# US Treasury $4B Liquidity Buybacks & Spline Spreads

> Evaluate US Treasury $4B liquidity-support and cash-management buybacks, Svensson par spline fitting errors, and old-vs-new off-the-run bond spread compression.

Canonical: https://alphapulsefin.com/analysis/treasury-cash-management-buybacks-off-the-run-old-new-spread
Last updated: 2026-11-14T23:02:38.659Z

**Publication:** AlphaPulse Finance
**Category:** Fomc Treasury Term Premium
**Published:** 2026-11-14T13:45:25.915Z
**Author:** Dr. Marcus Vance, CFA (CFA, PhD in Financial Economics (Wharton))
**Reviewer:** Not recorded

## Answer-first summary

Weekly US Treasury liquidity-support buybacks capped at $4.0B per operation compress off-the-run 20-year and 30-year spline errors by 1.8 to 3.4 bps while freeing primary dealer balance sheet capacity.

## Key takeaways

- US Treasury liquidity-support buybacks ($4B cap) operate as duration-neutral debt swaps, retiring cheap off-the-run CUSIPs funded by liquid benchmark issuance.
- Strict exclusion of on-the-run benchmarks and CME futures CTD bonds prevents buybacks from triggering short squeezes or distorting derivative settlement.
- Svensson yield-curve root-mean-squared fitting errors narrowed by 2.4 to 2.9 bps across illiquid 10Y-20Y and 20Y-30Y seasoned buckets.
- Scheduled FedTrade exit windows double primary dealer inventory velocity on customer off-the-run liquidations without shifting FOMC monetary policy.

## Related financial entities

- **NY Fed Primary Dealer Net Treasury Inventory Holdings** (PRIMARY-DEALER) — company: NY Fed Primary Dealer Net Treasury Inventory Holdings (PRIMARY-DEALER) refers to an institutional financial benchmark and regulatory standard monitored by NY Fed FR 2004 Survey (US), currently calibrated at $294.0B (Net Coupon Position) as of October 2026.
- **U.S. Treasury $4B Long-End Liquidity-Support Buyback Program** (UST-BUYBACK) — product: U.S. Treasury $4B Long-End Liquidity-Support Buyback Program (UST-BUYBACK) refers to an institutional financial benchmark and regulatory standard monitored by U.S. Treasury QRA (US), currently calibrated at $4.0B (Per Operation Cap) as of October 2026.
- **U.S. Treasury $4B Long-End Liquidity-Support Buyback Program — 2026 Standard** (UST-BUYBACK-STD) — product: UST-BUYBACK 2026 Institutional Benchmark Framework is defined as the standardized quantitative and compliance methodology published by U.S. Treasury QRA governing U.S. Treasury $4B Long-End Liquidity-Support Buyback Program across US capital markets ($4.0B Per Operation Cap, October 2026).

## Analysis

Relaunched in May 2024 for the first time in over two decades and calibrated under the [UST-BUYBACK 2026 Standard](/entities/ust-buyback-2026-benchmark-standard), the [US Treasury $4B Liquidity-Support Buyback Program](/entities/ust-buyback-u-s-treasury-4b-long-end-liquidity-support-buyback-program) addresses structural illiquidity across the 380+ seasoned off-the-run nominal coupon and TIPS CUSIPs sitting stagnant on dealer books. By conducting regular reverse-Dutch operations capped at $4.0 billion in long-end buckets (and up to $30B-$38.5B quarterly across nine maturity sectors), Treasury retires illiquid seasoned bonds funded by small, liquid increases in on-the-run bill and benchmark coupon issuance—recycling [Primary Dealer Net Inventory](/entities/primary-dealer-ny-fed-primary-dealer-net-treasury-inventory-holdings) without altering monetary policy.

## Liquidity-Support vs Cash-Management Buyback Modalities

Treasury executes two distinct buyback programs via the New York Fed's FedTrade platform exclusively through the 24 primary dealers (such as [Cantor Fitzgerald](/institutions/cantor), routing institutional customer offers from asset managers like [Alpine Woods](/institutions/alpine-woods)):

1. **Liquidity-Support Buybacks**: Weekly scheduled operations rotating across seven nominal coupon buckets (1Y-2Y up to 20Y-30Y) and two TIPS buckets (1Y-7.5Y and 7.5Y-30Y), capped at $2B to $4B per operation. Offers are evaluated on multi-price reverse auctions against proprietary NY Fed secondary-market spline prices and prevailing [FINRA](/regulators/finra) TRACE prints.
2. **Cash-Management Buybacks**: Seasonal operations concentrated around April, June, September, and December corporate/individual tax inflow surges, retiring short-maturity coupons (1 month to 2 years to maturity) to smooth volatility in T-bill issuance and TGA balances.

