BTC Spot EMA Conflict Price Guards

BTC Spot EMA Conflict: Price-Guard Research

Historical simulation only. Not investment advice.

Thesis and scope

This study tests the saved BTC Spot EMA Conflict strategy on Kalshi's KXBTC15M markets. It follows spot momentum when Coinbase BTC spot and the completed 15-minute EMA-26 disagree about the literal Kalshi strike. The original DSL, sizing, entry and stop rules, 30-second cadence, and pinned Coinbase edge snapshot were preserved.

The 10×10 grid changes only risk.price_floor and risk.price_ceiling. Floor values are 0.01, 0.35, 0.37, 0.39, 0.41, 0.43, 0.45, 0.47, 0.49 and 0.51. Ceiling values are 0.53, 0.55, 0.57, 0.59, 0.61, 0.62, 0.63, 0.64, 0.65 and 0.99. All 100 variants completed successfully, with no failed cells.

Results

97 of 100 variants have positive net PnL. Three have negative net PnL; none are exactly zero. The best original grid cell is variant 005, with floor 0.01 and ceiling 0.61: +$796.65 net PnL, 548 trades, a 53.6% win rate, Sharpe 0.41 and maximum drawdown −$264.30. Ranking below is by net PnL.

The three negative cells are:

Mean net PnL declines across the tested floor values, from $675.49 at floor 0.01 to $118.49 at floor 0.51. The ceiling marginal peaks at 0.61 with mean net PnL $589.50; the 0.99 ceiling has mean net PnL $186.66. Average degradation to the winner's adjacent cells is 3.35%. These are descriptions of this same historical grid, not independent validation. Win rate alone does not determine net PnL because trade gains, losses and costs also matter.

Permutation test and selection checks

The recovered permutation test completed 100 of 100 full-grid re-sweeps, preserving all 68 previously saved scores and computing the remaining 32. Both the original capture digest and input digest matched exactly before resuming. The seed remains 4413442631275866308.

Each re-sweep scrambles the edge-feed values through time while leaving the market price series fixed. The real winner is $796.65. The null winners range from −$280.83 to $706.57; none of the 100 null winners matched or exceeded the real winner. With the finite-sample correction, the upper-tail p-value is 1/101 = 0.009901.

This test supports a difference from this particular edge-timing null on the saved inputs. It does not test the predictive value of price-based conditions generally, establish future profitability, or substitute for an out-of-sample test.

The saved deflated Sharpe statistic is 0.754, with expected-max Sharpe 0.273 and the raw 100-cell trial-count assumption. These are separate statistics on different scales. The deflated Sharpe probability remains below 0.95, so the permutation result should not be read as a blanket pass of every selection-risk check.

Coverage and interpretation

Coverage remains partial. The sealed candle window is 2026-09-09 14:40:47 UTC through 2026-10-09 12:59:59.999999999 UTC; the requested end was 2026-10-09 14:40:47 UTC. The saved sweep coverage metadata reports actual bounds 2026-09-09 14:25:13.147 UTC through 2026-10-09 12:59:59.959 UTC. This recovery preserved those inputs and did not expand the dataset.

The study's contract sizes and configured risk capital remain unchanged. Dollar net PnL above is the comparison metric; displayed percentage returns use the saved strategy's configured risk capital and should not be interpreted as returns on a separately specified $1,000 bankroll.

The family is positive across 97% of this tested grid, with a cluster of strong cells around ceilings 0.59–0.62 and lower floors. That is encouraging within this sample, but the best cell was selected from the same 100-cell grid. Historical execution assumptions, costs, liquidity, drawdown, partial coverage and selection risk still limit what can be concluded. Live performance and durability remain unproven.

About this Kalshi strategy backtest

What question does BTC Spot EMA Conflict Price Guards investigate?

Does spot-following when the completed 15-minute EMA-26 disagrees with the literal Kalshi strike retain positive net PnL across global price guard choices, and exceed an edge-timing-shuffle null with market prices held fixed?

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 100 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.