From Wolf Street, February 22:
QT-1 blew up the repo market. This time around, the Fed wants to avoid that type of debacle so it wouldn’t have to “prematurely” end QT-2.
This has been in the works for a while and has come out in bits and pieces, but today the Fed made it official in the FOMC meeting minutes: It doesn’t know how far it can take QT without blowing up stuff, but it wants to reduce its balance sheet as far as possible without blowing up stuff. QT has already reduced the Fed’s balance sheet by $1.34 trillion. And so far, nothing has blown up. But last time it did QT, the repo market blew up.
The issue with withdrawing liquidity from the market via Quantitative Tightening is that this liquidity is taken out through a couple of drains that are close together – the Fed’s roll-off of Treasury securities and MBS – but liquidity has to flow there from all directions, and it may drain out faster in one corner, and that corner then runs out of liquidity and blows up, while there is still excess liquidity in other corners.
Yields solve that problem normally. The corner that is running out of liquidity will be willing to pay higher yields to attract liquidity from the corner that has excess, but the process is not instant; it can take too much time, and then something runs out of liquidity and blows up while at the other side of the markets, there is too much liquidity.
This hasn’t happened yet in QT-2, which is still running “smoothly,” as the FOMC minutes said today.
But it happened with QT-1 in September 2019, when the repo market blew out because banks were running low on excess liquidity and refused to lend to the repo market, and a repo-market panic ensued and threatened with contagion, and so the Fed stepped into the repo market and helter-skelter doused it with nearly $400 billion in liquidity, which raised its balance sheet again, thereby undoing a big part of QT-1. And that is to be avoided this time.
So the Fed’s solution seems to be two-fold, that’s what we see:
1. Withdrawing the liquidity slowly enough so it has time to flow to where it’s needed with minimal disruption, allowing QT to drain liquidity evenly until the balance sheet reaches the lowest comfortable level without blowing anything up. If something blows up, it would put a premature end to QT.
This topic was first broached in early January by Dallas Fed president Lorie Logan and dealt with at the time here in our illustrious comments. Slowing QT would reduce “the likelihood that we’d have to stop prematurely,” she’s said. And today, it was officially put on the table via the FOMC minutes.
2. A Standing Repo Facility that everyone knows will be there; and its mere presence, even if inactive, will tamp down on a panic. The Fed had an SRF through 2008, that was its classic measure to deal with market issues before QE, including during 9/11 when markets froze and were shut down.
Repos mature within a day or within a week or two weeks, or some relatively short term, and if they’re not rolled over, they then vanish from the balance sheet automatically. They don’t cling to the balance sheet for years or decades like QE assets.
But in 2009, the Fed skuttled its SRF because it wasn’t needed with the huge amount of QE the Fed was doing at the time. But it didn’t revive it when QT started, and in September 2019, after nearly two years of QT, with no SRF on standby, it had to scramble and improvise as the repo market was already blowing up.
So this time around, in July 2021, a year before QT started, the Fed revived its classic SRF in preparation for QT. So that’s done.
The minutes make the beginnings of a plan official.....
....MUCH MORE
Also at Wolf Street, party like it's 1989:
