CNBC’s Steve Linesman on Monday cited two unnamed senior Treasury officials to report that the US Treasury is considering using the money it stored in Treasury General Account (TGA), a checking account it holds at the Federal Reserve, to fund the upcoming upsized long-term treasury buyback program.
The title “Bessent could tap near $1 trillion Treasury General Account to fund bond buybacks” gave an impression that this is a powerful policy tweak. Close to USD 1 trillion of firepower available for the Treasury’s disposal, it “would limit any concern…that the Fed could be asked to help the Treasury,” as the report stated. With this “clever twist,” the Treasury also doesn’t have to sell short-term Treasurys, making this policy look less like an Operation Twist but more akin to a QE (quantitative easing)-lite. The perception of this so-called “Bessent Put” improved with this one tiny adjustment, didn’t it?
How Much Room Does Bessent Actually Have?
If only. The USD 1 trillion firepower? Of course they can’t just spend it all. Yes, the Treasury has full discretion over how to spend this deposit saved in the Fed, but at the minimum it is constrained by the debt ceiling considerations. That is because TGA has acted as a buffer against immediate default of government obligations. If previous debt ceiling behaviors were a guide, USD 300 billion of TGA balance seems to be a soft floor the Treasury prefers to maintain.
Another rule of thumb, as CNBC reported, was that “[when] Janet Yellen ran the Treasury, officials said the goal was to set the TGA at a ‘week ahead of cash needs.'”
TGA is best understood not as a pile of free cash sitting idle on the sideline waiting to be spent, but a checking account with a substantial amount of deposits and withdrawals daily. The below graph shows the daily inflow and outflow in the TGA, both hover at around USD 150 billion a day; USD 300 billion of TGA balance would mean around two days of gross cash outflow (it will likely be able to support the US government for more than two days of operations if the debt ceiling is hit, as it has other cash inflow than money received from bond sales).
More importantly, as Linesman suggested in his article, the US government is not expected to hit another debt ceiling until the end of next year. So, this potentially gives Bessent a free hand to spend more of the USD 1 trillion liquidity to buy back long bonds and compress the long term Treasury yield.
Why This Isn’t Really QE
Another constraint we also have to consider is how many long-term Treasurys they can buy. The upsized buyback program targets Treasurys with remaining maturity ranging from 10 years to 30 years. In terms of the outstanding stock of Treasurys in the market, there is around USD 4 trillion available. Plenty for the Treasury to purchase.
So, let’s unleash the “Treasury QE”? Of course they can try and maybe they should. Just don’t hold your breath and think it will give the Treasury ultimate control of the yield curve.
Even if we give a generous estimate of USD 700 billion “free cash” they can deploy, the expected effect would still be very different from, say, the Fed engaging in a USD 700 billion QE. One important difference is that the market knows very well that running down the TGA, to a large extent, won’t be a “permanent” increase in reserve supply. That is, the Treasury is expected, say, within a year, to have to boost Treasury supply to replenish TGA to the “normal” level.
In comparison, when the Fed buys long-term Treasury in a QE operation, it can create new reserves for the purchase. While quantitative tightening is ultimately expected, not all reserves created have to be reabsorbed out of the financial system. Moreover, the Fed doesn’t resell its Treasury holdings to the private market during QT.
So, the Treasury’s choice is then whether they will issue more short-term debt, which is just a delayed Operation Twist, or issue more long-term Treasurys, which means the deployment of cash in the TGA is a mere trick to lower the long-term debt supply for just several months.
So, no matter which way Bessent chooses to go with, it would only be a bridging program to help temporarily reduce long-term Treasury supply for at most several months.
Will Investors Even Play Along?
Also, while USD 933 billion can arguably be too much to keep in the government checking account, a prolonged extreme low level of ready-cash in the TGA can also create a perception that the US Treasury may have a higher risk of handling payment in extreme circumstances. This can in turn increase the risk premium embedded in Treasury yield, however small that will be.
More importantly, the question is whether investors will sell several hundreds of billions of long-term Treasurys back to the US government if they know in advance that the overall Treasurys supply will be shrunk and without clear knowledge of what and when replacement issuance would be available.
The standard buyback program is based on the assumption that after selling the off-the-run and slightly less liquid Treasurys to the US government, they will then replace the Treasury supply with new issues. Without a predictable and stable expectation of replacement issues, investors may hesitate to hand over a substantial amount of their holdings to the Treasury.
The Simpler, More Honest Alternative
Given all this, my humble opinion is that I don’t understand why Bessent would suggest a surprise and supersized buyback to tame the long term yield. To be honest, instead of hinting to run down the TGA for a large buyback scheme, why don’t they simply announce that they will reduce the size of 10- to 30-year Treasury issuance? The Treasury is currently scheduled to sell USD 39 to 42 billion of 10-year, USD 13 to 16 billion of 20-year, and 22 to 25 billion 30-year each month in the next quarter, plenty of room to cut and it is the same strategy in essence.
This simpler route may actually be more credible compared to the buyback gimmick he is currently pulling. Of course, one potential reason why Bessent prefers not to pursue this simpler route is his famous criticism of his predecessor Janet Yellen for doing the same. But seriously, it is hard to see how the buyback maneuver is any different in substance, except for the attention it has drawn from the market.
So far, that extra attention has not worked in Bessent’s favor. But still, he has no plan to change the issuance schedule. “We are going to continue with our regular program of auctions. So you will be hearing from us again at the beginning of next quarter,” Bessent said during the Iran sanction presser on Monday.
