Slowing or Stopping Fed QT

Could the Fed slow or stop QT and if so when is the greatest risk?

Figure 1: U.S. Treasury Holdings By The Fed (%)

Source: Continuum Economics/Datastream

Faster Fed Easing and Less QT

Fed chair Powell in the July FOMC press conference indicated that QT could continue even if the Fed were in an easing cycle. However, the context was that Powell was talking about an easing cycle that starts slowly to keep real interest rates from rising, as Yr/Yr CPI inflation slows. This could be at a quarterly pace of 25bps initially, which from the Fed's viewpoint does not cause huge frictions with ongoing QT.

However, QT is contractionary monetary policy, despite some seeing it as operating in the background. The shrinkage in the Fed balance sheet reduces reserves in the banking system and can also temper the willingness of banks to extend credit. QT is less powerful than QE though, as QT does not have a signal effect for interest rates and does not have the associated forward guidance of lower for longer than existed with QE.

At a quick pace of Fed rate cuts or alternative a longer duration of Fed easing, the Fed will likely ask the question of whether QT should be slowed or stopped for monetary policy reasons. This will also be framed we feel in a tension between below trend growth/slowly rising unemployment and inflation coming back to target, which will drive rate cuts and less restrictive policy. The market is also highly uncertain about the Fed reaction function, with 2024 rate cut forecasts varying from 50bps to over 200bps depending on growth and inflation views.

We look for a period of below 1% quarterly GDP growth, which we feel will be enough to get the Fed to cut interest rates, but if growth is only a bit below our forecasts then we would be moving towards stagnation and the Fed would likely become more aggressive in cutting rates. Thus though our baseline forecast would be that QT slows for monetary policy reasons in 2025, a skew exists that the alternative is that this could happen in H2 2024.This could also be a stop to QT rather than say reducing the pace by 50%.

It is also worth watching the 10-2yr yield curve in this context. As a switch occurs towards an easing cycle, 10-2yr normally swings to less disinversion and then a positive yield curve. If this was too quick then it could led the Fed to slow QT, as proper function of the U.S. Treasury market is in some way a monetary policy transmission issue.

Figure 2: Reserves Of Depositary Institutions (USD Blns)

Source: FRED/Continuum Economics

The other alternative issue for the pace and duration of Fed QT is when reserves reach an optimal level. Opinion is divided on this matter. The optimists argue that the Fed has plenty of room to technically continue QT and that a decision to slow or stop is not required until 2026. This is based not just on the reserves of depositary institutions (Figure 2) that remain above 2019 levels, but also that overnight reverse repo balances still remain at $873bln (Figure 3) and provide a further pool of liquidity for the financial system – this has fallen this year with a shift in new T bills. They also point out that liquidity withdrawn by QT at the current $960bln annual pace does not translate into an equivalent drop in reserves, as the Fed could be adding liquidity via other methods e.g. bank term funding program in the wake of SVB failure.

Figure 3: Overnight Reverse Repo Balances (USD Blns)

Source: Continuum Economics/Datastream

However, the Fed had to stop QT in 2019 when reserve balances got too low and caused tightness in money market rates. Some have argued that the Fed needs to be careful when reserve of depositary institutions get to around $2500bln (Figure 2) and that should be a watershed to consider slowing through not perhaps stopping QT. This could be reached in 2025. Some economists worry that reserves are unequally spread in the system and any new SVB style failure could quickly tighten the money market. This argument is producing some estimates that Fed QT should slow when reserve of deposit institutions fall below $3000bln, which is likely in H2 2024.

Thus Fed QT slowing for technical reasons is feasible in H2 2024/2025 or 2026. We tend to favor the earlier estimates, as we feel that the Fed can slow QT and see how the situation develops rather than repeating the mistake of 2019.