(FORTUNE Magazine) – In these turbulent times, it's a relief to ponder the things you really can count on. The recently departed, love-'em-or-hate-'em New York Yankees will roll into the playoffs again next season. Our foreign policy will drive the French to distraction. The new Harry Potter book will rule the bestseller list. Sunday night's big guy, Tony Soprano, will keep trudging through therapy. And your house will keep rising in value.

That last point is no trifling matter. Like everyone else, you watched your 401(k) shrink dramatically and cursed the bleeping broker who swore your tech stocks would soon roar back. But your five-bedroom colonial on half an acre doubled in price in just five years, and you're expecting it'll double again in the next five years, sure as Yankee pinstripes in October or Gallic pique.

It probably won't happen. And that's a damn good thing. Put simply, U.S. housing prices are stretching the outer limits of what's reasonable and sustainable. Instead of cooling down, prices keep hurtling upward, defying the laws of economic gravity just as grievously as those unmentionable dot-coms once did.

In other words, what looks like a gift to homeowners today is potentially a recipe for disaster later on: If the boom persists, housing will become so overheated it'll pull the entire economy into dangerous, fragile territory. In a year or two, prices will fall with a thud, unleashing a double-dip recession that will pummel home prices even more. "Every day prices rise, the risk gets greater that a bubble will form--and unwind in an ugly way," says Mark Zandi, an economist with consulting firm Economy.com.

From these heights, it's hard to look down. Since the boom began in 1995, housing prices have jumped 51%, or 32 points above inflation. The run-up has added $50,000 in wealth, on average, for every one of the nation's 72 million homeowners. In many markets the gains are even more extraordinary. In Boston, home prices have risen more than 110% since 1996, to an average of $398,000. In San Francisco and San Jose, a three-bedroom ranch will run you about $500,000, almost twice what it fetched seven years ago. Even in post-Sept. 11 New York City, housing galloped 11% in the year ended June 2002. And amazingly, at those nosebleed levels, prices keep climbing.

Fortunately, it's still too soon to announce the B-word. For the nation as a whole, no housing bubble exists. By FORTUNE's estimates, homes across the country are overvalued by a modest 5% to 10%. But to paraphrase former House Speaker Tip O'Neill, all housing is local: America is a mosaic of urban and suburban areas governed by their own quirky dynamics of employment, incomes, and desirability. That 5%-to-10% figure is a composite of a vast range of markets where values fall above, below, and precisely at the prices the fundamentals would dictate.

Right now, the biggest danger lies mainly with cities on the two coasts--the aforementioned Boston and the San Francisco Bay Area as well as New York, Miami, and Portland, Ore. Each is trapped in a classic paradox: Prices keep booming while jobs are vanishing. Clearly that can't last. The frothy markets are 12% to 22% overpriced, according to a study by Dr. Michael Sklarz of FNIS, a provider of real estate data to lenders and developers. Pricey? Yes. But we're still not talking Nasdaq 5000 or even the real estate market of the early 1980s, when homes in Boston and San Francisco and San Diego were 30% above their basic value. Those were bubbles. We're not there yet.

If housing acts sensibly from here, prices will adjust smoothly by dropping modestly in the strongest markets and moving sideways for two years or so in most places. As the economy recovers, interest rates will inevitably rise from current bargain-basement levels. Homeowners will react to steeper monthly payments logically--by refusing to pay ever higher prices for new homes. (That adjustment has already begun in the commercial market; see following story.)

What that means, of course, is that future returns on your home won't begin to approach the stupendous gains since the mid-1990s. Today most houses are like high-P/E stocks: They're already too expensive to make you a lot of money. The upward march of property taxes--they rose 6.5% last year--isn't helping matters. Over the next five years you'll do well if your house appreciates in the low single digits.

That's the upbeat forecast. If we do end up with a bubble, housing prices--and the economy as a whole--will suffer far more down the road. "If we do get a housing bubble and prices fall steeply, that could cause a double-dip recession,'' says Nicolas Retsinas, director of the Joint Center for Housing Studies at Harvard. "Then Americans who have borrowed to the hilt on their homes would lose their jobs and wouldn't be able to make mortgage payments." As foreclosures multiply, prices would collapse further, then scrape bottom for several years. So one way or another, the fun is ending. The best hope is that it ends soon.

For the economy, the housing story has been nothing short of miraculous. In the past, real estate has been a me-too sector: Construction activity and home prices rise during good times and drop when America heads into recession. But in the past two years housing's role has changed from follower to savior. As we'll see a little later, the housing market single-handedly kept last year's downturn mild and is now the strongest engine pushing the slow recovery. "Housing deserves credit for shoring up the economy," says Retsinas. "But that makes the economy highly dependent, maybe too dependent, on housing going forward."

