Hits: 4062

Well, the Case Shiller index has finally produced a positive print. Again, I will probably sound like a permabear, but this may not be all that it is cracked up to be. I have warned readers two years ago that the Case Shiller index, although an econometric marvel, is far from perfect in terms of determing the state of residential housing in the real world. 

The primary suspects are: 

  1. It ignores investment inventory which, when combined with poorly underwritten easy credit loans, was the catalyst for the housing bust in the first place. Investors simply walked away or were foreclosed upon, en masse. 
  2. It ignores multi-family housing, which is a significant portion of the stock in urban areas such as NYC. It is also a much higher risk loan that shows more defaults in mortgage portfolios.
  3. It ignores condos and coops. See "Who are ya gonna believe, the pundits or your lying eyes?" and "Who are you going to believe, the pundits or your lying eyes, part 2". for a clear understandng of how bad off this vital urban and suburban market actually is. The recent Case Shiller condo numbers show a statistical uptick, but as can be seen from the ground (reference previous links), the inventory story is simply atrocious, and their is plenty of additional inventory being built as I type this, which just adds to the foreclosure and existing sales inventory issues.My assumption is the government stimulus (which ends right about now) offering $8k tax breakes, seasonality (as this uptick occurred in the strongest historical selling season, the coming to market of larger apartments as inventory is completed (remember, it appears as if this index tracks gross prices, and not prices per square foot, which can be quite misleading in terms of actual price appreciation), combined with the GSE occupancy waiver (which very well may backfire as it brings back the easy mondy credit days of lax underwriting) is responsible for the trend. Will it last? We shall see, but the laws of supply and demand will apparently have to be suspended. 


Here is what others around the web are saying as sourced from RGE Monitor: 

Relevent Links

S&P says home prices on upswing in 2Q 09

  FHFA Home Price Index: Home Prices Decline Slightly in Second Quarter

Calculated Risk on Case-Shiller House Price Seasonal Adjustment and Comparison to Stress Tests

 Shiller (creator of the Case Shiller indes) on why housing prices may keep falling

PMI Economic and Real Estate Trends Q2 2009



Here is what others around the web are saying as sourced from RGE Monitor: 

Relevent Links

S&P says home prices on upswing in 2Q 09

  FHFA Home Price Index: Home Prices Decline Slightly in Second Quarter

Calculated Risk on Case-Shiller House Price Seasonal Adjustment and Comparison to Stress Tests

 Shiller (creator of the Case Shiller indes) on why housing prices may keep falling

