March 2005

Don't Ask, Don't Tell

CP (Pie In The Sky) I...As you know, the debate over inflation rages on.  At this point, it seems to be a matter of perspective.  It depends on where you're sitting, as can be seen in the compare and contrast comments above.  At least based on the ISM and ECRI commentary, if you happen to be in the trenches of actual daily business activity, inflationary pressures have been and sure appear to continue to be a real concern.  Alternatively, if you're sitting in the wood lined offices of the marbled floored Federal Reserve, inflation's not much of an issue at all apparently.  All right, let's get to the real meat of this discussion.  As always, no ranting and raving, we promise.  We want to spend a few minutes looking at the actual calculation of the CPI by the Bureau of Labor Statistics (the BLS).  We suggest it is illuminating, to say the very least.  As you know, a lot of folks tend to look at the headline number, as they do a lot of stats, and draw immediate conclusions.  But, as always, it's what's underneath the headlines that really counts.

Before getting started, let's take a very brief look at the inflation in household residential real estate values as per the Fed's own Flow of Funds reported numbers.
        

There is no question that the monetization of this inflation in real estate values you see above has been holding up domestic consumption as the rate of change in personal spending has been outstripping the rate of change in personal income growth in the US like clockwork for the last three years (the economic recovery period).  What follows is how this same inflation in residential real estate prices depicted above that is holding up the US consumer is influencing, or not influencing as the case may be, the reported inflation numbers in this country.  Specifically the CPI.    

First things first.  The following are the component "weights" the BLS uses to calculate the reported headline CPI each month.
      

As you can see, at 42+%, housing is THE single largest component in the reported CPI calculation.  It's wildly meaningful to the headline number, to say nothing of general perceptions regarding domestic inflation.  A few little tidbits before pushing forward.  In total, energy specifically only accounts for 7% of the overall CPI, with food and beverage prices clocking in at just about 15% of the complete weight of the index.  So, energy and food combined are just a little over one-half of the total weight that housing accounts for singularly in the CPI calculation.  Again, to suggest that the housing component of the CPI is meaningful is simply a gross understatement.  OK, let's dig a little bit deeper.  The housing component of the CPI is further broken down as follows:

As you can see, we're basically cutting to the chase here in terms of breaking down the housing component of the CPI in terms of individual input weights as a percentage of the total CPI calculation.  The single largest component input in the CPI is Owners Equivalent Rent (OER) at 23.4% of the total reported headline CPI.  Now, the BLS tells us that it polls on a monthly basis 50,000 landlords and tenants in coming up with the OER number.  OK, here come the facts of the moment.  In 2004, the year over year rate of change in Owners Equivalent Rent increased all of 2.2%.  And you know by now that the year over year rate of change in the headline CPI  number came in at 3.3%.  BUT, you also know that as of year end 2004, the actual homeownership rate in the US stood at a record 69.2%.  In other words, just less than 31% of the US population was actually renting as opposed to owning residential property.  So, when the CPI calculation is being made, 23.4% of the entire CPI number is picking up the "housing cost inflation" experience of only 31% of the total US population.  The housing cost inflation experience of the other 69% is largely being ignored.  C'mon, does this make sense? 

Alright, time for a little journey down memory lane.  In the following chart, we are graphing the prior three decade history of the reported change in the year over year CPI-Shelter number and the OFHEO (Office of Federal Housing Enterprise Oversight - the FNM and FRE regulator) year over year home price rate of change data.  A few quick notes of explanation.  The owners equivalent rent number is 71%+ of the reported CPI-Shelter component, so the Shelter numbers are a pretty darn good approximation of the influence of owners equivalent rent on the total CPI headline number.  The most recent year over year rate of change data that is as of 3Q 2004 came in at 13% for the US as a whole.  A far cry from owners equivalent rent at up 2.2% and total CPI-Shelter up 3% in 2004.  Moreover, as you might know, year over year home price acceleration in the US coastal areas was up between 15-20%.  The West Coast specifically was up near 23% in home price changes year over year.  The following is the longer term history of the change in OFHEO home price data alongside the very important and influential CPI-Shelter numbers.
         

