April 2004

Home Is The Heart (Of The Matter)

Home Is The Heart (Of The Matter)...The Fed Flow of Funds report for the period ended 4Q 2003 hit the Street a number of  weeks back.  You might remember that the 3Q report was delayed considerably into the early part of this year.  We always consider the Flow of Funds report to be nothing short of a treasure trove of data for anyone having the time and endurance to plow through approximately 140+ pages of nothing but numbers.  As per the recent Flow of Funds report, to suggest that 2003 was a year of significant credit driven reflation is simply an understatement.  Here are some very broad overview numbers to help put current circumstances into perspective.

Just as a very quick note, the doubling in nominal GDP since 1988 is not a typo.  Although most forms of systemic leverage have doubled in well under 10 years, it has taken nominal US GDP over 15 years to double.  The table above is clear in its message, households and the financial sector were primarily responsible for credit expansion and broader systemic liquidity creation in 2003.  On the asset side of the equation, the success of reflation in 2003 was simply more than evident as per the Flow of Funds report.  We included wages and salaries as well as payroll employment numbers for reference in the following table.

We ask you quite importantly and quite sincerely, just where would our economy be today had it not been for household real estate inflation over the last four years?  You probably do not want to know the answer.  All of the data in the tables above point to the same conclusion.  The leveraging of continuously inflating real estate values over the last four years has been absolutely crucial to the economy.  Whether Greenspan will ever admit it or not, this is exactly how the Fed "has successfully dealt with the aftermath of the stock market bubble".  All of these numbers come directly from the Fed.  The very same folks who simply cannot be ignorant of their meaning.  The problem, of course, being just what will or can the Fed attempt to inflate next if the residential real estate market runs into price turbulence at such a presently high altitude?  Especially given that the labor market appears a bit under the weather and wage and salary growth is not even keeping pace with lowball estimates of understated headline inflation.  As of the end of 4Q 2003, household real estate assets totaled approximately $15.1 trillion.  Household financial assets climbed to $34.3 trillion during the same period, but of this equities account for only about $8 trillion.  The bottom line is that residential real estate is almost twice as meaningful to households in terms of total household assets as are stock holdings specifically.  Although a number of stock market followers may be concerned about an equity crash or severe downturn ahead, maybe what they should really be worried about is US residential real estate values.  In terms of total household well being, emotional stability and forward perceptions, there is no other singular household asset class that means as much in dollar terms.  Not even pension entitlements.  Get the picture?
        

And, as you know, this is how we have treated this precious asset over the last half century.  We've levered it like there's simply no tomorrow.
      

If Tomorrow Never Comes...Although equity as a percentage of the (inflated) market value of residential real estate in 4Q was up very modestly from 2Q and 3Q of last year, on a year over year basis, we ended 2003 with the lowest percentage of collective residential real estate equity relative to the total market value of household real estate in US history.  Moreover, although it is very clear that nominal real estate prices continued their skyward ascent in favored areas such as the bicoastal regions of the country during last year, we find a statistic published by the Federal Housing Finance Board quite interesting and worthy of both current note and tracking ahead.  These folks publish the history of purchase prices of all homes financed with conventional single family mortgages.  This includes both new and existing homes.  What you see in the chart below is both the history of aggregate home prices and the accompanying monthly year over year rate of change.
      

It just so happens that January of this year (the latest data) was comping against a very strong like January period of 2003, hence the year over year rate of change is modestly negative.  Nonetheless, the recent trajectory of moving price rate of change is almost straight down from an annual rate of change cycle peak in January of last year.  Hard to believe to be honest.  Headline numbers regarding housing prices sure don't seem to support this declining rate of change data at all.  As you can probably barely make out above, historical experience in the current annualized rate of change area in which we find ourselves at present has been accompanied by relatively flat home prices in forward periods.  This type of occurrence was both obvious and extended in duration during the late 1980's and early 1990's.  In the early 1980's experience, prices grew at a much less accelerated rate than in the late 1970's when, once again, the zero annual rate of change environment occurred.  Are we potentially just at the entry point of a period of price flattening in very broad real estate prices as evidenced by this data?  Again, we do not see it in the rather sensational bicoastal price quotations, but we need to remember that cyclical real estate price weakness in the late 1980's did not begin in the bicoastal areas, but rather ended there.

