As prepared for delivery
Good morning and thank you for having me here at the Federal Reserve Bank of New York’s annual primary dealer meeting. I am Hunter McMaster, Director of Policy and Planning at the Treasury Department. I’m also performing the duties of the Assistant Secretary for Financial Markets. In this latter role, I oversee Treasury’s Office of Debt Management. I have worked with Secretary Bessent in various capacities going back to his time at Key Square, and I’m glad to be part of the team supporting him at Treasury.
As I’m sure you’ll remember, my predecessors have commonly remarked that the U.S. Treasury market is the deepest and most liquid market in the world. This is due, in no small part, to the individuals sitting in this room. Primary dealers perform critical functions that enable the Treasury market to thrive. These include, among other things, bidding a pro rata share in each of the more than 400 auctions that we conduct every year, intermediating client flow in the secondary market, and providing the official sector with invaluable insights on market functioning.
My remarks today will focus on three topics. First, I will review Treasury’s cash balance policy. Second, I will discuss how this policy informed our rebuild of the Treasury General Account following the increase to the debt limit in early July as part of the One Big Beautiful Bill. And, third, I will discuss Treasury’s buyback program, and the recent enhancements we’ve announced to improve its efficacy.
Treasury’s Cash Balance Policy
To begin, Treasury established the current cash balance policy in 2015 as a risk management tool. The policy entails maintaining funds in the Treasury General Account (TGA) that would be sufficient to at least meet our one-week ahead cash need. Our cash need is defined as net fiscal outflows plus the gross volume of maturing marketable debt. The policy ensures that, even if our ability to raise new funds were temporarily impaired, as might result from a cyber-attack or a natural disaster, Treasury would still be able to meet its obligations on time and in full.
When the One Big Beautiful Bill (OBBB) increased the debt limit on July 4, Treasury had $313 billion in the TGA. Although $313 billion in many contexts seems like a sizeable amount, let me discuss how this compares to our cash needs. First, consider that Treasury typically has $200 to $250 billion of Treasury bills maturing every Tuesday and Thursday, or, in other words, $400 to $500 billion maturing per week. Next, consider that we have coupon securities maturing both at mid- and end-of-month, and these maturities typically run between $100 and $150 billion each but can exceed $200 billion. Treasury must also be prepared to make interest payments on these maturity dates. On some days, interest payments are relatively small but, on others, much larger. For example, on August 15, each of our 10-, 20-, and 30-year nominal coupon securities maturing in February and August had an interest payment due. As a result, we paid $75 billion of interest on August 15 on securities with principal totaling more than $5.3 trillion. All this is to say, that between bill and coupon maturities and interest payments, $313 billion was well below our prudent policy level
Up until this point, I’ve discussed only flows associated with Treasury financing but some of our fiscal flows can be quite concentrated as well. For example, withdrawals associated with Medicare and Department of Veterans Affairs benefits tend to be at their heaviest around the turn of the month. Supplemental Security Income payments occur on the first of the month, and roughly one-fourth of Social Security payments are made on the third of the month.
This brief outline of our cash needs, I hope, clearly illustrates two points. First, in the current environment, $400 to $500 billion is the absolute bare minimum that Treasury needs to be compliant with its cash balance policy on any given day. Second, thinking about our cash needs in terms of averages is inappropriate because our cash flows meaningfully fluctuate over the course of the month, peaking just ahead of mid-month and month-end. At these times, our weekly cash needs can exceed $800 billion. I also want to reiterate that Treasury’s cash balance policy is not equal to a pre-determined amount of cash, such as $700 or $800 billion. Instead, Treasury calculates a tailored cash balance minimum based on the policy for every business day, linked to the projected outflows over the week following that business day.
Over the past several months, we’ve been aware of market speculation that our cash balance policy might soon change. For the avoidance of doubt, our cash balance policy has not changed. And a change in our policy is not currently under consideration.
Rebuilding the Treasury General Account (TGA)
Now, having reviewed our cash balance policy and why the $313 billion TGA level in early July was well below our prudent policy level, I will next review the actions we have recently taken to rebuild the TGA.
Since the passage of OBBB, Treasury has increased bill supply by $603 billion, and the cash balance has grown by about $500 billion. The increase in the cash balance was accomplished gradually over July to end the month at about $500 billion, and most of the remainder of the increase has occurred in September. We focused increased bill supply on shorter bill maturities, even auctioning $100 billion in a single security for the first time, and we monitored market conditions carefully along the way.
Our assessment, informed by perspectives from you and your clients, is that absorption of this additional bill supply has gone very well. Bill auctions’ bid-to-cover ratios have averaged nearly 3x since the debt limit was raised in early July. We’ve seen direct and indirect bidders collectively receive more than two-thirds of the bill auction awards over the same timeframe, which we assess as being indicative of strong end-investor demand.
