?

The company story is that its problems stem from unexpected increases in life expectancies, resulting in more people needing long-term care for longer periods of time. But increases in life expectancies have been going on for over a century. Didn't anyone at Manufacturers' notice this, before they went for Hancock? Did no one think to run some numbers, to look at Hancock's portfolio of long-term care contracts and to ask what the company might be worth under varying assumptions about future life expectancies?

Furthermore, these trends represent, from the insurer's point of view, correlated risks. Rising life expectancies raise the risk, i.e., the expected payout, associated with all the outstanding contracts. Like Acts of God and the King's enemies, these are the sorts of risks that insurers should not be covering. And indeed, Manulife seems to have concluded that belatedly. It is asking to be permitted to re-open existing contracts, some of them many years' standing, to transfer the unanticipated risks back to policy holders.

My … er … company is bigger than yours

Well, where were the actuaries? Were Manulife's people simply incompetent? Unlikely. Actuarial science is a pretty rigorous trade, and its practitioners are very cautious people by nature and by training. Were they overridden by the marketers, all full of irrational exuberance and dazzled by the size of the potential market and, anyway, never very strong in math?3 Maybe. But if so, where was top management? Surely, when more than half the company's capitalized value disappears, there should at least be some awkward questions for them. Who were they, and what were they thinking when the key decisions were made?

Dominic D'Alessandro was a central figure in "the boldest foreign acquisition in Canadian financial services history." In 2004, he became the president and CEO of the new Manulife Financial. The acquisition of Hancock made his company, again, the largest in Canada and number two in North America (fifth in the world) (Reuters 2003). In 2004 he was also recognized as Canada's Most Respected CEO, having been named Canada's Outstanding CEO of the Year in 2002. He oversaw the subsequent aggressive expansion of Manulife, reflected in its growing stock value and dividends. In 2008, he was named to the Canadian Insurance Hall of Fame. On his retirement in May 2009, he was handsomely rewarded for his services by a grateful board of directors, despite the reservations of some shareholders:

Manulife Financial Corp. on Thursday rejected shareholder appeals for a say on pay and awarded Dominic D'Alessandro a compensation package worth more than $25-million.

The chief executive of the country's largest insurer will receive in excess of $12.5-million for an "extraordinary performance" last year and another $12.5-million for working the five months until his retirement in May.

The award for 17-months' work is at the top of the range for Canadian financial executives and comes after shares in the company fell 49% last year and slid 26% this year following a $1.9-billion loss for the insurer last quarter.

The dismissal of investor appeals for a vote on executive compensation defies a trend that has seen the rest of Canada's top financial institutions each agree in the last month to give shareholders a voice on compensation. (Callan 2009)

Mr. D'Alessandro now serves as a director of a number of major Canadian companies.

Some shareholder bitterness remains. Online Globe and Mail reader "Venture Philanthropist" comments on Berman (2011):

Dominic D'Alessandro should be made to pay for his crimes (namely stupidity). … Amazing how much he took in before stepping away – I'm sure he feels bad, as he should, since even a fresh finance grad could have suggested the necessity of hedging.

Other online commentators vigorously recommend avoiding a stock whose board has shown such contempt for shareholders.

Whatever Mr. D'Alessandro's putative crimes, however, stupidity is surely not among them. And as for feeling bad, he can, like Liberace, cry all the way to the bank. Nor do his peers seem to see any disgrace in taking decisions that ultimately cost shareholders more than half the company.4

The Manulife story has some parallels to the much larger catastrophe of the 2008 financial meltdowns in New York and London. Top management in major financial institutions grew their companies by taking larger and larger bets on risks that they did not understand, and were richly rewarded for doing so. The eventual losses were borne by shareholders and taxpayers. One might at least be skeptical about any proposals to entrust the financing of significant parts of a health system to institutions displaying such major failures of governance, and offering such perverse incentives to senior managers.

At Manulife, we believe strongly that good corporate governance is critical to the company's long-term success and the protection of the interests of our many stakeholders. (Manulife Financial 2011)

Amen to that.

Know when to walk away, know when to run

Coming back to long-term care, however, the Hancock experience seriously undermines any claim that private insurance markets can support this growing sector. In a discussion of sources of "market failure" in private insurance markets, I once included "insufficient information for rate-making" (Evans 1984: 40). Predicting needs for and costs of care over a time horizon of years or decades is virtually impossible, for a number of reasons that Hancock and Manulife have since discovered. I argued that private insurers would therefore shun such markets.

