(FORTUNE Magazine) – For commercial property and casualty insurers, the World Trade Center disaster was what actuaries call an "extreme event." Insured losses are still unknowable, but they are estimated at $30 billion to $70 billion. Given that the commercial P&C industry, worldwide, came into this crisis with perhaps only about $300 billion in capital, the economic effects of this disaster are simply enormous.

But the business got thrown a bone. As extreme events often do, this one transformed the industry's prices. After Sept. 11, the premiums that insurers could charge suddenly rocketed upward into what's called a "hard market." Despite its biting name, that's a good market for sellers of insurance. It describes a period in which prices are high enough to potentially cover the industry's risks and provide it with something it almost never has: a decent, or maybe even good, return on equity. Hard markets are created by the addition of fear to the insurance equation--by the perception that risks have increased--and by a collective mindset among the sellers that they are now in the driver's seat, finally able to demand the prices they need.

Hard markets are extremely rare. But the moment that terrorism brought down the World Trade Center towers, it was obvious that insurance prices would jump. Capitalists react at such moments. They may be torn by grief, as most surely were by the disaster, but they remain opportunists. It did not hurt at this juncture, of course, that private-equity capital had been blistered by tech and telecom and was looking for new roosts. Here, unexpectedly, was this interesting henhouse called P&C insurance. Old hands in the insurance business knew exactly what to do, and they moved swiftly. Consider:

--Less than a month after Sept. 11, two P&C veterans met by hurried appointment in an Albany motel, midway between their homes. From Connecticut came Robert Clements, 69, chairman of an undistinguished insurer called Arch Capital but better known for his executive years at insurance broker Marsh & McLennan. There he made a career of starting commercial insurance companies--ACE and XL and Mid Ocean--in tax-friendly, bureaucracy-light Bermuda. Now, in that motel room, Clements pitched a new Bermuda fling to Paul Ingrey, 62, a retired and renowned reinsurance executive who'd driven from his summer home in upstate New York. Clements didn't need to do much talking. Two weeks later the two men publicly announced they were leaping into business under the Arch name with $800 million in new capital, most of it supplied by private-equity firms Warburg Pincus and Hellman & Friedman.

--In his lifetime, John J. "Jack" Byrne, 69, now chairman of White Mountains Insurance, a Bermuda company with executive offices in Vermont, has run more insurers than he can count (among them Geico and Fireman's Fund) and has repeatedly retired. But in October, hearing a proposal that he form a new Bermuda operation with Tony Taylor, 56, a well-known Lloyd's insurance underwriter, Byrne simply couldn't resist. A few weeks later, he took a suite in the Waldorf Towers in Manhattan, settled his big frame on a sofa--looking like a "pasha," said one of his sons--and received a stream of potential investors in a company to be called Montpelier. With some capital already in hand, Byrne had in mind raising $400 million more that day. But by noon he was well beyond that. "I had an afternoon of investors coming, and nothing to give them," Byrne says. Ultimately, he capitalized Montpelier with $1 billion, turning away, he says, an equal amount.

--You don't have to be in insurance to have heard of the family Greenberg: a father called Hank, and two of his sons, Jeffrey and Evan. Jeff, 50, CEO of Marsh & McLennan (which tragically lost 295 employees working in the WTC), moved fastest last fall, pushing a Marsh private-equity fund, Trident II, to race ahead with a new Bermuda company it had been planning. Trident moved so rapidly, in fact, that it flailed into an unartful name, Axis Specialty, for its venture. Axis became both the first new Bermuda company to announce its formation, on Sept. 28, and the biggest, with $1.65 billion in private-equity capital--about a quarter of the money investors wanted to give it, says the head of Trident.

Next, on Nov. 5, Evan Greenberg, 47, took battle position, becoming vice chairman and head of reinsurance at Bermuda's ACE Ltd., now prosperously grown up from the days when Clements started it. And then came the father, Maurice "Hank" Greenberg, 77, CEO of American International Group, who on Nov. 26 announced his own new Bermuda company, Allied World Assurance, capitalized with $1.5 billion. So suddenly, and maybe by wicked design on the part of D-Day veteran Hank (though he scoffs at the thought), we had shades of World War II: an "Allied" competing against an "Axis" (which was about to get nailed even further, by President Bush's phrase "axis of evil"). The more striking upshot was three Greenbergs, each in a competitive spot, in an industry caught up in convulsions.

The new Bermuda companies, of which there are six big ones (including, besides the aforementioned, Endurance and DaVinci), began with about $7 billion in capital. That hardly reshapes the industry--as Dean O'Hare, CEO of insurer Chubb, indelicately puts it, "That much capital is no more than a pimple on an elephant's ass." But last fall's race to Bermuda catches the excitement that abruptly clutched the normally drab P&C business. It was a gold rush; a feeding frenzy; an avuncular equivalent of dot-com mania. FORTUNE asked a friend of Paul Ingrey's why this reinsurance legend would want to come out of retirement. "I'd guess," answered the friend, "that he wants the chance to participate in one more rip-roaring hard market."

