By: John Straka
The U.S. housing market remains relatively weak, but it’s probably not as weak as you think. To what extent are home prices really falling again?
Differing Findings
Most recent media coverage of the “double dip in home prices” has centered on declines in the popular Case-Schiller price index; however, the data entering into this index is reported with a lag (the just released April index reflects data for February-April) and with some limitations. CoreLogic publishes a more up-to-date index value that earlier this month showed a small increase, and more importantly, CoreLogic also produces an index that excludes distressed sales. This non-distressed index has shown larger recent price increases, and it shows increases over the last 12 months in 20 states. Others basing their evidence on realtors’ listing data have concluded that there was some double dip last year, but prices have actually been rising now for several months (See Altos). These disparate findings belie overly simplistic media coverage, and they stress that “the housing market” is not one single market, of course, but a wide distribution of differing outcomes in very many local neighborhood home markets across the nation. (For a pointed view of this, see Charron.)
Improved Data Sources
Experian is now working with the leading source of the most granular and timely home market analytics and information, from nationwide local market data, and the best automated valuation model (AVM) provider based on these and other data, Collateral Analytics. (Their AVM leads in accuracy and geographic coverage in most large lender and third party AVM tests). While acknowledging their popularity, value, and progress, Collateral Analytics President Dr. Michael Sklarz questions the traditional dominance of repeat-sales home price indexes (from Case-Shiller etc.).
Repeat-sales data typically includes only around 20 to 30 percent of the total home sales taking place. Collateral Analytics instead studies the full market distribution of home sales and market data and uses their detailed data to construct hedonic price indexes that control for changing home characteristics. This approach provides a similar “constant quality” claim as repeat-sales—without throwing away a high percentage of the market observations. Collateral Analytics indexes also cover over 16,000 zip codes, considerably more than others.
Regular vs. Distressed Property Sales
Nationwide, some well-known problem states, areas and neighborhoods continue to fare worse than most others in today’s environment, and this skewed national distribution of markets is not well described by overall averages. Indeed, on closer inspection, the recent media-touted gloomy picture of home prices that are “falling again” or that “continue to fall” is a distorted view for many local home markets, where prices have been rising a little or even more, or at least remaining flat or stable.
Nationwide or MSA averages that include distressed-property sales (as Case-Shiller tends to do) can be misleading for most markets. The reason for this is that distressed-property sales, while given much prominence in recent years and lowering overall home-price averages, have affected but not dominated most local home markets. The reporting of continued heavy price discounts (twenty percent or significantly more) for distressed sales in most areas is a positive sign of market normality. It typically takes a significantly large buildup of distressed property sales in a local area or neighborhood home market to pull down regular property sale prices to their level.
For normal or regular home valuation, distressed sales are typically discounted due to their “fire sale” nature, “as is” sales, and property neglect or damage. This means that the non-distressed or regular home price trends are most relevant for most homes in most neighborhoods.
Several examples are shown below. As suggested in these price-per-living-area charts, regular (non-distressed) home-sale prices have fared considerably better in the housing downturn than the more widely reported overall indexes that combine regular and distressed sales(1).
Regular-Sale and Combined Home Prices in $ Per Square Foot of Living Area
and Distress Sales as a Pct of Total Sales
In Los Angeles, combined sale prices fell 46 percent peak-to-trough and are now 16 percent above the trough, while regular sale prices fell by considerably less, 33 percent, and are now 3 percent above the trough. Distressed sales as a percent of total sales peaked at 52 percent in 2009:Q1, but then fell to a little under 30 percent by 2010:Q2, where it has largely remained (this improvement occurred before the general “robo-signer” process concerns slowed down industry foreclosures). L.A. home prices per square foot have remained largely stable for the past two years, with some increase in distressed-sale prices in 2009. Market prices in this area most recently have tended to remain essentially flat—weak, but not declining anew, with some upward pressure from investors and bargain hunters (previously helped by tax credits before they expired). Double-Dip: No.
In Washington DC, single-family home prices per square foot have been in a saw- tooth seasonal pattern, with two drops of 15-20% followed by sizable rebounds in spring sales prices. The current combined regular & REO average price is 17 percent below its peak but 13 percent above its trough, while the regular-sale average price is just 12 percent below the peak and 10 percent above its trough. Distressed sales have been comparatively low, but rising slowly to a peak of a little over 20 percent in 2010, with some slight improvement recently to the high teens. Single-family prices in DC have remained comparatively strong; however, more of the homes in DC are actually condos, and condo prices have not been quite as strong, with the market data showing mixed signals but with the average price per square foot remaining essentially flat. Double-Dip: No.
