BTC Win Rate

BTC 15-Minute Markets on Kalshi: Exit and Entry-Band Comparison

Short disclaimer

This is historical simulation research on Kalshi BTC 15-minute markets. It is not investment advice and it does not predict future results. All numbers below come from a fixed three-variant batch over a 30-day window; no live trading was involved.

Intro / thesis

The question here is narrow: on KXBTC15M, what happens to win rate, net P&L, and drawdown when you tighten the per-fill take profit, widen the per-fill stop, and narrow the entry price band for the momentum rules?

All three variants share the same overall structure. They evaluate every 10 seconds, cap positions at 50 contracts, and use a stack of momentum and trend entries gated on time_to_expiry <= 5m, a spread cap, and Coinbase BTC signals (5-minute change, 1-minute velocity, price versus 1-hour VWAP, and a 12/1-minute EMA versus a 20/1-minute SMA). Each entry buys a fixed size of 25. The variants differ only in two ways: the per-fill exit offsets and, in one case, the price band on two of the momentum rules.

The base configuration uses a 0.15 take profit and a 0.20 stop, with the trend-following entries allowed in a 0.25–0.75 price band. The other two variants move to a 0.10 take profit and a 0.25 stop, and one of them also tightens those bands to 0.35–0.65.

Variant and strategy explanation

Three completed variants, labeled by their exits and bands:

1. Current: 0.15 take profit / 0.20 stop, bands 0.25–0.75. Rules 4 and 5 sell a fill once the executable bid is at least the fill's purchase price + 0.15 (before fees). Rules 2 and 3 sell a fill once the bid falls to the purchase price − 0.20, and a gap can fill lower. The two trend-following entries (rules 8 and 9) only fire when the market price is between 0.25 and 0.75.

2. Tighter profit, wider stop: 0.10 take profit / 0.25 stop, bands 0.25–0.75. Same entry logic and bands, but profits are taken at +0.10 and stops trigger at −0.25. Because the exit gates are per fill, each fill is closed on its own basis; unrelated fills are left alone.

3. Tighter profit, wider stop, narrowed bands 0.35–0.65: 0.10 take profit / 0.25 stop. Identical to variant 2 except rules 8 and 9 now require the market price to sit between 0.35 and 0.65, matching the tighter band already used by the earlier-momentum entries in rules 6 and 7.

A few mechanics matter for reading the results. Rules are evaluated top to bottom and the first match ends the tick, so the cancel, stop, and take-profit rules take precedence over entries. Entry orders are capped at 10 per market, and because partial fills and maker chunks can each count against that cap, the effective entry count can be conservative. The stops and take profits act on the executable bid, so realized exit prices can be worse than the gate on a gap.

Top results

The winner is the tighter-profit, wider-stop variant that kept the 0.25–0.75 bands. It produced $2,171.57 of net P&L over 2,059 trades with a 78.77% win rate, a Sharpe of 0.93, and a maximum drawdown of −195.08.

The narrowed-band version is effectively a tie on P&L: $2,171.16, four cents below the leader across two fewer trades. Win rate is essentially identical at 78.75%, and Sharpe and max drawdown are the same to the reported precision. Tightening the trend-following band from 0.25–0.75 to 0.35–0.65 changed almost nothing in this window. That is a useful finding in itself: the narrower band did not filter out much, and it did not improve the risk profile.

The headline relative to the base variant is the win rate. Moving from a 0.15/0.20 exit profile to 0.10/0.25 lifted win rate from about 74.8% to about 78.8% and Sharpe from 0.88 to 0.93, with a small increase in total P&L.

Bottom results

The base variant ranks last, but only just on P&L: $2,170.92, about $0.65 below the winner over 2,073 trades. Where it clearly differs is win rate and Sharpe, at 74.83% and 0.88 respectively.

It does have the best drawdown of the three at −186.77, roughly $8 shallower than the others. So the trade-off is visible: the base exits win less often and earn a slightly lower Sharpe, but they held up a bit better at the worst point. The tighter-profit variants win more often and score a better risk-adjusted return, at the cost of a modestly deeper drawdown.

One item is unresolved. The reported ROI percentages (around 4,342%) are enormous relative to the dollar P&L and drawdown figures, and nothing in the provided data explains the denominator behind them. I am flagging that as a discrepancy rather than accounting for it. I can't attribute it to a particular sizing convention or reporting basis from what's here. Treat the ROI column as directional only and lean on the dollar P&L, trade counts, win rates, and drawdowns, which are internally consistent.

The other thing worth stating plainly: across all three variants the P&L spread is under a dollar on a base of roughly $2,171. The ranking is real in the sense that it is what the simulation produced, but the gaps are tiny relative to the noise you would expect from a 30-day window. This batch compared three specific configurations. It did not run a sweep or any permutation test, so the ordering should not be read as statistical evidence that one exit profile is genuinely better.

Conclusion

If you care about win rate and Sharpe, the 0.10 take profit / 0.25 stop profile did what it was designed to do: it raised the hit rate by about four points and nudged Sharpe from 0.88 to 0.93, with a small edge in net P&L. If you care about drawdown, the original 0.15/0.20 exits were marginally better, at −186.77 versus −195.08.

Narrowing the momentum entry band to 0.35–0.65 was close to a no-op in this window. It removed two trades and changed P&L by forty cents, so there is no case here for preferring the tighter band on performance grounds. If anything, the wider band gives the strategy more shots without measurable cost.

Each successful variant is saved as a runnable Turbine strategy: the winner is btc-win-rate-18561a566969, the narrowed-band variant is btc-win-rate-5c044641402c, and the base is btc-win-rate-312a2158bb45. The right next step is to run these across different periods and risk budgets before drawing any conclusion, because a 30-day window with sub-dollar separation between variants is not enough to settle the question.

Long disclaimer

This report is research output from a historical simulation. It is provided for informational and educational purposes only and is not investment advice, a recommendation, or a solicitation to trade any market or instrument. Nothing here should be read as a promise or implication of future profits.

The results are backward-looking and were produced on a specific 30-day window of Kalshi KXBTC15M markets using a fixed set of three configuration variants. They do not account for the full range of market conditions, regime changes, liquidity variation, or fee and slippage structures you would encounter in live trading. Stops and take profits in the described strategies act on the executable bid, which means gaps can produce fills materially worse than the configured offset, and simulated exit prices may not be reproducible in practice. Spread caps and price bands can also behave differently when books are thin.

The batch compared three variants only. No parameter sweep, out-of-sample test, walk-forward analysis, or permutation test was run. The rank ordering of variants is therefore descriptive of this one simulation, not evidence of statistical robustness. Small differences in net P&L, drawdown, Sharpe, and win rate should be treated as within noise unless confirmed by broader testing.

The ROI figures reported alongside the dollar results appear inconsistent with the dollar P&L and drawdown magnitudes, and the cause is unresolved in this dataset. Do not rely on those percentages.

Simulation results depend on assumptions about fills, queue position, latency, data quality, and the behavior of the underlying Coinbase BTC signals. Errors or changes in any of these can materially alter outcomes. Past simulated performance is not indicative of, and does not guarantee, future

About this Kalshi strategy backtest

What question does BTC Win Rate investigate?

Compare exit and entry-band configurations on BTC 15-minute markets to see how win rate, net P&L, and drawdown trade off when the per-fill take profit is narrowed, the per-fill stop is widened, and the entry price band for the momentum rules is tightened.

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 3 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.