Existing home sales for January were released Friday morning by the National Association of Realtors (NAR report). On a seasonally adjusted annualized basis (or SAAR), sales were down 5.1% compared to last January and down 5.1% compared to the previous month (Dec. 2013). Once again the reported results missed Wall Street's consensus forecast. Notwithstanding the debate about the effect the weather has been having on economic activity, the housing data continues to validate my view that housing market activity is declining and that the bear market trend in housing that started in 2005 has resumed.
As detailed in previous articles, my analysis is predicated on month to month sequential comparisons rather than the year over year comps that get reported in the media headlines. The factors that drove home sales higher in 2013 vs. 2012 - historically low mortgage rates and a big surge in investor buying - in my estimation are non-recurring and are fading from their effect on sales. The month to month comparison enables us to track the current trend in sales, especially since the "seasonal adjustments" of the data by the NAR are specifically designed to "cleanse" seasonality from the data series and thereby enable an "apples to apples" comparison of the month sequential data.
While the year over year home sales for January (462k SAAR) fell 5.1%, there has been a steady monthly decline from July's peak sales of 538k, or a 14% drop. Again, please note that the data has been seasonally adjusted. This graph sourced from the NAR illustrates the sharp decline in sales (edits in red are mine):
As you can see, since interest rates bottomed in May, there has been a precipitous drop in existing home sales on a month to month basis starting in July. Please note that existing home sales are recorded once a home sale is completely closed and title has transferred. Because the average time from contract signing to closing is 40-60 days, the peak in sales seen in July would reflect contracts signed when interest rates were at their lowest in May.
Another troubling indicator in the data linked at the top is the trend in prices. While the year over year headline shows an 8.8% increase for January, there has been a steady monthly decline in the average sales price since June. Month to month from December to January, the average selling price declined 3.8%. But from June's peak to January, the average price dropped 9%, from $261k to $237.5k. Part of the narrative for slowing home sales has been "affordability." But from June to January the average priced declined by over $23k, yet sales declined 14%. For me this calls into question the "affordability" excuse that has been widely attached to waning home sales.
The other troubling reasoning given for the decline in home sales in January is the cold weather. However, as I have detailed before, as it turns out the average temperature across the country, according to this study, has been normal in relation to the last ten years. In fact, the west coast experienced above average temperatures in January and yet, in January, the west region experienced by far the largest percentage drop in home sales both month to month and year over year (data in the report linked at the top). In my view, while the unusually cold weather in the east may have had some small effect on sales, because of the seasonal adjustments applied to the data, I believe that more serious fundamental economic factors are at play.
One factor for sure affecting home sales is the decline of the first-time buyer. This demand cohort accounted for 26% of sales in January vs. 27% in December and 30% in January 2013. According to the NAR, this is the lowest percentage of first-time buyers since the NAR started compiling that particular data metric in October 2008. Per the NAR, 40% is considered normal. The biggest problem with this factor is that the "organic" housing market is dependent on a cycle of first-time buyers who then eventually become "move-up" buyers. If first-time buyer demand declines, it undermines the fundamental demand structure of the housing market. In that the investor/flipper demand cohort is widely anticipated to fade this year - see this Zillow.com report for instance - if first-time buyer demand does not turn around, the housing market is headed for a bigger drop in sales going forward than most Wall Street forecasts anticipate.
Based on my analysis of the housing market, I believe that the current market caps of the new homebuilder stocks reflect investor expectations for the housing market that are far too optimistic. Using the Dow Jones Home Construction index (DJUSHB) as a good proxy for the homebuilder equities, I first recommended shorting the sector in late January 2013, when the DJUSHB was at 515. Since that time, it has traded between 550 and 390. It closed today (Friday, Feb. 21) at 513 - basically flat from my original recommendation. In the same period of time, however, the S&P 500 has risen over 21%.
Given that homebuilder stocks are more volatile than the overall stock market, we would have expected based on a theoretic risk/return model that the risk of owning homebuilders would have been rewarded with a rate of return greater than that of a broad market index like the SPX when the overall stock market is rising. In my view, this significant underperformance of the homebuilder stocks is indicative of the deteriorating housing market fundamentals. When the stock market eventually experiences a significant correction, the homebuilder stocks will likely decline even more than the SPX. Based on this I continue to like selected homebuilder short-sell plays. Please see my previous articles for ideas.