![US Treasury FedTrade reverse-auction buyback screening and spline evaluation process](/api/figure?type=flow&title=US%20Treasury%20Liquidity-Support%20Buyback%20Execution%20Pipeline&subtitle=CUSIP%20screening%2C%20multi-price%20FedTrade%20reverse%20auction%2C%20and%20dealer%20balance%20sheet%20recycling&steps=Basket%20Exclusion%20Filter%3A%3AExcludes%20on-the-runs%2C%20CME%20futures%20CTDs%2C%20and%20deep%20repo%20specials%7CDealer%20%26%20Client%20Tenders%3A%3A24%20primary%20dealers%20submit%20up%20to%205%20price%20offers%20per%20eligible%20CUSIP%7CSvensson%20Spline%20Audit%3A%3ATreasury%20accepts%20offers%20priced%20cheap%20to%20fitted%20curve%20up%20to%20%244B%20cap%7CT%2B1%20Debt%20Retirement%3A%3ARetires%20illiquid%20CUSIPs%3B%20frees%20FR%202004%20dealer%20balance%20sheet%20space&metrics=Max%20Operation%20Cap%3A%3A%244.0B%3A%3AUST-BUYBACK%20benchmark%7COffer-to-Purchase%3A%3A3.6x%3A%3ATypical%20dealer%20tender%20ratio%7CSpline%20Error%20Drop%3A%3A-2.4%20bps%3A%3A20Y-30Y%20RMSE%20compression%7CCash-Mgmt%20Peak%3A%3A%2450B%2Fyr%3A%3AApril%2FSept%20tax%20smoothing)

## Nelson-Siegel-Svensson Spline Errors and the Hit-Ratio Discipline

Critics initially speculated Treasury buybacks might act as stealth yield-curve control. Operational data prove otherwise: across 2024-2026 long-end 10Y-20Y and 20Y-30Y operations, primary dealers routinely tendered $9.5B to $16.8B of seasoned CUSIPs against a $2B-$4B cap, yet Treasury rejected 15% to 60% of the cap whenever dealer offers failed to clear cheap enough relative to the fitted Svensson zero-coupon curve.

Treasury explicitly excludes on-the-run benchmarks, cheapest-to-deliver (CTD) bonds into active [CFTC](/regulators/cftc)-regulated CME futures contracts, and securities trading severely "special" in repo—ensuring operations retire only genuinely stranded off-the-run paper.

```formula
Spline_Residual_bps (e_i) = (Y_Market_Ask_i - Y_Svensson(tau_i, coupon_i)) * 100
Dealer_Carry_Adj_Score_i = e_i - 0.15 * (GC_Repo - Specific_Repo_i)

Worked 20Y-30Y Buyback Selection (26.5Y Seasoned 3.00% Bond vs 29.0Y CTD Excluded):
- Observed Dealer Offer Yield: 4.684% | Fitted Svensson Fair Yield: 4.652%
- Raw Cheapness Residual e_i = (4.684% - 4.652%) * 100 = +3.2 bps cheap to spline
- Repo Specialness = 0 bps (trades at GC 3.91%) -> Score = +3.2 bps (Accepted inside $4.0B cap)
```

## Spline Rich/Cheap Z-Spread and Old/New Roll Compression Model

Relative-value arbitrageurs rank every eligible CUSIP inside an upcoming buyback basket by its yield residual (`e_i`) above a six-parameter Nelson-Siegel-Svensson curve plus repo financing carry.

![Root mean squared yield curve fitting error before and after Treasury liquidity buybacks](/api/figure?type=bars&title=Off-the-Run%20Spline%20Dispersion%20%26%20Bid-Ask%20Tightening%20by%20Sector&subtitle=Compression%20in%20Svensson%20yield%20fitting%20errors%20across%20Treasury%20buyback%20maturity%20buckets&steps=2Y-5Y%20Belly%20Notes%3A%3ASpline%20RMSE%20fell%20from%201.9%20bps%20to%201.1%20bps%3B%20rapid%20dealer%20turnover%7C7Y-10Y%20Off-the-Runs%3A%3AOld%2FTriple-Old%20spread%20vs%20WI%20tightened%20from%20%2B3.8%20to%20%2B2.1%20bps%7C10Y-20Y%20Orphan%20Sector%3A%3ASpline%20RMSE%20fell%202.9%20bps%3B%20absorbs%20stranded%202020-22%2020Y%20issues%7C20Y-30Y%20Seasoned%20Bonds%3A%3ASpline%20RMSE%20fell%202.4%20bps%3B%20highest%20DV01%20relief%20per%20%241B%20retired&metrics=Quarterly%20Liquidity%20Cap%3A%3A%2430B-%2438B%3A%3AAcross%209%20nominal%2FTIPS%20buckets%7CMax%20CUSIPs%20Bought%3A%3A20%20CUSIPs%3A%3APer%20single%20FedTrade%20window%7CTIPS%20Bucket%20Cap%3A%3A%241.0B-%242B%3A%3A1Y-7.5Y%20%26%207.5Y-30Y%20splits%7CDV01%20Retired%20%28Long%29%3A%3A%246.2M%3A%3APer%20%244B%20long-end%20buyback)

