ETH/BTC Perp Momentum Engine

⚠ Not financial advice. Educational / simulation tool. Trading crypto perpetuals with leverage carries substantial risk of loss, up to and including your entire balance. Past or simulated performance does not indicate future results. You are solely responsible for any trades you place, manually or via this tool's automation.

Automation (optional)

BTC funding — ETH funding — Source Binance USDT-M Futures

Momentum readout — long the stronger asset, short the weaker one

Loading…

Fetching live perp candles.

Score gap (ETH − BTC)

—

Signal confidence

Candles used

—

BTC/ETH correlation (50 bar)

—

ETH/BTC ratio (perp mark)

—

Connect Bybit Demo Trading

Optional — turns engine actions into real orders against Bybit's live market price with virtual balance. Leave disconnected to stay in local simulation.

Not connected

Demo Trading only. In Bybit, log in → switch to Demo Trading (top right, it's a separate account with its own user ID) → API → create a key scoped to that demo account. Never paste a real-money mainnet key here — this page has no secure key storage, it lives in this browser tab's memory only.

Demo Trading API key

Demo Trading API secret

Trading engine

Local simulation · connect above to place real orders on Bybit Demo

Volume / scalp mode — tight take-profit, tight stop-loss, DCA off. Built for frequent trades aiming for a small clean win each time instead of riding one big move. Fee cost matters a lot here — this is designed for a zero/near-zero fee venue (Variational, Lighter), not Bybit's 0.055%.

Signal mode

Momentum bets a leading asset keeps leading. Mean-reversion bets the opposite — that a stretched ETH/BTC ratio corrects back toward its own average.

Z-score lookback (bars)

Window used to compute the ratio's own average and spread. Mean-reversion mode only.

Z-score entry (σ)

How many standard deviations the ratio must stretch before betting on reversion.

Direction

Auto lets the selected signal mode pick the direction. Forcing a direction skips it entirely — DCA, TP, and SL still apply normally.

Score diff threshold

Gap needed between BTC/ETH scores before a trade opens. Higher = pickier, fewer trades.

Entry margin / add ($)

Dollars committed per entry and per DCA add.

Max DCA entries / leg

Most times it will add to a losing position before it stops adding.

DCA trigger (%)

How far price must move against you before it adds another entry.

Take profit (% combined)

Closes everything once combined profit reaches this %.

Stop loss (% combined)

Closes everything once combined loss reaches this %.

Leverage

Multiplies position size — and risk — relative to margin used.

Starting balance ($)

Virtual balance for local simulation only.

Signal persistence (bars)

How many scans in a row must agree before it acts. 1 = acts immediately (off). Higher = slower but steadier.

Min regime strength (0–1)

Only trades when the market is trending, not choppy. 0 = off. Higher = stricter.

Min BTC/ETH correlation (0–1)

Pauses new trades if BTC and ETH stop moving together. 0 = off.

Taker fee per fill (%)

Exchange fee per order. Only change this to match your real fee tier.

Vol-adjusted leg sizing

Sizes ETH's leg by its volatility vs BTC's, instead of equal dollars on each side.

Balance (free)

$50,000

Equity

$50,000

Unrealized PnL

$0 (0%)

Realized PnL

No activity yet — start the engine to begin simulating.

Backtest

Replays the exact signal + DCA + TP/SL rules above against real historical price. Does this strategy actually work?

Uses the config values above and the interval selected on the chart. Needs at least 200 candles — pair short lookbacks (4h, 1 day) with a fine interval like 1m or 5m. Not run yet.

—

Every scan (default 60s): fetch perp candles for BTCUSDT and ETHUSDT → compute momentum scores → if no position is open and the score gap exceeds the threshold, open the stronger asset long and the weaker one short with the configured margin on each leg. While a position is open, each scan checks combined unrealized PnL against take-profit/stop-loss first; then checks whether momentum has flipped to the opposite asset (if so, closes everything and opens fresh in the new direction); then checks each leg for a DCA add — if a leg is down the configured trigger % from its last fill and hasn't hit the max entry count, it adds another entry at the current price, pulling the average price toward the market. All fills are simulated against the live Binance perp mark price — nothing is sent to an exchange. DCA-into-a-loser is a martingale-style approach: it improves average price on a rebound but increases exposure the further price moves against you, so the stop-loss is what caps worst-case loss.

Signal data: Binance USDT-M Futures public REST API. Execution (when connected): Bybit Demo Trading, which fills against Bybit's real mainnet price feed with a virtual balance. Requests are signed locally in your browser with the API secret you provide — that secret and everything else on this page is held only in this tab's memory and is gone on refresh; nothing is sent anywhere except directly to Bybit's API.
In local-simulation mode this does not place real orders and does not model fees, slippage, liquidation, or funding. In connected mode, orders are real orders against Bybit's live mainnet price feed and matching engine — the balance is virtual (Demo Trading) but fills, fees, slippage and liquidation risk on the position are real. Nothing here is investment advice.

ETH/BTC — Perp Momentum Engine

ETH/BTC ratio (perp mark)

Connect Bybit Demo Trading

Trading engine

Backtest

Terms & risk disclosure