Superseded by v2.2.5 — read this before loading a configuration
The live version is now Tokyo Drift v2.2.5, and the TradingView script has been updated to it.
Prop Safe has moved. There is a new Prop Safe built on the v2.2.5 defaults, with its own tab and TradingView export on the v2.2.5 page — use that one. The v2.2.4 Prop Safe below is kept for the record and is no longer recommended.
This page stays up because Split Sessions still lives here: it was not re-tuned for v2.2.5 and its numbers below are unchanged.
The one thing to know before you load a file from this page onto the updated script. v2.2.5 added a directional time gate, and these Load Parameters files were saved before it existed, so they do not contain those settings. Loading one does not switch the gate off — it leaves whatever the chart already had, and a chart that has been reset to the v2.2.5 defaults has new long entries blocked from 12:45 PM ET. The results below were measured with the gate off.
So after loading Split Sessions (or the old Prop Safe), open the Directional Time Gate
panel and set Stop taking LONG entries later in the session to off. On Prop
Safe the difference is small but real — profit factor 1.37 against 1.39, and the
January 2024 to May 2025 window drops from +$351 to +$73.
Three configurations, and which one to run
Every configuration on this page is the same script. Only input values differ, so moving between them is a settings change, not a different strategy. The Default is what the script loads with; the other two are Load Parameters files, copied from their tab below and pasted into TradingView’s settings dialog through Load parameters.
| configuration | shape | year (TradingView) | run it when |
|---|---|---|---|
| Default | one session from the Globex open to the New York close, VWAP anchored to the Globex open; breakeven, trail, MFE and MAE exits all on; no MACD filter | 574 trades, +$18,542, PF 1.52, drawdown $2,044 |
a normal account: it trades the most often of the two all-day configurations and keeps every loser small ($217 was the largest on the year) |
| Prop Safe | the Default with the MACD velocity filter on: fewer trades at the same exits | 457 trades, +$15,665, PF 1.57, drawdown $1,821 |
a funded evaluation or a small account: the least drawdown on the long history |
| Split Sessions | Asia, London and New York as three sessions; percent bracket; MAE exit only, no breakeven, trail or profit-side exit | 695 trades, +$21,279, PF 1.40, drawdown $2,643 |
you want the most net on the year and can carry the deeper history below |
Prop Safe is the one to run on a funded account. It has the smallest worst quarter of the three and is the only one positive in seven of the eleven quarters since 2024. Split Sessions earns the most on the year and gives the most back outside it. That trade-off is spelled out under Read the holdouts below, and it is the reason there are three tabs rather than one.
What changed since v1.5.1
The published TradingView version before this one was v1.5.1, so this is the accumulated work of nine versions. The full changelog is in the release note; the three that matter to how the strategy trades are these.
The MAE / MFE stop. A new exit group with two independent sides. Exit if IN LOSS cuts a trade that has been open at least N bars and is still at least X underwater, at the next open, ahead of the full stop. Exit if IN PROFIT takes a trade that has been open at least N bars and is at least X ahead but has stopped making progress. Each side has its own distance unit (points, percent of price, or a multiple of the block’s own ATR, frozen at the fill) and its own clock (bars or minutes). Both test the current unrealised move at each bar close, not the worst or best excursion since entry, and neither ever starts a reversal.
Why it matters for a funded account: a stop is a price. A trade that has been underwater for sixteen bars and is still 0.35% against you is, on this instrument, a trade that is not coming back often enough to pay for the ones that run to the full stop. The Default runs both sides, 0.35% after 22 bars in profit and 0.35% after 16 bars in loss, and together with the breakeven they took 152 of the year’s 573 exits. Split Sessions runs the loss side only.
Percent of price everywhere. The bracket has been percent-denominated since v1.3.0; the trail, the breakeven offset and the reversal bracket now share the unit, so a configuration cannot size its stop off price and its trail off ATR by accident. Percent transfers between instruments and across a doubling of gold’s price the way a point value cannot.
Readable orders. Every order on the chart and in the trade list says what it is: Enter Long, Flatten Target, Flatten Trail, Flatten Breakeven, Flatten Stop, Flatten EOD, MAE Exit, MFE Exit. The webhook payloads are unchanged.
