Wednesday, July 13, 2016

IntraGovernmental Holdings: What Are They, Where They Come From, & Is IG The New QE?

Some readers have highlighted the IG (IntraGovernmental) Reserve numbers in the Treasury's recent reports...that IG holdings have inexplicably surged in the past year since QE had ceased.  Simply put, IG and QE were both forms of Treasury demand (one non-marketable and one marketable, respectively).  But both reduced the bonds available to the open market and theoretically pushed yields lower than otherwise would have occurred.  And as IG began to wane in '09, QE was implemented.  In QE off years, IG again surged...strange timing?  The timing, source, and implications of IG holdings seem worthy of a little more exploration.

The short story...
The majority of all the growth in the IntraGovernmental fund since Jan 1, 2014 has come from the "other" category...the category used for all those funds so small as not to be named individually.

Since Jan 1, 2014...
  • total IG surplus + $730 b
    • Social Security surplus + $115 b
    • Highway Trust Fund +$74 b
    • DIF +$30b
    • Federal Employee Retirement Fund +$130 b
    • "Other" surplus has grown + $422 B
That is to say that almost 60% of the growth in IG since Jan 1, 2014 has come from invisible, tiny fund surplus' skyrocketing.  These funds represented just 15% of IG as of Jan 1, 2014.  The "other" category has almost doubled in that same time.  But this has reduced marketable Treasury issuance by $422 b...just as QE was winding down.  Coincidence?  This maintained nearly a $25 b/mo "QE" like bid since QE ended in late 2014.

The long story...
To gauge federal government spending, some look at the federal deficit but I prefer to watch the annual Treasury issuance.  To get the latest view, I'm showing July 1 to June 30 annual issuance vs. 10yr Treasury yield (below). 

For the twelve month calendar year just finished, the Treasury issued $1.23 trillion in new debt...more than double the previous twelve months issuance of $520 billion.  Despite this large jump in new supply, yields fell even absent the Fed's QE and nearly without a single net new purchase by "foreigners" (outlined HERE).
The chart below shows annual growth in the two sources of Treasury buying, Public vs. IG.  From 1990 through 2008, IG took down 53% of all Treasury issuance in GAS (Government Account Series, also known as non-marketable debt).  This left only 47% to be issued or auctioned off as marketable debt to the Public (Public=Federal Reserve, Foreigners, & Domestic buyers).  However, this changed radically from 2009 onward, when IG decreased, Treasury issuance increased, and IG took down just 10% of issuance through 2015.  This meant the Public did nearly all the Treasury buying led by Foreigners, the Fed's QE, but very little assistance from the Domestic Public (US based pensions, banks, insurers, private citizens).  But 2016 turned around as IG took down 30% of the resurgent Treasury issuance and the Domestic buyers the remaining 70%.  How and why US based insurers, banks, and pensions alike came up with $850 billion of cash over the past twelve months to buy record low yielding Treasury debt...a mystery.  How they did this with markets at record highs apparently without selling other assets to raise this cash and instead also buying stocks...let's just be honest...this is the stuff of banana republics.

And the chart below shows the interplay of the annual growth in IG vs. the annual growth in the Fed's balance sheet (QE).  As IG faded, QE was stepped up to fill the void...and when QE faded, IG again filled year more so than 2016 when the Fed's balance sheet receded by <$13> billion and IG hit it's $373 billion all time peak growth.

What is the IntraGovernmental Reserve?
The IG Reserve is made up of a myriad of Trust funds including Social Security (aka, OASDI comprising 52.5%), Federal Employees Retirement Fund (14%), Fed Hospital Insurance Trust Fund (3.5%), Highway Trust Fund (1.4%), Exchange Stabilization Fund (0.4%), etc. etc according to the most recent June 2016 Treasury Bulletin.  These funds are mandated to place their excess reserves into special GAS non-marketable Treasury's.  Of interest, the recent growth in IG has come not from any of the recognizable sources but entirely from the "other" category with a jump of $315 billion from October to November of '15.  How the "other" category grew by 30% virtually overnight is a mystery.  But no mystery that the automatic bid for $373 billion in GAS (government account series) meant an additional $373 billion in new marketable debt were not necessary to be issued.  Likewise, the "other" category seems to have supplied much of the 2012 jump in IG and most of the 2014 jump in IG?!?

What is clear is the largest funds historically supplying the vast majority of IG are facing decelerating growth or outright declines.  Below, the blue columns represent the annual excess of Social Security (OASDI) revenue over annual expenditures...and red line the total Social Security excess reserves held (again, 52.5% of total IG reserves).  Annual SS reserve growth peaked in '07 at +$190 billion/year and has now fallen to approx. +$45 billion/year...a 75% reduction in less than a decade and due to demographics, this will only worsen from here (outlined HERE).  Regarding all those "reserves", they are already spent and replaced with paper promises.  In order to be paid out, these "reserves" will need to be raised in some combination of new taxation or greater Treasury issuance.
Again for context, from 1990-->2009 Social Security (OASDI) surplus funds provided 64% of the IG reserves...but from 2010-->2016 SS is only supplying 30% of the IG Reserve growth.  Where the record IG growth is coming from and whether we should expect this continued "support" from fairly miniscule source(s) is a mystery buried deep in the "other" column?
In November, yields sat at about 2.35% but as the $373 billion of IG (nearly equivalent in $'s to Operation Twist) was worked in over the subsequent 3 months of auctions (decreasing the public issuance), yields fell to approx. 1.7% by early February.  Now with the 10yr to record lows of 1.35%, we'll have to wait to see the IG data for April through June (when it becomes available) to determine if IG (whatever it's source...?helicopters?) has been substituted for QE, partially or even perhaps totally replacing QE?