The efficient way to deploy your capital.
We move in sync with the high-frequency algorithms, running the same logic on the same clock. Liquidity and time reveal where the market is being drawn, and with the right timing, we can anticipate where and when the next swing point will form.
Skydweller.
Skydweller is a precision engine. It runs a fixed set of rules, takes no discretionary input, and acts only when every condition it was built around is present. The mandate is selection, not activity: only the highest-probability setups ever reach execution.
At its core, two systems work as one. An orderflow tracker reads how smart money is positioning. A liquidity map charts where resting orders cluster: stops above prior highs, stops beneath session lows, pools left untested by earlier sessions. Together they show where price is being pulled, and who is pulling it.
Calling the turn is its own discipline. The engine reads cross-market divergence across the three US index futures. That divergence marks the moment a cycle exhausts and pinpoints the exact turn.
Execution runs on a sub-one-minute timeframe, entering as close to the turn as structure allows. Tighter entries mean tighter stops, and tighter stops mean the best risk-to-reward a setup can structurally offer, with risk capped by construction on every trade. The engine also knows when not to work: it reads the economic calendar and stays offline around key data releases, and it trades only its proven window. Outside it, flat by design.
How a setup is selected
Five gates, in order. Every one must pass before the engine takes risk; most candidates die at the third.
Forward test
OUT-OF-SAMPLEReal-time, out-of-sample tracking against live market data. No capital at risk. Not a backtest, not live-money performance.
Forward-test trade log: engine fires, review period
Every engine fire in the review window, in order.
| Date ▲ | Time ET | Dir | Entry | Stop | Target | R | Exit | Note |
|---|---|---|---|---|---|---|---|---|
| 2026-05-14 | 09:12 | LONG | 29,460.75 | 29,448.25 | 29,523.25 | +4.00 | trail | Entered early, right side, +4R banked. |
| 2026-05-15 | 09:31 | SHORT | 29,210.25 | — | — | -1.00 | stop | False positive, no confirmation on the turn. Patched. |
| 2026-05-18 | 09:57 | SHORT | 29,275.75 | 29,308.75 | 29,110.75 | +4.00 | trail | Entered after reclaim, caught 4R of a 10R+ leg. |
| 2026-05-18 | 09:57 | SHORT | 29,245.75 | 29,308.75 | 29,052 | +1.50 | trail | Companion fire, same setup. |
| 2026-05-19 | 09:33 | SHORT | 28,977.25 | 28,999 | 28,868.5 | +1.50 | trail | Right side, early bar. |
| 2026-05-19 | 09:40 | SHORT | 28,985.75 | 29,068 | 28,802.75 | +1.50 | trail | ~30–60s after the ideal entry. |
| 2026-05-19 | 09:41 | SHORT | 28,956.5 | 29,068 | 28,802.75 | +1.38 | tp | Full TP on the draw. |
| 2026-05-20 | 09:35 | SHORT | 29,074 | 29,154.5 | 28,671.5 | -1.00 | stop | Should have been calendar-gated. Patched. |
| 2026-05-20 | 09:49 | SHORT | 29,142.5 | 29,194.75 | 29,023.25 | +1.50 | trail | Calendar gap, profitable anyway. |
| 2026-05-20 | 10:25 | SHORT | 29,125.5 | 29,233.75 | 29,008.5 | -1.00 | stop | Calendar gap. |
| 2026-05-20 | 10:34 | SHORT | 29,209.75 | 29,263 | 29,008.5 | -1.00 | stop | Calendar gap. |
| 2026-05-21 | 09:55 | SHORT | 29,292.5 | 29,315 | 29,180 | +1.50 | trail | Market flipped direction mid-session, wrong side, still trailed +1.5R. |
| 2026-05-26 | 09:36 | LONG | 29,833.25 | — | — | +81.40 | tp | First clean fire, right entry, full TP. +81.4R. |
| 2026-05-28 | 09:50 | LONG | 29,952.75 | 29,920.25 | 30,112 | +5.24 | trail | Entered on initial reclaim; a tighter retest would have paid more per dollar risked. |
| 2026-06-11 | 10:02 | LONG | 28,672.5 | 28,648.75 | 28,795.5 | +5.18 | tp | Full TP, flagged for hand review as outside the engine's core pattern. |
Backtest results
A historical walk-forward replay of the engine. Hypothetical by nature.
Same rules, 18 months
2026 YTD vs. the full backtest, same rules
| Metric | 2026 YTD (Jan–Jun) | Backtest (18mo) |
|---|---|---|
| Trades | 34 | 105 |
| Win rate | 67.6% | 72.4% |
| Net R | +293.3R | +520.1R |
| Avg R / trade | +8.63 | +4.95 |
Both columns are the same simulated replay. The forward-test record above (May 14 – Jun 11: 15 fires, 73.3%, +104.7R) runs separately and isn’t part of this backtest.
The crypto cubicle (BTC / ETH / SOL) trades live on Kraken today. Skydweller is the same discipline applied to index futures via Interactive Brokers. It's currently in forward test, ahead of live capital.
