Monetary Policy Is Now An All-Too-Real Depressant
By June 2003, the average 30-year fixed mortgage rate had fallen to a new low of about 5.25%. The pace of housing construction had risen to a rate of about 1.6 million new units annually, leading to overall employment in the construction industry of about 6.7 million workers. As the overall economy finally seemed to break free of the dot-com bust and transition to a more traditional-type recovery, monetary policy was adjusted in an effort to reduce inflation expectations before they got out of control. Over the next three years the mortgage rate retraced upward to about 6.8% by mid-2006.
Just as mortgage rates were at their lowest point, the Fed funds target, the overnight bank reserve interest rate the Federal Reserve uses as its primary monetary policy tool, was reduced to 1%, remaining at that historic low for another full year. The recession technically ended in 2001, but the Fed continued on its "stimulative" drive of increasingly "cheaper" interest rates in the banking system in an effort to cajole bank lending to make up the calculated shortfall in "aggregate demand". Based on their statistical models, the Federal Reserve observed what they call a positive output gap - a condition where real economic output is below some measure of aggregate potential, therefore requiring remediation in the form of financial manipulation.
In bringing the Fed funds rate down to the then-record low of 1%, the Fed was counting on the monetary transmission of Fed-created bank reserves to flow into the real economy to make up that calculated shortfall below potential. Targeting a 1% interest rate, the FOMC policy essentially committed the US central bank to create as much bank reserves as needed to meet any and all demand in the overnight US dollar wholesale market. Historically, the monetary transmission made its way from the Fed to wholesale money markets to mortgages, appearing as construction activity and consumer spending in the real economy. For adherents to aggregate demand, this debt-driven activity is welcomed since any economic activity regardless of origin or intent is believed to be fully interchangeable.
The US Fed funds market, however, was not the only wholesale money market in existence. By the early 2000's, eurodollars had become a parallel source of interbank funding. Though it largely existed as a separate concept due to national boundaries and therefore regulatory regimes (which were nearly non-existent in eurodollars, which is why that market was so attractive), the eurodollar market intertwines with the Fed funds market through regulatory acceptance. Since eurodollars are really just phantom dollars, or figments of the balance sheet accounting of global banks and their largely London-based subsidiaries, they have no actual "official" relationship with the Federal Reserve, or even the dollar as it is commonly understood. However, because the transfer of "dollars" between and amongst US or global banks operating in the US and US or global banks operating in the Wild West of eurodollar London is fully negotiable and acceptable as dollars to regulators, they are indistinguishable on the asset side.
Furthermore, there is a very close relationship between the published "cost" of eurodollars and the "cost" of Fed funds dollars. LIBOR and the Fed funds rate are highly correlated largely because these wholesale "dollar" markets are explicitly linked by various global bank subsidiaries operating in these different jurisdictions. It is an acceptable practice to move all of these forms of dollars between balance sheets in the various locations and markets with the keystroke of a computer. This means that, for all practical and legal purposes, eurodollars and Fed funds dollars were and are largely interchangeable substitutes, thus disregarding origin or primal source.
This was fairly common knowledge at the Fed since it had published several academic papers acknowledging as much. The eurodollar market even had a somewhat indirect official place in the US money supply since it was given a place in the "broad money" M3 series.
When the Fed held the Fed funds rate at 1% in June 2003, the LIBOR cost of funding for eurodollar arrangements went with it. Further, since LIBOR is the basis for the multi-hundred trillion "dollar" derivatives market, including the largest segment of that market, interest rate swaps, the historically low Fed funds rate had ancillary, spillover effects into every credit market trading dollars around the globe - including the shadow and synthetic markets for dollars.
The manipulative policies that were meant to forcefully shock the US economy out of the dot-com "jobless recovery" had turned into a full-blown asset bubble as financial firms had an ostensibly unlimited supply of "dollars" at historically low costs. Speculative finance through these new innovative means was given a hefty artificial boost by that interest rate target policy, and because the collateral impact on the real economy at the time conformed to theory it was even encouraged (Greenspan's comments about expanding floating rate mortgages, for example).
The Fed finally reversed its course in June 2004 and began to raise the Fed funds rate (at its preferred "measured" pace) lest "inflation" become a problem for the emerging recovery. Over the next two years, the FOMC would raise the Fed funds target by 4%, finally resting at 5.25% by June 2006. At nearly the same time, between 2003 and 2006, mortgage rates rose by a little over 1.5%. Mortgage spreads therefore compressed dramatically, indicating that the Fed's transmission of tighter financial conditions were not reaching the outer ends of the yield/risk curve. In reality, it was a very visible indication that there remained an oversupply of "money" in the mortgage market and further by that time the transmission mechanism was already largely inoperable (the transmission of monetary policy is supposed to work in both directions).
