One month ago, and roughly four weeks before Fed Chair Powell "stunned" markets by announcing the Fed would resume its permanent open market operations, i.e., bond purchases, i.e., permanent balance sheet expansion , i.e., QE, we laid out precisely what was about to happen in "The Fed Will Restart QE In November: This Is How It Will Do It." As we explained at the time, the dramatic recent blow out in the Fed Funds-IOER spread indicated that there was a roughly $400 billion shortage in reserves.
That shortage is what prompted Goldman to project that with repos clearly insufficient to restore liquidity to the short-term funding market, the Fed would launch a roughly $15bn/month rate of permanent OMOs, enough to support trend growth of the balance sheet plus some additional padding over the first two years to increase the size of the balance sheet by $150bn, restoring the reserve buffer and eliminating the current need for temporary OMOs. This strategy would result in balance sheet growth of roughly $180bn/year and net UST purchases by the Fed of roughly $375bn/year over the next couple of years.
In retrospect, the Fed's QE4, pardon "not a QE", plans appear to be far more aggressive when it comes to rebuilding its balance sheet, and as Goldman writes in a follow up note, Powell’s speech and comments "suggest the Fed may pursue a more rapid pace of balance sheet growth than we had initially anticipated."
"Powell’s comments yesterday suggested that the Committee may want to be proactive in addressing this issue between meetings, moving faster than implied at the September press conference" Goldman's analyst write, and note that Powell's announcement that a decision will come “soon” could potentially mean an announcement even before the October meeting (Goldman still expects this to be unveiled in November).
By mentioning “the purchases of Treasury bills we are contemplating” (which likely means both T-Bills and short coupons), Powell implied that the FOMC is willing to use a different composition of purchases to engineer the needed shift upward in the level of reserves. This, in turn, suggests to Goldman "that Fed officials would prefer to complete the adjustment quickly, rather than using the more gradual schedule we had previously assumed."
Here Goldman agrees with what we said previously, namely that while the Fed has been relying on temporary open market operations to inject reserves so far, "permanent open market operations should be a more effective tool over year-end, furthering the case for a faster permanent level shift."
As a result, Goldman now expects the Fed to announce additional monthly purchases of about $60BN for four months, split across Treasury bills and short maturity coupon Treasuries, in order to replenish the roughly $200bn reserve shortfall and support the pace of growth in non-reserve liabilities.
Exhibit 1 summarizes the bank's updated projections for the Fed’s gross Treasury purchases:
We expect the FOMC to announce additional monthly purchases of about $60bn for four months, split across Treasury bills and short maturity coupon Treasuries. In addition to the resumption of trend balance sheet growth (implying UST purchases of approximately $10bn per month to keep pace with the growth in non-reserve liabilities), our base case is that the Fed will engineer a one-off level shift of roughly $200bn over the course of four months (shown by the gray bars).
Such an aggressive expansion will return reserves to a level where they are "truly ample", allowing the Fed to cease relying on temporary open market operations as a regular tool. It is also worth recalling that both the “organic growth” and the “make up” purchases would be in addition to current purchases of USTs which amount to roughly $20bn/month which replace ongoing MBS paydowns, though the former would expand the balance sheet while the latter do not.
And while one can debate until one is blue in the face if the Fed buying $60BN in TSYs each month is QE, is "QE Lite", or is "not a QE" (narrator: it's QE, but the Fed doesn't want to call it that to avoid being seen as again doing Trump's bidding), a critical question is how much/quickly the Fed can purchase without disrupting the market?
According to Goldman, the monthly purchases in T-bills and short coupons required to achieve the level shift, trend balance sheet growth, and MBS reinvestment purchases would likely be around $66bn (with $6bn from MBS reinvestments). While unquestionably meaningful, as long as these purchases are spread across the course of the month, Goldman believes "they should be manageable." Why? Because - as a reference - average daily volume in T-bills has been around $130-140bn of late, while in coupons maturing in the next three years it has been around $110bn. As such, $66bn in purchases over the course of a month would only represent about 1.3% of total volume in the sector. One final point is that as Goldman concludes, "sizable Fed purchases in short coupons would if anything be a source of relief to the dealer community as primary dealer inventories in the sector remain historically elevated even after declining somewhat in the last couple of months."
As for the bigger picture, with the Fed loading up on short-term debt, it will then not only have enough "trial balloon" ammo to launch what even it will admit is a "normal" QE but to also activate Operation Twist in the not too distant future should the curve steepen too much in the coming months as bond traders position themselves for the coming recession.