How did we get here? Until two years ago the housing boom actually made perfect sense. Hit by Gulf war recession, home prices, adjusted for inflation, tumbled in the early 1990s and then remained flat until 1995. Next the planets aligned to create a golden age for housing. To promote home ownership, the Clinton administration pushed Fannie Mae, Freddie Mac, and the other big government lenders to extend credit to low- and medium-income borrowers, a policy that continues under President Bush. As the U.S. minted 17 million new jobs in the late 1990s, the hot labor market swelled household incomes, the key determinant in deciding to buy a house. Pumping the bellows was a raging stock market that added, albeit briefly, $8 trillion to the wealth of America's families, according to Wilshire Associates.

By far the most important catalyst in the run-up was and remains rock-bottom interest rates. During the mid- to late 1990s, big gains in productivity and the influx of low-cost imports pushed down prices despite the soaring economy. As inflation fell, so did interest rates. From mid-1990 to early 1999, the rate on a 30-year mortgage dropped from 10.5% to 6.8%. The impact of low rates, then and now, can't be overstated: They create an extraordinarily virtuous cycle. Even as prices rise, buyers take on larger and larger mortgages for the same, if not lower, monthly payments. That enables them to purchase ever bigger, more expensive houses without adding to their financial burden.

And purchase they did, driving up demand. But the supply of housing grew at a more measured rate (though it's quickening now). In the middle to late 1990s, housing starts averaged 1.1 million to 1.2 million units per year, far below the peak of 1.5 million in the late 1970s. Back then, developers flooded the market with new homes only to watch prices plummet when the early-1980s recession struck. Homebuilders learned their lesson, showing far more restraint in the late 1990s. "The homebuilding industry has consolidated," says Zandi. "Now it's dominated by big, publicly owned players like Toll Brothers and Pulte Homes that are far more conservative than the local builders." In addition, environmental restrictions and ever tighter zoning laws make it more difficult to turn raw land into building lots. For example, in 1997 it took developers in Northern Virginia an average of one year from buying the land to planting the first shovel. Now it takes as long as four years.

Zealous buyers and modest, disciplined new construction proved a combustible combination. From 1996 to the end of 2000, home prices jumped 33% on average, 20% more than inflation. Up to that point the market was merely doing what it usually does in the purple patch of an economic cycle. The red alert is what's happened since. In late 2000 the economy fell into recession. But amazingly, housing kept right on rocking through the downturn. "That is practically unheard-of," says Chris Mayer of the Wharton School. "Housing prices always drop in recessions. They don't hit new peaks as they did this time."

Prices are still racing ahead, behaving as if today's tepid upturn were a rip-roaring recovery. Since early 2001--despite a brief slowdown following Sept. 11--they've risen no less than 15%.

Why is housing thumbing its nose at economic trends? For two reasons, one logical, the other irrational and potentially disastrous. The irrational force is the speculation that's creeping into the market with a fervor that threatens to unhinge prices from fundamentals. But at least part of the current run-up has a logical explanation: Those low mortgage rates we mentioned a moment ago are even lower today. Even after the recovery began early this year, rates kept falling. With the huge losses in shareholder wealth and the vast overcapacity in everything from semiconductors to fiber-optic systems, inflation is almost invisible. So the Fed can hold down rates with no fear of igniting an upward spiral in prices.

Since mid-2000, 30-year mortgage rates have dropped from 8.5% to a 40-year low of 5.95%. Remember that virtuous cycle of the late 1990s? Well, it's gotten even more sanctifying. For around $2,400 a month, you can now own a $600,000 house; in 2000 that same mortgage payment got you a $424,000 house.

With business investment in a rut, housing has been the economy's main prop. Last year, home sales hit a record of 5.3 million units, sparking strong sales of appliances, carpets, and furniture. Construction, subdued in the late 1990s, is now thriving--a bad omen for home prices down the road but good for the economy today. In the 12 months that ended in August, spending on building and remodeling hit a record $371 million, 7% above the same period in 1999-2000, when the economy was booming.

More significant, housing has lent epic support to consumer spending. In fact, surging home values are counterbalancing the impact of plunging stock prices. At first that seems illogical. Since March of 2000, Americans have lost $8 trillion on their stock portfolios and gained $2.7 trillion on their homes. But as a stimulus for spending, homes are a far more powerful force than shares. A recent study by professors John Quigley of the University of California at Berkeley, Robert Shiller of Yale, and Karl Case of Wellesley shows that for every $1,000 their house appreciates, homeowners spend $60, or 0.6%. (The inverse is also true.) For stocks, the figure is between 0.2% and 0.3%. By that reckoning the increase in housing prices since 2000 will unleash at least $180 billion in spending, slightly less than the approximately $200 billion drop caused by stock market losses.