Some insurance folks remember the dates of hard markets maybe better than their kids' birthdays. John Sinnott, a Marsh & McLennan senior executive who searches for well-priced insurance for his corporate clients, recalls the hard market of almost 20 years ago: "It lasted from Oct. 1, 1984, to March 31, 1987. Most people say it began on Jan. 1, 1985, but it didn't." The cause that time was an extreme shortage of liability coverage for such corporate risks as product defects and pollution. The only other hard market since then developed in 1992, after Hurricane Andrew, and was limited to property insurance.

The reason hard markets quickly soften is that the sellers of insurance begin to cave, lowering their prices so that they can keep getting premiums on which to earn investment income, meanwhile telling themselves they have pricing room to spare. This perception gets to the proposition that insurers are always dealing with what they expect to pay in losses. They can't know what they'll pay, because they're covering the future. To an extent not true of any other industry, this one sells its product today and finds out tomorrow (or maybe decades hence) what its costs will be, as the losses that have been estimated--and booked on the income statement--are tested by reality.

The entire scenario has given this industry one of the worst profit records going (see chart), brought about by an ugly, repeated pattern: As competition increases and softens the market, salespeople sell with mounting disregard for rates, even as costs come in above expectations. The industry knows, of course, that it will periodically get hit by natural disasters. But they are often bigger than actuaries anticipate, and then there are the special shocks: asbestos and environmental exposures, toxic mold, litigation explosions generally, and now terrorism.

The historical record of short hard markets and prolonged soft markets should kill optimism, but it never does: "I'm the world's worst optimist," Chubb's O'Hare said recently. "You have to be, to be in this business."

Acts of war are not customarily covered by P&C policies, but the industry declared after Sept. 11 that whatever the attacks were called, it would pay the losses they caused. Insurers know, however, that they don't have the resources to cope with repeated acts of large-scale terrorism, so they have urgently sought legislation that would make the government the principal insurer for any future assaults. Congress has bogged down on this issue, failing to agree on what the law should provide. As a result, the insurance industry has been forced to treat terrorism as a business problem.

At first, after Sept. 11, it looked as if both primary insurers and their reinsurers would, to the extent possible, flee from covering any losses terrorism might cause in the future. But that hasn't happened. Said Donald Kramer, a vice chairman of ACE, in late April: "Is terrorism insurable? Everybody's said no. Yet everybody's coming out with terrorism products."

There are, to be sure, acts of terrorism that almost no insurer or reinsurer wants to cover: nuclear, chemical, and biological risks, known in the industry by the initials NCB. But some companies are perfectly willing, at a dear price, to cover other kinds of terrorism. An example, paradoxically, would be the kind of destruction by aircraft that took down the World Trade Center towers but that, considering security measures since taken, is regarded today by insurers as an acceptable risk.

Other primary insurers would like to be out of the terrorism business entirely but can't easily escape. Usually that's because they're stuck with policies that haven't expired (and that don't pay them for the risk they now recognize they're covering) or because they are handcuffed by state regulations and statutes.

The statutory handcuffs are definitely painful in the case of workers' compensation: Companies writing this important line of insurance cannot exclude any kind of risk--not even war--from their policies. In another handcuff example, about 30 states require commercial insurers writing property insurance, even when it specifically excludes terrorism losses, to cover "fire following," meaning the fires that would ensue, say, from the terroristic destruction of a building. And then there is the ultimate stricture: Four states that want their business communities to have the protection of insurance--Florida, Georgia, New York, and Texas--simply haven't allowed insurers to write terrorism exclusions into their policies. In those states, therefore, any insurer issuing a policy not only is covering garden-variety terrorism (if you will pardon the oxymoron) but is on the hook for NCB risks as well. Thus, an insurer has a choice: Take on the full panoply of risks, which could lead to a fast death, or go for the crippling illness of not writing policies. One insurer CEO, talking about that "squeeze" recently, said, "I feel I could wake up tomorrow and find that I had lost my entire company."

The reinsurers in this industry are in a different position. The bad part of their story is that they came into Sept. 11 loaded with risks. The economics of that day show the consequences: Though reinsurers have only about 40% of the insurance industry's capital, they are likely, when all the sands have been sifted, to absorb 60% to 80% of the losses. Even now, the four biggest reinsurers--Munich Re, Swiss Re, Berkshire Hathaway, and GE's Employers Re--have assessed their Sept. 11 net losses at more than $7 billion.

