Performance
Base return and scaled return are listed side by side, year by year, from period-by-period funding snapshots in the backend as of 2026-08-25. This page also states the methodology — the same data under a different method differs by a factor of two, and the flattering one is inflated.
- 124.296%
- Average annual, 2021–2025 (scaled)
- Base funding 12.43%, principal scaled 10x interest-free
- 885.961%
- Cumulative, 2020–2026
- Summed year by year, not compounded
- 375.621%
- Best cell (ETHUSDT, 2021)
- Bull markets give frequent and rich positive periods
- 7.873%
- Worst cell (ETHUSDT, 2022)
- In a bear market this is how thin the strategy gets
Cumulative return
The axis shows cumulative return after principal is scaled. It rises monotonically not because good years were selected — funding is credited period by period and does not get given back, so under a "how much has been earned to date" reading the line can only rise or go flat. Years when funding thinned out show up as a flat stretch, not a reversal.
Year by year
Both columns have to be shown together. Given only the scaled column, the 15.898% in 2026 reads as "funding paid more than three hundred percent", when it is 1.590% scaled 10x.
| Year | BTC base | BTC scaled | ETH base | ETH scaled |
|---|---|---|---|---|
| 2026Excluded from the average | 1.590% | 15.898% | 0.960% | 9.603% |
| 2025 | 5.133% | 51.327% | 4.935% | 49.352% |
| 2024 | 11.980% | 119.805% | 13.019% | 130.185% |
| 2023 | 7.842% | 78.421% | 8.238% | 82.379% |
| 2022 | 4.165% | 41.649% | 0.787% | 7.873% |
| 2021 | 30.635% | 306.352% | 37.562% | 375.621% |
| 2020Excluded from the average | 18.765% | 187.646% | 31.581% | 315.810% |
Averages use 10 annual cells (5 years × two symbols). The ugliest cell is 7.873% for ETHUSDT in 2022, and it has not been removed. Figures are a back-test over historical funding rates; future returns may differ.
Methodology: how these numbers are produced
Cumulative means summed, not compounded
The backend can produce either, and the compounded figure looks better. But compounding assumes each funding payment is immediately redeployed, and this system opens positions by manual command with no automatic reinvestment — presenting the compounded number as a track record inflates it. The previous version of this site made exactly that mistake: its hand-written table used compounded values from an older dashboard, which do not reconcile against the same period-by-period data summed.
The average covers only the 5 complete, verifiable years
The average computed in the backend spans seven years, folding in one year with an unknown period count and the current year that is only half done. Writing "annual average" over a set that includes half a year is not a precision problem, it is a methodology problem. The excluded years stay in the table and still count toward the cumulative total — that money did land, and dropping it would be the real omission.
"Base" and "scaled" differ by one multiple
The base column is the funding rate itself; the scaled column is the result after principal is scaled 10x interest-free. The multiple comes from the exchange's own position rules rather than borrowed money, so there is no hourly-interest liability on the account. It scales the size of the principal, not the directional exposure — matched notional on both legs is unchanged.
The ugliest cell stays
ETHUSDT returned 0.787% in 2022 (7.873% after scaling), and it is the most persuasive part of this table. Faced with a table containing only good years, a reader assumes the bad ones were hidden — and then stops believing the good ones too.
Risk disclosure
Figures above are a historical back-test of funding rates recorded period by period, as of 2026-08-25. They describe how funding behaved under past market conditions, are not a forecast of future returns, and are not investment advice. Averages cover only the 5 complete years (2021–2025) that can be verified settlement by settlement; 2026, 2020 remain in the table but are excluded from the average. Funding tracks market sentiment: positive periods are frequent and rich in bull markets, and fall sharply in flat or bear markets — ETHUSDT returned just 7.873% across 2022. The strategy removes directional price risk. It does not remove exchange risk, contract-rule changes, or liquidity and execution risk in extreme conditions.