FuturesAI MACD Filter: How to Check It Survives Fees
Short answer: FuturesAI’s published MACD + Volume Profile backtest won 58.7% of 218 BTC 1H trades in 2024. Its payoff ratio works out to about 1.17. With no costs, the break-even win rate is about 46%. At Bitget’s 0.06% market fee on both sides, it rises to about 53.8%. Add 0.05% slippage and it reaches about 57.0%, close to the 58.7% result.
A 58.7% win rate sounds like plenty of room. It isn’t always. The cushion depends on how big the wins are compared with the losses, and on how much each round trip costs.
This worked example takes the numbers FuturesAI published in its MACD + Volume Profile insight (February 20, 2026) and runs them through Bitget’s fees as listed on FuturesAI’s Bitget payback page. It is not investment advice, and leveraged futures can lose the full principal.
Contents
- What did FuturesAI’s MACD + Volume Profile backtest show?
- What is the break-even win rate before fees?
- How much do fees and slippage raise the break-even?
- How much does filtering signals save in fees?
- How do you run this check on your own strategy?
What did FuturesAI’s MACD + Volume Profile backtest show?
FuturesAI’s February 20, 2026 insight reports that adding a Volume Profile filter to MACD crossovers on BTC 1-hour candles in 2024 cut signals from 624 to 218 and raised the win rate from 41.2% to 58.7%.
| BTC 1H, 2024 (FuturesAI backtest) | MACD only | MACD + Volume Profile |
|---|---|---|
| Total signals | 624 | 218 |
| Win rate | 41.2% | 58.7% |
| Profit factor | 0.94 | 1.67 |
| Sharpe ratio | 0.31 | 1.42 |
| Max drawdown | -28.4% | -11.8% |
| Annual return | -3.2% | +42.8% |
Source: FuturesAI, “MACD + Volume Profile: A Combined Signal Approach,” February 20, 2026. Sample: BTC 1H candles, calendar year 2024.
When I checked the insight page, I noticed it never says whether fees or slippage were deducted. So the safe reading is gross. Everything below adds costs on top.
The combined rules are simple enough to copy: a MACD crossover near a Point of Control or High Volume Node for longs, near a Low Volume Node acting as resistance for shorts, with two or more bars of rising histogram as confirmation.
What is the break-even win rate before fees?
For FuturesAI’s MACD + Volume Profile backtest, the payoff ratio implied by a 1.67 profit factor at a 58.7% win rate is about 1.17, which puts the break-even win rate before any costs at about 46.0%.
Profit factor is total wins divided by total losses. So the average win divided by the average loss equals profit factor times the losing rate divided by the winning rate: 1.67 x 0.413 / 0.587 = 1.175. Break-even is 1 / (1 + 1.175) = 46.0%.
That’s a cushion of about 12.7 points. Sounds comfortable. Now the costs.
How much do fees and slippage raise the break-even?
At Bitget’s 0.06% market fee on entry and exit, a 0.12% round-trip cost raises the break-even win rate of FuturesAI’s MACD + Volume Profile backtest from about 46.0% to about 53.8%, using an average loss of about 0.71% per trade.
That average loss comes from the published numbers. 42.8% over 218 trades is about 0.196% per trade, treating the annual return as a simple sum at 1x full-size positions. Expectancy per trade is 0.587 x 1.175 – 0.413 = 0.277 average losses. Dividing gives an average loss near 0.71% and an average win near 0.83%.
With a round-trip cost c, every win shrinks by c and every loss grows by c. Break-even becomes (loss + c) / (loss x (1 + payoff ratio)).
| Cost per round trip | Round-trip cost | Break-even win rate | Does 58.7% clear it? |
|---|---|---|---|
| Limit both sides, effective (0.005% x 2) | 0.01% | 46.6% | Yes, by 12.1 points |
| Market both sides, effective (0.015% x 2) | 0.03% | 47.9% | Yes, by 10.8 points |
| Limit both sides, before payback (0.02% x 2) | 0.04% | 48.6% | Yes, by 10.1 points |
| Market both sides, before payback (0.06% x 2) | 0.12% | 53.8% | Yes, by 4.9 points |
| Market before payback + 0.05% slippage (assumed) | 0.17% | 57.0% | Yes, by 1.7 points |
| Market before payback + 0.10% slippage (assumed) | 0.22% | 60.2% | No, short by 1.5 points |
Fees from FuturesAI’s Bitget payback page (limit 0.02% / market 0.06% before payback; 0.005% / 0.015% effective). Slippage rows are assumptions, not measured values.