But the reinsurers' situation started to improve dramatically as 2002 arrived and they began renewing their contracts, many dated Jan. 1. Unlike primary insurers, reinsurers are not regulated by the states and therefore have no obligation to cover terrorism in any way. So, with the new year, most began writing terrorism exclusions into their contracts.

Strikingly, though, it is the reinsurers and their close cousins, the companies that write "excess" coverage for corporations--meaning layers of insurance above, say, $100 million--that have chosen to offer the new terrorism products that ACE's Kramer was referring to. These sellers include such U.S. companies as AIG and Berkshire; Britain's Lloyd's; and a string of Bermuda insurers, including ACE, XL Capital, RenaissanceRe, and several of the new companies set up since Sept. 11. For all these sellers, the lure is the big prices they can get today on policies and their belief that they are good "underwriters," meaning they can skillfully judge which risks are acceptable. In particular, they're focused on avoiding "aggregations" that would geographically bunch their risks, putting them in the same kind of danger that aggregations in hurricane or earthquake country do.

It's uncommon for insurers to spell out the details of their terrorism coverage. But in the 2001 Berkshire Hathaway annual report, Warren Buffett gave some facts about four contracts exposing Berkshire to terrorism risks. One new property catastrophe policy that Berkshire has taken on, for example, leaves it providing "significant coverage" on Chicago's Sears Tower once losses there pass a threshold of $500 million. In another instance of terrorism tolerance, Bermuda's RenaissanceRe, a master at using sophisticated simulation models to write natural-disaster catastrophe reinsurance, has put the models to use in filling, at prices that have soared, today's demand for workers' comp catastrophe reinsurance.

When they can get terrorism out of their minds, P&C insurers are loving the market they're in right now. This year's first quarter was virtually free of natural disasters, and rates are generally sky-high (though sometimes it seems that way only because they were so far down to begin with). Major companies renewing their insurance since Sept. 11 have been obliged to pay big increases, and some of the hikes have been huge--50%, 100%, 200%, and even higher. The screw has turned, too, on which risks call for the highest price increases. Once insurers put factories at the top of their high-risk list. Today what leads is a skyscraper office building holding thousands of white-collar employees.

Trying to meet their own budgets, corporate buyers of insurance have typically held on to more risk, by accepting larger retentions (that is, deductibles) and forgoing "excess" insurance they would normally buy to cover nightmare contingencies. In short, many are under-insured from where they'd like to be. The buyers have lost coverage, too, for some kinds of risks. In the soft market that prevailed in the late 1990s, for example, insurers were so anxious to get business that they'd throw in coverage for such risks as "employment practices" (among them harassment and discrimination). Today getting those risks covered is a large expense, if the coverage can be gotten at all.

Not surprisingly, many of the buyers are mad and mutinous. They're stuck with new risk profiles that could endanger their earnings--though few investors have caught on to this--and they consider the insurers to be unreasonably profiteering. In a survey done jointly this year by UBS Warburg and Mactavish Research, nearly 70% of the corporate risk managers polled called the rate increases they were seeing either "unjust" or, worse, "outrageous."

The buyers do not dwell, naturally, on those days when insurance was so cheap and cost-effective that risk managers would have been foolish not to load up with every bauble available. Nor do they spend much time thinking about the economics of the WTC disaster. "What do I say to customers who complain that the prices they're paying today are unjustified?" asks one reinsurance executive, heatedly. "I say, 'Look, it's very simple. We as an industry sold you a product that may cost us $70 billion, for which we charged zippo.' " Now, he's saying, it's payback time.

The effects of all this on insurers' earnings are complex. The industry did its bookkeeping for Sept. 11 last year, taking its estimated losses as 2001 expenses. It still hasn't paid out all those losses by any means, but payouts will not hit earnings unless the estimates prove to be low.

This year's first quarter, therefore, was free of Sept. 11 effects, and the profits of most big American and Bermuda P&C companies were up substantially. But the fact is that earnings of the industry don't yet reflect the rates being charged in today's hard market. That's because of the way insurers keep their books: Today's rates show up in an informational item called "written" premiums, but what goes on today's profit and loss statements as revenues are "earned" premiums, which for the most part come from policies written before Sept. 11. It is only as this year progresses and as 2003 comes along that today's written premiums will become earned premiums, and P&L statements will begin to reflect the rich rates now being charged.

Naturally, though, the stock market discounts what it perceives about the future, and its attitude toward commercial P&C stocks has been revealing. After Sept. 11, to go back a little, investors knocked down P&C stocks with the general market, then boomeranged them upward to levels recognizing the arrival of a hard market. But investors know, from bitter experience, that this industry is inclined to shoot itself in the foot--in other words, to lapse into price cutting at the least provocation. So many of the industry's stocks have just been hanging around this year, not moving much up or down. This translates into watchful waiting about the really fundamental question: Just how long can this hard market last?

