VWAP Momentum Wide
When BTC spot price is above 1h VWAP with positive 5-minute momentum, KXBTC15M YES contracts tend to be undervalued; when below with negative momentum, NO contracts tend to be undervalued. The wider 0.10-0.90 price band captures more entry opportunities compared to tighter bands.
Historical research only. Not investment advice.
Top strategy variants
Bottom strategy variants
Research Report: Kalshi KXBTC15M — VWAP & Momentum Strategy Simulation
Short Disclaimer This is a historical simulation study. Past performance does not guarantee future results. All figures are hypothetical and net of friction assumptions.
Intro / Thesis
We tested whether combining BTC spot price relative to its one-hour volume-weighted average price (1h VWAP) with short-term 5-minute momentum creates a systematic edge in Kalshi’s 15-minute BTC binary options (KXBTC15M). The core idea: when BTC trades above 1h VWAP and has positive 5m momentum, YES contracts are perceived to be undervalued; below VWAP with negative momentum, NO contracts appear undervalued. We also wanted to see if a wide entry price band (0.10–0.90) captures more opportunities than narrower bands without degrading trade quality.
Variant and Strategy Explanation
The base strategy uses live Coinbase data (spot price, 1h VWAP, and 5-minute percent change) refreshed every 10 seconds. It runs a 30-second decision loop on KXBTC15M contracts.
Rules are simple:
- Buy YES when BTC > 1h VWAP, 5m change > +0.05%, contract price within 0.10–0.90, spread ≤ 0.03, and more than 2 minutes until expiry.
- Buy NO when BTC < 1h VWAP, 5m change < –0.05%, same price/spread/time constraints.
- Stop loss at –$25 unrealized PnL on any open position.
- Forced exit when ≤2 minutes to expiry.
100 variants were generated from this base. Each variant modifies parameters like entry thresholds, spread tolerance, price band edges, stop-loss levels, or momentum values. All runs are complete; 100 strategies were simulated, producing top and bottom cohorts measured by ROI, Sharpe, drawdown, total PnL, win rate, and trade count.
Top Results
The top-performing variants are disappointing by any conventional measure. Every single top-ranked strategy lost money — heavily.
| Rank | Variant | ROI % | Sharpe | Max DD % | Win Rate | Total PnL | Trades |
|---|---|---|---|---|---|---|---|
| 1 | v017 | –304.2 | –1.58 | –93.32% | 10.37% | –$76.05 | 3,787 |
| 2 | v018 | –304.2 | –1.58 | –93.32% | 10.37% | –$76.05 | 3,787 |
| 3 | v019 | –304.2 | –1.58 | –93.32% | 10.37% | –$76.05 | 3,787 |
| 4 | v020 | –304.2 | –1.58 | –93.32% | 10.37% | –$76.05 | 3,787 |
| 5 | v037 | –306.4 | –1.68 | –94.07% | 10.55% | –$76.60 | 3,849 |
| 6 | v038 | –306.4 | –1.68 | –94.07% | 10.55% | –$76.60 | 3,849 |
| 7 | v039 | –306.4 | –1.68 | –94.07% | 10.55% | –$76.60 | 3,849 |
| 8 | v040 | –306.4 | –1.68 | –94.07% | 10.55% | –$76.60 | 3,849 |
These strategies all share similar characteristics: win rates around 10–11%, massive drawdowns exceeding 90%, and deeply negative Sharpe ratios. The wide 0.10–0.90 price band did generate high trade volumes (3,700–3,800+) but each additional trade appeared to pile on losses. Despite capturing many entry opportunities, the edge was negative — the market consistently priced these contracts correctly or against our signals. The stop loss at –$25 was triggered frequently, locking in small losses that compounded into large drawdowns across thousands of trades.
None of these top variants represent a viable trading approach. They are labeled “top” only by relative rank within a uniformly loss-making population.
Bottom Results
The worst performers are even more destructive, though the gap between “top” and “bottom” is narrow.
| Rank | Variant | ROI % | Sharpe | Max DD % | Win Rate | Total PnL | Trades |
|---|---|---|---|---|---|---|---|
| 93 | v061 | –1248.8 | 0.0 | –76.13% | 10.96% | –$62.44 | 3,348 |
| 94 | v062 | –1248.8 | 0.0 | –76.13% | 10.96% | –$62.44 | 3,348 |
| 95 | v063 | –1248.8 | 0.0 | –76.13% | 10.96% | –$62.44 | 3,348 |
| 96 | v064 | –1248.8 | 0.0 | –76.13% | 10.96% | –$62.44 | 3,348 |
| 97 | v081 | –1250.2 | 0.0 | –76.27% | 10.89% | –$62.51 | 3,374 |
| 98 | v082 | –1250.2 | 0.0 | –76.27% | 10.89% | –$62.51 | 3,374 |
| 99 | v083 | –1250.2 | 0.0 | –76.27% | 10.89% | –$62.51 | 3,374 |
| 100 | v084 | –1250.2 | 0.0 | –76.27% | 10.89% | –$62.51 | 3,374 |
Win rates here are marginally higher (around 10.9%) than the top group, and total dollar losses are slightly lower ($62 vs $76), but ROI percentages are catastrophic — the denominator effect of smaller average position sizing or account equity makes the percentage loss look worse. Sharpe ratios register as zero due to consistently negative returns with near-zero volatility of that negative drift. These bottom variants likely used slightly different stop-loss or size parameters that reduced absolute dollar loss but still produced uniformly negative results.
Conclusion
The VWAP-plus-momentum thesis on KXBTC15M contracts fails dramatically in historical simulation. None of the 100 variants produced a profit. Win rates hover stubbornly around 10–11%, suggesting the binary outcomes move against our signals roughly 90% of the time. The wide 0.10–0.90 price band succeeded in increasing trade frequency but merely amplified losses — more entries meant more losing trades, not more winners. Stop losses at –$25 acted as a death by a thousand cuts. Tight spreads and proximity to expiry offered no shelter.
This concept, at least in this exact formulation on Kalshi’s 15-minute BTC binaries, does not demonstrate an exploitable edge. The market appears efficient relative to this signal set. Each successful variant from the simulation is saved as a runnable Turbine strategy for audit and further study, but we do not recommend live deployment of any variant in this cohort.
Long Disclaimer This document is produced for research and informational purposes only. All results are derived from historical simulations conducted in a controlled backtesting environment. They do not represent actual trading, live account performance, or future outcomes. Simulated results are subject to limitations including, but not limited to, imperfect modeling of liquidity, execution latency, exchange fees, and market impact. Any strategy discussed here — including saved runnable variants — carries significant risk of loss. Past simulated performance is not indicative of future results. This is not financial advice, a solicitation, or a recommendation to trade any instrument on any venue. Consult a qualified professional before engaging in speculative trading.
This report is generated from historical simulations. Backtests can be wrong or incomplete, and live trading can differ materially because of liquidity, fees, slippage, latency, market resolution, outages, and data quality. Do your own review before running any strategy.