Charting The Fed's Monetization Vs Treasury Issuance Mismatch
The long-end of the curve has recently suffered a major whammy: first, contrary to expectations, the FRBNY announced that it would do just 6% of the total buybacks in the 10-17/17-30 Year area; and second, the overall size of the announced monthly monetizations ($75 billion excluding MBS reinvestments) ended up being far less than the $100 billion hoped for (again excluding reinvestments), although the market has conveniently so far ignored this major surprise. Yet what is interesting is that despite the presumed disappointment focusing on the long-end of the curve, the chart below from Morgan Stanley shows just why the long-end is about to experience a fresh surge in buying interest, undoing all the positive work the Fed has done on behalf of the banks by steepening the curve, and leading to yet more flattening. Another observation is that, as we have said many time before, the Fed's existing plan leaves just under $250 billion in the form of a demand gap that has to be closed by foreign central banks - and once QE 2.1 or greater is announced, America's will become completely independent of foreign monetary retaliation: even if foreigners go on strike, and like US stock investors, refuse to touch the market, the Fed will still monetize every single cent of deficit spend funding. In one brilliant stroke, Bernanke made America completely impervious to international retaliation.
Some narrative from Jim Caron on the above chart:
- Size of $850-900 billion Treasury debt through end-2Q ($110 billion per month, 8-month horizon)
- QE2 @ November FOMC: $600 billion through end-2Q ($75 billion per month, 8-month horizon)
- SOMA reinvestment @ August FOMC: $250-300 billion through end-2Q ($30-35 billion per month, indefinite horizon)
- SOMA holdings per security will be allowed to rise above the 35% threshold but only in modest increments
- Fed purchases on track to absorb 100% of the monthly Treasury issuance (F2011 total of $1.15 trillion)
- Largest distribution mismatch between Treasury issuance and Fed purchases is in the 7-10y sector
The two jarring observations from the chart top right, is that the Fed will monetize 50% more in the 7-10 Y bucket than the Treasury will issue. Expect for the right side of the belly to be the next to see its yield plunge. Additionally, even with only 7% of purchases in the 10-30Y bucket, the Fed will still gobble up about 80% of the total issuance. And with the 30 Y the only real source of yield left, we expect fund flush with case to begin front-running the next Fed monetization event by buying up precisely this sector, which will be the one most focused on by the Fed should it really lose control over LT inflation expectations as determined by 30 Y yield indications and forward swap spreads (contrary to TIPS curves).
Indicatively, here is how the various Fed purchases have looked like in the past by bucket. We certainly expect the Fed to revert to its historic stance of
acquiring more 30 Years than 10 Years as soon as June of next year. Which means let the frontrunning begin.
- 2009 duration of purchases: 5.24 years
- 2010 YTD duration of purchases: 5.33 years
- 2010/11 duration of purchases: 5-6 years (Fed has some room to extend out the curve from the current purchasing pace – mainly via the 7-10y sector which the Fed now projects to receive 23% of the purchases (vs. 11% and 4% in 2009 and 2010, respectively)