Wednesday, March 28, 2007

Bear markets: how deep, how long?

Using the S&P/Case-Shiller® Home Price Index for comparable sales of single family homes in the Boston MSA, the peak of the late 80's / early 90's market occured in July 1988. Plotted below in black are the percent price declines in the subsequent 106 months relative to that peak.


The market did not bottom until ~36-48 months after the peak. One way to interpret this plot is that on aggregate, a home purchased in July 1988 would be sold at a loss for the next 106 months (>8.5 yrs). A home purchased when half of the ultimate price decline was complete (-8% @ ~24 months) would be sold at a loss during roughly the next 56 months (>4.5 yrs). These are NOT inflation adjusted prices. This does NOT include commision costs.

There is no reason to expect history to repeat itself exactly, but the yellow bars on the plot show very similar price declines from the recent September 05 market peak. The arrow indicates the most recent data release for January 2007. The final five bars in the plot correspond to the CME futures for the Boston MSA for the next 12 months (as of March 27).

Only time will tell whether the current bear market wil match or exceed the severity of the previous cycle. But the plot does provide cautionary evidence that purchasing in a falling market is no protection from potential losses for extended durations. Next time someone tells you "now is a great time to buy," keep this in mind.

Wednesday, March 21, 2007

February Market Wrap

MAR and the Warren Group reported market stats for MA this week. For once their numbers were in reasonable agreement. See the Boston Globe for the overview (here). Below are plots of SFH and condo median prices, as reported by MAR for the past few years:

One thing that I have been noticing is that MAR is continually revising their stats from past months/years. I don't have the time to keep up with their changes, and they don't bother to report the changes when they make them, so I will have to acknowledge that the numbers I'm plotting are "as originally reported." The differences are typically small, so not too big of a deal I guess. Doesn't do much for my confidence. I plan to spend more time from now on using the S&P/Case-Shiller® Home Price Indices, which have a number of advantages (see previous post here).

Months inventory continued to run below year ago levels in February:
December/January were unusually warm, and mortgage rates dropped, likely contributing to unusual seasonal strength in Jan/Feb closings. In Feb the weather turned and rates rose, and I expect the March closings and months inventory will suffer by comparison to last March (which was relatively strong).

Wednesday, March 14, 2007

Comparing the previous and current real estate cycles

The Case-Schiller Home Price Index is available for a number of markets around the US. This index tracks the repeat sales of single family homes, and thus serves as a unique measure of changes in house prices across time, without distortions from new development, changes in sales mix, etc. A detailed discussion of the index methodology is available in PDF format (here). From this document:

The S&P/Case-Shiller® Home Price Indices are designed to be a reliable and consistent benchmark of housing prices in the United States. Their purpose is to measure the average change in home prices in a particular geographic market. They are calculated monthly and cover 20 major metropolitan areas (Metropolitan Statistical Areas or MSAs), which are also aggregated to form two composites – one comprising 10 of the metro areas, the other comprising all 20. The indices measure changes in housing market prices given a constant level of quality. Changes in the types and sizes of houses or changes in the physical characteristics of houses are specifically excluded from the calculations to avoid incorrectly affecting the index value.

The historical indices go back to 1987, and are available in Excel fromat from Standard & Poor's (here). I've extracted and plotted the data for the Boston MSA to emphasize a few points. The Boston MSA consists of these counties: Essex MA, Middlesex MA, Norfolk MA, Plymouth MA, Suffolk MA, Rockingham NH, and Strafford NH. Here's how prices in this MSA have behaved since 1987:

The prices are indexed to a value of 100 in January 2000, thus this plot shows relative price changes indexed to that date. Interestingly, the Chicago Mercantile Exchange (CME) now allows futures trading based on this index (here). The last five bars in the plot correspond to the contract prices for futures over the next 5 quarters, recorded on 3/12/07. The predictive value of these contracts are suspect, given that they are thinly traded, but they do give some indication if where bets are being placed.

To put the price changes into a more accesible format, below I've plotted the year over year change in Boston MSA prices starting in 1988, along with the CME futures contracts:

This plot makes clear the previous bear market starting in 1989, followed by a slow recovery, eventual multi-year bull, and the current bear market beginning in 2006.

Zooming in on the end of the last bull run, one can see the growing downtrend in prices (trending at -5% yoy for the most recent available data), and the CME futures for the next five quarters. Clearly, little immediate recovery is anticipated in the futures market, which are predicting continued deprectiation in SFH prices at around -5% yoy through next winter.

An interesting comparison can be made by examining the YOY % Change in the CS HPI for Boston MSA starting in the last bull year of each cycle (1988 and 2005). Here the black bars report % price changes in the 1988-1991 window, and the yellow bars % price changes so far since 2005, with the CME futures also included.




One striking point that comes out of this side by side comparison is how much sharper the onset of the current bear market in SFH prices has been relative to the last cycle (note that the comparison for condos would likely result in the opposite conclusion). Thus far the CME futures are not predicting anything like the severe price drops observed over the fall/winter of 1990-91. For reference, the light blue shading represents the early 90's recession which coincided with the depths of the bear real estate market. Given all the recent talk about recessions, it is interesting to ponder how the current cycle would fare in a similar scenario. It is also interesting to note that during the last bear market recovery, the yearly change in the Boston MSA CS Index did not exceed 5% until 1997, 5 years beyond the end of this plot. A long and slow recovery indeed.

Thursday, March 01, 2007

January Market Wrap

The Warren Group results for January (reported as yoy change):

SFH sales up 5.1%
SFH prices down 3.4%
Condo sales up 2.7%
Condo prices flat


MAR numbers were much stronger for January:
SFH sales up 12.6% yoy
SFH prices down 2.4%
Condo sales up 5.3%
Condo prices down 0.7%

Here are MAR reported median prices for SFHs and Condos relative to recent history:

In the January press release MAR no longer reports inventory broken down by SFH and condo, so I will omit those plots. The increase in sales, and decline in inventory from last January to this January led to a major decline in Months Inventory from 14.1 to 10.7 months:
Overall, these numbers reverse or halt many of the trends that have been consistent since the summer of 2005 (e.g. declining sales, increasing inventory). However, median prices continue to decline. The coming months will certainly be interesting. My sense from reading the MAR press release is that February numbers will be much weaker, likely due to the turn in the weather in January, and the surge in mortage rates that began in December (which should affect closings with a 2-3 month lag).