In September 2019, the system creaked as the Fed had gone too far with bank reserves reduction. We’re facing a similar challenge in the coming months as the Treasury replenishes its cash buffer. We identify a range from $2.5tr to $3tr within which bank reserves can settle, likely at the upper end. And if there’s an issue, the Fed can always rebuild themHere’s Why US Bank Reserves Today Are So Much Higher Than They Were. And Why They Are Now FallingBefore the Great Financial Crisis (GFC) bank reserves were typically around $0.2tr, as banks only held cash at the Fed for regulatory reasons – the Fed paid zero percent interest on them. The GFC prompted a number of phases of Fed bond buying – quantitative easing (QE) – which meant banks built reserves (other side of the QE trade).To facilitate this, the Fed remunerated banks at a newly introduced rate on reserves (currently 4.4%). Fast forward to the subsequent pandemic-induced QE, and the US banking system has become structurally characterized as a large reserve volume one.At the same time, recent years have allowed the Federal Reserve to slowly deflate the excess of liquidity, predominantly through the reverse repo facility. This facility allows market participants to post excess liquidity at the Federal Reserve. At its peak in 2022/2023, some $2.5tr of overnight liquidity routinely went back to the Fed.Mostly as a consequence of post-pandemic quantitative tightening (QT), these balances are now running at below $50bn. More recently, it’s been given a further push lower, as the lift of the debt ceiling allowed the US Treasury to issue more bills for the purpose of rebuilding its cash buffer.With reverse repo facility balances approaching zero, and with the US Treasury continuing to rebuild its cash buffer, the likely next phase sees this show up in reductions in bank reserves. Bank reserves are currently running at $3.3tr. They actually peaked at over $4tr in 2021 as the Fed’s QE programme saw a crescendo.Since 2022, bank reserves have been broadly steady, albeit within a wide $0.5tr range. Why? Because up till now, most of the unwind of post-pandemic excess liquidity happened through reductions in balances going back to the Fed on the reverse repo facility.Ahead Bank Reserves Can Lurch Lower as the Treasury Rebuilds Its Cash Buffer at the FedGoing forward, bank reserves will be far more responsive to liquidity variations. The policy push here is the c.$20bn that continues to roll off the Fed’s balance sheet through what is now quite a tame QT-lite.But the biggest direct liquidity push of late is coming from the US Treasury, as the build in their cash buffer has a counterpart in lower bank reserves. That Treasury cash buffer is now at $550bn. It could get to $800bn (a staging area in the past). If it did, it would mean bank reserves falling by $250bn, pulling bank reserves down towards $3tr. Should we worry about this?The last time the Fed actively allowed bank reserves to fall was during the QE unwind (QT) that finally concluded in 2019. In September of that year, the Fed discovered that they had gone too far. Bank reserves had been halved to $1.4tr by then, and at that month end there was a severe disruption as the market struggled to deal with a dearth of liquidity, manifesting in a huge spike in repo rates.The system basically haemorrhaged, partly driven by a relatively moderate corporate tax payment need of some $120bn. So what did the Fed do in response? They went ahead and rebuilt bank reserves through repo, and ultimately by buying T-bills. That’s the recipe; rebuild reserves if seen to be too low. The Fed has likely learnt from this experience.The Big Question Is Whether This Can Become a Problem, and at What Level Does the System Creak (As It Did Before)The follow-on question is what level of bank reserves are comfortable? To help answer this question we can refer back to the 2019 experience, which can at the very least identify a floor where reserves would be too low. Far from perfect, but one thing we can do is make a judgement on reserves based off them a percent of GDP.In September 2019, that percent hit a low of 6.5%. That was forced back up towards 8% as the Fed rebuilt reserves. That type of level should be considered an absolute floor.Calibrating this to today and projecting over the coming six month, that would equate to a bank reserves balance of some $2.5tr (at 8%). Then 9% reserves would be $2.8tr, while 10% would be $3.1trn. Simplistically, if 8% is a floor, then 9% offers some comfort to that floor, while 10% is likely very comfortable. The current level of bank reserves is $3.3tr, so we’re comfortably above the most conservative level of $3.1tr.What if the US Treasury goes ahead of adds a further $250bn to its cash buffer? Basically, this would bring bank reserves to the conservative 10% area. We should be fine there. At the same time, QT is shaving reserves by some $20bn per month (note that the $35bn cap on MBS roll-offs is rarely hit).That can further eat into reserves. That said, it seems to us that even then, circumstances are relatively comfortable. It may in fact also turn out that 9% is a doable level of bank reserves as percent of GDP, equating to $2.8tr of reserves.Identifying the Range $2.5tr (tight) to $3tr (Comfortable)The top chart is bank reserves plus the reverse repo balancesBottom Line, the System Can Handle a Fall in Reserves. Worst Case, the Fed Rebuilds ThemIn the end though, the Fed is likely to be super conservative here. The US is a high reserves banking system, and the Fed will not want to be seen to be leaving the market short. If there was the evolution of a shortage, the Fed would engage in a rebuild of reserves.They could do this in a temporary fashion through the Fed’s standing repo facility (currently broadly unused), or in a more permanent fashion through buying bills. They could also do this through buying bonds, but might prefer not to, as it could be construed as QE.By all means watch this space, as it’s important for the proper functioning of the system. But we would be of the opinion that the Fed is on the case, and will manage to avoid a repeat of September 2019.Disclaimer: This publication has been prepared by ING solely for information purposes irrespective of a particular user’s means, financial situation or investment objectives. The information does not constitute investment recommendation, and nor is it investment, legal or tax advice or an offer or solicitation to purchase or sell any financial instrument. Read moreOriginal Post