NEVER before have we seen this type of rate of change differential between the rate of change in the OFHEO home price data and the rate of change seen in the housing costs implicit in the CPI-Shelter data.  NEVER.  Is it fair to say that NEVER has the headline CPI been less reflective of real US residential housing price inflation?  Or alternatively, is it fair to say that NEVER has the CPI been so understated relative to the actual accelerating cost of US residential real estate?  In all sincerity, we believe the answer to both questions is a resounding yes.

Some very quick and simplistic math.  Given that the year over year rate of change in CPI-Shelter was 3% last year, and that the CPI-Shelter component makes up about 33% of the total CPI, the shelter component added about 1% in absolute terms to the total 3.3% CPI headline number for 2004.  If we used the 13% OFHEO housing price increase number as opposed to the 3% CPI-Shelter price increase number to adjust the total CPI calculation, the absolute percentage contribution of CPI-Shelter to total CPI would have been about 4.3%.  Cutting right to the bottom line, as opposed to a 3.3% CPI rate for 2004, using the OFHEO data as a substitute for the CPI shelter number would have caused headline CPI rate of change to jump over 6% in 2004.  Are we saying that 6+% is really the true inflation rate in the US?  We won't go quite that far, although it's sure tempting.  After all, 31% of the folks in the US are renting.  We prefer to characterize the situation as one where we believe it is absolutely fair to say that the reported CPI of the moment is meaningfully understating the true rate of consumer inflation in the US if one wishes to capture the true inflation in residential real estate values.  And what we believe backs up this comment entirely is the fact that every single number used in the analysis or calculations above came directly from a government entity or agency.  Every single one.

One last comment on how the CPI is interpreted and often used in mainstream analysis.  As you know, we often see analysts use the "core" rate of CPI as the real indicator of consumer price pressures at any point in time.  The core rate, of course, being the headline number stripped of the influence of food and energy costs.  As we stated above, food and energy components of the CPI account for roughly 22% of the total weight of the index.  If we assume that the owners equivalent rent component is also being understated quite meaningfully in the current environment in terms of the true inflation in residential housing costs, we need to remember that this component again represents a substantial 23%+ of the total headline CPI weight.  So, although this might sound like a stretch, when we're looking at the supposed "core rate" of inflation, we're really stripping out food, energy, and by academic definition of being understated meaningfully, housing.  In other words, the current "core rate" quotes are really excluding the true costs of food, energy and housing.  That being the case, how can we really consider this a "consumer"  price index when it is essentially excluding the true nature of the three largest and most important consumables, so to speak, in any consumers life?  Although this is really a point in time comment more than anything else, the current CPI is telling us very little about real world cost pressures at the consumer level.  And, as you know, we haven't even ventured down the path of exploring the hedonic price adjustments likewise influencing reported CPI data to the downside.

Don't Ask, Don't Tell...Let's get to the matters at hand that affect us as investors each day.  Also some comments about what may lie ahead.  First, do you really believe that the Fed and the Administration are unaware of the extremely simple data and calculations presented above?  Do you really think they are that out to lunch?  Do you really think they are complete idiots?  Of course you don't.  After all, everything above just happens to be their data.  Although we're not wild fans of Fed policy over the last decade+, we indeed do give them credit for being able to carry out simple addition, subtraction, multiplication and division.  That we're pretty darn sure of.  The Fed knows that residential real estate prices are inflating meaningfully and rents are not.  They know the CPI at the moment is not indicative of true inflationary pressures when it comes to consumer prices in the here and now.  But they and the Administration have a vested interest in appearing ignorant.

