While we weren’t at the most recent Value Investors Congress, we have heard about a number of the ideas that were presented. As usual, it sounds as if most of the ideas were thoughtful and well researched. The one idea that we would take the other side of, though, was Bill Ackman from Pershing Square’s idea to be long, literally, U.S. housing, which was unveiled in a presentation titled, “How To Make a Fortune”. To state it bluntly, we think Ackman is wrong on housing.
We understand his thesis on U.S. housing focuses on a few key points. First, affordability is at its highest (so most affordable) in decades due to low low mortgage rates. Second, household formation will rebound and go back to long term trends, which suggest growth in demand. Third, supply of housing, which he admits is high, will start to decline as builder production rates are as low as they have ever been. Finally, he believes the downside in housing is limited because at a price institutions could step in and soak up the excess inventory. To be fair to Ackman this is a secondhand summary of his thesis, but we wanted to address some of these key points and highlight where this thesis falls short.
The oft used point to support a bottom in U.S. home prices is affordability, which is underscored by the fact that mortgage rates are at all time lows and make the financing costs as low as they have ever been. While we can’t disagree that mortgage rates are at all time lows (that is just a fact), we do disagree on affordability. First, credit standards have increased and lenders typically require larger down payments and more upfront points, which increase the “all-in cost” of a house. (If the consumer can get a loan at all.) Second, we believe, based on our supply and demand models, that home prices have anywhere between 15 – 30% more downside, which implies that the U.S. housing stock is actually overpriced. Finally, and most importantly, our analysis actually shows as houses get “cheaper”, or more affordable, demand goes down.
As we’ve highlighted in the chart below, based on our proprietary census work, household formation has turned negative for literally the first time ever. This is attributed to the fact that individuals are getting married at later and later ages. In fact, we have seen a 1000 basis points growth in unmarried people in the aged of 25 – 34 over the last decade. As these people get married less and later, it has a commensurate impact on household formation. In the shorter term, unemployment is also a key negative catalyst for household formation. If the long term trend in the chart below tells us anything, we shouldn’t look at history as a guide for future household formation.
Despite new homebuilding rates being at all time lows, we have seen no meaningful improvement in the national housing inventory overhang. In the chart directly below, we highlight months of supply of homes on the market. Currently, there are almost 11 months of supply on the market, which is near the highs of 2008. Specifically, there are now 4.04 million housing units on the market, a number which has accelerating throughout the year. So, while it would be nice if low new homebuilding rates had an impact on this massive inventory overhang, they do not. The primary reason for this is, of course, that new home sales are a small percentage of the overall housing market. (As an aside, there are also an estimated 6 million houses that are not on the market, but are considered “shadow” inventory.)
Institutional buying of houses
While on a limited basis, small funds have been created to buy housing stock, we have not seen institutions stepping up on a larger scale to buy houses. The primary reason for this is that individual homes don’t lend themselves to purchases of scale due to their localized nature. Each and every neighborhood is unique and has its own attributes from which value must be determined and researched. As well, the management of single family housing as an asset is labor intensive as it relates to managing the rentals of these properties. Further, a wide-scale institutional buyout of housing stock would require banks to suffer massive losses on their loan books – a scenario that they have been avoiding the entire time, as evidenced by the growth in average number of days homeowners spend in foreclosure (which is also impacted by other factors such as moratoriums and litigation).
Our Financials Sector Head Josh Steiner and his associate Allison Kaptur have done the bulk of our work on housing, include a 101 page presentation, and as noted above, one of the primary reasons that we remain bearish on housing is that we expect future prices to decline anywhere between 15 – 30%. Specifically, supply of housing is in the top two deciles of inventory levels, and historically prices have typically fallen more than 15% over the following 15 months when at these levels. The correlation on an r-squared basis of supply and future pricing is 0.83.
As our CEO Keith McCullough likes to say, facts don’t lie, people do. As the housing river cards continue to show their data, it is becoming increasingly that U.S. housing in no bargain.
Daryl G. Jones