The story goes that quantitative easing was instituted to “bridge” the financial and economic systems across a “rough patch”. Via QE, the Federal Reserve conjured new “money” into existence (if we did this, its called forgery…not QE) & exchanged this new “money” to the largest banks for US Treasury bonds and financial assets (Mortgage backed securities). Banks were left flush with cash, the Fed holding $4.5 trillion in Treasury’s and MBS. The suggested goal of this exercise was that QE would lower interest rates and this would settle the economic system and subsequently boost asset valuations.
How’d it work out? As far as asset prices, generally they have soared 300% to 400% but the part about lowering interest rates…not so much. The chart below shows the Federal Reserve holdings of over 5yr to 10yr US Treasury debt versus the yield on the 10yr Treasury. The Fed increased its holdings of 5 to 10yr US debt from $100 billion in early 2009 to nearly $900 billion by early 2013…and then lowered it back to just $290 billion currently. And the correlation on the 10 year yields…essentially zero.
#1 – January 2009
Fed held $97 B 5+ to 10 year US Treasury debt vs. 10 year debt yielding 2.36%
#2 – January ’09 to March ’10
Fed adds $120 billion, yields rise 160 basis points…wrong
#3 – March ’10 to June ’12
Fed adds $500 billion, yields fall 250 basis points…right
#4 – June ’12 to May ’13
Fed adds $160 billion, yields rise 23 basis points…wrong
#5 – May ’13 to October ’13
Fed ceases adding but maintains peak holdings of $890 billion, yields rise 122 basis points…wrong
#6 – October ’13 to July ’16
Fed rolls off $440 billion, yields fall 161 basis points to a modern day record low of 1.37%. This is so contrary to the rationale offered by the Fed to initiate QE…wrong
#7 – July ’16 to May ’18
Fed rolls off $160 billion, yields rise 174 basis points…right
Which is to say, there seems to have been no correlation or even a negative correlation of the Fed conjuring nearly $4 trillion with which to buy (and sell) unprecedented quantities of Treasury’s (particularly the 5 to 10 year variety), and the direction of interest rates.
In fact, the lowest rates seen on the 10 year were while the Fed was systematically rolling off $450 billion in 5 to 10 year debt in mid 2016! Which leaves a very uncomfortable question...is the Fed filled with the dumbest smart people in America who just haven’t seemed to realize this…or did they have a different goal all along?
QE did have massive impacts, just not where it was “supposed to”. If the Fed’s goal to was transfer power and wealth to an ever shrinking cohort, seems the Fed has achieved its goal. My best guess of what QE was all about and who benefitted and continues to benefit…HERE and HERE.
As long as we are discussing the Fed, might as well take a look at the balance sheet. Last week the Fed reduced its total balance sheet by $20 billion and the total reduction is about 2.8% since official “normalization” began. But the real story continues to be the how. Fed Treasury holdings fell by $8 billion last week in a convoluted process of rolling off $31 billion of 1yr to 10yr Treasury’s while buying $10 billion in 10+ year debt and $12 billion in less than 1 year short term holdings.
As the chart below details, the Fed holdings of over 5yr to 10yr Treasury holdings continues declining, now down a massive 68% from peak holdings.
While the holdings of over 10 year Treasury’s is down less than 6% from peak holdings and essentially unchanged since normalization began.
The change in the short term holdings is fascinating, the Fed fully round tripping, “normalizing” the short end to what it held pre-GFC (as of ’07). Since Operation Twist ended, the Fed has been on a tear purchasing over $400 billion of short term debt. The massive short end bid coupled with the huge roll-off of 10yr debt should have been pushing the short end rates considerably higher and middle down…perhaps pushing toward an inverted yield curve?!?
The trend of fast shrinking 1 to 10 year Fed holdings, significant buying and growth in holdings of less than 1 year, and maintaining nearly all long bonds.
A quick aside, showing the Fed Treasury holdings versus the 10yr minus 2yr spread.
Finally, just to round out the picture, last week the Fed’s holdings of MBS was essentially unchanged while the Fed has sold just 2% of its MBS since peak holdings.
But interestingly, while the Fed has rolled off $60 billion in over 10yr MBS, it has simultaneously purchased $30 billion in 5 to 10yr holdings…for a net total decrease of $30 billion since peak holdings.
Make of all that what you will.