What is a Bitcoin signal backtest?

A signal backtest replays a trading rule against real historical BTC/USD price data and reports how it would have performed. This page runs the btclyzer STANDARD and ENHANCED algorithms on actual daily, weekly and intraday candles, then reports total return, win rate, maximum drawdown, Sharpe ratio and the equity curve vs a passive buy & hold benchmark — so you can see whether the rating logic would have outperformed simply holding BTC.

What is the difference between STANDARD and ENHANCED in the backtest?

STANDARD is the free multi-factor algorithm: RSI, MACD, EMA 20/50/200, Bollinger %B, Stoch RSI, volume, Fear & Greed Index and CBBI fused into a single BUY / SELL / HODL rating per timeframe. ENHANCED is the PRO version of the same algorithm boosted by Trap line detection (volatility-trap pattern recognition) and weekly COT (CFTC) institutional positioning data. The backtest lets you compare both side by side on identical history.

Which timeframes can I backtest?

1H, 4H, 1D, 1W and 1M — the same five timeframes used by the live dashboard. Each timeframe uses its own weighted rule set (mean-reversion indicators weigh more on 1H/4H; trend indicators weigh more on 1D and above), so the backtest reflects exactly how the rating would have behaved at that timeframe.

Does the backtest include fees, slippage or taxes?

No. The backtest reports the pure rule-based outcome — no exchange fees, no spreads, no slippage and no capital-gains tax. Real-world results will differ because of those frictions. For a realistic estimate, subtract approximately 0.1–0.5 percent per round-trip trade from the reported return.

How is the equity curve constructed?

Starting from an initial $10,000 position, the backtest flips between fully invested in BTC (rating = BUY) and fully in USD (rating = SELL); HODL keeps the previous state. Each rating transition is filled at the candle close of the bar that produced the signal, with no leverage. The buy & hold benchmark is the same $10,000 placed into BTC at the first candle and held through the last candle, used as the comparison line on the equity chart.

What does the Sharpe ratio mean here?

Sharpe ratio measures risk-adjusted return — how much excess return the strategy delivered per unit of volatility. Higher is better. A Sharpe above 1 is generally considered good, above 2 is excellent. Because Bitcoin itself is volatile, even buy & hold often shows a Sharpe below 1; a strategy that beats buy & hold on Sharpe is doing meaningful risk management, not just chasing returns.

Is this investment advice?

No. Past performance does not guarantee future results. The backtest is an educational tool that illustrates how an algorithmic rule would have behaved historically — it is not a recommendation to trade. Cryptocurrency trading carries substantial risk of loss. Always do your own research.

Free tool

Bitcoin Signal Backtest

Replay the btclyzer STANDARD and ENHANCED algorithms on real Bitcoin history. Pick a timeframe and window — see total return, win rate, max drawdown and equity curve vs buy & hold. Long-only strategy: BUY enters, SELL exits to cash.

How the backtest works

The backtest replays the exact same computeSignal() / computeEnhancedSignal() used by the live dashboard, candle-close by candle-close. There is no curve-fitting and no "tuned for backtest" variant — what you see is the real algorithm running on real history.

Strategy

BUY on a closed candle → enter 100% of equity long at that candle's close price.
SELL on a closed candle while in a long position → exit to cash at that close.
HODL (rating not changed) → keep whatever position you have.
No fees, no slippage. Both the algo curve and the HODL curve are computed identically so the relative outperformance stays meaningful.

Metrics

Algo return — total % gain or loss of $1 invested at the start of the window following the algo's signals.
vs HODL — algo return minus the buy-and-hold return over the same window.
Win rate — share of closed trades that exited in profit.
Max drawdown — worst peak-to-trough equity drop during the window.
Sharpe — mean trade return / std dev of trade returns, annualised by the timeframe's candle frequency. Good for ranking variants; not a portfolio-management number.

Frequently asked questions

What's the difference between "trades" and "flips"?

A trade is one closed long position — BUY at one candle close, SELL at a later one. A flip is any signal change (BUY↔HODL↔SELL). A typical BUY → HODL → SELL → HODL → BUY → SELL cycle is 2 trades but 4 flips. The dashboard's history tab lists every flip, so it will normally show roughly 2× the trade count seen here.

When will ENHANCED backtest be available?

ENHANCED algorithm backtest is coming soon. ENHANCED adds Trap-line detection (price action) plus the weekly CFTC COT institutional-positioning index on top of STANDARD — we're finishing a tighter pass on the historical-data plumbing before exposing the backtest. Live ENHANCED runs on the dashboard right now for PRO users.

Are fees and slippage included?

No. Both the algo curve and the HODL curve are computed without fees or slippage so their direct comparison stays honest. Real execution costs depend on your exchange and order type — a reasonable rule of thumb is to subtract 0.1–0.3% per round-trip trade from the algo result.

Why don't shorter timeframes always show more profit even when win rate is high?

1H and 4H produce many more signals than 1D / 1W. Each trade is exposed to slippage and to the algorithm's imperfect signal-to-noise ratio at the shorter horizon. High win rate with small per-trade gains can be wiped by a single deep drawdown trade. The dashboard's free tier broadcasts 4H precisely because it sits in the sweet spot.

Why does my backtest result match someone else's exactly?

Results are cached server-side for 1 hour per (timeframe, algorithm, window) tuple — so a popular configuration only triggers one real run per hour. The signal logic is deterministic given the same kline data, so two users running the same config see the same numbers down to the decimal.

How far back can I run the backtest?

Up to 365 days for any timeframe. Beyond that the 1H/4H kline windows would require multiple page-fetches from Bybit and the marginal value is small — by then you're testing a regime the live algorithm wasn't even running through.

Is this financial advice?

No. Past performance does not guarantee future results. btclyzer is an educational technical-analysis platform. Cryptocurrency trading involves substantial risk of loss.

Early access · limited spots

Become a tester — get PRO free for life

btclyzer is pre-launch. The first testers who try it and send honest feedback keep PRO for life — no card, no catch.

Watch the algorithm live

The backtest shows what the algorithm would have done — the live dashboard shows what it's doing right now, across all five timeframes, free.

Launch the dashboard →