SOFR spread tracker

This tracker subtracts a comparison benchmark from SOFR on every date both were published and reports the gap in basis points, with a chart of the spread, a chart of the two rates and summary statistics for the range you choose. It is meant for treasury, funding and rates staff who watch repo conditions, and for anyone pricing a SOFR loan or swap against fed funds who wants the historical basis in numbers rather than impressions.

Features

• Three comparison rates. EFFR for secured versus unsecured funding, TGCR and BGCR for the repo segments that SOFR contains, all from the same New York Fed snapshot.
• Date-matched spreads. Each spread uses two prints for the same value date. Days when only one rate was published are excluded and listed, never filled in.
• Statistics with dates. Mean, median, widest and narrowest spreads with the dates they occurred, the share of days SOFR was above, equal to or below the other rate, and the latest reading.
• Two linked charts. The spread on its own, and both rates on one axis so you can see whether a wide spread came from SOFR rising or the comparison rate falling.
• Largest-days table. The five biggest gaps in the range, which in practice is a quick list of the month-ends, quarter-ends and stress days.
• CSV export. Every matched date with both rates and the spread, ready for a spreadsheet or a regression.

How to use the SOFR spread tracker

1. Pick the comparison rate. Choose EFFR to see secured against unsecured funding, BGCR to isolate what the cleared bilateral (DVP) repo trades add to SOFR, or TGCR to see what GCF Repo and DVP trades together add to tri-party repo.
2. Choose a range with the quick buttons or type a start and end date. Any dates from April 2, 2018 to the latest snapshot date work; weekends and holidays inside the range are skipped automatically.
3. Read the mean and median first, then the widest and narrowest days. A mean far from the median means a few extreme days are pulling the average.
4. Hover over, tap or drag across either chart to read a date on both at once, and check the table for the days that drove the extremes.
5. Download the CSV if you want to test the pattern yourself, for example by flagging month-end dates.

How the spread is measured

The formula

All four rates are published rounded to the nearest basis point, so every daily spread is a whole number of basis points; the mean is shown to one decimal. Each day counts once: unlike a compounded rate, a Friday print is not weighted for three days, because the question is how the markets priced, not how much interest accrued.

Dates are matched exactly. SOFR, TGCR and BGCR are not published on days SIFMA recommends a full close for government securities, and EFFR follows a different calendar, so in this snapshot EFFR has 12 dates with no SOFR, most of them Good Fridays. The tracker drops those days and lists any that fall in your range.

What SOFR − TGCR and SOFR − BGCR signal

The three repo rates are nested. According to the New York Fed's methodology, TGCR covers overnight tri-party general collateral repo on Treasuries; BGCR adds the GCF Repo trades; SOFR adds bilateral Treasury repo cleared through FICC's delivery-versus-payment (DVP) service, after trimming the lowest-rate 20% of DVP volume to remove some specials. All three are volume-weighted medians. SOFR − BGCR therefore roughly reflects the effect of adding the cleared bilateral (DVP) segment, and SOFR − TGCR the effect of adding GCF Repo and DVP trades together. Because each rate is a median of a pooled set of trades, not a sum, neither spread splits exactly by segment.

A positive SOFR − TGCR gap says that cash lent against Treasuries in the centrally cleared GCF and DVP segments was priced above tri-party cash. A common explanation for a widening gap is balance-sheet pressure on dealers around month-end and quarter-end reporting dates; the tracker measures the gap but cannot prove the cause. In the data the gap is small but persistent: SOFR was at or above TGCR on 97.1% of days since April 2018, and the yearly mean has ranged from 0.3 bp in 2021 to 2.5 bp in 2018. Since 2025 the median day has been +2 bp.

The extremes are worth knowing. The widest SOFR − TGCR reading in the data is 13 bp, on October 1, 2024 (SOFR 5.05%, TGCR 4.92%). The only stretch of negative values came in 2021 and 2022, with a low of −4 bp on April 21 and June 14, 2022. BGCR matched TGCR on 2,076 of the 2,129 days in the data and never differed by more than 1 bp; the mean SOFR − BGCR spread was 1.73 bp against 1.75 bp for SOFR − TGCR. GCF Repo rarely moves the median, so most of the SOFR − TGCR gap comes from the DVP segment.

