FuturesAI Scoreboard: Does +0.01% a Call Survive Fees?
Short answer: FuturesAI’s scoreboard showed an average return of +0.01% per graded prediction on November 10, 2026, across 313 calls. A single round trip costs between 0.010% and 0.088% of position size on the three exchanges on its payback page. Every fee tier is bigger than the average gain. Measured the way the scoreboard measures it, the typical call does not cover its own fees.
Plus 0.01% sounds like a rounding error, and it is.
The scoreboard also says plainly that fees are not subtracted from that number. So the real question is what is left once you pay them. I ran the sum for six fee tiers, and the sign flips every time.
This is a calculation, not a verdict on the service. If you want the break-even hit-rate view of the same data, see FuturesAI Scoreboard: Is 45.5% Above Break-Even? and the stop-distance version in FuturesAI Win Rate: What Fees and Slippage Do to Break-Even.
Open the FuturesAI payback page
Educational content, not investment advice. Leveraged trading can lose the full principal. Past hit rates do not guarantee future returns.
Jump to a Section
- What does the +0.01% average actually mean?
- How much does one round trip cost on each exchange?
- What is the average call worth after fees?
- Does leverage change the answer?
- What does this calculation leave out?
- How do you rerun this on your own numbers?
What does the +0.01% average actually mean?
The FuturesAI scoreboard, checked on November 10, 2026, reports an average return of +0.01% per graded prediction, calculated by assuming every one of 313 predictions was traded with the same amount, bought at the price when the call was recorded and sold at the grading price.
When I checked the page, the other headline numbers sat right next to it. The AI hit rate was 46.0% (144 of 313). The payoff ratio was 1.18. Up calls averaged +1.70% across 193 predictions and down calls averaged −2.71% across 120.
Add all 313 results together and the page shows about +2.0%. Divide that by 313 and you get roughly +0.0064% per call. That is the unrounded figure I use below, because rounding to 0.01% would flatter the result by more than half.
The page states one more thing that matters here: fees are not deducted.
How much does one round trip cost on each exchange?
On FuturesAI’s payback page as of November 10, 2026, the effective fee after discount and payback is 0.005% (limit) and 0.015% (market) at Bitget, 0.016% and 0.040% at OKX, and 0.016% and 0.044% at Bybit, so a full round trip of entry plus exit costs 0.010% to 0.088% of position size.
A round trip means two orders, so I doubled each rate. Both legs use the same order type in this table. Mixed limit-in and market-out trades land between the two columns.
| Exchange | Limit, one side | Limit round trip | Market, one side | Market round trip |
|---|---|---|---|---|
| Bitget | 0.005% | 0.010% | 0.015% | 0.030% |
| OKX | 0.016% | 0.032% | 0.040% | 0.080% |
| Bybit | 0.016% | 0.032% | 0.044% | 0.088% |
These are the effective rates the page shows in brackets after applying fee discounts and payback. If you trade without the affiliate link, your headline rate is higher and every number below gets worse.
What is the average call worth after fees?
An average FuturesAI scoreboard call earning about +0.0064% before fees nets between −0.0036% and −0.0816% per round trip after the fees above, which adds up to between −1.13% and −25.54% across the 313 graded predictions.
The table subtracts each round-trip cost from +0.0064%, then multiplies by 313 to show the whole sample. It uses the page’s own +2.0% total as the starting point.
| Fee tier (round trip) | Cost per call | Net per call | Net over 313 calls (summed) |
|---|---|---|---|
| Bitget limit | 0.010% | −0.0036% | −1.13% |
| Bitget market | 0.030% | −0.0236% | −7.39% |
| OKX limit | 0.032% | −0.0256% | −8.02% |
| Bybit limit | 0.032% | −0.0256% | −8.02% |
| OKX market | 0.080% | −0.0736% | −23.04% |
| Bybit market | 0.088% | −0.0816% | −25.54% |

The best case, Bitget with limit orders both ways, still lands slightly under zero. It is the closest to flat, and I would not call that an edge.
Does leverage change the answer?
Leverage does not change the sign of a futures trade’s expected result after fees: at 10x, the +0.0064% gross average and the 0.030% Bitget market round trip both scale by 10, so the net stays negative at about −0.236% of margin per call.
It changes the size of the swings. Leverage multiplies gain, fee and loss alike, and the scoreboard’s average losing call is −2.71% on the down side, which at 10x is a 27.1% hit to margin before fees.
Honestly, this is the part people skip. Higher leverage makes a thin average look more exciting while the fee drag grows in the same proportion.
What does this calculation leave out?
This calculation leaves out slippage, funding payments and the gap between the scoreboard’s buy-at-recorded-price assumption and the price you would actually get, so a real account can come out better or worse than the figures above.
- It treats every prediction as a full-size trade with equal money, as the scoreboard does.
- It ignores funding on perpetual contracts, which matters for crypto holds that run for days.
- Stocks and crypto sit in one pool. I noticed that the page splits them: domestic stocks 40.3% (52 of 129), US stocks 45.9% (51 of 111), crypto 54.5% (48 of 88). Crypto is the only market above 50%, and it still trailed its own market benchmark in that window.
- 313 calls over about seven weeks is a small sample. A swing of a few calls moves the average more than any fee tier does.
How do you rerun this on your own numbers?
To check any signal against fees, take its average return per trade before costs, subtract twice your one-side fee rate, and multiply by the number of trades, using the same units for all three.
- Write down the average gross return per trade, unrounded if the source lets you.
- Find your effective fee for the order type you really use, after discounts and payback.
- Double it for the round trip.
- Subtract. If the result is negative, the signal has to improve before it is worth trading.
Try it with the tight-stop rows from the earlier posts and you will see the same pattern. The cheaper your fee, the more of a thin edge survives.
Tips
- Use the unrounded average. A +0.01% label hides anything between +0.005% and +0.015%.
- Price limit and market orders separately. The gap is more than 2x at OKX and Bybit.
- Save the date you read the scoreboard. The numbers move daily.
Warnings
- A scoreboard average is not a track record of anyone’s real account.
- Fee rates and payback shares can change. Check the live page before you rely on a rate.
- Leverage can lose the full principal. Nothing here is a promise of profit.
FAQ
Does the FuturesAI scoreboard include trading fees?
No. The page says fees are not deducted from the per-prediction return. It also notes that the market index it compares against uses closing prices, so a day of timing difference is possible.
Why use +0.0064% instead of +0.01%?
The page rounds the average to +0.01% but also shows a summed result of about +2.0% over 313 calls. Dividing the two gives about +0.0064%, which is a more careful starting point for fee math.
Would the result be positive with a bigger sample?
Nobody can say from this data. Seven weeks and 313 calls is too few to separate skill from luck, and the average could move either way as more predictions are graded.
Sources
- FuturesAI, AI prediction scoreboard (read November 10, 2026)
- FuturesAI, exchange fee payback and effective fee rates (read November 10, 2026)
๐ Related guides
Not investment advice. Figures are as shown on FuturesAI’s public pages on November 10, 2026 and change over time. Fact-checked based on public sources as of November 10, 2026.