PMI Economic and Real Estate Trends Q2 2009


  • The S&P/Case-Shiller 20-City Composite Index fell 15.4% y/y in June 2009 after declining 17.1% y/y in May. Of the 20 cities covered in the survey, 18 showed an improvement in the y/y return in June over May, though the y/y decline continued to be in double digits for 15 cities. The m/m return in June was positive for 15 of the 20 cities after accounting for seasonality. As of June 2009, average home prices are at similar levels to what they were in early 2003. From the peak in mid-2006, the 10-City Composite is down 32.5% and the 20-City Composite is down 31.4%.(S&P)
  • David M. Blitzer, Chairman of the Index Committee, S&P: " This is the first time we have seen a positive quarter-over-quarter print in three years. Both the 10-City and 20-City Composites posted monthly increases, as did most of the cities. As seen in both seasonally adjusted and unadjusted data, as well as the charts, there are hints of an upward turn from a bottom. However, some of the hardest hit cities, especially in the Sun Belt, show continued weakness." (August 25, 2009)
  • Ajay Rajadhyaksha and Glenn Boyd, Analysts, Barclays Capital: Even seasonally adjusted home-price data has been skewed higher during the spring months of this year and last year by an “amplified” version of typical patterns. More homeowners sell their properties during those months, cutting the share of homes being offloaded at distressed prices, as new buyers focus on “desirable neighborhoods” where values hold up better. (via Bloomberg; July 31, 2009) I agree.
  • Calculated Risk Blog: "So far house prices are tracking the baseline scenario, but I believe the seasonal adjustment is insufficient and prices will decline faster in the Fall." (July 28, 2009)
  • Professor Robert J. Shiller, Yale University: Prices may continue to fall or stagnate into 2010 or 2011. There is much less demand for housing than is supply. Even if there is a quick end to the recession, the housing market's poor performance may linger. (via NYT; June 7, 2009) Cities with relatively smaller declines like Boston could show signs of rebound, after the pace of foreclosures has fallen. Given the inventory on the markets and shadow inventory of houses, a scenario where home prices decline for years is more probable. (via Yahoo Finance video; July 14, 2009)
  • Professor William C Wheaton, MIT: The supply side fundamentals of the housing market point to a bottoming out and prices should stabilize soon and resume rising. The excess inventory of housing should be back below normal in 2011. Most US houses are valued at or under replacement cost and decades of house price decline, as was the experience in Japan, is unlikely. (via VoxEU; July 18, 2009)
  • Robert Mellman, JPMorgan: "Declines in the Case-Shiller index probably are moderating, but the latest monthly reading overstates the extent...the number of forced sales entering the market is expected to increase substantially during the second half of this year, a source of significant further downward pressure on house prices." (July 31, 2009)
  • The PMI Group: The risk of home price declines in most Metropolitan Statistical Areas (MSAs) increased in Q1 2009 as per the PMI U.S. Market Risk Index, driven by unemployment and foreclosure rates. 60% of the nations top 50 MSAs have a greater than 50% chance of declining home prices in two years time, with the risk highest for MSAs in Arizona, Nevada California and Florida.


Thoughts on Case Shiller - June 2009 from the BoomBustBlog community

Table of Contents
  • General
  • Quantifying Seasonality - 58% of the uptick
  • Data anomaly? - Seems unlikely
  • Attractive price? - No, homes are still quite overvalued!
  • Attractive Financing? - Yes this is probably the other 42%
  • Animal Spirits? - No.
I believe this supports the belief that we have another 20% at least of home price declines to go from here.


Case Shiller for June 2009 came out today (source). The figures were strong. June 2009 Composite-20 was 141.86, up 1.4% from May's 139.84. People are now calling a firm bottom, that we are off to the races again (source).

Last month there was some probability that this was pure seasonality and nothing else, with the most likely case that we were seeing some underlying stabilization in pricing occur. This now puts to rest any probability of pure seasonality. There is clearly strength here.

The question is, "where is this strength coming from and what will happen to those drivers?"

Possible sources:
  • Data anomaly
  • Seasonality
  • Attractive price
  • Attractive financing
  • Animal spirits
I will dig into these in turn. As this analysis will show, this is a combination of seasonality and the effects of the tax credit in concert with the FHA's 3.5% downpayment requirement. The credit was put in place earlier but it expires 12/1/2009. Like Cash for Clunkers, some people are making sure to get this done before the program expires.

Prices are unambiguously still too high. I also present a coherent argument against some of the traditional arguments:
  • Affordability is high
  • Home price movements are like the Titanic, and they are now positive
The fact that a normal housing market has positive serial autocorrelation does not mean that one can artificially prop a market, get a bounce, then remove the prop and expect the market to continue along an upward trajectory. Economics doesn't work like that. Yet that is what many otherwise intelligent people are trying to argue right now.

The answer to these questions are a big deal. If home prices rebound strongly and the housing market spontaneously rebounds, this will be very supportive for our economy.

Quantifying Seasonality - 58% of the uptick

In the April 2009 I had noted the following:
"Last month I had noted the following regarding the seasonality of Case Shiller data:
"There appears to be some seasonality in the data. This happened in 2008 and in 2007, plus or minus a month - just check out those m/m declines. The decline eases off from March to June before picking back up again. We appear to be tracking this."
With this in mind, consider the following chart of the Case Shiller Composite 20 Index: Observations:
  • Clear seasonality. I annotate to highlight the very clear seasonality in this graph. It always tends to come b/w Jan and June, give or take. We saw this in 2005, in 2006, in 2007, in 2008, and it appears we are again seeing it in 2009.
  • Lost in March, gained in April. Note too the lack of a seasonal uptick in March 2009 after a clear seasonal uptick in February 2009. We then saw the m/m decline shrink to an apparent -0.6% (7.4% annualized). Logic would suggest that whatever factor caused March 2009 to be worse than expected, caused April to be better than expected.
Well, make that through June, as it was in 2008.

