SOFR vs EFFR: What the Two Overnight Rates Measure

The Secured Overnight Financing Rate (SOFR) and the Effective Federal Funds Rate (EFFR) are often displayed side by side because both are U.S. dollar overnight benchmarks. Their shared tenor does not make them interchangeable. SOFR is built from Treasury-collateralized repurchase agreement, or repo, transactions. EFFR is built from overnight federal funds transactions, which are unsecured dollar borrowings in the federal funds market.

The distinction answers a basic question about what each number represents. SOFR is a broad measure of the general cost of financing Treasury securities overnight. EFFR is a transaction-based measure of the federal funds market's overnight borrowing rate. The New York Fed publishes both as separate reference rates on its reference-rates page, with separate market inputs and methods.

Neither benchmark is the interest rate on every overnight loan in the United States. SOFR does not describe every repo contract or every Treasury security. EFFR is not a survey of every bank's funding cost, a consumer deposit rate, or an unsecured rate available to any firm. Start with the transaction pool before using either rate to answer a question.

The federal funds market consists of domestic unsecured U.S. dollar borrowings by depository institutions from other depository institutions and certain other entities, primarily government-sponsored enterprises. These are unsecured overnight dollar borrowings, rather than loans secured by Treasury collateral. The New York Fed calculates EFFR from overnight federal funds transactions reported by domestic banks and U.S. branches and agencies of foreign banks through the FR 2420 Report of Selected Money Market Rates.

EFFR is a volume-weighted median of those reported transactions. In practical terms, the calculation orders reported trades by rate and finds the rate at the midpoint of transaction volume; it is not a simple average of all reported rates. The published rate is rounded to the nearest basis point. The New York Fed's EFFR page and reference-rate methodology describe the input data and calculation.

EFFR is related to, but distinct from, the federal funds target range. The FOMC sets that target range; it does not set the rate of each completed overnight loan. The Federal Reserve uses administered rates, including interest on reserve balances and its overnight reverse repurchase facility, to guide market rates toward the target range. The Fed explains these roles in its policy-rate overview and IORB FAQs. EFFR is an observed market rate, not the target itself.

A repo exchanges cash and securities under an agreement to reverse the exchange later. In the trades behind SOFR, the borrowing is collateralized by U.S. Treasury securities. The New York Fed describes SOFR as a broad measure of the general cost of financing Treasury securities overnight, rather than the rate on one negotiated repo trade.

The published SOFR calculation combines transactions used for the Broad General Collateral Rate (BGCR) with additional bilateral Treasury repo transactions cleared through the Fixed Income Clearing Corporation's Delivery-versus-Payment service. BGCR itself includes tri-party Treasury general-collateral trades and General Collateral Finance repo trades. These segments bring together transaction data from different parts of the Treasury repo market; they are not a census of every private Treasury financing trade. The New York Fed summarizes the components on its reference-rates page and details them in its methodology.

Collateral matters because Treasury securities can be wanted for more than their value as a way to secure cash. In the cleared bilateral DVP segment, a trade identifies specific securities for settlement. When a particular issue is scarce or useful, a cash provider may accept a lower cash return to obtain it; that specific security is commonly described as trading special. This securities motive can affect repo rates even when the cash borrowing is overnight.

EFFR and SOFR both use a volume-weighted median and are rounded to the nearest basis point. That common statistical summary does not make their inputs alike. EFFR draws on reported unsecured federal funds transactions. SOFR draws on defined Treasury repo segments, where the borrower receives cash against Treasury collateral. The counterparty pool, collateral, trade purpose, and market structure differ.

The SOFR methodology also adjusts how the cleared DVP segment contributes to the benchmark. To reduce the influence of transactions considered “specials” on a measure of general Treasury financing cost, the New York Fed removes 20 percent of the lowest-rate transaction volume from that segment each day, after excluding relevant affiliated trades. The methodology says this removes some, but not all, transactions in which securities are trading special. SOFR is therefore a broad Treasury financing measure with a stated filter, not a pure general-collateral rate.

For either rate, the New York Fed reviews transaction data for apparent errors or unusual trades and publishes accompanying information about the rate calculation. Its methodology describes exclusions, possible same-day revisions, and contingencies if data for a market segment are unavailable. A published benchmark is a carefully defined summary of the available input data, not a promise that every transaction or every lender was captured without later correction.

The two rates are published on different schedules. SOFR is typically published on business days at about 8:00 a.m. Eastern Time and reflects overnight Treasury repo activity for its value date. EFFR is published at about 9:00 a.m. Eastern Time for the prior business day. The New York Fed's reference-rate page, EFFR page, and methodology pages set out their publication details.

