The speculative housing bubble of the mid-2000s has widely been considered the scapegoat for the ensuing financial crisis. I am not going to debate the relative importance of housing versus unbridled securitization, suffice it to say the key word is “unbridled”.
Because housing is “the” culprit, a number of news outfits have been created to chronicle its demise and forecast its premature death. Zillow is now a new tool in the appraiser / loan officer toolbox. RealtyTrac is the “source”. And CNBC’s Diana Olick is the megaphone. In the process, they are perpetuating the spiral. This is human since it is the spiral that put them in business in the first place, God forbid the situation should improve! Even seasonal improvements are suspicious – see Diane’s June 30th warning:
Would everyone please stop ignoring the fact that the recent monthly increases in home prices are largely seasonal!
Well, seasonal or not, will prices increase at some point? Diana’s article title was “Excluding Distressed Sales, Are Home Prices Just Fine?” I have contributed my two-cent worth to the subject in my own post last week, and given the number of page views and comments, it is clearly a controversial topic. Interestingly, this same topic, distressed sales and their impact on housing prices, was addressed by Chairman Bernanke as the last question of his FOMC press conference on June 22nd – fast forward to minute 47:24.
To add fuel to the fire and food for thought, I am going to give a shot at the black swan, a.k.a. The shadow inventory, as the distressed sales inventory has become known.
Any comparison between today’s crisis and the 30’s is scary. For good reason: Between 1929 and 1933, personal income fell 44%, real output fell 30%, and unemployment reached 25%. With regard to housing, David Wheelock, now Vice President of the St. Louis Federal Reserve Bank, penned a very interesting article in the Review of May/June 2008. Here is a quote from page 139:
Thus, at the beginning of 1934, approximately one-half of urban houses with an outstanding mortgage were in default (Bridewell, 1938, p.172). For comparison, in the fourth quarter of 2007, 3.6 percent of all U.S. residential mortgages and 20.4 percent of adjustable-rate subprime mortgages had been delinquent for at least 90 days.
To update this number, I looked at the New York Fed data. The peak for the national average of mortgages delinquent for 90+ days was 8.9% in Q1, 2010. In Q1, 2011, it was down to 7.5%. While much higher than the 2007 number, it never came close to the 1934 level.
The pundits will say: You forget Nevada! I will add: What about Florida? Both are respectively at 24% and 18%. Two things here. One, while these two states are way above the national average, all others are at 9% or below, including California. Two, while Nevada is indeed in trouble, it ranks 35th in population at 2.7 million, i.e. 0.8% of the total U.S. and therefore not the norm. So, we have just established a major difference between the 30’s and today. Then, 50% of mortgages were delinquent at the peak, today the number currently stands at 7.5%.
Next I looked at foreclosures. Some headlines claimed they were above the Great Depression levels. In the 30’s, they peaked in 1933 at 1.3%. During the four-year peak period of 1932 to 1935, the cumulative foreclosure rate was 5%, and 9.9% for the 1929-1938 period – by then, it had fallen back to where it had started, 0.6%.
They reached 1 million in 2010. Out of a total of 126 million units, this equated to 0.8%. The year 2011 is going to peak at 1.2 million, or 0.95%, according to the “source”, RealtyTrac. Just a note here. There is a big difference between what RealtyTrac tracks as foreclosure filings - 2.8 million in 2010 and 2009, 2.3 million in 2008, etc., and the actual number of foreclosures. While the trend is interesting to follow, reality means distressed sales, which mostly encompass foreclosures, a.k.a. REOs, are around 80% of the total, along with short sales.
When is this going to end? It’s the shadow inventory, stupid! Here is what we know:
There are about 3.9 million existing homes for sale reported by the NAR, and a measly 167,000 new homes inventory (Hanley Wood estimate), for a total visible inventory of 4 million, round numbers.
First American Core Logic (CLGX) estimates the shadow inventory at 1.7 million units, for a grand total inventory of 5.7 million units. I have not been able to corroborate the number but I will use it a ballpark estimate.
Since 2009, distressed sales have accounted for about 30% of total sales, with an uptick to 35% in Q1, 2010. Total sales have averaged 5 million units per year over this period.
Since 2005, there were 36.7 million existing home sales and roughly 7 million distressed sales.
From 2004 to mid-2007, the bubble years, there were 23.1 million existing home sales and 3.8 million new home sales, for a total of 26.9 million.
Nevada is a data point, albeit statistically meaningless.
Therefore, if Core Logic is closer to reality than Realty Trac, then:
The shadow inventory will have disappeared in about 13 months.
Distressed sales are discounted by 25% on average. So to be conservative, this means that when all will be said and done, prices will have gone up by 8% on a static basis, i.e. by simply looking at the average mix going from index 92 to 100.
However, on a dynamic basis, prices may well go much higher. If index 92 reflects the current mix – 4 million visible inventory at 100, 1.7 million Shadow Inventory at 75 -, what will the index be when total inventory will have gone down, in a year’s time, from the current 5.7 million to 4 million? That is, from a 13.7 months of supply to a 9.7 months of supply. While longer term, the slack is made up by an increase in new homes supply, I doubt the homebuilders are ready to bank on this scenario just yet.
Lastly, and to test the assumptions made here, the 7 million distressed sales since 2005 represent 5.5% of the housing stock, a number comparable to the 7% of the 1931-1936 period. I would say the number is high, considering where we were in the 30’s. It looks even higher when compared to the 26.9 million total sales of the bubble years. Are we saying that 26% of these sales went bad? Worse, since only two out of three sales were financed by new mortgages, based on census 2010 data, are we saying that 39% of these mortgages went bad? And do we have another 10% to go? (for those of you who lost the math here: 26.9 / 3 * 2 = 17.9; 7 out of 17.9 = 39%; 1.7 out of 17.9 = 10%). Remember, the national delinquency average is 7.5%, and 85% of the outstanding mortgages originated since 1995… The math does not check out as 49% is way too far from 7.5%.
My hunch is that we are much closer to the end of the shadow inventory than even Core Logic thinks. Here is where RealtyTrac comes in handy. In May 2011, RealtyTrac numbered 215,000 properties with foreclosure filings, i.e. one out of 605. This is a dramatic improvement over the Q3, 2009 number of one out of 136, and a 33% decrease from a year ago. Looking at the NY Fed data on mortgages, aside from the improvement from the 2009 peak delinquency rate, I find the same improvement in the so-called Quarterly Transition Rate for 30-60 Day Late Mortgage Accounts – this is the percentage of accounts that go from this category to either current or more delinquent, 90-Day Late. From 2006 to mid-2009, the percentage of accounts getting worse went from 10% to 45%, while those getting better went from 50% to 20%. Since mid-2009, the reverse is occurring: we are now at 28% for those getting worse, and at 31% for those getting better. I do not need much more to call this a bottom in housing, whether from an economic or stock market standpoint.
To digress for a second, I was looking at my Dow Jones short term model, trying to pick the next move. Some charts were uncharted – I had to go back to the long term. IBM (IBM) is up 200% from its 2000 low; United Technologies Corp. (UTX) is up 230%; McDonald's (MCD) is up 550%. Interestingly, these were three different sectors – technology, capital goods, and consumer durables. Is housing next? And if it is, should I be patient enough to buy and hold - my nose? My answer is: Yes and yes. I may be off by a few months, but I would rather be early.
Disclosure: I am long TOL, JOE, PHM.