Back in October, when the repo market started cracking under the weight of the Fed’s gradual reduction in Reserves and Reverses, and funding spreads briefly exploded, we quoted BofA STIR expert – and former NY Fed repo guru – Mark Cabana, who said that according to his estimates, the Lowest Comfortable Level of Reserves (or LCLOR) is around $3-3.25 trillion given “(1) bank willingness to compete for large time deposits and (2) reserve / GDP metrics (back in 2019, the repo market locked up once reserves dropped to about 7% of GDP not too far from where they are now).”
We bring this up because in the latest weekly Fed balance sheet update, we find that the amount of Fed reserve balances as of Wednesday, Jan 1, tumbled by a whopping $326 billion – the second biggest drop on record – pushing the total from the comfortable level of $3.218 trillion to $2.892 trillion, the lowest since November 2020.
Normally, this liquidity-draining plunge would have been sufficient for funding spreads blowing out and stocks being flushed into a liquidation frenzy, however in this particular case there was a footnote: the fact that it took place at year, means that there was a mitigating factor, namely the flood of reserves went largely into reverse repos, which as we noted previously, soared by a near-record $213BN to $474BN, the highest since June, after previously dropping below $100BN.
Reverse repo soars by a near-record $213BN to $474BN, the highest since June, on year-end window dressing and funding shortage. Expect this to drop to $80BN by the weekend. pic.twitter.com/NhLLvIaIdq
— zerohedge (@zerohedge) December 31, 2024
This, as frequent readers know by now, is largely for window dressing purposes with banks scrambling to pad their books for regulatory purposes at quarter and year-ends, in the process draining substantial liquidity out of the system… then just days later reverting to normal, and sure enough on Wednesday we already saw a huge drain in the reverse repo balances, which dropped by 50%, or $233BN in one day, to $240BN, an amount which we are confident will slide aggressively, and back below $100BN, in the coming days.
That also explains the market’s sanguine take on the recent plunge in reserves: it is quite confident the numbers will increase in the coming days.
There is another reason why what would otherwise be a concerning drop in systemic liquidity has yet to manifest itself in any funding spreads blowing out, and for that we have to thank Democrats for refusing to kick the can on the debt ceiling, which as discussed previously, will hit within days and restart the familiar 4-6 months liquidity flood that traditionally precedes the next debt ceiling crisis.
As Goldman’s William Marshall writes in a Thursday note (available to pro subs), the start of 2025 brings the end of the debt limit suspension period. And while the treasury likely won’t have to dip into its extraordinary measures until the middle of January, as Janet Yellen noted recently, from that point it will generally have to operate under the constraints of cash on hand plus available extraordinary measures until there is a resolution. At that point the countdown to the next debt ceiling crisis begins, and the deadline for the next D-Day, or debt limit action, is likely not until July or August 2025.
Some more details from the Goldman report:
We estimate Treasury will start out with slightly more than $1tn in headroom, reflecting the sum of the Treasury’s cash balance plus extraordinary measures available up front. As Treasury Secretary Yellen noted in a letter to Congress, it likely won’t be until the middle of January that Treasury actually has to start dipping into its extraordinary measures thanks to redemptions of nonmarketable securities on January 2. In addition to the capacity available to begin with, we assume some incremental headroom will come available each month alongside a more sizeable boost at mid-year.
There is uncertainty as to how exactly Treasury will manage its available levers—i.e. the pace at which it draws down the TGA versus depleting extraordinary measures via higher net marketable borrowing. Still, the overall effect will be less bill supply and higher levels of broad liquidity (thanks to a lower TGA) in the system than the counterfactual of no constraint. Exhibit 1 illustrates the average change in T-bills outstanding and cash balances normalized to past debt-limit resolutions (using a sample since 2011).
On average bill supply falls by about $130bn and the TGA drops roughly $225bn in the 6 months into an agreement. The subsequent rebuild tends to be somewhat faster—on average bills outstanding and the TGA both reverse that decline within the month or two following a resolution.
As noted above, the TGA will be starting off at a considerably higher level compared to when Treasury has reached the debt ceiling in the past (highs of approximately $450BN in 2021 and 2023). Goldman expects this year’s TGA drawdown to be comparable to the 2021 and 2023 experiences when cash balances fell roughly $425bn in the window between reaching the debt ceiling and resolution.