Holiday treatment and data contingencies also matter. The Treasury repo reference rates follow the publication schedule described for U.S. government securities market closures, while EFFR follows the New York Fed's holiday schedule. If data for a SOFR market segment are not available, the published methodology allows the New York Fed to use adjusted prior data for that segment in specified circumstances. Published rates can also be corrected under the stated revision rules when errors or later-arriving data are found.

When comparing two daily observations, match the value dates, not just the dates or times shown on two screens. A display may label a rate by its publication date, value date, or the date a vendor received it. Around weekends, holidays, early closes, or a revision, those labels can obscure which overnight transactions a number summarizes. The methodology pages are the source for resolving that record.

The New York Fed also publishes compounded SOFR averages for 30-, 90-, and 180-calendar-day periods and a SOFR Index. These are period measures built from daily SOFR, not another name for the one-day fixing. If the question is the cost of one overnight Treasury-backed loan, a compounded average answers a different question; if a document names an average or index, use that exact series and its stated dates.

The two rates can differ because they summarize transactions with different collateral, counterparties, and financing purposes. Conditions affecting the supply of cash or demand for unsecured federal funds need not move in lockstep with conditions affecting Treasury repo financing. A change in demand for a specific Treasury issue can also influence the rates on special repo trades, while EFFR has no Treasury collateral component.

Specialness is one possible source of a difference, not a complete explanation for any observed gap. The New York Fed trims the lowest-rate 20 percent of DVP transaction volume to limit the influence of specials on SOFR, but its methodology says some special trades remain. Their effect on the published median depends on the full transaction distribution and the other included repo segments. A daily SOFR–EFFR difference cannot identify specialness by itself.

The reporting and publication process can add another layer. The rates use different transaction reports and market-segment data, are published at different times, and may follow different holiday or correction procedures. These details can affect which trades enter a particular day's benchmark or how a displayed observation is dated. They do not establish a cause for a specific divergence without examining that day's official data and methodology notices.

There is no fixed spread or permanent ordering implied by the definitions. The rates may be influenced by shared monetary and liquidity conditions, but each represents a different market. Do not infer a policy change, funding stress, or arbitrage opportunity from a gap alone.

Use SOFR when the question is about a broad measure of overnight cash borrowing secured by U.S. Treasury securities across the repo segments specified by the New York Fed. It is not the actual borrowing rate for every institution, every Treasury security, or every repo contract.

Use EFFR when the question is about the effective rate on overnight unsecured transactions in the federal funds market. It is not the FOMC target range, the rate on Treasury-collateralized borrowing, or a promise of the rate a particular bank or business can obtain. If the question is about the policy setting, consult the FOMC target range separately.

For a contract or model, use the rate the governing document actually names. A contract may use daily SOFR, a compounded SOFR average, EFFR, or another specified series, with its own observation dates and fallback terms. The fact that two rates are both overnight benchmarks does not make their definitions or contractual use interchangeable.

Before interpreting a difference, write down the rate name and the exact question it is intended to answer. Confirm whether the observation is SOFR or EFFR, the value date, publication date, units, and whether a rate was revised or published with a methodology notice. Use the New York Fed's official series rather than a rounded screenshot when the exact observation matters.

Next, keep the transaction bases visible. EFFR represents reported unsecured federal funds trades. SOFR represents a defined set of Treasury-secured repo trades, including a DVP segment with a specified special-trade filter. Note that both rates use volume-weighted medians, but don't mistake a shared calculation statistic for a shared market.

Finally, describe the difference before explaining it. “SOFR was above EFFR for the same value date” is a comparison. “Treasury specialness caused the gap” is a causal claim that needs evidence beyond the two published values. Check the relevant methodology, transaction summaries, holiday schedule, and any revision notice before assigning a reason. The strongest comparison keeps the market, collateral, data date, publication process, and limits of the benchmark in view.

A compact comparison record can list the benchmark name, transaction market, collateral, value date, publication time, calculation method, and any revision note. For SOFR, identify the daily fixing or the specified compounded series and remember that the DVP filter applies to one segment. For EFFR, identify the prior-business-day unsecured federal funds observation. Keeping those fields together makes a measured difference easier to reproduce and prevents a date label or abbreviated rate name from standing in for the underlying data.

For the policy target and how EFFR relates to bank lending rates, see Federal Funds Rate vs. Prime Rate. For secured repo funding mechanics, see the Repo vs. Reverse Repo guide.