As you know, the headline CPI number is the key ingredient in annual Social Security COLA's (cost of living adjustments).  Moreover, the CPI is the principal adjustment factor in annual TIP (Treasury Inflation Protected securities) total rate of return payouts (some protection, right?).  Military and civil service wage COLA's depend on the CPI.  And so do levels of Federal Income tax tables in terms of potential "bracket creep".  (A lower CPI does not allow absolute dollar thresholds of incremental tax brackets to "creep" up much at all, academically meaning a higher total absolute dollar level of taxes collected.)  The bottom line is that the higher the CPI rises, the more the absolute dollar cost to the government annually.  Let's face it, we have a massive Federal deficit already (both on and off balance sheet).  Does the Fed or the government really want to see that move meaningfully higher due to a CPI calculation that perhaps would be a bit shocking if indeed it reflected reality?  And finally, a substantially higher headline CPI could, in all sincerity, push the Fed into raising short term interest rates much faster than has been the case up to this point due solely to financial market perceptions.  The Fed knows the current US economy is highly levered.  They know a rapid Fed Funds rate acceleration would absolutely cause some real damage and perhaps immediately burst a number of financial and real world bubbles they themselves fostered in the first place.  From our standpoint, they and the Administration (responsible for the CPI calculation) are playing a game of "perceptual chicken" with the markets when it comes to reported inflation.  They are practicing a "don't ask, don't tell policy".  Dangerous?  Yes.  But what is their alternative at this point given their failure in terms of lack of financial discipline many moons ago?

This brings us to what we believe is one last important issue.  And an extremely important issue it is.  What about the trillions and trillions of dollars sitting apparently quite complacently in the US bond markets at the moment?  Are the majority of current US bond holders complete idiots?  Can they not do the simple calculations presented above?  Are they willing to accept implicit negative real rates of return across the entire Treasury curve of the moment based on the reality of CPI numbers adjusted to real world facts as opposed to the headline reports?  Or are they also adhering to the financial mantra of "don't ask, don't tell"?  Although we're trying to factually break out the real world data and are literally attempting to quantitatively "prove" that the current CPI numbers are not reflective of current economic reality, there have been plenty of voices on the Street suggesting the same for some time now.  Bond "superstar" Bill Gross has been singing this song for so long that he's being dismissed by many as a "broken record".  All of the Blue Chip Economists in each Business Week survey of the last few years have been predicting higher long term interest rates.  Are inflation paranoid bond vigilante's simply asleep at the wheel?  Or were they shot dead in the bond market version of the OK Corral some time back?

We won't belabor the point, but this little simplistic CPI adjustment calculation discussion again simply reinforces in our minds that the current day bond market is not about anticipating or discounting real world inflation at all in terms of setting fixed income prices.  Or anticipating deflation either, for that matter.  The US bond market of the moment is all about playing interest rate spreads.  It's all about levered speculation.  Believe us, the Fed has much bigger problems than an understated CPI at the moment.  Much bigger in terms of total financial market leverage of the moment.  Until the leveraged speculating and interest rate spread driven carry trade game comes to an end, we sure as heck would not be looking to the bond market for clues as to the true nature of inflationary pressures, or lack thereof, in the real economy.  Levered bond players playing the spread game could care less about inflation.  Theirs is a world of basis point spreads, yield differentials, mathematical algorithms and interest rate protection derivatives overlays.  In other words, this ain't your father's bond market anymore.  Get it?

Again, we suggest being aware of the dynamics at work and the reality of the macro economic numbers at present is our best investment offense in looking ahead.  At some point, the markets will reflect economic reality.  For now, unprecedented systemic liquidity creation and resulting leverage has skewed the connection between the fixed income markets and the underlying reality of inflationary pressures in the real economy.  Hopefully, if we can already "see" it, we can act accordingly when change ultimately descends upon the markets and the economy (as it always has in the past and will again down the road).  But in the meantime, you know the game implicitly being played by the mainstream - don't ask, don't tell.