Real Estate Market Recovery?
By Dr. Alfred J. Gobar
Chairman, Alfred Gobar Associates
As illustrated in Exhibit A, the recent recession was considerably deeper in percentage terms than the one that affected Southern California after June 1990. Sadly, the recovery from the most recent recession is a good deal more anemic than was the previous recovery, suggesting happy times for real estate properties (except for those who know how to profit from adversity) are still in the distant future.
Occasionally people wonder why nonagricultural wage and salary employment, as illustrated in Exhibit A, figures in so many of the graphs we use. The reason for that is shown in the graph in Exhibit B. The little triangles are estimates of the number of occupied units in the United States based on a statistical model, the major input to which is nonagricultural wage and salary employment. The little squares in the graph are actual households as reported by the U.S. Bureau of the Census.
For those of you with an unhealthy interest in statistics, the statistical model has estimated the number of occupied units nationwide with an R2 coefficient of correlation of 0.9867. In economics, this type of correlation usually implies the analyst is cheating. As can be seen, the estimate of number of households nationwide based on statistics has been flat since 2005. It is not expected to turn sharply upward in the immediate future—based on the trends in Exhibit A. The number of occupied units probably will rise above the estimate as the data evolves. Anecdotal information indicates that many households are living in homes on which they are no longer making payments on the mortgage—i.e., economic recovery is often less than physical occupancy in times of financial duress.
With more specificity with regard to Southern California (Ventura, Los Angeles, Orange, Riverside, San Bernardino, and San Diego Counties), growth in nonagricultural wage and salary employment in the twelve months ended February 2012 totaled 62,600 jobs. This is well below the long-term average in the six-county area during periods of normal housing markets—approximately 150,000 new jobs a year. Actually, employment growth in nonagricultural wage and salary employment over the most recent twelve months was considerably less than the comparable figure for the twelve-month period ended February 2011. Southern California’s economy (and therefore its real estate market) is not recovering rapidly. In fact, these statistics indicate the opposite.
Categories of nonagricultural wage and salary employment experiencing continued decline include construction, certain portions of the financial sector, and local government, as well as the information sector as newspapers become technologically less efficient than they were prior to the Information Age. Somewhat surprisingly, manufacturing employment grew during the twelve-month period contrary to the long-term trend over the last 15 years of a secular decline in manufacturing employment throughout the United States and with special regard to Southern California.
In 1969, analysts at Alfred Gobar Associates noticed the relationship between nonagricultural wage and salary employment and housing market trends. This led the number crunchers at the consulting company to develop a plethora of algorithms which incorporate generally available time series economic data published by the government and other sources into “models,” which simulate housing market conditions for each Metropolitan Statistical Area in the United States overall. The efficacy of these models with regard to the housing markets in each of these Metropolitan Areas was tested against the Census data for 1970, 1980, 1990, and 2000. Since Dr. Gobar’s retreat from consulting, less attention has been paid to the statistical simulation models, although kindly old Dr. Gobar did fund an analysis of the efficiency of the models in terms of estimating the number of occupied units by Metropolitan Statistical Area as of the date of the 2010 Census.
The output of the most recent update with extensions to Third Quarter 2011 is a basis for illustrating housing market conditions in each of the Metropolitan Statistical Areas that make up the Southern California market in which the Inland Empire is such an important segment.
The conventional output of the models, which was used by such investors as PMI, GE Capital, Nationwide Builders, etc., from the 1970s to the 1990s consisted of a total of 17 pages for each market. In the interest of simplicity, many of the outputs were combined into indices. The simplified index of housing market conditions for the Inland Empire is shown in exhibit C.
Although the statistical simulations of current market conditions overall in the Inland Empire are not quite as bad as they were at the depth of the 1990’s recession, the current trend is still down. The indices for this market have deteriorated fairly consistently since Third Quarter 2006. Housing price in the Inland Empire (as well as Southern California overall) has historically been high relationship to the ideal price structure. Because of declining demand derivative of the recession, even with the decreasing prices in the Inland Empire, the relationship between price and income is still not comfortable.
