It appears the Fed's balance sheet "trial balloons" using primary dealers as intermediaries have begun.
After yesterday's unexpectedly explicit guidance on the future of the Fed's balance sheet, which prompted Goldman, Citi, and various other banks to suggest they may bring forward their estimates for when the Fed will announce the start of "renormalization", moments ago JPM's Michael Feroli, traditionally the analysts "closest" to the Fed, did just that when he issued a report stating that that there is now "chance of a September start" to renormalization, with the values for monthly roll-off caps and phase-in period to be "revealed at the June FOMC meeting."
According to Feroli, JPM continues to look for normalization to commence at the December FOMC meeting but "there is some chance of a September start, though this would not have a material difference for our projections on a multi-year horizon. At the meeting at which normalization starts we expect the Committee to announce a set of monthly roll-off caps for the following year, which increase regularly every three months.
"Our best guess is that the initial caps are $4 billion a month for MBS and $8 billion a month for US Treasuries. In the preannounced schedule, these caps would be augmented each quarter by $4 billion and $8 billion, respectively, until at the end of the year they are $16 billion and $32 billion. Consistent with yesterday’s minutes, even after the normalization process is fully phased in the monthly caps will still be in place, though in most months after the full phase-in they would cease to bind."
And here are the finer details which the Fed may or may not have leaked to select banks, in an attempt to prepare for what is coming, and talking down the equity bubble:
The Committee has yet to communicate values for the monthly caps or the length of the phase-in period. Presumably they will do this in the minutes to the June FOMC meeting. It is less clear that the Committee will have decided on a monetary policy implementation framework by the time roll-off begins (i.e. the current ratesetting system vs reverting to the pre-2008 system) and hence whether they will have decided on an ultimate amount of excess reserves available when the balance sheet is fully normalized. We have assumed a $500 billion target for excess reserves in our projections below. The other key assumption on the liability side of the Fed’s balance sheet is currency growth, which we have penciled in at 4% per year.
Under these assumptions the balance sheet is fully normalized in late 2021 at a level close to $3.0 trillion, down from about $4.5 trillion now. After normalization the Fed would turn to become a net buyer of Treasuries, at a pace of around $400 billion per year, partly to meet growing demand for currency and partly to replace MBS which will continue to roll off the balance sheet. During the normalization process we see the funds rate as the tool of first resort for adjusting policy to both economic strength and weakness, unless the funds rate returns to the effective lower bound around zero, at which point roll-offs would stop. This latter eventuality is a risk for normalization being completed later than we anticipate.
Mechanistic questions about the impact of this a Quantitative Tightening on asset prices aside (it will be very bearish as the market will realize soon), we remind reader what former Fed governor Kevin Warsh said about the Fed's normalization "policy" several weeks ago.
I am confused by the Fed’s ‘normalization’ strategy in monetary policy. Its preferred sequencing of rate increases and balance sheet reductions differ markedly from what was agreed when we conceived QE in the ’war room’ amid the crisis. There might be good reason. But, the transmission mechanisms of rate changes and balance sheet adjustments are markedly different than projected. So too are the distributional effects. This merits a more robust public explanation.
Alas, a public explanation will not be provided.