Perpetual funding rates and annualized yield: a 2025 comparison

One year, four different funding profiles

We recalculated the calendar-year 2025 funding histories stored in an existing research cache. Each asset has 1,095 unique recorded settlements, from 1 January through 31 December UTC. The signed sums are BTC 5.1264%, ETH 4.9287%, SOL 0.3507% and BNB −2.0849%. These are sums of the published rates, not compounded or realised portfolio returns.

BNB illustrates why frequency and amount must be separated. Only 18.08% of its recorded settlements have a negative rate, yet the yearly sum is negative. The size of negative observations can outweigh a larger number of positive or zero observations. SOL’s rate sum is close to zero despite many positive settlements.

The denominator determines the meaning

At each settlement, funding cash flow depends on the applicable signed rate and the position’s settlement notional. Account return divides the resulting cash flows and all other profit or loss by the relevant invested capital. Summing rates assumes a constant notional for interpretation; a real position’s notional can change with price or quantity.

Margin, collateral and any spot inventory also occupy capital. Fees, spread, market impact, basis changes and collateral valuation can affect account results. A rate quoted relative to contract notional cannot be relabelled as a return on total capital without stating these components.

Funding is a market price, not a guaranteed yield

Positive funding generally means longs pay shorts; negative funding reverses that direction. The sign and magnitude vary with contract demand and the venue’s rules. A short settlement sample annualised into a headline yield can conceal negative periods, concentration in a few observations and changes in settlement schedules.

The table uses a complete past calendar year to avoid presenting a recent run of favourable rates as an enduring return. It is not a forecast. Comparisons across years should use the actual recorded settlement schedule and distinguish missing records from zero rates.

How to read funding claims responsibly

Ask which venue and contract supplied the data, whether the rate is observed or predicted, how much capital is in the denominator, and whether costs and negative payments are included. Funding income can coexist with trading losses or liquidation risk. No entry condition, position-sizing rule, leverage choice or execution method is supplied here.

What is a crypto perpetual funding rate?

Funding is a periodic payment between perpetual long and short positions under the venue’s settlement terms. Its direction can change. A rate is not the same as a return on all capital required to maintain a position.

Can annualized funding rates predict a guaranteed yield?

No. Annualizing a short interval assumes a rate continues; actual funding, position value, fees and margin needs can change. The historical comparison reports recorded settlements rather than a guaranteed investment return.

Limitations

The rate histories were checked for unique timestamps and finite values and their counts were reconciled for the year. They were not newly fetched from the exchange. No account ledger, actual fees, trade fills or strategy performance is published or inferred.

Sources and data access

• Binance funding-rate history documentation

Cited sources document external facts or data formats; they do not endorse our findings. The public snapshot contains the table and methods. Source-file fingerprints are included for recalculated datasets. Full archived raw files are not distributed here; fingerprints identify the files used, rather than independently proving the results.

Aggregate data snapshot (JSON) · Markdown · Methods and limitations

Live prior 30-day hourly baseline — separate from the historical study

Only the current watch list is shown, up to 12 assets. Missing or stale estimates remain unavailable. Venue and quote currency are kept separate; beta needs matching BTC data. Hourly statistics are not minute-move probabilities.

Observation cutoff (UTC): 20/10/2026, 15:01

Related research