What's more, the American home has become a virtual ATM. In 2001 and 2002, homeowners took $350 billion in cash from their houses by refinancing their mortgages and obtaining home-equity loans. So far they've spent about $70 billion of that money, on everything from vacations to cars. According to the Federal Reserve, they'll probably use another $120 billion to pay off credit card and other debt. That leaves an astounding $165 billion in their checking and money-market accounts from all that refinancing. Zandi of Economy.com predicts that homeowners will spend most of that money. If so, the wealth effect will reach well over $200 billion, exceeding by far the drag from the stock market's implosion.

All told, housing's impact on economic growth has been unprecedented. In 2001, the sector added 0.5 percentage points to GDP; without it, the economy would have shrunk by 0.2% instead of expanding by 0.3%. So far this year housing is contributing a percentage point--or fully one-third--of the nation's 3% annualized growth. Its work isn't finished either. Homeowners will give the economy an additional lift by spending part of that refi cash over the next two or three years. "Housing still has lots of firepower," says Zandi. Obviously, though, the wreckage from a burst bubble would destroy whatever benefits of the wealth effect remain.

Given housing's new make-or-break power over prosperity, it's critical to recognize the hallmarks of a bubble. Here's the main one: "In a real estate bubble,'' says Martin Barnes of the Bank Credit Analyst, an investment research firm, "money keeps flowing into housing on the assumption that because prices rose in the past, they're bound to keep rising." Then, instead of buying homes to live in, people start treating their houses as pure money-making investments. They drive prices higher and higher, until sooner or later the bubble bursts.

That's not just theory. It happened in a few big, torrid markets in the early 1980s and early 1990s. A decade ago home prices in San Francisco and Boston plunged 27%.

Both eras provide a primer for spotting future bubbles. When prices become unglued from two benchmarks for fundamental value--household income and rents--they eventually crash. That's precisely what happened in the early 1980s and early 1990s. And if today's froth keeps building, it could happen again.

Let's start with household income. "The price of homes must be related to the incomes of the people living in them," says Ingo Winzer, a consultant who tracks U.S. housing prices. If you think about it, that makes sense: As families' earnings rise, they buy proportionately bigger, more expensive houses. When breadwinners are unemployed, they trade down or rent. Household incomes impose a kind of gravitational pull on home prices.

That's the main reason houses usually aren't a great speculative investment. Household income grows about 2% a year in real terms, trailing GDP by a point. Adding inflation, that's 5%. Over the past 30 years home prices did just what household incomes would predict. They rose 5% per year, on average--a decent return, though hardly spectacular.

But home prices don't track household income in lock step. Sometimes prices rise a lot faster, especially in hot markets. If the ratio of price to income hovers far above the historical norm, prices are poised to fall. Those historical averages vary wildly from city to city. According to John Burns Real Estate Consulting, New York families pay almost 5.3 times their salaries for their homes. It's 4.8 times in California's Orange County, and 5.5 times in San Francisco. By contrast, you'll spend just over two times your pay on a manse in Houston or Atlanta.

For the entire U.S., the ratio of prices to incomes is high but not alarming. It's currently 3.5, about 6% above its 21-year average. But in Boston the ratio stands at 7.0, 59% above its normal level. San Francisco, San Diego, San Jose, Fort Lauderdale, Denver, and New York City are all at least 25% higher than normal.

The second benchmark is rents. They reflect the big fundamentals--household income and supply growth. Housing has two sources of value. The first is that you've got to live somewhere. The second is the capital gain you get when you sell. The most important, by far, is a home's value as a residence. Or it should be, unless you're speculating. The value of your home as a place to live is equivalent to what you could rent it for, or the rent on comparable homes. Thus rents dictate the value of homes the way earnings govern stock prices. If the price-to-rent ratio on your house grows huge, watch out. It flashes the same warning as an inflated P/E--namely, that prices are heading for a fall.

But in hot markets prices frequently race ahead of rents, just as they outrace household incomes. That's dangerous, because rents adjust to the changes in the local economy much faster than do housing prices. "In the short term, rents simply reflect the economic basics better than prices," says Andrew Clark, senior research analyst at Lipper, a mutual fund analysis firm.

In many markets today, rents are telling a far different story than prices. Since 1996, U.S. rents have increased just 10%, adjusted for inflation, one-third of the 30% jump in housing prices. Once again, the biggest gaps exist in the hottest markets. FNIS measures ratios of prices to apartment rentals. In Los Angeles, the figure stands at 28, compared with 17 in 1995. Over the same period, New York has grown from 18 to 25; Boston from 19 to 25; and San Francisco from 21 to 23.