On March 26, 2006, the Fed discontinued its M3 tracking series, explaining:
"M3 does not appear to convey any additional information about economic activity that is not already embodied in M2 and has not played a role in the monetary policy process for many years. Consequently, the Board judged that the costs of collecting the underlying data and publishing M3 outweigh the benefits."
Given all that has transpired since 2006, perhaps the Federal Reserve should have spent the money on monitoring what has been the epicenter of crisis since August 2007. The economists at the Fed may have judged eurodollars and repo markets as unimportant to monetary policy but history's judgment will be far less kind. Instead of receding in importance, funding arrangements had simply become as bespoke as some of the most complicated structured finance products (the notorious SIV's, CDO^squared, and synthetic structures). Not only were funding arrangements trending toward customized one-off agreements, they largely followed the rest of the marginal credit system off-exchange and off-balance sheet. It's not hard to imagine why outdated theories anchored in irrelevant data series of money and credit would miss or ignore the importance of this banking "evolution".
The IMF has retroactively tried to track exactly how many "dollars" are caught up in the vortex of eurodollar markets, intertwined into the balance sheet magic of modern finance, but even they can only estimate within a broad range since most of this information has to be disentangled from the footnotes of individual bank reporting statements. This is called the dollar overhang, or shortage, but it really amounts to an entire class of semi-official, but utterly complex, "money". The range of the dollar overhang, according to the IMF study, is between $1.5 trillion to $6.5 trillion. In reality, like the Fed's retreat from M3, nobody really knows for sure.
The complication of that complication goes well beyond the academic exercise of tracking and tabulating. Again, if the Fed's primary concern in 2004, after several years of historically low interest rates, was heading off inflation it really needed to have a good idea of not only how much "money" was in existence, but even what actually defined "money" in addition to why and how that new money was being used (and abused). A dollar basis swap is every bit as "real" as a physical dollar, but for financial firms operating in these global wholesale markets there is a world of difference. The "cost" of a basis swap is not always the same as the "cost" of a physical dollar, reserve dollar on "deposit" at the Fed, or overnight eurodollar acquired via repo.
This added layer (actually multiple layers) of complexity meant that the simple inflation calculation made and contained within the inappropriately constructed mathematical models that enumerated the output gap and expected/desired aggregate demand was essentially blind. We know this very clearly by what actually happened in the real economy; the mortgage/real estate bubble got worse despite a 400 basis point rise in Fed funds over two years. The transmission mechanism was broken years before the first appearance of ZIRP and all the unconventional monetary excursions.
As late as the middle of 2006, the US was building more than 2.1 million new housing units annually, up an astounding 30% from the 2003 inflection in mortgage rates. The construction industry added another million jobs over that same period. The personal savings rate, at the time, was estimated to have been negative (it has since been adjusted and re-modeled to about 1%, still a historic low) as the scale of mortgage-based money flowing through the household sector was enormous. Between 2003 and 2007, US households added a staggering $4.5 trillion in mortgage debt.
The reverse trend, the market's natural correction to the Fed's mistakes, is the economic equivalent of a natural catastrophe. But there isn't anything natural about it; in other words the catastrophe is not the housing collapse nor the Great Recession. The catastrophe was the hubris that monetary policy could use blunt tools such as mortgages and real estate construction to manage "aggregate demand" without any regard to collateral effects on individual economic agents or even the distortion and perversion taking place in the financial economy. And then willfully ignoring those distortions as "not appearing to convey any additional information about economic activity".
In every meaningful sense, the market's correction to that inflationary episode has actually been the most logical and appropriate - the mass of vacant structures in bubble regions stands as testament to the lack of true demand for real estate apart from that oversupply of "money". The market is merely correcting the artificial imbalance of both the supply of debt money and the misdirection of real economic resources in the absence of a true tightening mechanism for money creation before 2007.
That correction has meant a collapse in the real economic output, particularly in the sectors most closely related to monetary artificiality, i.e., the transmission mechanism. The pace of new unit construction permits hit a "recovery" high in July 2012 of 812,000 (annual rate), which was roundly celebrated as some sort of milestone. Though that rate was only about one-third of the pace of 2006, it did represent a 29.5% increase over July. Even the pace of permits expanded for the depressed single-family home segment, rising 23% year-over-year. Despite that appearance of a potential housing/real estate recovery in permits and starts, however, overall completions of construction projects are only up 7% over July 2011 due solely to multi-family structures. Completions of single-family units, in contrast, are 6% below July 2011.
Some see that as a resurgence of real estate and housing, but it really is a market that refuses to be used as a monetary tool again - the real economy is rejecting these attempts at monetary intrusion. Matching the completions data, the level of construction jobs has remained largely unchanged over the past two years. Between the mid-2006 peak of 7.7 million jobs and the "bottom" reached in early 2011, the industry lost about 2.3 million jobs (a 29% decline). Since that trough eighteen months ago, the construction industry has replaced only 60,000 of those jobs on net.
