Trade-Strategy.comSignal Desk for BTC/USDCBTC/USDC backtest - 1-hour, 2026-02-23 to 2026-08-08
BTC Intraday Regime Pulse
Intraday engine reading the 1h regime, 4h structure, volume anomalies, taker flow and breakout quality.
Key observations
- The window covers 167 days (2026-02-23 to 2026-08-08). BTC/USDC moved from 66,686.2 to 64,880.3, a -2.71% change over the period.
- The engine opened 20 long and 21 short positions - 1.7 trades per week on average - with a mean holding time of 0.9 hours.
- The best trade gained +1.57R (+162.00 USDC) on 2026-06-01; the worst lost -17.14R (-24.00 USDC) on 2026-07-03.
- The deepest equity drawdown reached -381.00 USDC between 2026-07-05 and 2026-07-13.
- Longs closed in profit 80% of the time; shorts 66.7%.
Backtest chart
Last 236 1-hour candles with entries (arrows) and exits (circles - green win, red loss).
Monthly breakdown
| Month | Trades | Win rate | PnL | Total R |
|---|---|---|---|---|
| 2026-07 | 13 | 61.5% | +345.00 USDC | -17.43R |
| 2026-06 | 7 | 100% | +900.00 USDC | +7.70R |
| 2026-05 | 12 | 66.7% | +723.00 USDC | -0.47R |
| 2026-04 | 9 | 77.8% | +3,288.00 USDC | +4.20R |
Recent trades
| Date | Side | Entry | Exit | Result | PnL | Exit reason |
|---|---|---|---|---|---|---|
| 2026-07-27 | SHORT | 64,647.6 | 64,142 | +0.90R | +210.00 USDC | stop |
| 2026-07-26 | LONG | 65,129.5 | 65,144.19 | -2.55R | -18.00 USDC | stop |
| 2026-07-23 | SHORT | 65,452.7 | 65,384.42 | +0.23R | +6.00 USDC | stop |
| 2026-07-20 | LONG | 64,899.9 | 65,022.74 | +0.58R | +33.00 USDC | stop |
| 2026-07-20 | SHORT | 64,181.6 | 63,780.09 | +1.57R | +162.00 USDC | target |
| 2026-07-17 | LONG | 63,895 | 64,233.2 | +0.88R | +126.00 USDC | timeout |
| 2026-07-17 | SHORT | 63,359.2 | 63,093.23 | +0.81R | +102.00 USDC | stop |
| 2026-07-13 | SHORT | 62,758.4 | 62,940.23 | -1.29R | -111.00 USDC | stop |
| 2026-07-06 | LONG | 64,258.9 | 63,857.87 | -1.14R | -213.00 USDC | stop |
| 2026-07-06 | SHORT | 62,730.6 | 62,797.96 | -1.75R | -57.00 USDC | stop |
| 2026-07-05 | LONG | 63,153.7 | 63,352.58 | +0.75R | +69.00 USDC | stop |
| 2026-07-04 | LONG | 62,796.1 | 62,973.54 | +0.72R | +60.00 USDC | stop |
| 2026-07-03 | LONG | 62,353.9 | 62,350.81 | -17.14R | -24.00 USDC | stop |
| 2026-06-22 | LONG | 64,675.1 | 64,832.73 | +1.21R | +48.00 USDC | target |
| 2026-06-17 | SHORT | 65,263.1 | 65,174.39 | +0.41R | +18.00 USDC | stop |
| 2026-06-15 | LONG | 66,564.9 | 66,827.82 | +0.80R | +93.00 USDC | stop |
| 2026-06-14 | LONG | 64,053.1 | 64,921.74 | +0.94R | +384.00 USDC | stop |
| 2026-06-01 | SHORT | 71,401.8 | 71,220.23 | +1.23R | +51.00 USDC | target |
| 2026-06-01 | SHORT | 72,591.2 | 72,141.04 | +1.57R | +162.00 USDC | target |
| 2026-06-01 | SHORT | 73,197.6 | 72,785.38 | +1.54R | +144.00 USDC | target |
Why these parameters
This engine treats BTC/USDC like a regime machine. It first classifies the current state of the market on the 1h timeframe - trending up, trending down or consolidating - using an EMA 50 filter and the 4h price structure. It then watches for volume anomalies (a signal candle at least one standard deviation above average volume) and taker-flow imbalance (threshold 0.12) to time breakouts out of that regime.
Only setups that score at least 65 points on the engine's breakout-quality score produce a signal, and the strategy runs in global position mode: one position at a time with a 15-minute cooldown after every exit. The 1h timeframe keeps the sample meaningful while avoiding the noise of minute-level data.
Default execution parameters: 30,000 USDC position notional at 3x leverage (10,000 USDC capital), 24 bars maximum holding, and transaction costs of 0.04% per side plus a 0.03% daily funding charge.