This is an updated graph of the m/m change:
Last year, in the midst of continued deterioration, we saw an increase in the m/m decline of 2.1%. Feb 2008's m/m decline was 2.6%, while June 2008's m/m decline was 0.5%.

This year, Feb 2009 came in at (2.2%) while June 2009's m/m growth was 1.4%, improvement of 3.6%. So, I would tend to think seasonality accounted again for about 2.1% of the 3.6%, or 58% of the "improvement".

Not everything but still is quite large.

Data anomaly? - Seems unlikely

One may ask, "could this be due to some sort of a shift to higher priced homes as the crisis moves from subprime to prime?" I would posit no. The Case Shiller index methodology is available here (source), with the short form here (source). They use a value weighted "matched pairs" methodology and a 2 month lagged, 3 month moving average. They focus only on single family homes. This is a stable methodology.

One may disagree with Shiller's conclusions or policy prescriptions, but I wouldn't doubt his index.

Attractive price? - No, homes are still quite overvalued!

This is a double header chart, showing a long term historical chart of real housing prices starting in 1900 which clearly establishes mean reversion of late, and an update to 6/09 using recent Case Shiller and CPI data:

Yes, homes are cheaper, but are they cheap? No. They are still inflated.

And then there is a view of home prices relative to household income:
I would strongly consider those questions about mean reversion, the next stage of the housing crash, financing, and demand.

I believe this is especially the case given the following considerations:
  • We have over 2 years of housing inventory, trying to get sold.
  • We have a bunch of banks that are increasingly realizing that foreclosure is their best option at this time, as they come to the conclusion that preventable foreclosures are few in number and costly to attempt to seek out.
  • The suburban home model is in the process of dying, as we transition out of a peak cheap oil world.
  • We have people losing jobs like we have never seen before (see chart below).
  • We have falling wages and declining consumer credit.
  • We have high job uncertainty.
  • Mortgage rate spreads and financing terms are still artificially loose, held down by our socialized and insolvent mortgage system which I endearingly call "Fannie, Freddie, and FHA". However, financing terms are in the process of normalizing.
  • Treasury rates which drive mortgage rates are higher than they were, but are still quite low relative to their inherent credit risk.

Attractive Financing? - Yes this is probably the other 42%

Here is where some of the "juice" may be coming from, in addition to seasonality.

For one, something like 95% of all mortgage originations are being done through governmental mortgage operations - FNM, FRE, FHA, GNM. Consider this WSJ article (WSJ 8/12/09). Some of the facts on the FHA are damning:
  • FHA accepts a trifling 3.5% downpayment, which itself can be whittled down to 2% through "tricks" (implying that with the tax credit, no downpayment is actually made)
  • Ginnie bundles, guarantees and sells FHA insured mortgages, bearing the risk of FHA underwriting standards.
  • Ginnie's mortgage portfolio has doubled since the end of 2006 and has grown rapidly recently.
  • At least 9 out of every 10 mortgages now carries a guarantee from one of these governmental housing organizations.
  • FHA's reserve fund, their equity buffer against loss, leaves them with an implied 33x leverage ratio vs 16x in 2007.
FNM/FRE were put into conservatorship. Losses there are growing, after which we heard they are being wound down. Then FNM/FRE's regulator quit. Then the head of GNM quit. Huge losses are sitting within these institutions that are not being recognized. They will be, at some point. Right now they are being hidden.

Once again I will note the first time homebuyer tax credit, with a lead-in provided by CalcRisk (CR 8/24/09).