What SOFR − EFFR signals

EFFR is the volume-weighted median of overnight federal funds transactions that banks report on the FR 2420. Comparing it with SOFR sets a secured rate against an unsecured one, which is why the spread is used as a gauge of how plentiful cash is relative to the Treasury collateral that needs financing. When cash is abundant, lenders accept less to lend against Treasuries than banks pay to borrow unsecured in the fed funds market, and SOFR prints below EFFR. When collateral piles up or balance sheets tighten, repo rises above fed funds.

The yearly figures in the snapshot follow that pattern closely. SOFR averaged 4.0 bp below EFFR in 2021 and 4.5 bp below in 2022, when it was under EFFR on every single day of the year. The spread drifted back toward zero in 2023 (mean −2.0 bp) and 2024 (+0.1 bp), turned positive in 2025 (+3.0 bp, with SOFR above EFFR on 62.2% of days) and has averaged +0.2 bp in 2026 so far. Choose "Since 2018" above to see that shift on the chart.

The calendar shows up too. Across all quarter-end dates since April 2018 the spread averaged +6.7 bp, against −0.4 bp on ordinary days that were not month-ends. From 2025 on the contrast is sharper: +9.9 bp at the seven quarter-ends against +1.5 bp on days that were not month-ends. The largest SOFR − EFFR reading ever recorded is +295 bp on September 17, 2019, when SOFR printed 5.25% against EFFR at 2.30%. That day TGCR was also 5.25%, so SOFR − TGCR was zero: the whole repo market moved, not just the segments SOFR adds to tri-party. The guide to SOFR spikes and month-end covers that episode, and the secured vs unsecured rates guide explains the market mechanics.

Worked example: the October 2025 month-end

The month-end premium over fed funds was about five times the premium over tri-party repo, so most of the gap was a secured-versus-unsecured effect. EFFR also stepped down from 4.12% to 3.87% on October 30, the first EFFR date in the new 3.75%–4.00% target range, and SOFR fell with it; the spread stayed wide across the change. A spread is more informative than either level when the policy rate moves. Enter October 27, 2025 and November 5, 2025 above to reproduce these numbers.

Using the spread well

Who uses this, and for what

Borrowers choosing between a fed funds and a SOFR reference use the SOFR − EFFR history to see what the basis has been in different conditions. Funding analysts track SOFR − TGCR around reporting dates, with the percentile and volume data showing whether a wide day was a broad move or a thin tail.

For the level of SOFR itself over a period, use the SOFR history lookup. For what a given move in basis points does to interest on a balance, use the basis point calculator or the rate shock calculator. The current side-by-side readings for all four rates are on the benchmarks page.

Limits of a daily spread

Frequently asked questions

Is SOFR normally above or below EFFR?

It depends on the period. Since April 2018 SOFR was below EFFR on 57.0% of matched days, above on 31.7% and equal on 11.3%, for an overall mean of −0.2 bp. But that average hides regimes: every day of 2022 was negative, while in the 12 months to October 8, 2026, 62.4% of days were positive.

Why is SOFR almost always above TGCR?

Because SOFR contains every TGCR trade plus GCF Repo and cleared bilateral DVP trades, and those added segments have usually priced higher. SOFR was 1 to 3 bp above TGCR on 70% of days since 2018 and below it on only 2.9%, all in 2021 and 2022.

Why does the spread jump at month-end and quarter-end?

A common explanation is that dealers have less room to expand their balance sheets around reporting dates, which raises the price of repo relative to unsecured bank funding. The tracker measures the size of the effect; it cannot prove the cause.

Why does the latest spread differ from what I see elsewhere?

Check the date. This site uses the New York Fed snapshot through October 8, 2026. SOFR is published at about 8:00 a.m. ET for the previous business day and can be revised once, at about 2:30 p.m. ET the same day, if the change exceeds one basis point, so an early copy and a later copy of the same day can differ.

Can I compare SOFR with Term SOFR or LIBOR here?

No. CME Term SOFR requires a license from CME Group, so this site does not display its values, and all USD LIBOR panel settings ceased on June 30, 2023. The comparison rates here are three of the other overnight benchmarks the New York Fed publishes.

Privacy and data

The spreads and statistics are computed in your browser. SOFR, TGCR, BGCR and EFFR come from a dated snapshot of Federal Reserve Bank of New York data served from this site, and the dates you choose are not sent anywhere. The methodology page explains how the snapshot is collected and refreshed.