A similar index for Los Angeles County, which continues to be the largest economic entity in Southern California, is as follows in exhibit D. Currently the index on an overall basis is lower than at any time shown. The index suggests that spillover demand from Los Angeles County is not soon likely to be a major element in housing demand in the High Desert.
The exhibt E index for Ventura County shows overall market conditions about similar to what they were during the worst of the 1990s:
Between Third Quarter 2010 and Third Quarter 2011, Ventura County’s economy improved enough that incremental demand exceeded the incremental supply of new housing based on very feeble building permit activity in prior months in Ventura County in 2009 and 2010. The most recent index point is actually up a little bit from the two previous index points for the Ventura area.
Another of the Southern California economies that has suffered less than the Inland Empire is Orange County, see the graph in exhibit F. The index for Third Quarter 2011 was higher than in Third Quarter 2009 or 2010, indicating a very modest improvement in relative supply and demand conditions in housing in the Orange County Metropolitan Statistical Area.
By far the strongest recovery of the housing market in Southern California is in San Diego County where the index has been improving for about three years, and incremental demand has exceeded incremental supply, see exhibit G
As shown, the low point of the index in recent years is well above the comparable indicator for the 1990’s recession.
This column has referred in the past to economics as the “dismal science.” In a vain hope to counter this probably accurate definition of my life’s work, we have included a comparable graph (Exhibit H) for a market that is currently in much better condition:
This market appears to have been immune to the recession and, in fact, from 2006 to 2011 the index for the market increased year by year, while the price index shows that housing prices in the market are less than they need to be in terms of the consumer support levels driven by the local economy.
Another regional economy which reflects stronger-than-typical housing market conditions is illustrated by the exhibit I graph.
Although there has been some decrease in overall market strength (i.e., demand exceeds supply), it still remains a very viable real estate market in comparison with much of the U.S. Perhaps Joseph W. Brady wants to sponsor a contest for readers of this column to identify these two mystery markets which are apparently doing well despite the abysmal conditions in much of the rest of the U.S. (at least the statistics are a whole lot better).
An interesting anomaly related to the most recent analysis of these statistical data concerns the composition of employment growth. As noted in the early part of this column, nonagricultural wage and salary employment grew less between 2011 and 2012 than it did in the prior twelve months. Another source of employment data, however (based on the household survey), reflects a humungous increase in employment as reported by individual households. Typically, the difference in estimated employment between the household survey and the establishment survey relates to contract workers, small entrepreneurs, underground employment, etc. For long periods of time, the employment levels estimated on the basis of the employer surveys have been about ±87.0 percent of the employment levels estimated on the basis of the household survey. The most recent data available, however, suggests that reported change in nonagricultural wage and salary employment based on the employer survey was 58.0 percent of the level of employment change estimated on the basis of the household survey. Either an awful lot of people are working “off the books” or there is a glitz in the reporting or our interpretation of the numbers. We have been interpreting these numbers for roughly 40 years and have never seen this kind of relationship before. Since it is a one-time event, it could have been a typographical error, a simple transposition, or some other statistical screw-up. Not too much should be inferred from this unusual circumstance. Nonetheless, it is consistent with what we expected to happen with the onset of a highly-regulated society—an increase in informal employment as the market system worked to side-step a stifling bureaucracy. The late Jack Kyser and I discussed the implications of the Obama Administration’s ideology at a meeting a little over two years ago and hypothesized that the informal economy would grow faster than it has in the past in response to overregulation, mitigating to some extent the negative effects on growth of a highly-regulated economy.
If, in fact, the informal job growth was actually as great as the statistics seem to indicate, the overall condition of the market should be better than the graphs shown above suggest.
Let’s hope so.
Readers interested in reviewing Dr. Gobar’s observations about the statistical relationship of real estate market behavior to conventional economic time series data are referred to his book which is available from Alfred Gobar Associates.