That said, a process that drags into Q3 could see TGA fall significantly further still as July and August tend to be seasonally large deficit months, and thus GS estimates meaningful bill paydowns in the $400 to $600bn range over the first two quarters in this scenario, on par with the paydown observed in 2021
Paradoxically, and as we explained in early 2021, the period preceding a debt ceiling D-Day (which will takes place some time in Q3) tends to be very beneficial for risk assets, not because markets fail to discount a potential looming political crisis, but because overall liquidity levels soar.
Indeed, as Goldman confirms, “debt ceiling limitations mean that until there is a deal there will be less bill supply and higher levels of broad liquidity than would have been the case otherwise.” Less Bill supply means the Treasury is forced to draw down on its Treasury General Account (i.e. cash balances) as it can’t roll Bills, and meanwhile the liquidity from maturing Bills is allocated to other risk assets lifting them in the process despite what is clearly another looming crisis.
To wit, the bank writes that “balance sheet runoff means, however, that a given TGA drawdown is likely to translate to a smaller build in overall liquidity in the system (measured as reserves plus RRP) — we expect this would rise roughly $150-250bn through mid-year under our QT baseline assuming no resolution ahead of then.“
Ironically, as with the scope for a larger TGA drawdown, a more protracted process that spills into the second half of the year could see a sharper rise in overall liquidity that exceeds what was seen in 2023 when injections from the Bank Term Funding Program also helped offset the impact of QT. Conversely, a later tapering and/or end to QT than baseline expectations of an H1 end, would dampen any potential rise in liquidity although should Trump pursue an activist Fed to boost overall liquidity, we don’t expect this happening. Ultimately, absent any debt limit constraint, bill supply would be roughly flat alongside a roughly $350bn draining of overall liquidity from the system in the first half of the year.
In other words, the good news is that with just days to go until the countdown to the next debt ceiling fiasco begins, the Treasury cash balance is historically high compared to past instances when the debt ceiling has been reached as discussed above, and If there is no resolution in the first half of the year, the TGA drawdown would more than fully offset the draining of liquidity via QT alongside meaningfully negative bill supply.
Not surprisingly in light of this imminent liquidity flood, Goldman thinks that “the higher starting level for the TGA and magnitude of any potential bill paydown should at least support a more benign first half of the year in dollar funding conditions than would have been the case otherwise, dampening the tightening bias in swap spreads that have prevailed in prior debt ceiling episodes.”
Goldman’s rule of thumb is that a $100bn rise in liquidity from current levels is worth 0.5bp to SOFR-FF, while a $100bn drop in bills outstanding is worth about 0.1bp. That said, the market prices that in to a large degree, with most SOFR-FF tightening backloaded into the second half of the year…
… which leaves vulnerability to a faster resolution of the debt limit (or a later/slower ending of QT). It’s also worth noting that balance sheet capacity constraints (rather than the overall level of liquidity) have played a greater role in driving volatility in funding markets over the last year—these may ease now that year-end is behind us, but not because of the debt limit.
Putting it all together, Trump could not have asked for a better “debt ceiling crisis”, because while Democrats have been hoping to pull the rug from under Trump as soon as he is in the White House (expect the reality of the US jobs market to be unloaded by the deep state apparatchiks at the BLS like a ton of bricks), the accelerated drain of some $750BN in Treasury cash will provide a generous buffer for risk assets to maintain their levitation well into the second half of 2025 much to the delight of the 47th US president.
Ironically, it would be the end of the debt ceiling fiasco that is bearish for markets! That’s because, the Treasury tends to rebuild the TGA fairly quickly upon resolution, suggesting risks of a swift undoing of any tailwinds in funding and spread markets.
That said, the intersection of this episode with the more mature phase of Fed QT may see Treasury proceed somewhat more cautiously in eventually returning the TGA to target, however, particularly with Treasury’s “steady-state” TGA target somewhat higher than it was pre-pandemic.
According to Goldman, the August-September 2019 repo crisis is perhaps the closest analog: QT had ended, and the mid-September jump in the TGA coincided with the surge in funding market volatility that prompted Fed liquidity injections. While the end point would ultimately be the same, a more gradual replenishing of Treasury’s cash balance would reduce the risk of excessive volatility that could arise from quickly withdrawing liquidity from the system.
In other words, stocks – and liquidity sppoppnges like Bitcoin surge – for the duration of the TGA drain, then risk tumbles once the debt ceiling resolution drains $750BN from the market some time in Q3, and then – if the stress from said drain is too much – we get another “Not QE” from the Fed, which triggers the next inflationary shock into late 2025 and onward.
More in the full Goldman note available to pro subs.
Tyler Durden
Fri, 01/03/2025 – 13:25