In juxtaposition to what you see above, the most widely followed home price data in the US comes from the OFHEO (Office of Federal Housing Enterprise Oversight).  What you see below is the annualized quarter over quarter change in their HPI (Housing Price Index).
         

Essentially breathtaking quarterly change in 4Q 2003.  But just who are we to believe when it comes to the macro trend in housing prices, the OFHEO or the Federal Housing Finance Board data?  First, the OFHEO data, despite being the most widely watched, is a bit quirky.  During periods of extraordinary refi activity, it could very well be that the OFHEO data is understating the rise in housing prices as many refi's are being done with no new appraisal.  Mortgage providers are simply resetting terms for a nice fee for current clients.  Subsequently when refi's die down, the OFHEO data is simply playing catch up ball.  That's exactly what we think happened in the 4Q data you see above.  Another further explanation in the apparent discrepancy between these two housing price data sets lies in the relationship between fixed and adjustable rate mortgage lending.  Adjustable rate mortgages as a percentage of total mortgages outstanding spiked over the last year without a coincidental spike in interest rates in general.  As you can see in the following chart, historically the use of adjustable rate mortgages usually accelerates when interest rates rise, as would seem rational.  But recently, those taking on new mortgages are going adjustable at close to the lowest fixed mortgage rates of our lifetimes.  Clearly, this phenomenon is happening for one of two reasons.  Either home buyers are absolutely convinced interest rates are going even lower, or a good portion of current buyers are no longer able to qualify for higher cost fixed mortgages, despite their incredibly low absolute rate levels.  Does this hint at the last marginal buyer influencing current prices?
          

Could it be that the current popularity of lower cost adjustable rate mortgage debt is skewing the OFHEO housing price data to the high side, given that the Federal Finance Housing Board data does not include prices of homes financed with adjustable mortgages?  Could be to a point.  In one sense, it would certainly be a tribute to what in large part we would characterize as the "minimum monthly payment economy" of the moment.  After all, tomorrow may never come, right?

Is Everybody In Yet?...For now, the home ownership rate in the US sits at an all time high as of the end of 2003.  What we believe is important to note is that as the home ownership rate flattened between 1988 and the mid-1990's, we also experienced coincidental flattening of home prices as per the Federal Housing Finance Board data in the chart above.  Seems simply common sense.  Perhaps one clue as to a potential near term peaking in residential real estate prices will be a distinct flattening in the aggregate US home ownership rate somewhere ahead.  In looking at the chart below, the rate of annual change in the US home ownership rate has been slowing over the past few years relative to the rocket launch in rate of change during the mid-1990's.  But to declare even a short term peak quite yet is still a bit premature.
      

What we also find interesting and perhaps very telling about just where we are in the present residential real estate cycle is the historical complexion of real estate buyers as categorized by age demographics.  The following table documents the long term homeownership rate in this country by age classification:
        

Although it's not wildly surprising by any means, it's absolutely crystal clear that the younger folks predominantly drove the macro US homeownership rate higher in the late 1990's and early into this decade.  It's also clear that homeownership rates among those under 45 are more cyclical than is the case with other age groups.  And this is the same age classification of younger folks driving current home ownership rates higher whose current trajectory of unemployment demographics look a whole lot different than the direction of the headline aggregate unemployment rate of the moment.
      

As you know, what you see above is quite a contrast to the official headline unemployment rate trends of the last year or so.  Quite a contrast indeed.  So it seems pretty obvious that the drivers of the increase in the home ownership rate since the mid-1990's, and more so since year end 1999, are those clearly bearing a good portion of the brunt of current labor market weakness over the past four years.  Does this suggest that housing prices have elevated based on a rock solid wage driven financial footing, especially in the case of new younger buyers over the last 5+ years?  You probably didn't need us to drag you through all of this primary data.  It's simple.  Financial leverage is driving housing demand among those least able to afford it and among those quite susceptible to current labor market weakness as is displayed above.  When it comes to the current housing cycle, that's the heart of the matter, shall we say.

