TurbineMangrove
POWERED BY TURBINEFI
All StrategiesResearchCommunity
Pricing
Kalshi TradingPolymarketAI Strategy BuilderBacktesting EngineSandbox RuntimeEdge Data FeedsStrategy LibraryBacktested StrategiesLive Performance
BlogXDiscord
AffiliatesMerchDocs
Pricing
← Strategies
← Strategies

Two-Sided Spread Capture

Starter is $19/month with 10 deploys per week. Upgrade as you grow.

This strategy operates in the KXBTC market on Kalshi, checking every 30 seconds. When the gap between the best buy and sell prices reaches 5 cents, it places up to three maker orders on each side to capture that difference, but only within prices between 5 and 95 cents. The total contracts held are capped at 52, and it only adds orders as a maker, refreshing them after any fill.

Kalshi·May 12 to Jun 11·Created 3mo ago
Share to XShare to Reddit
Turbine Studio

Your next trade is a sentence away.

Describe an idea, review the backtest, and start from the same workspace.

Build a bot
Net P&L
+$1,448
Return +2785.1% on risk capital
Sharpe
2.28
Return Sharpe
Win rate
100.0%
41160 trades
Max drawdown
-$44
Peak to trough

What this backtest shows

Over the May 12 to Jun 11 window, this spread capture strategy on Kalshi turned in +$1,448 of simulated profit (+2785.1% on its configured risk capital), at a 2.28 Sharpe. It placed 41160 simulated trades and won 100.0% of them — a high hit rate — against shallow worst peak-to-trough drawdown of -$44.

Under the hood it simulated 704 Bitcoin (BTC) markets, closing 392 winning and 0 losing positions after $321 in modeled fees, an average of 1372.0 trades a day. That trade-by-trade detail, the equity curve above, and the full rule set below are what separate this page from a one-line leaderboard entry.

Net PnL is the headline here; the Sharpe is unannualized over this short window, so read it as a within-sample texture of the equity curve rather than an industry-standard risk score. Because every figure comes from a single 30-day historical replay, it is best treated as a hypothesis to pressure-test rather than a forecast — the same rules can behave very differently once live fills, API latency, and shifting volatility enter the picture.

Market and data window

This backtest runs against Bitcoin (BTC) markets on Kalshi's hourly series across 30 days (May 12 to Jun 11). These are short-horizon contracts that open and settle on a fixed hourly cadence, so the strategy is measured across many independent events rather than one long trend. Rules are evaluated once per hour, and a signal can fill no earlier than the next tradable candle at top-of-book prices, net of Kalshi-style taker fees.

Related backtests

Compare other Bitcoin (BTC) hourly strategies backtested on Turbine:

  • Kalshi KXBTC Spread Capture — 2.89 Sharpe, +$1,551 simulated PnL
  • Kalshi KXBTC Spread Capture — 2.84 Sharpe, +$1,355 simulated PnL
  • Kalshi KXBTC Spread Capture — 2.88 Sharpe, +$939 simulated PnL

A few notes on these numbers

Methodology
Backtests replay supported historical market data, usually a 30-day Studio window, and evaluate strategy rules on the available cadence. Signals can fill no earlier than the next tradable step. Taker fills use top-of-book prices or available L2 depth, with Kalshi-style fees where modeled. Queue position, latency, and future liquidity are not fully modeled. Simulated P&L is net of modeled fees over the window.
Starting capital
Kalshi backtests use the strategy's configured max position as the risk-capital denominator. Return is computed from that configured capital, while total P&L is still shown in dollars.
Past performance
Backtested results are hypothetical. They are derived from historical data and do not reflect live execution risks like slippage beyond the orderbook, partial fills, API latency, or venue outages. Past performance does not guarantee future results.
Not financial advice
Nothing on this page is investment, legal, or tax advice. Prediction markets carry real risk of loss. Only trade with money you can afford to lose, and do your own research before deploying any strategy live.