Housing Stock and Housing Stocks
Prior to the go-go years when people would be flipping Miami condos a dozen a day, the only thing sexy about the housing industry was shag carpeting. All of the businesses associated with the industry are dull, but I think there is a strong likelihood that there are gems in the rubble. In the past few months, a basket of stocks I would broadly define as housing related has dropped even more than the market (30-40%), so I’ve begun poking around. This has much more to do with cheapness relative to some vague conception of normalized earnings power than some macro call.
For those who look beyond Berkshire’s insurance operations, another major contributor to the company is housing-related industries: Shaw Industries (carpet), Johns Manville (insulation), Acme Brick, MiTek (connectors), Clayton (trailers), and Benjamin Moore (paints). If you look at the 2000-2006 letters, the housing related businesses get mentioned as easy to understand businesses that make attractive returns on tangible capital over an economic cycle. For once, I don’t think he is making it sound a lot easier than it really is.
With the exception of Clayton, all of these businesses contribute a small but important part to the final product. Instead of paying Michael Porter to reveal a secret of the business world, I’ll just throw it out there for free. When someone produces a product that is a very small but crucial element to an entire system, it creates a competitive advantage. Whether or not the competitive advantage is exploited will reveal itself in the financials. Shaving 30% of your costs off an input that is 1% of your total cost is meaningless to the bottom line if it means your end product is crap. Such a cost differential becomes increasingly irrelevant as the expected life of the product increases. This is probably why stuff like backyard furnace steel is usually not a hit. Nobody wants to drive over a bridge or sleep in a building made with that stuff.
Cheap imports exist for sure, but there is a degree of brand equity and reputation for certain products. Never doubt that there are people always looking to cut corners and eke out a couple basis points more of margin. In at least one instance, the relative cost of a slightly more expensive initial input is vastly overshadowed by the legal costs and damages stemming from issues like Chinese drywall (links, links, links if you don’t remember reading about this). The concentration of these issues in Florida might suggest purchases of inferior inputs are linked to more speculative builders and the broader industry is reliant on reputable suppliers, although perhaps it is a self-serving interpretation. One might say that it means domestic producers are facing pressure. Not every input for a building is supplied by someone with a competitive advantage. MiTek, which makes truss connectors, has a competitive advantage because people don’t care about relative savings of $200 on a $100,000 building if it means the roof will cave in.
I don’t think the above narrative that every housing related company makes a small part of a larger system is correct, but I think it is the right qualitative aspect to seek in the sector. All of the write offs and restructurings over the past five years should have them on solid footing by now in terms of cost structure. A house that is worth $100k or $800k still needs sinks, cabinets, doors, etc., so one isn’t betting exclusively on a recovery in home prices, even though that will probably be the evidence that many people point to as housing recovering. Certain industries such as gypsum or concrete may not have obvious competitive advantages that stem from brands, but they can be low-cost producers and be able to survive just a little longer than the other guy. All of these factors should minimize the need to fret over the macro picture, in addition to buying the companies cheap.
To bring this back to what Buffett is talking about with housing, nobody really knows exactly when it recovers. The rocket science is in recognizing people need places to live and there are going to be more people tomorrow than there are today. The calculus is fudged since household formations are being pushed out by poor economic prospects. Perhaps I’m being stubborn in assuming people will not always want to eventually live without their parents or friends. The average in annual housing starts from 1959-2010 was about 1.5 million and recent years have been well below that — about .6 million in 2010. Is there a huge risk in betting on a reversion to the mean in the next five years with a skewed risk/reward scenario due to the prices of the relevant businesses?
So even though incremental demand exceeds incremental supply in theory, the lack of household formations isn’t tipping the balance. There’s a chicken and egg problem with people talking about how housing will get the economy going again, when the inverse is equally true that getting the economy going again will boost housing. As Sam Zell pointed out in a recent interview, there are actually more single family homes rented than apartments in the U.S. Home ownership rates and house prices are not the only things that matter, since people need places to live. There are many dueling data points. Time is an oft-ignored factor that has an effect on the economy, but nobody talks about it since there is nothing they can do about it — no articles to write when all people should do is wait, no sales commissions to be generated, no political talking points.
I can’t calculate any of this with precision, but feel comfortable stating that the housing industry will revert to the mean sometime over the next 5-10 years. If there is a business that trades at 5 times normalized earnings and will survive, you have a double or triple bagger within the next 5-10 years if it eventually trades at 10-15 times normalized earnings, which produces a nothing-to-scoff-at 15-30% annualized return. A bias to avoid is just picking earnings from 2004-2006 and thinking that’s “normalized” — housing starts were well above the 1.5 million average then (easily adjustable graph of all the data). Can we just pluck earnings from years when housing starts were 1.5 million, then? No. Plenty of companies have closed down facilities in the past five years, so perhaps differing competitive dynamics will emerge with better positioned suppliers. It's not so simple that tossing around multiples and gathering evidence to support self-serving normalized earnings will generate investment returns.
I don’t know what will fix housing and I don’t know what will happen. I do think the current macroeconomic rumblings from everyone are myopic and present an opportunity. Well capitalized, cash flow positive, housing related stocks with a demonstrated history of attractive returns over a cycle, and then wait, sounds like something that might work. This is an edge for long-term value investors since it requires patience and a temporarily contrarian perspective.