BTC Entry Bands

Kalshi KXBTC15M Entry-Band Floors, Run to Settlement

Short disclaimer: This is a historical simulation write-up, not investment advice. Past simulated results do not predict future outcomes. Nothing here is a recommendation to trade.

Intro / thesis

The KXBTC15M market on Kalshi is a 15-minute Bitcoin contract: you take a side, and the contract settles against whether spot finishes above or below the strike. The strategy family being tested here is a strike-distance ladder — it looks for moments when Bitcoin spot is meaningfully on one side of the strike, momentum agrees, and the corresponding side of the contract is already trading rich.

The specific question in this batch was narrow and practical: does the floor of the entry ask band matter? If we only buy when the contract is already priced at or above some minimum (0.85, 0.89, 0.93, or 0.95), do we get a meaningfully different result once we remove the 0.995 take-profit rules and let every position run to settlement?

Four variants were built and completed, one per floor. Each candidate has its own complete rule set, market selection, and 30-day window; the base DSL is context only. Each successful variant from the batch is saved as a runnable Turbine strategy, so any of the four can be picked up and re-run as-is.

Variant and strategy explanation

All four variants share the same skeleton, and only the entry ask floor changes. The shared engine:

• Venue and market: Kalshi, series KXBTC15M, most_liquid selection. Evaluation loops every 10 seconds.
• Position and price limits: max_position 50 contracts, price floor 0.05, price ceiling 0.99.
• Edge input: Coinbase BTC-USD with price, change_5m, velocity_1m, and ema_26_15m, refreshed every 5 seconds. Two derived values are computed: spot_vs_strike (signed distance of spot from the strike) and ema_vs_strike (signed distance of the 26-period EMA from the strike).
• Rule order: Rules fire top to bottom, first match wins.

Rules 1 and 2 are the hard stop-losses: if a YES position exists and the YES best bid is at or below 0.65, sell YES; same idea for NO at or below 0.65.

Rules 3 and 4 are settlement-adjacent exits: if a YES position exists and spot is now below the strike, sell YES; if a NO position exists and spot is now above the strike, sell NO.

Rules 5 and 6 are the take-profit rules from the base DSL — sell YES if time to expiry is 15 minutes or less but more than 2 minutes and the YES bid is at or above 0.995; same for NO. These are removed in all four variants. That removal is the deliberate change under test: without them, positions are held through to settlement rather than being closed out near 1.00.

Rules 7 onward are the entry ladder. Each level requires: a position below 50, no working order, edge.btc.price > 0, a time-to-expiry ceiling that tightens as the level deepens (8m, 5m, 4m, 2m, 1m, 30s, 25s), a minimum signed spot distance from the strike that also tightens (100, 85, 70, 46.7, 35, 23.3 points for the late levels), ema_vs_strike on the correct side of zero, change_5m on the correct side, velocity_1m at or beyond zero on the correct side, the entry ask within the band, and a spread at or under 0.02 (0.03 at the deepest level). Action is buy_yes or buy_no at size 50, executed immediate-or-cancel.

The four variants differ only in the lower bound of the ask band:

Everything above the floor — the ladder structure, the strike-distance gates, the momentum filters, the stop rules, and the removal of take profit — is identical across all four.

Top results

Ranked by ROI over the 30-day window:

1. Entry band 0.85–0.99, no take profit — ROI 512.84%, total PnL 256.42, max drawdown -165.84, Sharpe 0.40, 629 trades, win rate 88.89%. Saved as strategy btc-entry-bands-1764fb7c2726.
2. Entry band 0.89–0.99, no take profit — ROI 479.80%, total PnL 239.90, max drawdown -137.72, Sharpe 0.38, 601 trades, win rate 90.38%. Saved as strategy btc-entry-bands-18ed3c6fba7a.
3. Entry band 0.93–0.99, no take profit — ROI 434.92%, total PnL 217.46, max drawdown -91.05, Sharpe 0.65, 524 trades, win rate 93.15%. Saved as strategy btc-entry-bands-1ea144fc3b1e.
4. Entry band 0.95–0.99, no take profit — ROI 149.80%, total PnL 74.90, max drawdown -129.37, Sharpe 0.23, 484 trades, win rate 92.95%. Saved as strategy btc-entry-bands-55862c8ccea4.

The pattern is worth naming plainly. The loosest floor — 0.85 — produced the highest raw ROI and the highest total PnL, because it took the most trades (629) and captured more edge events. But it also carried the deepest drawdown (-165.84) and the lowest win rate of the four (88.89%). Tightening the floor progressively cut trade count, raised win rate, and reduced drawdown — until the 0.95 floor, where the trade count dropped to 484 and returns collapsed to 149.80% ROI.

