Banner year for housing
admin | November 22, 2019
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.
The recovery from the housing bust of a decade ago has been slow and in some cases incomplete. Housing demand stalled late last year but has revived in 2019. In fact, housing on a number of measures has hit new highs since before the financial crisis.
Setting new post-crisis highs
Housing demand has strengthened this year. Industry sources primarily attribute the improvement in 2019 to a rise in affordability, reflecting both a drop in mortgage rates and a moderation in the pace of home price appreciation. A number of major housing indicators have hit new highs this year (Exhibit 1).
Exhibit 1: New Highs for Housing Indicators
Builders have cranked up their activity recently. Housing starts hit a cycle high in August, while building permits jumped to a new recent high last month. Single-family starts in October were only the third highest of the cycle but have risen for five straight months and seem likely to exceed the January high-water mark soon.
Meanwhile, new home sales exceeded 700,000 units annualized in August and September, the first such back-to-back readings since 2007, though the latest observation is somewhat below the June figure. Similarly, the Mortgage Bankers’ Association Purchase Index has achieved levels this year not seen in a decade.
The outlier in this story is existing home sales and the related pending home sales series. The NAR’s measure of resales has been improving this year but remains well off the best levels seen in 2017. A frequent refrain from NAR releases, repeated with the October data released Thursday, has been that sales have been constrained by a lack of supply. In fact, the NAR chief economist explicitly and forcefully argued in the most recent release that sales would be higher if there were more existing homes on the market.
The process of recovering from the housing bust has been long and onerous. In fact, a number of key housing sector indicators arguably remain depressed by historical standards. Exhibit 2 compares current levels for key housing sector variables with the extremes of the Housing Boom and Housing Bust, as well as the 1995-2000 average. The late-1990s average could arguably represent something close to a fair value for a strong economy, though it is worth noting that the population has grown considerably over the past 20 years. In theory, housing activity should be considerably higher than it was in the late-1990s, all else equal.
Exhibit 2: Housing indicators levels
Housing starts and building permits are currently running roughly halfway between the boom highs and the bust lows. The latest levels remain well below the average of the late-1990s, even though the number of households in the U.S. is up about 20% over the past 20 years.
The results are even more striking for single-family starts and permits. Multi-unit construction has been robust in the recovery years since the bust, partly because many households have decided in the wake of a steep drop in home prices during the bust that renting is better than owning. Single-family starts and permits are running 20% to 25% below the pace of the late 1990s. Indeed, as I discussed in a piece earlier this year, the pace of starts is running well below the rate of household formation, so that a shortage of available housing has developed.
In contrast, existing home sales exceeded the late-1990s pace by early in the current decade and are currently running about 15% ahead of the 1995-2000 average. In fact, this measure returned to normal just a few years after the depths of the housing bust. There are many possible explanations for the divergent performance of new and existing home sales, but one prominent one would be that prices for existing homes were so depressed by the massive foreclosure-related overhang in supply in the aftermath of the housing bust, that builders were essentially priced out of the market for a while. In other words, for a period of time, the cost of building a new home was significantly higher than the market price for a comparable existing home, especially in depressed markets.
While that dynamic should have dissipated by now, builders widely report that the costs of new construction have risen sharply in recent years, reflecting zoning laws, environmental requirements, and, to some extent, high commodity prices. Finally, the mix appears to be normalizing somewhat this year, as new home sales have increased considerably, while existing home sales are flat (year-to-date average vs. 2018 average), constrained by tight inventories.
There are many different ways to interpret the data in these exhibits. From a broad, macroeconomic perspective, my primary takeaway is that the behavior of the housing sector has been far different in this expansion than in prior business cycles. Typically, housing is one of the first sectors to turn up in a recovery and down in a recession, and the magnitude of cyclical swings has historically been steep. Coming out of the housing bust, there were many structural impediments to a violent housing recovery. In addition, builders were undoubtedly in a “once bitten, twice shy” mode. Thus, in the current business cycle, the housing sector has played more of a “tortoise” role than the traditional “hare” pattern. The good news in this is that the housing sector, even more than a decade into the current expansion, does not appear to exhibit the typical strains and excesses one might normally expect after such a long cycle. In fact, there is arguably plenty of headroom, so that, if the underlying fundamentals for demand remain robust, housing activity can continue to advance for some time.