Key facts:
  • Tax credit: 10% of the purchase price up to $8000
  • Downpayment deductibility: Yes, can be applied towards downpayment
  • Home purchase date range: 1/1/09 - 12/1/09
  • Required adjusted gross income: $135k or less ($260k or less joint filed)
  • Required lack of owned home beforehand: 3 years
  • Required duration staying in new home: 3 years at least
The combo of FHA and tax credit allow for no money down homebuying, like the good old days. CalcRisk notes the following:
"Anecdote: I've spoken with two younger guys (30 ish) who told me they had no down payment, but edit: were looking to buy a house NOW. They are using the tax credit and FHA to buy. I think that conversation is happening in many places."
We have quantified a large flurry of first time homebuyers doing this. This is almost certainly creating a strong patch of demand which should not recur. These buyers are getting a nice discount, which adjusts the downside calculation for the downwards.

CalcRisk believes it is highly likely that this will get renewed by Congress, though he believes its effect will wane over time. His rationale is that the homebuilder lobbies are "pulling out all the stops". Is that all it takes these days, to get free bailout money?

Maybe it is just me, but I do not share his optimism. You can't have it both ways. On the one side are price and volume increases. On the other side are massive deficits, that are generating increasing attention. There is rightfully a bias against pure consumption subsidization, on the back of all taxpayers. If we keep this up, the argument the homebuilders will have to resort to is "we need to keep this in place to keep these numbers up, and if we don't they will crash". And they will be right. But like with CIT, and with Cash for Clunkers, government will be thinking "well, we've given them a boatload of our money, deficits are a bigger issue, it looks like this recovery is for real, and that program has gone on for almost a year in any case. Maybe it's time to cut the umbilical cord."

When people talk about home price affordability being at multi-decade highs, this is the sort of stuff that they are referring to. You can buy an overpriced home with artificially cheap financing. Cheap debt servicing does not make a home cheap. Again, no one could make me buy GGP stock at 30, even if I could borrow at 0% to do so. The equity value is not worth that price, yet I will have to pay back that loan principal.

Animal Spirits? - No.

One could also make the argument that we could see something in housing akin to what we saw after the Tech bubble, where lots of stimulus prompted the stock market to not fall as it normally would before rising up normally again.

Look at the deviation from "trend" in this graph:
As Janszen has noted though, housing is different from an equity bubble. Housing bust is death by a thousand cuts. Home prices tend to be strongly serially autocorrelated, moving in the same direction for long spans of time.

Equities are put in a separate risk bucket. They tend to crash out. Conversely, they are more prone to wild snap backs.

A housing bull would likely say that even though home prices have not reached "cheap", they did go down a bunch, the financing is good (and may not persist for too much longer), money is cheap, and home prices have now moved up 2 months in a row. Home prices are like the Titanic, and the Titanic has shifted into "moving up" phase.

To this I would respond yes, home prices are like the Titanic. But home price bubbles do not crash out in 3.5 years then jerk upwards again, as we are seeing now. And that dynamic does not mean that one can artificially prop a market, get a bounce, then remove the prop and expect the market to continue along an upward trajectory. Economics doesn't work like that.

First consider this chart of Japanese home prices:

Japan asset bubble us asset bubble


US and Japan housing bubbles

Their prices peaked in the middle of 1991 and have declined ever since. We have now seen 18 years of decline.

One may then counter that Japan is a one off case due to their poor monetary policy. Well, then what about Los Angeles?

Consider the following chart:
Note, our bubble was bigger, stronger and longer than theirs. Ours has also has considerably more stimulation than theirs. Yet periodically throughout that bubble we saw seasonal upticks, and they were also during the March-June/August time frame.

Our housing prices peaked in December 2005. Through June 2009, that's 3.5 years. Even relative to this smaller bust, we are still in the crash phase, which is then followed by a long tail of lower prices at a diminished rate of decline. Given our bubble was much bigger and longer in the making, I contend this will be longer if anything, not shorter.

And again, if you subsidize 10% of the purchase price of a home and then subsidize the financing, you will get a bump. But to use the logic that organic home price movements are like the Titanic, so ample stimulus that is then removed in the context of overvaluation will create a self-sustaining bump, is just silly.