The all-day session, anchored to the Globex open
Split Sessions trades the three sessions Tokyo Drift has always had. The Default and Prop Safe do something the earlier versions could not: they run one window from 6:45 PM to 3:00 PM ET on the Asia slot, with London and New York switched off, and anchor the VWAP to the Globex open rather than to the session.
The effect is a single continuous VWAP for the whole trading day, and a trend that begins in Asia can be traded through London and into New York on the same reference line rather than being cut at each session boundary and re-anchored to a VWAP that has only a few bars of history. On the year-long export the exits show what that buys: with the full stack on, the trail took 129 of the 573 exits, the target 125, the stop 119, and only 26 trades reached the end-of-day flatten. Split Sessions, which cuts at each session close, sent 165 of 691 trades to the flatten.
The cost is an all-day position budget rather than three, nine trades and three losses per day, and a day long enough that the breakeven, trail and profit-side exits have to be on to keep the average loser small. That is what makes the Default the conservative configuration and Split Sessions the higher-net one.
Results
Verified on TradingView. Micro Gold, 5-minute, one contract, $50,000
account, $1.20 per side and one tick of slippage. The table is the
TradingView trade-list export for each configuration over the last 365 days
(16 September 2025 to 15 September 2026). The equity curves for both this
window and the selection window are on each tab below.
| configuration | trades | net | win rate | PF | max drawdown | largest loss |
|---|---|---|---|---|---|---|
| Default | 574 | +$18,542 |
65.7% | 1.52 | $2,044 |
$217 |
| Prop Safe | 457 | +$15,665 |
65.9% | 1.57 | $1,821 |
$339 |
| Split Sessions | 695 | +$21,279 |
52.9% | 1.40 | $2,643 |
$348 |
The selection window, 31 May to 15 September 2026, is the range the configurations were tuned on and is the best case, not the expectation:
| configuration | trades | net | win rate | PF | max drawdown |
|---|---|---|---|---|---|
| Default | 181 | +$11,672 |
73.5% | 2.43 | $744 |
| Prop Safe | 147 | +$10,032 |
75.5% | 2.69 | $536 |
| Split Sessions | 213 | +$17,822 |
67.1% | 2.70 | $659 |
Read the holdouts
The year above contains the selection window. To see each configuration on data it was never tuned on, the engine twin runs it on the year before the export (June 2025 to May 2026) and on everything before that (January 2024 to May 2025). These are engine figures, not TradingView runs, at the same cost model.
| configuration | tuned window PF | year before PF | 2024 – mid-2025 PF | quarters positive | worst quarter | drawdown since 2024 |
|---|---|---|---|---|---|---|
| Default | 2.34 | 1.24 | 0.97 | 6 / 11 | −$1,084 |
$2,200 |
| Prop Safe | 2.73 | 1.22 | 1.03 | 7 / 11 | −$654 |
$2,300 |
| Split Sessions | 2.75 | 1.06 | 0.84 | 6 / 11 | −$3,034 |
$8,289 |
Split Sessions is the highest-net configuration on the year and the weakest outside it. Every quarter of 2024 is negative for it, and its drawdown since 2024 is four times the Default’s. The mechanism is not mysterious: with no breakeven, no trail and no profit-side exit, a loser runs to the full 0.45% stop, and the loss-side MAE exit cannot reach a trade that gets there inside sixteen bars. In the 2024 regime enough of them did. The Default and Prop Safe cap those losers and keep a profit factor near or above one on both holdouts. Size by this table, not by the selection window.
Verification
Each configuration is a paired export: the TradingView trade list and the Save Parameters file it was generated under, so the settings under test are the settings that were live. The Python engine twin reproduces the Default’s fills at 96.2% recall and 97.0% precision entry-by-entry over 573 trades, with the entry price exact on 99% of matched trades and the same exit bar on 99%. Split Sessions matches at 97.3% and Prop Safe at 95.6%. The residual is the continuous-contract data difference described on the v1.5.0 page, not an engine gap.
Known limits
- The tuned window is the last hundred days. Every configuration was selected on the 31 May to 15 September 2026 range, and the holdout table is the honest statement of what to expect beyond it.
- One instrument. All three configurations are Micro Gold. Percent brackets travel; the anti-liquidation and candle-range filters are in ATR and travel too; the session hours and the MACD settings do not.
- The holdouts are engine figures. The twin is verified at 96.2% entry-by-entry, but only the year table is a TradingView run.