TradingViewPine ScriptBTC · —

CME Gap-Fill v4.4.1

When the CME session closes and Bitcoin keeps trading, the futures chart reopens away from its previous close, and those gaps have a well-documented tendency to fill. The strategy tracks up to five open gaps at once but does not trade one on sight: it waits for a later candle to print a wick back inside the gap zone, which is the evidence that the fill is actually underway, and only then enters in the direction of the fill — a gap up gets faded short, a gap down gets bought. There is no take profit. The position rides until the stop takes it out, with the stop pulled to break-even once price has covered half the gap and trailing 1% behind the best price from there. Gaps that go unconfirmed for four hours expire untraded, which is what keeps the system out of the gaps that were never going to close.

How it trades

Gap source: CME:BTC1! front-month Bitcoin futures, requested with lookahead disabled
Gap definition: Body-to-body — clear space between the previous and current candle bodies
Minimum gap: 0.15% of price
Confirmation: A later candle must print a wick inside the gap zone
Expiry: Unconfirmed gaps are discarded after 4 hours; up to 5 tracked at once
Direction: Gap up → short, gap down → long
Entry: Market, on the close of the confirming candle
Stop: Distance sized so that being stopped out costs 1% of equity
Target: None — take profit disabled; the stop is the only exit
Break-even & trail: Stop to entry +0.1% at 50% of the gap covered, then trails 1% behind the best price
Costs modelled: 0.04% commission per trade and 5 ticks of slippage
Sizing: Fixed 1 unit per trade — position size does not compound

Performance

Net return: 93.20%
Profit factor: 3.30
Max drawdown: 16.51%
Win rate: 36.05%
Total trades: 233
Expectancy: $40.00
Sharpe: —
Recovery factor: 3.40

Test conditions

Period: 2025-01-01 → 2026-08-28
Initial deposit: $10,000
Risk per trade: 1% of equity per trade
Modelling: TradingView tester, orders on bar close
Spread: 0.04% commission + 5 ticks slippage

How to read these numbers

The gap source is a continuous contract, and that needs checking before the number is quoted anywhere. CME Bitcoin futures are monthly, and CME:BTC1! splices one contract to the next without back-adjusting, so every roll puts a price step into the series that was never a tradeable gap. Roll basis on Bitcoin futures regularly exceeds the 0.15% minimum gap, and roughly twenty rolls fall inside this test window. Some share of the 233 trades may be chasing splice artefacts rather than real gaps.
The equity curve is not compounded. Position size is fixed at one unit per trade for the whole run, so +93.20% is the sum of fixed-size outcomes against the starting $10,000, not a compounding return. The stop distance, by contrast, is recalculated from live equity on every trade — so dollar risk per trade grows with the account while position size never does. The two halves of the risk model are on different footings, which is worth resolving deliberately rather than by default.
The trailing stop is what makes the money, not the gap fill. The average winner is $159 against an average loser of $27, even though the initial stop is sized to lose about 1% of equity. Most losing trades are therefore being cut near break-even by the trail rather than stopped out in full. That is the edge — and it is also the dependency: the result assumes break-even and trailing stops fill at their levels, which on a 24/7 asset with five ticks of modelled slippage is optimistic during a fast move.
In its favour, the result does not rest on a few lucky trades. TradingView attributes −$539 of P&L to outlier trades, meaning the outliers were net negative and the profit comes from the body of the distribution. Both directions are independently profitable (+$4,932 short, +$4,387 long) across 233 trades, and the plumbing is clean — both data requests disable lookahead, entries are gated on confirmed bars, and orders process on bar close, so there is no obvious forward-looking bias.

These are historical backtest results, not live trading returns. Backtests exclude slippage variance, requote behaviour, and swap changes that a live account would experience. Past performance does not guarantee future results.

Equity curve

Account balance and drawdown across every closed trade.

Unmodified TradingView strategy report for CME Gap-Fill v4.4.1

Unretouched TradingView report — the source of the figures above. TradingView's Strategy Tester exports no per-trade journal or equity series on this plan, so the published report is the record rather than a reconstruction.

Trade journal

Every closed position from the test, in order.

TradingView's Strategy Tester exports no per-trade journal or equity series on this plan, so the published report is the record rather than a reconstruction.

← Gold EMA Cross Scalper Chandelier Exit Strategy v3.1 →

CME Gap-Fill v4.4.1

How it trades

Gap source

CME:BTC1! front-month Bitcoin futures, requested with lookahead disabled

Gap definition

Body-to-body — clear space between the previous and current candle bodies

Minimum gap

0.15% of price

Confirmation

A later candle must print a wick inside the gap zone

Expiry

Unconfirmed gaps are discarded after 4 hours; up to 5 tracked at once

Direction

Gap up → short, gap down → long

Entry

Market, on the close of the confirming candle

Stop

Distance sized so that being stopped out costs 1% of equity

Target

None — take profit disabled; the stop is the only exit

Break-even & trail

Stop to entry +0.1% at 50% of the gap covered, then trails 1% behind the best price

Costs modelled

0.04% commission per trade and 5 ticks of slippage

Sizing

Fixed 1 unit per trade — position size does not compound

Performance

Net return: 93.20%
Profit factor: 3.30
Max drawdown: 16.51%
Win rate: 36.05%
Total trades: 233
Expectancy: $40.00
Sharpe: —
Recovery factor: 3.40

Test conditions

Period: 2025-01-01 → 2026-08-28
Initial deposit: $10,000
Risk per trade: 1% of equity per trade
Modelling: TradingView tester, orders on bar close
Spread: 0.04% commission + 5 ticks slippage

How to read these numbers

The gap source is a continuous contract, and that needs checking before the number is quoted anywhere. CME Bitcoin futures are monthly, and CME:BTC1! splices one contract to the next without back-adjusting, so every roll puts a price step into the series that was never a tradeable gap. Roll basis on Bitcoin futures regularly exceeds the 0.15% minimum gap, and roughly twenty rolls fall inside this test window. Some share of the 233 trades may be chasing splice artefacts rather than real gaps.
The equity curve is not compounded. Position size is fixed at one unit per trade for the whole run, so +93.20% is the sum of fixed-size outcomes against the starting $10,000, not a compounding return. The stop distance, by contrast, is recalculated from live equity on every trade — so dollar risk per trade grows with the account while position size never does. The two halves of the risk model are on different footings, which is worth resolving deliberately rather than by default.
The trailing stop is what makes the money, not the gap fill. The average winner is $159 against an average loser of $27, even though the initial stop is sized to lose about 1% of equity. Most losing trades are therefore being cut near break-even by the trail rather than stopped out in full. That is the edge — and it is also the dependency: the result assumes break-even and trailing stops fill at their levels, which on a 24/7 asset with five ticks of modelled slippage is optimistic during a fast move.
In its favour, the result does not rest on a few lucky trades. TradingView attributes −$539 of P&L to outlier trades, meaning the outliers were net negative and the profit comes from the body of the distribution. Both directions are independently profitable (+$4,932 short, +$4,387 long) across 233 trades, and the plumbing is clean — both data requests disable lookahead, entries are gated on confirmed bars, and orders process on bar close, so there is no obvious forward-looking bias.