Axonic Insights: Don’t Fight the TED

During a Bloomberg roundtable in June 2024, then-citizen Scott Bessent accused then- Treasury Secretary Janet Yellen of having “taken control of monetary policy” through Treasury bill issuance and said the change in issuance composition had “eased financial conditions substantially.” Yellen rejected the accusation that Treasury was managing its borrowing strategy to support the economy ahead of the election. Bessent now appears to be employing a similar strategy.

The difficult thing about analyzing complex multivariate economic systems is isolating variables. Isolating the effect of Bessent’s announcement of the soon-to-be twist on the yield curve is difficult given the offsetting move higher in oil that occurred at the same time. The correct question? Where would yields be without his communication? If we think about this, it is not only a powerful tool signaling the presence of a Treasury ‘put’, but it is also a stealthy form of Fed-backed QE. Why? The Treasury issues bills to buy notes and bonds. The Fed will then buy those bills as part of its reserve management program (“RMP”). While one step removed, the Fed is creating reserves that serve as the ultimate source of funds for the long end purchase.

Treasury’s Recent Actions and Signaling

In recent months, there have been a number of policy signals and actions designed with the implicit or explicit intent to control the long end of the US treasury curve.

• Action. Treasury’s unusual intervention in Japanese currency markets was aimed at preventing Japan from selling US Treasuries to fund Yen purchases in defense of the currency. Our largest treasury buyer owns $1.1T of US debt, and its sale of US Treasuries would be significant enough to move long yields higher.
• Signal. The treasury has announced an upsize to its yield curve ‘twist’ program. While the Treasury Buyback program is almost entirely signal at this point, it is a powerful one… but it may have limitations.

• • Limitation. Bill issuance drains reserve balances because banks move out of reserves held at the Fed and into higher-yielding bills. There are no longer any reserves available in the ON RRP program to absorb outflows from bill issuance, so bill issuance to fund buybacks drains aggregate reserves. In turn, this requires the Fed to continue RMP.
  • Limitation. The underlying cause for higher note and bond yields is fundamental: strong growth, higher-than-target inflation, large AI CapEx demand creation, and flagrant government deficit spending. This combination of fundamental factors makes treasury’s attempts at yield curve control (YCC) difficult.
• Signal. Recent talk of a full SLR exemption of US Treasuries indicates such action is still possible. It would be another meaningful tool to help manage yields. Banks would be more willing to keep Treasuries on balance sheet with an exemption in place.

Last month on August 19^th, the US Treasury doubled the size of its long-end buyback operations, from $2B to $4B each effective September 9^th. There are 4 operations per quarter, so this totals $64B in buybacks per year. In the context of the size of the long-end (10-30s) market’s ~$800B in issuance annually, it is small, but it serves as a powerful signal. Moreover, the marginal buyer sets price. Its direct intervention in the Yen market and its announced intervention directly in the US market appear to signal that Treasury has crossed the Rubicon and will take all measures in its power to contain long-end sovereign yields.

To be clear, the Treasury’s buyback program is a ‘sterilized’ intervention. Unlike the Fed, Treasury cannot create reserves (a special form of money that only depository institutions can hold) to pay for what it buys. Rather, it must fund every dollar of bond purchases by first selling a dollar of bills to the same private sector that just sold it the bonds. That is a type of Operation Twist run by the fiscal agent (Treasury) rather than the monetary one (the Fed), at a fraction of Twist’s size. For context, the 2011 Fed Twist was $400B, extended in 2012 by an additional $270B.

A subsequent follow-up announcement on August 24^th, indicating that the Treasury may use the Treasury General Account (“TGA”) to fund buybacks, had some impact with 10-year yields, moving from 4.73% to 4.70% on the news. This indicates that the Treasury may be willing to spend down a portion of its TGA to fund upsized buybacks, which simply means this iteration of buybacks would not require bill issuance, but only because Treasury believes it had some excess cash/cushion in its savings account. The TGA is the Treasury’s main Fed checking account by which it makes payments to the private sector and saves money collected via taxes.

(Please see the Appendix for a timeline of recent interventions and additional discussion.)

A Dangerous Precedent

The Treasury and Fed are supposed to remain independent, but what’s happening now may risk that separation and look more like consolidated balance sheet quantitative easing (“QE”). Only under exigent circumstances has Section 13(3) of the FRA been invoked to break down that barrier. Here’s what’s happening now.

1. Treasury buys long bonds funded by its own bill issuance. Privately held duration falls (i.e. – 10-year notes are replaced by bills).
   1. 1. When Treasury issues a bill, the buyer’s payment settles into the TGA at the Fed. (The TGA is not a reserve balance at the Fed.) A dollar sitting in TGA is a dollar not available to the banking system. Therefore, bill issuance drains system reserves.
2. The Fed creates new reserves to buy bills under RMP. Since December the FOMC has directed the Desk to buy bills to keep reserves ample. While not technically QE, when paired with Treasury’s bill issuance and note buybacks it results in duration transformation that is QE.
3. Outcome: The private sector has now swapped a 10 or 30-year bond for an overnight reserve balance, and the government has retired long-duration debt and replaced it with overnight money. That is QE, split across two agencies, each of which can truthfully deny doing it. While neither leg is QE, the composite result is.

Bill supply drains reserves, drained reserves trigger RMPs, and RMPs create new reserves. Will the Fed increase the size of the RMP as it plays with twist? We believe it will. We note that the Fed made no mention of ending the program during its most recent meeting.

Conclusion

While none of Treasury’s actions appear extraordinary, we may be witnessing the slide down a proverbial slippery slope towards a consolidated balance sheet. As investors, we must look past our visceral reaction and recognize that there will be a bid for duration that may not be fundamental but that is just as powerful. Ultimately, will Stan Druckenmiller be right? Will the experiment fail? Perhaps, but for now we are not prepared to fight the TED (Fed + Treasury) and will look to extend duration given what we perceive to be asymmetry to long yields in a sad world where fundamentals won’t matter – until they do. We will admit that the size of the RMP and Treasury twist is not yet large enough to correct for disaster were it to occur, but the direction of travel is clear. Treasury (with the Fed in waiting) is ready to act.

¹ The September 2026 FOMC rate hike was essential to preserving Fed credibility. On its own, the 25bps hike may not have been enough to curtail inflation fears, but a hawkish move in the 2026 dots for another hike this year – coupled with eight members looking for another hike next year – reinforced this new Fed’s commitment to fighting inflation in a ‘timelier’ fashion. This should help control an inflation-driven move in the long end.