How this backtest works
This backtest replays the BTC Intraday Regime Pulse strategy over BTC/USDC perpetual candles between 2026-02-23 to 2026-08-08 - the most recent 1-hour candles available in the market data feed. The simulation runs server-side and every entry decision uses only information that existed at the signal time, so there is no lookahead bias in the results.
Positions are sized at 30,000 USDC notional with 3x leverage (10,000 USDC capital) for the base variants, or at 2% take-profit / 1% stop-loss of that deployed position for the standard TP/SL variants. Exit rules execute on the next candle after the signal, and a trade closes when the target or stop is hit, when the maximum holding time expires, or when the data window ends.
Transaction costs are modeled as 0.04% on entry and 0.04% on exit, plus a 0.03% daily funding charge for positions held overnight. The 0.08% round-trip cost is realistic for a retail perpetual-futures account on major venues.
Every trade produced by the engine is shown on this page - nothing is filtered or excluded. The trade table lists the most recent 20, the chart marks every trade in the window, and the monthly table aggregates all of them.
Reading the results
Win rate
73.17% of the 41 trades closed in profit (30 winners of 41). With a 1.8R target multiple the breakeven win rate is about 36%. This result sits 37.2 percentage points above that breakeven point.
Profit factor
A profit factor of 0.82 means the engine lost more than it made in R terms, which usually signals that a few large losses overwhelmed many small wins.
Average R and total R
Each trade is measured in R - multiples of the initial stop distance. The average trade on this page returned -0.15R, and all trades together accumulated -6.00R. A positive total R over a window of 167 days is the most honest single summary of whether the edge held up.
Max drawdown
The equity curve fell -19.85R from peak to trough in R terms (2026-06-22 to 2026-07-13) and -381.00 USDC in USDC terms. Drawdown is the price of every strategy - the question is whether the recovery took days or months.
PnL vs R
R multiples scale profit against the distance to the stop, while USDC PnL is the fixed-notional result after costs. The two can diverge when stops are wide (a big stop makes each R worth more USDC), which is why both are shown - the R curve is the fairer comparison across different stop distances.
Tips for using this backtest
- Volume and taker-flow spikes are this engine's fuel - signals are most meaningful during liquid sessions (London/NY overlap) where the underlying data is clean.
- The 15-minute cooldown is part of the edge: it stops the engine from re-entering a breakout that just failed.
- A 166-day window is a sample, not a promise - review the live track record in the app before committing funds.
Limitations and risks
- The window is a sample, not a promise: results cover 2026-02-23 to 2026-08-08 and shift as new candles arrive and older ones drop out of the window.
- Slippage is not modeled beyond the 0.04% per-side cost assumption - live fills on fast moves can be worse than assumed.
- The engine and its default parameters were developed on past data. A change in BTC market regime (volatility, funding rates, liquidity) can invalidate an edge that worked in this window.
- Funding is charged as a flat 0.03% per day of holding; actual funding varies continuously with market conditions and can be higher in crowded trades.
- This is educational software. Nothing on this page is a recommendation to buy, sell or hold any position, and past performance never guarantees future results.
Other versions of this backtest
Other strategies on 1-hour candles
FAQ
What does "regime" mean here?
The engine classifies the 1h market state (trending up, trending down or consolidating) using an EMA 50 filter and the 4h price structure, then only takes breakouts that score at least 65 points on its quality score.
Why is there only one position at a time?
The strategy uses global position mode - the whole signal budget goes into one position and a 15-minute cooldown follows every exit. That limits overtrading on repeated breakout failures.
What is the volume Z-score threshold?
It is set to 1: a signal candle's volume must be at least one standard deviation above its average. The taker-flow threshold of 0.12 additionally requires a clear imbalance between aggressive buying and selling.
What does the profit factor mean?
Profit factor divides gross wins by gross losses, measured in R multiples. Above 1 the strategy made more than it lost; a value like 10+ usually reflects a small number of losses rather than a huge win rate.
What costs are included?
The simulation applies 0.04% per side (buy and sell) plus a 0.03% daily funding charge for positions held overnight - 0.08% round trip plus funding on longer holds.
How is this backtest calculated?
The server replays the strategy over the last 1-hour BTC/USDC candles (2026-02-23 to 2026-08-08). Entries only use information available at signal time - no future data - and the full trade log is listed on this page.
Why does the result differ from the live app?
This page uses the strategy's default settings over a fixed historical window. The app runs live signals with your chosen settings, and results change as new candles arrive.
Is this investment advice?
No. Backtests are historical simulations with fixed parameters. Past results do not guarantee future performance, and Trade-Strategy.com is educational software.
Backtests are historical simulations of past BTC/USDC candle data with fixed parameters. They do not guarantee future results and are not investment advice.