The United States housing market recently wrapped up its best selling season in three years. Prices of existing homes, as measured by the Case-Shiller 20-city index, now stand 3.6% above the lows set in April; unit volumes of existing home sales are 15% above their November 2008 low; production of housing was 25% higher on a seasonally adjusted basis in August than at its April low; and some homebuilder order books have even started to grow for the first time in about a dozen quarters. In all, a pretty good couple of months that surprised most market participants in its direction as well as its magnitude.

As the market heads into the fall, there is considerable difference of opinion as to the sustainability of the current path of home prices. Is this recent strength the beginning of a sustainable bottom, or just a dead-cat bounce? Well, I can say with some certainty that prices are weakening as I write, but not by any more (and maybe even less) than in a normal year at this time. Seasonality is strong in housing, buoying prices in the spring and weighing on them in the later part of the year. In truth, it won't be until next spring that we have a good idea about the answer to the above question, but activity in 2009 suggests something is afoot. Buyers came out of the woodwork in places like California, and to a lesser extent in Phoenix and some areas of Florida, suggesting that prices in the harder-hit areas are attracting bargain hunters in very large numbers. This hasn't been the case for a few years, indicating the ballgame may have changed (with some generous help from Uncle Sam).

Head Winds
It's well-known that there are still significant head winds, the largest of which is the massive number of homes that are going to be coming back to the market over the next several years. In fact, we think it's a foregone conclusion that at least 4 million units will be foreclosed upon in the U.S. over the next couple of years (recent estimates by other researchers are as high as 7 million). With annual sales currently running at a bit more than 5 million annually, these 4 million-7 million units have the potential to cause some problems if dumped on the market at a rapid rate. Indeed, many of the housing pessimists have seized upon this as sure-fire evidence of lower housing prices going forward. Massive amounts of "supply," coupled with a tight lending environment, the expiration of a U.S. government handout, and a shrinking job base, means that prices absolutely have to go down in order to meet a market-clearing price, the theory goes.

This may be the case, but I doubt it. The above analysis sounds elegant, but like many elegant-sounding theories it is probably wide of the mark because it improperly categorises bank-owned properties as "new supply" when it's really nothing of the sort. The overwhelming majority of this inventory was built years ago and therefore is already part of the existing supply. In some cases, it may even already be included (or has been included at some time) in the "for sale" statistics. Authentic new "supply" (housing starts) currently hitting the market is actually quite low. In fact, it hasn't been this low since at least 1959 (and probably a lot longer, but we don't have data prior to then). This year, builders will begin construction on somewhere around 620,000 units, a number that's about 32% below the old record low of 906,000 set--you guessed it--last year. Taking the analysis one step further, let's assume that U.S. housing production posts a miraculous 50% increase next year to 930,000. The three-year total, at about 2.4 million units, will be roughly a million units less than the three-year total that occurred at the 1991 bottom and about 1.5 million less than both three-year periods that occurred at the 1975 and 1982 bottoms. At the same time, there are currently 18% more households than in 1991, 34% more than in 1982, and 56% more than in 1975.

More appropriately, it's likely the foreclosure inventory represents beginnings of a mix shift as many former owners become renters while some former renters become owners. But at what price are these transactions likely to occur? We've maintained for a while that U.S. home prices will find support at levels where investors can purchase the foreclosed properties and rent them back to the people who used to be owners at attractive returns. Prices in many markets are already there. The "15 times annual rent" rule is a back-of-the-envelope method for determining fair home prices based upon the area's rental rates. By this admittedly simple metric, homes look cheap in many regions. For example, the U.S. department of Housing and Urban Development (HUD) tells us that the median rent for a three-bedroom property is currently $1,697 per month in Riverside, CA. This translates into a fair value of roughly $305,000, or almost 30% higher than that area's $235,000 current median listing price according to Housingtracker.net. This calculation also resonates with first-time buyers. Cheap prices, combined with some help from Uncle Sam in the form of an $8,000 tax credit, have revved up the lower end of the market where these buyers are active. Of course, if the credit goes away and rates climb significantly, the math gets more difficult. But rates would have to climb considerably for payments to appear expensive on a historical basis at today's prices.

To be clear, I'm not calling for rapid and continued price increases anytime soon, but I do think stabilisation is definitely possible. It's true that rents are going to be pressured for as long as the employment base shrinks and there are more housing units than households. And we agree with the pessimists that the combined housing market (both owned and rental) is currently oversupplied. However, it may not be by as much as many think. In truth, nobody really knows how big the glut is, but it's important to note that the overbuilding that erupted in the 2000-05 time period was contained to the "owner" market. Production in the rental market was fairly subdued throughout that period. As a result, the overall amount of housing put in place (both owned and rental) wasn't as far out of line as some think. We estimate there were as many as 1.8 million excess housing units in place as late as 2006, but the market has been undershooting potential demand by quite a lot this year and likely next year as well. Even assuming lackluster household formation due to economic conditions persists for the next quarter or two, it's very possible the excess number of homes could be soaked up by the end of next year if the employment base stops shrinking.

Of course, prices will firm well before the housing glut is completely absorbed, which is what may have happened earlier this year. Places like California have actually overshot inventory equilibrium as buyers' thirst for distressed properties hasn't been near quenched. Surprisingly, several cities in that state now boast of an undersupply.

Stocks
It's a bit late to capitalise on firming home prices by purchasing homebuilders or other companies tangentially connected to housing. Like most of the U.S. market, these names have rallied significantly from their March lows and don't currently offer a wide margin of safety. Plus, it's likely many stock prices will struggle over the next several months as the U.S. market enters its seasonally weak period this fall. Nonetheless, we generally think several of the builders are in pretty good shape for the next couple of years and might be interesting purchase candidates at lower prices. All of the U.S. names except Beazer enjoy the strongest financial positions they've had in decades, if not ever. Some of the names, like KB Homes and Meritage, are enjoying great success in the entry-level market, where their homes are price competitive with foreclosures. Builders like MDC Holdings, NVR, Standard Pacific, and Lennar all are enjoying gross margins nearer to normal levels, due in part to land bases that are now properly stated on the balance sheets. Finally, land deals are starting to happen in several regions of the U.S., indicating many builders are now worried about having too little land, as opposed to too much.

Investors interested in the space need to be aware, however, that the perils of buying a homebuilder still persist. Historically weak barriers to entry may have been strengthened in the short term through a lack of financing, but this too shall pass. Eventually, entrants will invade the business, just as they have in the past. Cyclicality is still problematic, even though the massive downdraft is likely over. And finally, it's unlikely that homebuilders have shed their capital-intensive ways, although several boast of a new asset-light strategy. For these reasons, investors should look to homebuilders not as high-quality long-term holdings, but as opportunistic purchases when the names are way out of favour.