The Flow Down...From the Fed's Flow of Funds data, we find the following graphical data quite revealing.  What you see below is the quarter over quarter growth rate in household real estate holdings based on the Fed's version of market value.  Notice anything funny?
        

Quarterly growth in household real estate values at market per the Fed's data in the fourth quarter of last year has no parallel over the last three years at least.  A period of significant housing price inflation and some of the lowest mortgage rates on record.  Of course the 4Q 2003 experience comes literally one quarter after the for now bottom in mortgage finance rates.  In conjunction with this price burst, total household net worth climbed to a new high as 2003 came to an end.  In the following chart we document the historical components of household net worth.  Once again, it's clear as to what has driven household net worth over the last four years as financial assets are not yet back to their prior annual peak set in late 1999.
     

Although 2003 was one heck of a snap back year for equities in this country, household holdings of equities grew $1.1 trillion in 2003 while the market value of household real estate apparently jumped $1.4 trillion.  Again, when we get right down to the bottom line, which is more meaningful to households, real estate or common stocks?  If you ask us, any so-called wealth effect is being primarily driven by ascending real estate values.  Real estate values very heavily dependent on the existing US credit bubble.  And dependent on its continued expansion.

Again, although timing is the ever present uncertainty, it would be very easy for us to conclude based on the factual data that the current US residential housing cycle has been pushed to extraordinary lengths based on credit bubble dynamics.  In fact the final straw may very well be the recent spike in adjustable rate financing without any corresponding spike in fixed rate costs.  In like manner, even a leveling off or minor price retracement in this asset class would have very significant consequences for US consumer spending and broader systemic liquidity creation.  It's certainly no secret that the Fed is clearly in the "inflate or die" mode at the present.  And the key asset in terms of household participation in the "inflate or die" campaign is residential real estate.  Up to this point, the Fed has been completely unsuccessful in "reflating" payroll employment growth as well as wage and salary growth.  Hence, residential real estate inflation has rested almost entirely on the health and continued expansion of the credit bubble, as well as mania thinking on the part of buyers.  We suggest that unless meaningful job and wage growth is clearly evident directly ahead, residential real estate is skating on very thin ice.  Ice that could easily be broken by continued payroll weakness accompanied by the coming conclusion of direct consumer stimulus, to say nothing of potential upward movement in interest rates.

Maybe rather than suggesting that US residential real estate prices are a credit dependent a time bomb with an already lit fuse, let us briefly have a look at one last chart.  Here's the experience of year over year change in new dwelling construction in Japan from 1988 through to year end 2003.  Obviously it encompasses the post stock market Nikkei bubble peak in late 1989.  Clearly enthusiasm for residential real estate continued, at least for a while, even as the Japanese Nikkei and economy began sinking significantly in the early to mid 1990's.  Of course necessarily accompanying this spurt in Japanese residential construction were large declines in Japanese interest rates post the stock market bubble peak.  As the Nikkei crested in early 1990, the Japanese discount rate was approximately 5.5%.  By the time the residential construction boom ended in 1996, the Japanese discount rate was 0.5%.  It's just a good thing that our present experience in the US is nothing like what occurred in Japan a decade back, right?
          

At The Wire...Just prior to publishing, we received some info we believe indispensable in terms of suggesting where we are in the residential real estate cycle of the moment.  You may remember that our home perch is the San Francisco Bay Area.  In other words, the outer limits when it comes to unbelievable price activity in residential real estate.  We can certainly attest to you that present conditions in the SF Bay Area are nothing short of a frenzy, and we're not being melodramatic in that characterization.  From the office of the State of California that issues real estate licenses, here it comes.  Real estate licenses issued by the State increased 44% during the 2002-2003 period (47,000 new licenses issued).  Relative to the 2000-2001 period?  Licenses issued are up 95%.  That's right, a doubling in less than four years.  Remember, this is not the number of total realtors, but rather the number of new licenses issued.  If this isn't a testimony to feverish mania, then what is?  As you may remember, we saw the same thing with CFA test applicants in the late 1990's about ten seconds prior to the equity peak.  As always, everyone wants a ticket to the promised land.  At least until it turns into a dust bowl, anyway.