Superseded by v3.6.1
The live version is now Drift VWAP Pullback v3.6.1, with new defaults: targets, stops, breakeven and trail in % of price, and Max Extension on in both sessions. If you update, your saved target, stop, breakeven and trail values reset to the new defaults — see updating from v3.4.1 before you do. The figures below are v3.4.1’s and are kept as the record.
New York is opt-in from this version
Asia and London are on by default; NY ships off. This is not a claim that NY loses money – on the settings above it runs a 2.54 profit factor over the test window. It is that NY’s weakest trades cannot be screened out in advance.
Its losing bucket is a specific, measurable thing: trades whose favourable excursion never reaches the breakeven trigger, which then run to the full stop. Over the test window, 27 such trades cost $4,547 while the other 147 made $12,039. Every available lever was measured against a parity-validated engine over two separate windows – all five entry filters and their combinations, five tighter stop levels, and five holding-period caps. Every one of them removed more value in winners than it saved in avoided stops, several while leaving the number of full stops unchanged or higher.
Asia and London do not behave this way, which is why they stay on. If you want NY, enable it deliberately in the NY panel.
What changed vs v3.1.0
- Reversal Engine, per session. After a stop-out the strategy can wait a set number of bars and take the opposite side with its own bracket, breakeven and trail. Which exit type arms it is selectable. Its budget is independent of Max Trades / Max Losses – a reversal is armed by a stop-out, so sharing the loss budget would make it unreachable at the default of one loss per session.
- Independent trail trigger. A trail no longer has to wait for breakeven. “Needs BE first” and “Trail Trigger (pts)” are separate gates, either skippable. London now uses this on both directions.
- Per-session Enable Longs / Enable Shorts.
- Per-session Max Extension Filter – blocks an entry already stretched past N standard deviations from that session’s VWAP. Available on all three sessions, off by default on all three.
- Guardrail attribution fix. Trade and loss counters were credited to whichever session was current when a trade closed rather than the one that opened it – reachable on shipped settings, since London’s EOD lands on the exact minute NY’s window opens.
- Pullback/reversal race fix, plus a quality-of-life pass: 201 inputs greyed out when inert, the trail input relabelled Trail Distance, three new chart plots, and a unified session dashboard.
Results
MNQ, 5-minute, 2026-05-17 to 2026-08-30, 1 contract, $0.62 per contract commission, 1 tick slippage. Figures are from a real TradingView “List of Trades” export, not a simulation.
| default (Asia + London) | with NY opted in | |
|---|---|---|
| trades | 214 | 393 |
| net profit | $10,228.64 | $17,832.18 |
| win rate | 86.92% | 83.46% |
| profit factor | 3.777 | 3.071 |
| max drawdown (closed trades) | $287.96 | $405.22 |
Per session, same run:
| session | trades | net | win rate | profit factor |
|---|---|---|---|---|
| Asia | 77 | $4,740.02 | 96.10% | 13.77 |
| London | 137 | $5,488.62 | 81.75% | 2.66 |
| NY (opt-in) | 174 | $7,492.74 | 79.89% | 2.54 |
The Reversal Engine accounts for 14 of those trades and $1,293.64 – a small sample, and it should be read as a starting point rather than a tuned result.
Parity
Checked entry by entry against the TradingView export – timestamp and direction matched individually, not aggregate totals, because a matching trade count and net P&L can coexist with entries on almost entirely different bars.
383 of 393 trades matched exactly on entry timestamp and direction: 97.5% recall, 98.2% precision. Of those, 368 agree on net P&L within $1 once TradingView’s $1.24 round-trip commission is accounted for, and 356 match the exit price exactly.
Four of the ten remaining differences are a data artifact rather than a strategy difference: the continuous-contract series used for verification rolls from the June to the September contract on a different date than TradingView’s, leaving the two sources roughly 300 points apart for about four days in June.
Tuning maturity
Asia is dialed in. London improved materially in this version through the new independent trail trigger. NY is profitable but is the session with a known, unsolved entry-quality problem, which is exactly why it now ships off. One instrument, one ~15-week window – not a forecast.