Looking at what the industry needs, it ought to last a long time. As noted, the industry's profits have been dolefully substandard. Furthermore, in the late 1990s profits weren't even as good as they appeared, because many companies, carrying out their task of estimating their future claims, were clearly low-balling these so as to make their earnings look as decent as possible. Eventually, though, an insurer's claim costs move from being estimates to actual money paid out. That means all sorts of companies are covering not only today's costs but also others from their cheating past. Analysts commonly refer to this problem as a fix-it matter, saying, "The industry needs to repair its balance sheets."

Beyond all that, the plunge in interest rates and the lackluster stock market is dramatically hurting returns on insurers' huge investment portfolios. Then there are today's accounting scandals, which threaten to raise claims on directors and officers liability policies. So there is a whole battery of reasons insurers should be urgently vigilant about not doing themselves in with price cutting. But they've had that need before--and have still cut prices.

In the mystery of what will happen to prices, those six new Bermuda companies, despite their alleged pimple of capital, could be important players. They arrived at this hard-market party with one profit edge: clean balance sheets that don't need repair. But they also have investors--those private-equity types--who expect annual returns of perhaps 20% and maybe an exit in three years, by means of IPOs or mergers. These rules of the game mean that the new companies must push hard for premiums but not lay themselves open to outsized losses.

What's transpiring so far in premiums, it appears, is a matter of opinion. Going by a regulatory rule of thumb, the $7 billion in capital that the six companies possess might allow them to comfortably write business bringing in annual premiums of maybe $7 billion to $10 billion. Today, a half-year after their founding, they're only at $2 billion. But Bob Clements of Arch says that the $600 million of business it has written is "a gratifying number." And Hank Greenberg of Allied World Assurance (commonly called AWAC), which has done $250 million, said in April that the company had made a "great start." Greenberg even salted his description by saying AWAC had an edge over the other new Bermuda ventures because at "most of the other companies you can't find anyone in the office when you call."

In that same April week, though, an executive of a big investor in AWAC, Swiss Re, said at a UBS Warburg conference that AWAC had gotten off slowly. His opinion was supported by two insiders in still another of the Bermuda companies. "I think," one said, "that all the companies are behind plan." The other insider envisioned that one or more might even soon decide they had raised too much capital and move to return some to their investors. Returning capital would be "good," he said, as opposed to the alternative of the companies' chasing premiums by cutting prices.

That's the fear: that the new Bermuda companies, with capital to employ and impatient shareholders, will be unable to resist shaving prices to build their businesses. On the other hand, maybe they won't do that because so many of their executives are experienced enough to know better. "We're old and grizzled," says Michael Morrison, 72, operating head of AWAC," and hopefully won't act precipitously to start cutting rates."

But the equation of the startups is unquestionably tough. If an investor is to earn an annual return of 20% over three years, the dollar he put in last fall must grow to $1.73 by the end of 2004. And the companies trying to make that happen, says Patrick Thiele, CEO of an established Bermuda company, PartnerRe, must leap high hurdles: "You've got to get things up and running pretty quickly; you've got to make maybe 18% on the capital you've put in; you've got to hope there's no big loss coming that will blow you away; and then you've got to sell at the right time when retail investors are willing to pay a big price for a company in a semi-commodity industry." And if you don't hit that window, he says ominously, "you're going to get stuck in the next cycle."

As that suggests, Thiele is not banking--nor are many other insurance executives--on a hard market that lasts and lasts. Indeed, how could such a phenomenon be expected, given this industry's low-profit history? Another implicit prediction about the hard market came this spring from Citigroup's Sandy Weill, who chose not to put money into the P&C industry but rather to take it out--by selling 20% of Citi's Travelers operation. And General Electric, no slouch at timely trades either, has been talking about selling its biggest P&C operation, Employers Reinsurance. At the least, that would get a special piece of volatility off GE's books: In the past few years, Employers Re has had a dismal profit record.

No one can really know, of course, how long today's hard market will last. But one person with an Olympian view, Donald Watson, a Standard & Poor's managing director specializing in insurance ratings, says that he does not think the rate increases being realized this year will be followed by further increases in 2003. There's plenty of capital around, he says, "and the equation is looking a bit lopsided, with far more supply than demand." That's an omen, he thinks: "It doesn't take too much to assume that we will begin to see some old-fashioned price competition begin to creep back into the market by the end of the year."

As golfers say, "Every putt makes someone happy." If price competition comes back late this year, the buyers will love it. The sellers can reflect that at least they had a great time for a while. And if terrorism returns, the calculations for all will be drastically changed.

FEEDBACK cloomis@fortunemail.com