When prices become uncoupled from rents, economists say, a bubble can't be far behind. "It usually means that people are buying simply because they think these are great investments," says Edward Leamer of UCLA. "They've been burned by stocks. They think houses will just keep appreciating." And they expect to make a killing when they sell. It's just another form of speculation. The signs of frenzy are unmistakable. Houses get multiple bids, often from all-cash buyers. People get out of stocks and buy more expensive homes. In the current run-up, we've been seeing some of that.

Today's growing gulf between incomes and rents on the one hand and prices on the other could have another plausible explanation: It's possible that low interest rates are causing a totally logical jump in home prices because they make buying expensive houses so much more affordable. So is it reason or speculation that's driving up housing prices?

A look at our old friend San Francisco provides part of the answer. What's happening in the Bay Area today is reminiscent of the euphoria there in the late 1980s. From 1984 to 1989 prices rose 80% in real terms, only to fall 27% over the next six years. Since 1996 the increase has been 70%. Prices started falling in 2000 in what looked like a perfectly sane adjustment, but incredibly, they've picked up again. In the second quarter of 2001 they rose 3% on an annualized basis in San Francisco and 6.6% in San Jose.

The upward spiral is pulling the Bay Area into bubble territory. "The worst situation is where prices are already too high and the local economy is going bad," warns Winzer. That's definitely the case there. Since 1999 the region has lost 180,000 jobs, many of them high-paying positions at once-thriving tech players like Cisco and Hewlett-Packard. Nor is it comforting that rents in the area are falling drastically.

Despite the job losses--and the formidable prices--buyers' zeal is waning only slightly. "It's still a seller's market," says Clay Duncan, a broker at RE/MAX in San Francisco. The market between $500,000 and $800,000 is still extremely strong--people are bidding $50,000 to $70,000 over the asking price." Starter homes are also getting scooped up. "Anything under $500,000 is selling like hotcakes," says Michael Freethy of Prudential California Realty. In a sure sign the market's still tight, the Bay Area has just four months' inventory of homes for sale, meaning that the number of houses on the market is so low that the entire stock should sell in four months. In the bear market of the early 1990s the inventory level stood at nine months.

It's clear from markets like San Francisco that America's faith in houses has reached at least slightly irrational levels. "The decline in rates isn't enough to sustain the increase in prices," says Leamer of UCLA. The reason is simple: Rates are far below their historical averages. As the economy improves, they're sure to bounce back, even if they don't quite reach the 8% that prevailed over the past decade. Then new homeowners will be stuck with much bigger mortgages, at rates not too different from what they'd pay during a typical recovery. Let's say rates return to 7.5% in year or so, the same level as in mid-1999. Then buyers would be spending $2,400 a month on a $480,000 house, the same monthly payment they now make on that $600,000 home.

Sorry, but prices need to fall, especially in the hot markets. Hopefully the adjustment will be gradual. A home isn't like shares of Cisco: If your house isn't appreciating, it's still valuable to you as a place to live. As buyers disappear, sellers at first stick to their asking prices. Then buyers make lowball offers. More and more houses go unsold, swelling the inventory of listings. It may take six months to a year before the seller agrees to lower the price. Then prices are more likely to drift down over three or four years than to drop sharply. In addition, household income should rise, a factor that will mitigate the fall in prices. In San Francisco--or Boston or San Jose, for that matter--prices will decline just as they did in the early 1980s. But this time they'd only have to fall 2% a year for several years to reach their correct level.

That's hardly a disaster for the economy, especially with all the pent-up purchasing power that's left over from the refinancing boom. What would be a disaster is a genuine bubble. They aren't pretty. We've all seen the damage they can do when they burst.

Of course, even if we don't get a housing bubble, we may still wind up with a double-dip recession. An oil shock or a financial implosion could also pulverize the economy and slam housing prices. Barring that, however, the most likely outcome is that the market works, and housing prices quickly cool. Then as the recovery picks up speed, interest rates will rise, but their impact will be offset by a surge in new jobs. Housing will enjoy a soft landing. The former engine of the economy will become a drag on growth, but with any luck only a slight one. "Housing will operate against the rest of the recovery cycle," says David Wyss, chief economist for Standard & Poor's. "It will make this recovery quieter than past expansions." The reason: Most people did their buying during the recession and its aftermath, the reverse of what usually happens. So by the time the economy starts rolling again, their hunger for housing will be gone.

We know you think of your house--your special slice of suburbia, gambrel roof, sauna, and all--as anything but ordinary. Unfortunately, as an investment, it's almost certain to be just that.

FEEDBACK: stully@fortunemail.com

REPORTER ASSOCIATES Christopher Tkaczyk, Noshua Watson