It takes a brave man to buck the trend. Warren Buffett is such a man, as he demonstrated last week when he took potshots at the overhyped Chinese stock market.

Referring to a market that has increased more than five times in the past two years, Mr. Buffett suggested that stock valuations in China might just have gotten a bit ahead of themselves. Thank heavens. I’m hoping Mr. Buffett takes a look at contemporary art one of these days, or maybe the euro.

Meanwhile, where might a brave man want to invest these days? How about the housing sector — is anyone ready to take the plunge? Bradley Gendell of the registered investment adviser Cumberland Associates was brave enough to stand up before clients last week and suggest that Ryland Homes (RYL $28) and Toll Brothers (TOL $23) were stocks to buy. Mr. Gendell admittedly was bruised after suggesting to the same gathering a year earlier that another homebuilder, Centex (CTX $26), was attractive at $52. Undaunted, he made a positive case for the two home building stocks.

The story is based primarily on the notion that the stocks have been beaten up so badly that they are cheap, and that expectations are finally unrealistically low.

At the end of the day, as Mr. Gendell pointed out, there are actually families out there who want to own a home. Home starts have been way above trend for some time, but are now at the lowest level in 14 years, helping to undo some of the prior excesses. Mr. Gendell points out that home builders have had successful “sales” weekends of late, where price reductions have helped thin out bloated inventories. The point is, he says, there is demand for housing at the right price.

Where are we in that demand cycle? According to data from the Harvard Joint Center for Housing Studies, household formations, the bedrock of housing demand, have trended down in each decade since the 1970s, when formations averaged 1.7 million each year.

In the 1980s the figure dropped to 1.5 million, and then to slightly less than 1.3 million in the 1990s. In the current period, the so-called baby boom echo has slightly boosted household formations, to about 1.3 million. That is, the children of boomers are now old enough to get married and buy homes, turning around the 30-year decline.

This demographic upturn, combined with idiotically easy money, pumped up demand for new homes to unsustainable levels. This year, Mr. Gendell estimates that starts will total 1.4 million units, considerably behind last year’s total of 1.8 million units and 2005’s level of 2.1 million.

Where is demand headed? Putting together data from several sources, a case can be made that normalized demand over the next decade will run about 1.8 million starts a year. This forecast includes 1.3 million homes for the expected growth in households, about 400,000 new homes to replace “removals,” and 140,000 units that on average are vacant at any time.

Further out, the projections build to as much as 2.4 million annually, with the increment coming from 300,000 “second” or vacation homes and another 300,000 units for immigrants.

The opportunity for investors is that certain home building companies are well-positioned to earn a lot of money as demand recovers. Ryland and Toll, according to Mr. Gendell, have taken their lumps, and in each case appear poised to stage a recovery over the next 12 to 18 months. The stocks, according to Mr. Gendell, always rally well in advance of the fundamentals.

Both companies have excellent, seasoned management and healthy balance sheets. Both are diversified geographically and have what Mr. Gendell refers to as “distinguished” land strategies. Toll seeks out locations near urban centers, where building permits are hard to secure, and uses its expertise to develop such properties.

Ryland, on the other hand, traditionally has bought land as it needed it, so that the company is not holding huge overpriced inventories of property. In the last quarter, Ryland took a write-down of its land holdings that was less than expected.

Owning both companies also gives investors some diversification on price point. The average home price for Toll is $700,000, while the average price for Ryland is $300,000. The companies are addressing different segments of the market, but neither, according to Mr. Gendell, is dependent on the subprime buyer.

Mr. Gendell says that historically the home building stocks have sold between one and two times book value, and when they hit the lower end of that range, they are attractive. Today the group is selling, by his calculation, at 0.7 times book value. In the case of Ryland, book value is now $33 a share, well above the share price. The company will post an estimated loss this year of $2.95 because of the $5 a share write-down of property inventories.

On an operating basis, Ryland will earn about $2.30. As the industry recovers to a more normal level of 1.5 million to 1.7 million housing starts, the company should be able to earn $5 to $5.50 a share.

Under the same circumstances, Toll should be able to earn about $3.25 a share. The company is currently operating in the red and will likely post full-year earnings of about $0.65 a share.

It should be said that some investors liken buying the home builders today to catching a falling knife. Comparisons with book value are only useful if book value stands still, which may not be the case, as the industry may be forced to reduce further the carrying value of land inventories. The industry has to digest several years of excess production, in the midst of a looming consumer pullback and chaos in the mortgage markets.

The chief economist for the National Association of Home Builders, David Seiders, reckons the peak-to-trough drop in single-family house starts from the first quarter of 2006 to the second quarter of next year will be 50%. He is predicting that the trend will turn positive by the third quarter of next year.

Alarmingly, the Census Bureau reported last week that 2.1 million homes are for sale in the country, up from 1.9 million one year ago, and the National Association of Realtors reported that September sales were the lowest since it began reporting the data in 1999.

In other words, it is a brave move to buy the home builder stocks today, but one that could prove extremely profitable.

Mrs. Peek is an investor with Cumberland Associates. peek10021@aol.com