In the Miami area, the combined average home price per square foot fell by 48 percent peak to trough and is now just 1 percent above the 2009:Q2 trough. The regular-sale average price already experienced an earlier double-dip, falling by 32 percent to 2009:Q2, then stabilizing for a couple of quarters before falling another 9 percent relative to the peak; since 2010:Q3 this average has been choppy but basically flat, now 3 percent above that second trough. Prices in Miami have been among the weakest in large metro areas, but average prices have been largely flat for the past year, without any sharp new double dip. Distressed sales as a percent of the total peaked at 53 percent in 2009:Q1, but then fell to a little under 30 percent by 2010:Q2; since then there has been some return to a higher distress share, in the mid to upper 30s (but all of these figures are about 10 percentage points lower for condos). New Double-Dip: No.
The Dallas area has seen some of the strongest prices in the nation. The combined price per square foot had an earlier peak and fell by 31 percent peak to trough, but it is now 33 percent above the trough. The regular-sale average price fell briefly by 22 percent peak to trough, but it has since risen by 32 percent from the 2009:Q1 trough to where it is now 3 percent above the peak. The increases have occurred in a saw-tooth seasonal pattern with spring prices the highest, but prices here have been largely rising considerably. Distress sales as a percent of the total peaked at 22 percent in 2009:Q1 but have largely fallen since and now stand at just 11 percent. Double-Dip: No.
Here You Can See 47 More Examples of Where Double-Dips Are and Are Not:
» Midwest
To summarize this information and gain a little more insight into the general area conditions for most homes and individuals in the U.S., we can add up the number of homes and the total population across the counties examined. To be sure, this information is not a rigorous random sample across homes, but I have tried to include and show the details of both stronger and weaker metro-area counties throughout the U.S. As shown in the tables below, the information used here has covered 51 metro-area counties, including a total population of over 15 million homes and nearly 75 million individuals(2). These results may be regarded as suggestive of findings from a more thoroughgoing study.
Based on these reviews of the market price averages and other data, my assessment is that a little over half of the counties examined are not currently or recently experiencing a double-dip in home prices. Moreover, these counties, where home prices appear to be at least flat or relatively stronger, encompass almost two-thirds (65%) of the total affected U.S. population examined, and nearly three-fifths (58%) of the total properties covered by the data studied.
Conclusion
This is, on balance, good news. But there are remaining concerns. One is the continued high, or more recently rising, shares of distressed sales in many markets, and the “shadow inventory” of distressed sales now being held up in the current foreclosure pipeline. But it is also interesting to see that many of the reductions in the distressed-property shares of total sales in high-stress areas occurred before the foreclosure processing slowdowns. Another interesting observation is that most of the recent double-dips in prices have been relatively mild compared to the previous original peak-to-trough meltdown.
While, to be sure, there are plenty of reasons to remain uncertain and cautious about U.S. home prices, home markets in general do vary considerably, with significant elements of improvement and strength as well as the continuing weaknesses. Despite many reports today about “the beleaguered housing market,” there really is no such thing … not unless the report is referring to a very specific local market.
There definitely are double dips in many areas, and reasons for continuing overall concern. But the best available evidence suggests that there are actually double-dip markets—most relatively moderately so, stable markets, and stronger markets, with markets affecting a majority of homes and individuals actually in the stable and stronger categories.
Note: In a next installment, we’ll look at some more granular micro market data, to explore in greater depth the extensive variety of home-price outcomes and market conditions in weak pockets and strong pockets across various local areas and home markets. This will highlight the importance of having very good information, at sub-county and even sub-zip code levels, on local-neighborhood home markets.
Source of Home Price and Market Information: Collateral Analytics HomePriceTrends. I thank Michael Sklarz for providing the extensive information for this report and for comments, and I thank Stacy Schulman for assistance in this posting.
__________________
(1) Based on analysis by Collateral Analytics, price/living sq ft is a useful, simple “hedonic” measure which typically controls for around 70 percent or more of the changing characteristics in a housing stock and home sale mix. Patterns in home prices without dividing by the square footage are generally similar, but not always.
(2) The property inventory counts are from Collateral Analytics, while the population estimates are from the 2010 U.S. Census.