I found that slippage, not the fee, is what closes the gap. At 0.10% of assumed slippage on top of market fees, the backtest’s 58.7% falls short.
The small average move is the reason. When the average loss is only about 0.71%, a 0.12% cost is a big slice of it.
How much does filtering signals save in fees?
Cutting FuturesAI’s BTC 1H 2024 signals from 624 to 218 removes 406 round trips, which at a 0.12% round-trip market fee is about 48.7% of equity a year at 1x full-size positions.
| Cost assumption (1x, full-size) | MACD only: fee drag | MACD only: return after | MACD + VP: fee drag | MACD + VP: return after |
|---|---|---|---|---|
| Market, effective (0.03% per round trip) | 18.72% | -21.9% | 6.54% | 36.3% |
| Market, before payback (0.12%) | 74.88% | -78.1% | 26.16% | 16.6% |
| Market + 0.05% slippage (0.17%) | 106.08% | Below -100% (account gone) | 37.06% | 5.7% |
Simple subtraction from FuturesAI’s published annual returns; ignores compounding and funding payments.
The unfiltered strategy was already slightly negative before costs. With fees it gets much worse. The filter matters twice: better trades, and far fewer of them.
Notice the effective-fee column too. Moving from 0.12% to 0.03% per round trip, the rate FuturesAI lists for Bitget after its discount and 50% payback, takes the filtered strategy’s fee drag from about 26.2% to about 6.5%.
How do you run this check on your own strategy?
To check whether any backtest survives fees, derive its payoff ratio from the win rate and profit factor, estimate the average loss in percent, then add your round-trip fee plus a slippage allowance and recompute the break-even win rate.
- Write down the win rate, profit factor, number of trades and total return from the backtest.
- Payoff ratio = profit factor x (1 – win rate) / win rate.
- Average loss = (total return / trades) / (win rate x payoff – (1 – win rate)).
- Round-trip cost = entry fee + exit fee + a slippage allowance you’re honest about.
- Break-even = (average loss + cost) / (average loss x (1 + payoff ratio)).
- Compare with the backtest win rate. Under 3 points of cushion is thin.
Our earlier walk-through, FuturesAI Strategy Kit: How to Add Real Fees to a Backtest, shows how to add fees inside the strategy kit itself. For the simpler version of this formula, see FuturesAI Payback: Break-Even Win Rate After Fees.
Tips
- Treat any backtest that doesn’t state its fee assumptions as gross.
- Use limit orders where the strategy allows; the round-trip cost drops from 0.12% to 0.04% before payback on Bitget.
- Re-run the math with a slippage figure from your own fills, not a guess.
- Check FuturesAI’s payback page again before relying on the 0.015% effective fee.
Warnings
- The 58.7% win rate is a 2024 backtest on 218 BTC 1H signals, not a live record.
- Average loss and win here are derived estimates from published totals, not reported figures.
- Funding payments on perpetual futures are not included.
- This is not investment advice. Leveraged futures can lose the full principal.
FAQ
Does FuturesAI’s MACD + Volume Profile backtest include trading fees?
FuturesAI’s insight page, dated February 20, 2026, does not say whether fees or slippage were included in its BTC 1H 2024 results. Treat the 58.7% win rate and 42.8% annual return as gross until you add your own costs.
Why does filtering out 65% of MACD signals matter for fees?
Every futures round trip pays a fee on the way in and again on the way out. Cutting signals from 624 to 218 in FuturesAI’s BTC 1H 2024 backtest removes 406 round trips, which at a 0.12% round-trip market fee is about 48.7% of equity at 1x full-size positions.
Is a 58.7% backtest win rate the same as a live win rate?
No. A backtest replays history with fixed rules, while a live record includes real fills, slippage and changing markets. Use the backtest to test the math, and check live results separately before risking money.
Sources
- FuturesAI: MACD + Volume Profile: A Combined Signal Approach (Feb 20, 2026)
- FuturesAI: Bitget payback (fee before and after payback)
Next step: take the last backtest you trusted, run the six steps above with your real fee tier, and write down the cushion. If it is under 3 points, test with smaller size first. For a risk-adjusted view, see FuturesAI Sharpe Ratios: How to Recompute Them After Fees.
Figures checked against FuturesAI pages as of October 30, 2026. Not investment advice; leveraged trading can lose the full principal.
🔗 Related guides
