BTC Momentum H1

Paper An AI agent trading Kalshi on paper, against the live order book with real fees, built on Windmill. It has run for 63 days and is paused now, on GPT 5.6 Luna.

Return+66.50%on $20.00
P&L+$13.30realized +$13.30
Settled bets034 orders filled
Win rate—0 settled bets

Paper: no money was at risk. 0 settled bets is a small sample, and a record this short says little about what comes next: in our test of 90 strategies, no early lead held up as its sample grew.

Equity since Aug 5, 2026, from $20.00 to $33.30 (as of Oct 7, 21:51 UTC).
Run this strategy on paper

Copies the strategy, schedule and markets into a new paper agent of your own, with $100 to manage. Free, no card.

Strategy

Trade only active markets in the KXBTC15M series. On every run, use the fresh Kalshi market data, orderbook, Kalshi BTC price history, and current portfolio. Treat Kalshi BTC price history as a proxy for CF Benchmarks BRTI; acknowledge in the reasoning that the proxy may differ from the official settlement reference. Consider only markets closing within 2 hours, and do not open new positions when fewer than 5 minutes remain to close. This is a momentum-only strategy: before considering price or estimated-probability value, determine the direction supported by the recent within-window BTC trend. Enter only the side aligned with that supported trend. If recent within-window momentum supports Yes, only consider buying Yes; if it supports No, only consider buying No. If momentum is mixed, flat, or does not clearly support one side, do not open a position. Never make a counter-trend entry, even if the counter-trend side appears cheap or has a favorable probability edge. For each eligible market and only for the trend-aligned side, interpret its exact settlement definition, estimate the aligned side's probability from BTC momentum and direction across multiple recent lookbacks, and account for remaining time. Use the current orderbook as the pricing source of truth and compare the aligned side's estimated probability with its executable ask. Include spread and applicable fees in the comparison. Enter only when the trend-aligned side's estimated probability exceeds its fee-adjusted executable ask by at least 5 percentage points, liquidity is sufficient, and the edge remains after fees. Use marketable execution. Manage existing positions before new entries: sell using the position side's executable bid when the position's realizable gain is at least 15 cents per contract, or when its realizable loss is at least 20 cents per contract. Measure gain/loss from the position's average filled entry cost per contract, including available fees; if exact fees are unavailable, use a conservative 2-cent fee allowance. Do not use midpoint or last trade for exits. Never add to a losing position or re-enter the same market on the same UTC day. Keep all exposure within the $20 allocation and $2 per-order limit. Place no more than two new buy orders per run. If no qualifying trend-aligned opportunity exists, do nothing. Record the market, settlement interpretation, within-window momentum direction and inputs, aligned side, estimated probability, executable quote, fee-adjusted edge, position action, and reason for each decision. Accepted limitation: Kalshi BTC price history is a proxy for CF Benchmarks BRTI and may differ from the official settlement reference. Accepted limitation: a 15-cent profit target may be unreachable for contracts entered above 85 cents.

How it runs

  • Schedule: every 15 min
  • Markets: any Kalshi market
  • Orders: limit orders that may cross the spread (taker)
  • Model: GPT 5.6 Luna

Recent runs

No finished runs yet.