This observation was obviously incorrect in the case of long-term care. But a number of US insurers are in fact running away from the business. Manulife management and shareholders might wish that they had, too. They had insufficient information for rate-making, took a leap in the dark and have come down to a hard landing. They are now betting heavily on recovering their losses by raising prices to past as well as future customers.

Despite the American experience, there appears to be continuing optimism about the potential for such a market in Canada. But insurers will need to screen applicants very carefully to weed out those at highest risk. It will thus be at most "a niche market … a very defined market" (quoted in Perkins 2011). That makes good commercial sense. The most fundamental principle of private health insurance is: "Sell coverage only to those least likely to need it."

How, then, will most long-term care be financed? No prizes will be given for the correct answer.


L'effet Hancock : Manuvie et les limites de l'assurance maladie privée

Résumé

L'assurance pour soins de longue durée (SLD) est le rêve de tout vendeur. Des millions de baby-boomers bien établis, soucieux de protéger leur bien contre l'expropriation aléatoire de la dépendance institutionnelle: quel marché! Mais pour Manuvie, qui perd 1,5 million de dollars par jour en réclamations pour SLD auprès de sa filiale John Hancock, ce type de soins devient un véritable cauchemar. Les porte-parole de la compagnie jettent le blâme sur un accroissement inattendu de l'espérance de vie. Mais l'erreur de gestion fondamentale a été d'assurer les risques de corrélation. La mise en commun des risques ne fonctionne que si les risques individuels ne sont pas en corrélation. La croissance de l'espérance de vie affecte l'ensemble des contrats. Manuvie a commis la même erreur en vendant des rentes liées aux valeurs boursières de pair avec des seuils de garantie – essentiellement une assurance contre le déclin du marché boursier. Pour les actionnaires, les résultats ont été catastrophiques. Entre-temps, la haute gestion récolte les honneurs et est richement récompensée.

References

Berman, D. 2011 (May 7). "An Unloved Stock That Could Have Some Life Left in It." The Globe and Mail. Retrieved July 6, 2011. <http://www.theglobeandmail.com/globe-investor/markets/markets-blog/an-unloved-stock-that-could-have-some-life-left-in-it/article2013272/>.

Callan, Eoin. 2009 (March 27). "Manulife CEO Awarded More Than $25M." Financial Post. Retrieved July 6, 2011. <http://www.ottawacitizen.com/story_print.html?id=1432125&sponsor=>.

Evans, Robert G. 1984. Strained Mercy: The Economics of Canadian Health Care. Toronto: Butterworths.

Manulife Financial. 2011. "Corporate Governance." Retrieved July 6, 2011. <http://www.manulife.com/public/home/index/0,,lang=en,00.html>.

Perkins, T. 2011 (May 7). "Manulife's Long-Term U.S. Headache." The Globe and Mail. Retrieved July 6, 2011. <http://www.theglobeandmail.com/globe-investor/manulifes-long-term-us-headache/article2013496/>.

Reuters. 2003 (September 29). "Manulife to Buy John Hancock." Los Angeles Times. Retrieved July 6, 2011. <http://articles.latimes.com/2003/sep/29/business/fi-hancock29>.

Venture Philanthropist. 2011 (May 8). Comment on D. Berman, "An Unloved Stock," The Globe and Mail. Retrieved July 6, 2011. <http://www.theglobeandmail.com/globe-investor/markets/markets-blog/an-unloved-stock-that-could-have-some-life-left-in-it/article2013272/comments/>.

Footnotes

Notes

1 A bank, any bank, would have been a much better investment. BMO, for example, is up 50% over the last 10 years, and Scotiabank has more than doubled.

2 The "underwriting gain" (or loss) defined by premiums less payouts is also augmented by the insurer's investment earnings. Premiums are paid over a period of time during which the insurer can invest them. When interest rates are high, insurers can still make a profit even if the underwriting "gain" is negative; if interest rates fall, premiums must be raised.

3 "Initially you had an industry of dreamers, including insurance agents and companies, that looked and said, 'Six million baby boomers in the United States are all going to buy this product and we all have to get out there because look at the market potential,'" said Jesse Slome, executive director of the American Association for Long-Term Care Insurance." Quoted in Perkins (2011).

4 Well, perhaps not ultimately. The Hancock story is still very much ongoing. If present management decides to sell and leave, the write-downs could be pretty impressive.