The 0.93 floor occupies a distinctly different spot than its ROI rank suggests. It had the best Sharpe in the group (0.65), the smallest drawdown (-91.05), and the highest win rate (93.15%), while still producing 217.46 in total PnL. By return-adjusted measures it's the strongest variant in the batch; by raw ROI it sits third.

Bottom results

Read in reverse ROI order, the same four variants line up as:

• 0.95 floor: lowest ROI (149.80%), lowest PnL (74.90), lowest Sharpe (0.23), and a drawdown (-129.37) that is deeper than the 0.93 floor's despite far fewer trades.
• 0.93 floor: third on ROI, first on Sharpe and win rate, smallest drawdown.
• 0.89 floor: second on ROI, with the second-deepest drawdown.
• 0.85 floor: top on ROI and PnL, but deepest drawdown and lowest win rate.

The 0.95 variant is the clear laggard. It filters out a large share of opportunities — 484 trades versus 629 on the 0.85 floor — and the trades it does take don't compensate: win rate is high (92.95%) but the payoff per trade is evidently worse, dragging ROI down to roughly a third of the top variant's. That is the intuitive cost of demanding the market already price the contract near certainty: you pay up for a small edge.

One thing that doesn't resolve cleanly: the 0.95 floor has a deeper max drawdown than the 0.93 floor (-129.37 vs -91.05) despite fewer trades and a higher win rate. By trade count and win rate alone you'd expect the shallower book. This discrepancy is unresolved in the data presented here, and I won't invent a mechanism to explain it.

Similarly, Sharpe and ROI don't rank in the same order — the highest-ROI variant has the second-lowest Sharpe (0.40) and the third-ranked variant has the best (0.65). That is what it is; no additional accounting story is warranted from what's here.

Conclusion

Within this batch, the entry-band floor is a real lever, and it doesn't move results in a straight line.

If the goal is raw cumulative return over this 30-day window, the 0.85 floor wins: it is the most permissive, takes the most trades, and generates the most PnL, at the cost of the deepest drawdown. If the goal is return-adjusted quality, the 0.93 floor is the interesting one — best Sharpe, smallest drawdown, highest win rate — and it gives up meaningful total PnL to get there. The 0.95 floor is the weakest by essentially every measure in this run.

Two caveats belong on the front page rather than buried:

First, this was an explicit four-variant batch, not a sweep or permutation test. Nothing here establishes statistical robustness, and no confidence interval around these ROI figures is being claimed. A different 30-day window, a different risk budget per trade, or even a different set of strike-distance gates could easily reorder these four.

Second, removing the 0.995 take-profit rules changes the character of the strategy far more than the floor change does. Positions now run to settlement, so PnL is dominated by final settlement outcomes rather than early exits near 1.00. That makes these results structurally different from any version of this ladder that still carries take profit

About this Kalshi strategy backtest

What question does BTC Entry Bands investigate?

Compare four entry ask-band floors (0.85, 0.89, 0.93, 0.95) on the KXBTC15M strike-distance ladder, each with the 0.995 take-profit rules removed so positions run to settlement.

Read the analysis alongside the individual variant results. Check which entry, exit, price-band or risk settings changed, which settings stayed fixed, and whether the report documents enough historical coverage to support its comparison. A different parameter value defines a different simulation; it does not establish a new source of live trading returns.

How many strategy variants are included?

This report lists 4 completed and 0 failed strategy variants. Failed comparisons do not count as successful tests with zero profit. Review trade counts and drawdown alongside net P&L, win rate and Sharpe, and read any coverage or execution caveats in the analysis. A result based on very few trades needs particular care, even when its return looks high.

Does a profitable Kalshi backtest predict live performance?

These are historical simulations. Live results can differ because of liquidity, fees, slippage, latency, market resolution, outages and data quality. Selecting the best result after comparing many variants can also overstate how reliably it will perform on new data. Compare nearby settings, inspect losing variants and test on a separate period before considering a live run. The report is research, not a promise of profit or investment advice.

How can I inspect or test a strategy from this report?

Use an available Open strategy link to inspect that variant. Review its rules, market, date window, fill assumptions, fees and trade history before changing parameters or rerunning it. Confirm which data and execution assumptions the selected market actually supports; a replay does not establish that every historical order could have filled live. For the workflow, read the Kalshi and prediction market backtesting guide. The Kalshi backtesting software page explains the product's supported testing workflow.