Housing: A Tale of Two Time Periods
The next couple of quarters will be tough, but the longer term is different.
I'm often asked about my outlook on the housing sector. Lately, my immediate reply has been to ask, "Over what time period?" The reason is that today, more than ever, one needs to delineate between the near term and that which lies beyond several months from now. The outlook is not great for the next couple of quarters, but the longer-term trend can be nowhere but up.
There's no doubt that the housing market is currently suffering a John Daly-sized hangover from the government-induced stimulant that expired in April. The chart below shows the May and June declines in pending home sales are likely to cause a sizable drop in closings over the coming few months. In fact, we wouldn't be surprised to see the seasonally adjusted selling rate, or SAAR, of existing homes fall back into the low- to mid-4 million range from June's 5.37 SAAR at some point this summer.
In addition, our pricing and inventory indicators are not enjoying the same positive dynamic they did at this time last year. Institutional investors will remember we were one of the only shops to correctly call for a multimonth sequential increase in the highly influential Case-Shiller home price index in July 2009. We determined a reversal in its steep three-year downtrend was likely, based on an aggregation of our real-time listing data. Unfortunately, the model, shown below, isn't forecasting the same strength this summer. It's easy to see the model was forecasting strong gains for the middle part of 2009, which turned out to be exactly the case. Today, however, our indicator has turned down, displaying much less strength than a year ago. True, our model has gotten less accurate over the past 12 months, so we may be picking up more of a mix shift than anything (that won't negatively affect the Case-Shiller data). Even so, we'd be surprised to see the strength in the Case-Shiller averages we saw last year in the months ahead. It's likely the best investors can hope for in the near term is flat sequential pricing (more likely slight declines), even though to date the Case-Shiller indexes are acting better than our model predicted in 2010.
In this case we hope we remain wrong, yet inventory dynamics probably affirm this negative short-term thesis. Both 2008 and 2009 were anomalous in that they didn't display the seasonal build in inventories typically seen in the spring months. As the chart below shows, inventory across most of the major metros in America was held to a 2% or less sequential growth rate in most of the seasonally strong spring months in both years. This was much different than in 2006 and 2007, where inventories grew more than 6% sequentially for a few months. We think some buyers were caught off guard by this lack of inventory growth last year, prompting a bit of a positive price reaction in the spring and early summer months. This year clearly has shown a different dynamic, as the spring inventory bump has again reappeared, although fortunately for what looks to be a shorter period than in years past. Bottom line, the reintroduction of seasonal inventory after a two-year hiatus makes pricing strength more difficult in the near term.
In addition, the talk among builders and industry participants is that demand has fallen off even more than most had expected after the April expiration of the tax credit, spurring some discounting. We don't think that such discounting is likely to be widespread and persistent, as the new-home market is devoid of standing inventory. At just 210,000 empty finished new homes in June, the new-home market has the least amount of inventory since 1968, a situation that bodes well for prices.
Bottom line, we expect sales, production, and prices of both new and existing homes to languish for some time as the market works off the effects of increased listed inventory and the expiration of a demand-shifting tax credit. That said, we feel both prices and production levels are well past their downward spirals.
Long Term Is a Different Story
When one thinks of years instead of months, some excitement is warranted, as we believe the housing market is setting itself up for number of strong years once a recovery begins in earnest. What's more, it's a matter of when, not if. As the chart below shows, the new-home market has been hit disproportionately in the current cycle, as its SAAR, at roughly an average of 300,000 units over the past two months, now resides 73% below peak levels (in 2005) while existing sales are only off about 25%. Even at the low 4 million mark, which, as mentioned above, is where we expect it to end up for a few months, the existing home sales rate will be off of peak levels by a comparably modest 34%.
The drastic draw-down in new-home production (both primary and rental) is important because it indicates the market is doing what needs to be done to mop up the oversupply created in the bubble years. In fact, the true amount of new supply introduced into the market over the three years ending in 2010 will be the lowest three-year total, by far, since the government started publishing the data in 1959. As much as people equate foreclosures to excess supply, they're really not. The units currently being foreclosed upon have been in the existing supply for some time. When pricing enables investors to buy the units and rent them back to former owners at a reasonable profit, prices stop going down. Many markets are currently at or below that level, meaning another drastic leg down in pricing across most of America is unlikely.
The Tricky Part Is Determining if and When a Jobs Recovery Might Begin in Earnest
With total home starts likely to average somewhere around 550,000 annual units for 2009 and 2010, in addition to the 905,000 units started in 2008, it's likely the current oversupply can be absorbed in short order. Importantly, the country has the potential to form roughly 1.2 million-1.5 million households under normal economic conditions with today's demographics. This, in conjunction with the 300,000 or so homes that are taken out of the supply every year, indicates a need for at least 1.5 million total housing units annually.
The problem is that most households require jobs to support themselves financially, and the lack thereof means household growth will remain well below potential until employment revs up. Importantly, employment is not a leading indicator of housing production; the data are very clear on this point. However, job growth is critical to sustaining any nascent recovery, and lack of jobs is currently the housing market's biggest problem. We're not sure when material job growth will return, but note there are a number of factors in place that suggest such a condition may not be far off. Productivity levels are at a point that would normally invoke hiring, the ISM employment survey and industrial production both indicate more full-time hiring is needed, and temporary hiring in several sectors has shifted into higher gear. Lastly, as the chart below shows, annual growth in the average work week of production workers, traditionally a very good leading indicator for private sector employment, has shot up to near unprecedented levels.
Not in 40 years has this metric been at such a level without impressive job growth shortly thereafter. If business heads can ever gain confidence that future business conditions will support payroll expansion, hiring will result. This of course will require cooperation from a federal government that to this point hasn't provided much, in this analyst's opinion. We're hopeful that after the midterm elections a healthier dynamic can ensue. If so, this should boost household formation somewhat, providing the critical ingredient needed to sustain any eventual housing rebound.
Stocks for Your Radar
As always, the question for investors is, "what is priced in and how does that differ from what's likely to transpire over a reasonable investment horizon?" At this point, we think more of the negative short-term dynamics are represented in the prices of many names than the long-term potential, although it appears several analysts are now warming to our view. Many of the homebuilders sell for less than, or right around, their tangible book values once deferred tax assets are added back. For instance, DR Horton (DHI), Pulte (PHM), and Toll Brothers (TOL) currently trade roughly equal to our estimate of their adjusted tangible book values, while KB Homes (KBH) and Lennar (LEN) both trade well below. NVR (NVR) trades well above book value, but its model is such that it deserves a much larger valuation than its competitors. In addition, management has just resumed the company's stock repurchase program after 10 quarters of dormancy, an event that's likely to boost earnings-per-share growth more than most think. All five builders enjoy extremely strong balance sheets, such that they'll easily be able to wait out the current downturn. What's more, the competitive position of the big builders relative to the small ones has likely never been better, as many of the latter remain in dire financial straits. As a result, it's likely the big builders will gain market share very early in any recovery.
Generally a low-quality business, the homebuilder group is not one we'd council investors pursue for buy-and-hold strategies. Yet sometimes the group offers compelling risk/reward opportunities for those willing to question the consensus. We think this may be one of those times.
Eric Landry does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.