It hasn't been for lack of trying on the Federal Reserve's part. The Fed has been consistent in its adherence to theory despite all evidence that points to ineffectiveness (at best). In 2008, mortgage rates remained largely stable between 5.8% and 6.4%. That changed in December 2008 with the ominous introduction of ZIRP - the first of the temporary emergency measures that are now permanent fixtures (another very visible sign of theory gone awry). By January 2009, average mortgage rates had fallen a full percentage point to a new low of about 5%.
After the announcement and anticipation of QE 1 in 2009 (which saw the Fed buy mortgage bonds directly), mortgage rates fell to 4.8%, remaining steady at that level until the summer of 2010. With the announcement and anticipation of QE 2 (which would not purchase mortgage bonds), average mortgage rates fell again to 4.2% by the actual launch of QE 2, before rising back to 4.8% near the end of that monetary program. Throughout the summer of 2011 and into 2012, mortgage rates have again declined, now below 4%.
The entire time mortgage rates were falling, as mentioned above, housing and construction have been totally and completely unresponsive. Interest rates have had no discernible impact on the real economy - as if the real estate market no longer responds to monetary inputs of any size, scale or method. Given that disconnect, there is some obvious cognitive dissonance surrounding this month's QE 3 (or QE 5) announcement and implementation. The preferred transmission of monetary projection into the real economy is not at all favorable to the transmission of monetary projection, yet here we go again.
Some of that disconnect is straightforward - lending standards on mortgages are nowhere near what they were in 2006. The pool of potential borrowers has been diminished with the welcome destruction of subprime. That pool has been further eroded by the collapse in real estate prices since home equity levels have almost uniformly disappeared in total disregard to geographic location. In other words, the housing bubble was largely concentrated in five sunny states, but the housing collapse has not been at all concentrated. There is much less accumulated equity for the Fed to work with as a channel or outlet for monetary largesse. The flow through HELOC's and home equity loans to fuel new spending is much diminished in a world of 812,000 permits vs. the 2006 world of 2.2 million. The ability to use homes as a monetary substitute for closing the "output gap" has been disabled by natural circumstances alone.
On the complex side, the various definitions and costs of "money" are playing a primary role in thwarting these monetary designs. The shortage of interbank "money" (financial collateral) has been arguably the biggest problem for monetary flow. That was something that does not appear to have made its way into monetary thinking as late as the end of QE 2 in the US and LTRO 2 in Europe. Monetary policy has not meaningfully adjusted to the reality that money is not a uniform term in the modern system, and further that cost of money is well beyond posted interest rates (liquidity and regulatory leverage are, perhaps, even more relevant to the cost of money in the financial economy right now).
The real economy's response, as in the real estate market, demonstrates that aggregate demand in those monetary models and aggregate demand such that it might actually exist in the real world economy are far apart both functionally and conceptually. Instead, the scale of what the Fed thinks economic potential on a macro level and the aggregation of what individual economic agents (consumers and businesses) are actually able to achieve without credit-induced asset inflation are separate and mutually exclusive concepts, yet easily understood by common sense. At what new historically low mortgage rate of interest will suddenly induce construction of two million housing units or return two million construction jobs? Apparently 3.8% wasn't low enough, so maybe 3% will do the trick (as an aside, I would wager a hefty amount that mortgage rates of 10% or more would actually create more lending for real estate).
In the big picture of economic history, the primary problem of monetary philosophy in the enlarging era of broken transmissions has been the elevation of temporal correlations into ironclad laws. The cabal of mainstream economists has extrapolated certain ideas of particular time periods to be timeless. They have modeled specific data sets and formed lasting conclusions as if the relationship of macro variables will be the same forever. Low interest rates, at certain points in history, such as the 1990-2006 period, have transmitted policy objectives into the real economy through home equity-financed consumption and real estate construction. That is why most economists believe you can't have a recovery without housing.
The truth is far more simple. We cannot have a manipulated, centrally planned recovery without housing. A market recovery on its own is entirely possible, feasible and even likely, but that would require recognition that after four years of massive monetary intrusion to rebuild the worst of the artificial world of 2006 these economic "laws" are no longer applicable. QE 3 cannot possibly succeed as traditionally defined, but it is not a neutral proposition without a downside. As the real economy slides into re-recession, thus panicking the Fed into appealing one more time to the gods of real estate, monetary policy is no longer a stimulant but an all too real depressant. Paradigm shifts are always messy, but they are more unsettling when those that hold the keys to our currency flat out ignore it, or, perhaps worse, aren't even aware of it.