Further corroborating this assessment, we’ve seen Treasury bills perform well compared to matched-maturity SOFR OIS, even with the sizable increase in supply. Rather than cheapen compared to SOFR, we’ve seen bill auction stop-out rates either remain flat or richen on spread. In addition, Treasury staff have been keeping a close eye on repo rates within the context of broader money market conditions. Whereas SOFR on average printed slightly below the effective Federal Funds rate (EFFR) during the three months preceding the increase in the debt limit, we’ve seen SOFR on average print around 7 basis points above EFFR in September. On September 15, $78 billion of net Treasury coupon supply settled, and the mid-month corporate and non-withheld tax date contributed to a $150 billion day-over-day increase in the TGA. Although there were signs of modest repo market pressure as SOFR printed at 4.51 percent, or 18 basis points above EFFR, anecdotal market reports indicated that this event was well-absorbed and short-lived, as SOFR declined by 12 basis points the very next day.
With quarter-end arriving tomorrow and with nearly $100 billion of net Treasury bill and coupon auction supply settling, we, like all of you, will continue to monitor repo market conditions carefully.
Treasury Buyback Enhancements
Turning to our next topic, on Treasury buybacks. As this audience is aware, Treasury introduced a regular buyback program in May of 2024. From our perspective, this program has two use cases—cash management and secondary market liquidity support. Since its inception last year, we’ve bought back $113 billion for cash management and over $115 billion for liquidity support.
I am going to focus most of my remarks on the operations for liquidity support. From the beginning, we believed that providing the market with a regular and predictable opportunity to sell less-liquid securities back to Treasury at a fair price would enable market intermediaries to recycle risk into additional market-making opportunities. The condition “at a fair price” is key. This is a discretionary program, and we are a price-sensitive buyer. Over the total 85 buyback operations, on average we executed approximately one quarter of a tick cheap below mid-market prices. If we don’t receive sufficiently aggressive offers, we can, and on a somewhat regular basis we do, buy back less than our stated maximum. We do not consider the lack of a “full fill” as inherently problematic but rather a reflection of the robust liquidity conditions in that segment of the market at the time.
The feedback that we’ve received, both from you and from your clients, is that the buyback program, although small, is meeting our goals. However, in the spirit of constant improvement, we’ve been actively evaluating potential enhancements that would improve the program’s worth. Some of these enhancements were announced as part of our most recent quarterly refunding policy statement.
First, we doubled the frequency of buybacks in the long end of the curve, for the 10- to 20- and 20- to 30- year sectors, and, by extension, doubled the amount we’re willing to buy back in those segments. Our operations in these segments have resulted in the most consistently favorable execution, which justifies this change. I want to reiterate that the change to long-end buybacks is not about attempting to alter the weighted-average maturity (WAM) of the Treasury debt, which currently stands around 6 years. Buybacks are small relative to the overall size of the debt, and the purchases we’ve done to date have changed the WAM by only a few weeks, all other things being equal. In addition, the change to the frequency of long-end buybacks will have only a marginal effect, a matter of days, on the WAM.
Second, we are increasing the aggregate size of cash management buybacks from $120 billion per year to $150 billion. Although cash management buybacks are designed principally with Treasury’s cash needs in mind, we have heard anecdotally that our purchases at the very front end of the curve have been helpful in removing some of the more balance sheet-intensive Treasury holdings from your books.
And, finally, we plan to offer direct buyback access to a limited number of additional counterparties, beginning in the first half 2026, based on their participation in Treasury auctions. While Treasury encourages primary dealers to facilitate submission of client offers, the current set-up can be challenging and costly to intermediate. Broadening direct access to a select number of additional counterparties can enhance the level of liquidity support we provide to the market and improve our trade execution.
We provided guidance to the market regarding the eligibility criteria for direct buyback access in a supplemental announcement published several days ago, and we have contacted eligible counterparties. To identify potential additional counterparties, Treasury ranked the auction bidders by total nominal coupon and TIPS auction awards, duration adjusted to 10-year equivalents, during the first half of 2025. Due to operational considerations, the initial list of eligible additional counterparties consists of auction bidders that are not primary dealers and that had more than $35 billion in awards during the first half of 2025.
Treasury’s New Auction System
Finally, I wanted to alert all of you that Treasury is in the process of upgrading its existing Treasury Automated Auction Processing System, or TAAPS. We are targeting for this upgrade to occur by the end of next year. As a result, the interface is going to look a little different than what you’re used to. With that in mind, my colleagues at Treasury’s Bureau of the Fiscal Service will begin reaching out over the next couple of months to arrange overview sessions so that you can begin gaining familiarity with the new system.
Conclusion
In conclusion, thank you all for your continued support of the U.S. Treasury market. You serve an invaluable role in ensuring that ours remains the deepest and most liquid market in the world and I look forward to continuing to work together to enhance these characteristics of the U.S. Treasury market.