ETH Entry Band

ETH 15-Minute Entry Band Narrowing: A Kalshi Crypto Research Note

Simulated historical results only. Nothing here is a forecast, a recommendation, or a claim about future performance. Prediction markets can and do lose money.

Intro / thesis

The base strategy for Kalshi's KXETH15M series enters a YES or NO position only when a pile of conditions line up near expiry: ETH, BTC, and SOL all moving the same direction on 5-minute change, ETH price on the correct side of both its 1-hour VWAP and 12-period 1-minute EMA, MACD histogram agreeing, the quoted price sitting inside a band, and the spread tight. The band is the thing we can actually tune without touching the signal logic. The thesis under test: if we pull the entry band in toward 0.50 — where the market is genuinely undecided — do we get better outcomes than the current 0.35–0.70 range, holding everything else fixed?

Nine band variants ran over a 30-day window each, weekdays 09:00–17:00 Pacific, one position at a time, 50 contracts per entry, with the same stop rules in every variant: flatten at 10 seconds to expiry, exit a YES leg if its best bid falls to 0.25 or below, exit a NO leg on the same logic, and a hard dollar stop at −7 unrealized. The only thing that changes across variants is the two price thresholds in the entry condition. This was an explicit batch of the requested comparisons, not a parameter sweep, and no permutation test was run.

Variant and strategy explanation

Every variant shares rules 1–4 as exits (expiry flatten, two bid-based stops at ≤0.25, and the −7 dollar stop) and rules 5–6 as the entries (buy_yes and buy_no, size 50). What differs is the price floor and ceiling inside those two entry rules.

The candidate set walked inward from the current band and outward from it:

• Baseline 0.35–0.70 — the existing configuration, included as the reference point.
• 0.30–0.70 and 0.33–0.67 — wider or near-baseline bands that mostly add more low-priced entries.
• 0.38–0.62, 0.40–0.60, 0.42–0.58 — moderate narrowing.
• 0.44–0.56, 0.45–0.55 — tight, centered bands.
• 0.45–0.60 — an asymmetric band, tight on the cheap side but keeping more expensive YES entries in play.

Because entries fire only when position_size == 0 and time_to_expiry <= 12m, a tighter band doesn't just change which trades you take; it changes how many trades you take at all. A narrow band near 0.50 is a much rarer condition than a 0.35-wide door, so the trade counts differ materially across variants. Each successful variant is saved as a runnable Turbine strategy, so any of these can be pulled up and inspected directly rather than reconstructed from this note.

One mechanic worth stating plainly, because it shapes how to read the numbers below: the entry band only governs entries. Exits are untouched. A tighter band narrows what you're allowed to buy, not how you're allowed to get out.

Top results

Sorted by return on investment over the 30-day simulation:

The headline: the widest band tested (0.38–0.62) finished first at 875.8% ROI on 128 trades, and the tight centered band (0.45–0.55) finished seventh at 716.65% on just 70 trades. If the thesis was "narrowing toward 0.50 helps," the ordering doesn't support it in a clean way — the winner is a moderate widening of the current band, not a tightening. Two of the narrower bands (0.45–0.60 and 0.42–0.58) did land above baseline, so the picture isn't "wider is always better" either.

The 0.45–0.55 row is the interesting one for anyone drawn to the original thesis. It posted the best win rate in the batch (0.657) and by far the smallest drawdown (−35.97), but the lowest Sharpe of the set (0.56) and only 70 trades. Those facts sit awkwardly together: a strategy that wins most of its trades and draws down least, yet scores worst on a risk-adjusted return measure. That discrepancy is unresolved in this data. It isn't something I can attribute to a specific accounting detail or window-length artifact, and I'd rather flag it than explain it away.

Bottom results

The bottom of the table is where the results get hard to read as a story. The tightest centered band, 0.44–0.56, came last at 701.77% — only about 17 points behind the winner — and it did it with a mid-pack drawdown (−46.25) and a respectable 0.600 win rate. The two variants that produced the most trades, 0.30–0.70 (180 trades) and 0.33–0.67 (164 trades), produced only mid-table returns, and the 0.30–0.70 band had the lowest win rate of the batch at exactly 0.500.

Notice also that the entire field is compressed: the spread from best to worst ROI is roughly 174 percentage points on a base that spans 700–876%, and the Sharpe ratios only range from 0.53 to 0.68. These are not dramatically separated results. With one 30-day window per variant and no sweep or permutation test, ranking order this tight is not a reliable signal.

Conclusion

Does narrowing the ETH 15-minute entry band toward 0.50 improve results? On this 30-day simulation, not decisively. The top of the table is a moderate widening (0.38–0.62), the two tightest centered bands landed seventh and ninth, and the current baseline finished sixth of nine. Several narrower bands did beat baseline on ROI, but the tightest one did not, so the relationship isn't monotonic and the thesis isn't cleanly confirmed or refuted by this batch.

What the data does show clearly:

• Trade count moves with band width. Wide bands produce more entries; tight bands produce fewer. The 0.45–0.55 band took less than half the trades of the widest variant.
• Win rate and risk-adjusted return don't move together. The highest win rate and lowest drawdown went to a band with the lowest Sharpe.
• The field is tightly clustered. Differences this small, from a single window, shouldn't be treated as a ranking.

The 0

About this Kalshi strategy backtest

What question does ETH Entry Band investigate?

Does narrowing the ETH 15-minute entry price band toward the middle (nearer 0.50) improve results versus the current 0.35-0.70 band, when all other entry conditions are unchanged?

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 9 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.