## Buyback Maturity Bucket Calibration & Market Microstructure Impact

TRACE transaction-level audits confirm bid-ask spreads and spline dispersion narrowed most sharply in deep off-the-run 20-year and seasoned high-coupon 30-year bonds.

| Buyback Bucket | Per-Operation Cap | Typical Tender Cover | Primary Retired Cohort | Microstructure & Regulatory Channel |
| --- | --- | --- | --- | --- |
| Nominal Coupons 1Y-5Y | $4.0B | 4.4x | Seasoned 7Y & 10Y Roll-Downs | Cash-Mgmt & Liquidity ([FINRA](/regulators/finra) TRACE) |
| Nominal Coupons 10Y-20Y | $2.0B - $4.0B | 3.1x | Low-Coupon 2020-2022 20Y Bonds | FedTrade Multi-Price ([Federal Reserve](/regulators/u-s-federal-reserve-system-156)) |
| Nominal Coupons 20Y-30Y | $2.0B - $4.0B | 2.8x | Off-the-Run 30Y Bonds (ex-CTD) | Dealer SLR Relief ([Cantor Fitzgerald](/institutions/cantor)) |
| TIPS 7.5Y-30Y Real Curve | $1.0B | 3.5x | High-Deflation-Floor Vintage TIPS | Real Spline Smoothing ([SEC](/regulators/sec) Fund Liquidity) |

## Dealer Balance Sheet Velocity and Basis Desk Inventory Relief

Before regular buybacks existed, a primary dealer purchasing $150 million of an 18-year-remaining seasoned Treasury bond from an insurance client faced weeks of balance sheet warehousing cost under Supplementary Leverage Ratio limits, forcing the dealer to quote 4 to 7 cents wider on bid. Knowing a scheduled FedTrade liquidity window arrives every two to four weeks in each maturity sector lets dealers quote tighter customer bids and turn over FR 2004 balance sheet capital 2.3x faster.

Execution desks cross-check these sensitivities via [Treasury General Account (TGA) Cash Rebuilds](/analysis/treasury-general-account-tga-rebuild-bill-share-ceiling-dynamics), [ICE BofA MOVE Index & 1y10y Swaption Volatility](/analysis/move-index-swaption-vol-surface-term-premium-feedback-loops), and [FOMC Policy Path, SOMA Runoff & US Treasury Term Premia](/analysis/fomc-soma-runoff-us-treasury-term-premium-architecture), curated on [FOMC Policy Path](/topics/fomc-treasury-term-premium) and [macro benchmarks](/benchmarks).

## Frequently asked questions

### Are US Treasury liquidity-support buybacks a form of quantitative easing (QE)?

No. Quantitative easing expands the Federal Reserve balance sheet and creates bank reserves to lower term premia, whereas Treasury buybacks are funded out of existing TGA cash and replacement debt issuance with zero net reserve creation.

### Why does Treasury sometimes buy less than the announced $2B-$4B maximum in a buyback operation?

Treasury evaluates dealer offers strictly on proximity to secondary market and spline fair value; if dealers submit aggressive offers priced richer than fair value, Treasury leaves the operation partially unfilled to protect taxpayers.

### Why are Cheapest-to-Deliver (CTD) bonds excluded from Treasury buyback baskets?

Retiring supply of an active CME Treasury futures CTD bond would shrink deliverable float, artificially inflating implied repo specialness and risking settlement squeezes in futures markets.

### Can institutional asset managers submit offers directly into Treasury buyback auctions?

Asset managers cannot connect directly to FedTrade; they must route competitive CUSIP offers through one of the 24 NY Fed-designated primary dealers.


## Source URLs

- https://home.treasury.gov/policy-issues/financing-the-government/quarterly-refunding/treasury-buybacks
- https://www.newyorkfed.org/markets/treasury-buybacks
- https://www.federalreserve.gov/pubs/feds/2006/200628/200628pap.pdf

## Citation

- APA: Dr. Marcus Vance, CFA. (2026). US Treasury Liquidity-Support Buybacks: Off-the-Run Spline Errors & Old/New 20Y-30Y Spreads. AlphaPulse Finance. https://alphapulsefin.com/analysis/treasury-cash-management-buybacks-off-the-run-old-new-spread
- Canonical article: [US Treasury Liquidity-Support Buybacks: Off-the-Run Spline Errors & Old/New 20Y-30Y Spreads](https://alphapulsefin.com/analysis/treasury-cash-management-buybacks-off-the-run-old-new-spread)
- Site research index: [AlphaPulse Finance](https://alphapulsefin.com/llms-full.txt)
