Level 5 + 10

BTC 15-Minute Kalshi Strategy: Restoring Level 5 and Widening Distance Gates

Short disclaimer: This is historical simulation research on Kalshi crypto markets. It is not investment advice, not a live track record, and not a promise of future results.

Intro / thesis

The BTC 15-minute strategy on Kalshi is a short-horizon, strike-distance system. It looks at where the Coinbase BTC spot price and a 15-minute EMA sit relative to the contract's strike, and it only takes a position when price is far enough from the strike, momentum agrees, and the contract still has room to move.

Earlier work had trimmed this strategy down. Two entry levels — the 8-minute entry (referred to as level 5) and the 25-second entry (level 10) — were removed, and the strike-distance gates had been narrowed. The thesis for this batch was simple: put the level 5 and level 10 entries back, and widen every strike-distance gate by 16.5%, so the strategy can act earlier and from further out. Two variants were tested:

• Variant 1 restores both level 5 and level 10, with all distance gates widened by 16.5%.
• Variant 2 restores level 5 and widens the gates identically, but leaves level 10 off.

The only difference between the two is the presence of the final, no-distance-filter entry that fires in the last 25 seconds. That makes this batch a clean A/B test of whether that last-second entry helps or hurts.

Variant and strategy explanation

Both variants share the same skeleton, which is what the base strategy already did:

• Evaluate every 10 seconds on KXBTC15M, the most liquid contract.
• Position limit of 5 contracts. Entry prices are confined to the 0.90–0.99 ask band, with a spread cap of 0.02 (0.03 for the level 10 entry).
• Exits run top-to-bottom and are checked before entries each tick. In order: a stop-out if the bid on the held side falls to 0.35 or below; a spot-vs-strike invalidation if spot crosses back through the strike against the position; and a take-profit if the bid reaches 0.995 while between 15 minutes and 2 minutes to expiry.
• Entries require the spot price to be positive, order_count == 0 (no live orders), and position_size < 5.

The confirmed entry ladder, with the widened gates, is:

The directional logic is unchanged: a YES entry needs the EMA above the strike, positive 5-minute change, and non-negative 1-minute velocity. A NO entry needs the mirror image. The widened gates expand the universe of qualifying strikes at every timed tier — the level 7 gate, for instance, moves from roughly 40 to 46.7 points of spot distance — while levels 8 and 9 loosen from 30 to 35 and from 20 to 23.3 respectively.

Each successful variant in this batch is saved as a runnable Turbine strategy, so the configurations above can be re-run or inspected directly rather than reconstructed by hand.

Top results

Variant 1: level 5 restored, all distance gates +16.5%, level 10 restored — rank 1

• Total PnL: $31.59
• ROI: 631.8%
• Max drawdown: -14.26%
• Sharpe: 0.35
• Total trades: 563
• Win rate: 93.6%

This was the batch winner. It is the more active of the two variants — 563 trades versus 481 — which is expected given that level 10 adds a late-window entry with no strike-distance requirement at all. Despite the looser structure, it also posted the higher return, and the higher win rate, at the cost of a deeper drawdown.

Bottom results

Variant 2: level 5 restored, all distance gates +16.5%, no level 10 — rank 2

• Total PnL: $28.87
• ROI: 577.4%
• Max drawdown: -11.21%
• Sharpe: 0.32
• Total trades: 481
• Win rate: 93.3%

Dropping level 10 cost about $2.72 of PnL and 54.4 points of ROI versus Variant 1. It also reduced drawdown from -14.26% to -11.21%. That is the expected shape of a trade-off: the extra entry adds both volume and risk.

One thing worth flagging honestly: the per-trade math does not reconcile in an obvious way. Variant 1 earns more total PnL than Variant 2 on only 82 more trades, and its win rate is higher, yet adding trades normally pulls a win rate toward the new entrants' average. Whether level 10 trades happened to cluster in favorable windows, whether sizing or fill assumptions differ between the runs, or whether something else in the simulation accounts for it is unresolved. I'm not going to invent an accounting story for it. Treat the win-rate and PnL relationship between these two variants as an open question, not a settled result.

A second caveat on comparability: both variants ran 30-day windows, but that is a short sample for a strategy trading several hundred times. The ROI figures above are large because they are computed against a small simulated capital base, not because the strategy is producing that return on a realistic account.

Conclusion

Both requested changes held up in this 30-day simulation. Restoring level 5 and widening the distance gates by 16.5% produced a strategy that trades actively and wins most of its positions, and re-adding level 10 on top of that was net positive on both PnL and Sharpe — at the cost of a deeper drawdown.

The practical read: if you care about absolute return, Variant 1 is the higher-ceiling configuration. If you care about drawdown, Variant 2 gives up roughly a tenth of the return to cut max drawdown by about three points. Neither is obviously correct; that is a risk-preference call.

What this batch does not establish is robustness. This was an explicit two-way comparison with no parameter sweep, no permutation testing, and no walk-forward validation across different market regimes. A 30-day window in one volatility environment does not tell you how either variant behaves in a different one. The batch winner should be read as "this configuration performed best here," not as "this configuration is best."

Both variants are saved as runnable Turbine strategies and can be re-run against other windows for a more honest read.

Long disclaimer: This report describes historical simulation research conducted on Kalshi crypto prediction markets and is provided for informational and research purposes only. Nothing in this document constitutes financial, investment, legal, or tax advice, nor an offer or solicitation to buy or sell any security, contract, or financial instrument. All figures — PnL, ROI, Sharpe ratio, maximum drawdown, trade counts, and win rates — are the output of a backtest or simulation over the stated 30-day period and do not represent actual trading results. Simulated results are subject to inherent limitations: they rely on assumptions about fills, liquidity, spreads, order execution, and market data that may not hold in live conditions; they do not account for slippage, fees, latency, market impact, or capital constraints in any way that reflects a real trading account; and they may benefit from hindsight in the selection of rules and parameters. Past performance, whether actual or simulated, is not indicative of future results. Prediction markets carry risk of total loss of capital, and markets can become illiquid or halt. Kalshi operates under U.S. regulatory oversight, and access to its markets is restricted to eligible participants in permitted jurisdictions; nothing here should be read as an endorsement of trading on any venue by any person. Forward-looking statements, if any, are speculative and inherently uncertain. Readers should conduct their own independent due diligence and consult a qualified professional before making any trading or investment decision. No warranty is made as to the accuracy, completeness, or timeliness of the data or analysis presented, and the author assumes no liability for any loss arising from use of this material. Some figures or relationships noted in this report could not be reconciled and are flagged as unresolved; they should not be relied upon.

About this Kalshi strategy backtest

What question does Level 5 + 10 investigate?

Add back the level 5 and level 10 entries to the BTC 15-minute strategy and widen the strike-distance gates by 16.5%, testing two variants: one with level 10 restored, one without.

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 2 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.