AutoPilot Trader - THT V4.0: What Changed, and Why There Are Now Two Builds
July and August were rough. If you were running AutoPilot Trader through that stretch, you felt it. THT V3.3 gave back a meaningful chunk of gains across those two months, and I want to be direct about that before saying anything else: THT V4.0 exists because of what happened there, not despite it. Every significant improvement to APT has come out of a drawdown. That has been true since V1, and it was true again this summer.
Today, V4.0 ships for all active AutoPilot Trader members. V3.3 is retired. There are now two selectable builds: Core and Shield. This article walks through what changed mechanically, how the two builds differ, the simulated performance record you can audit yourself, and the honest caveats you need before you put any of this in front of real capital.
What We Changed in V4, and Why
Trade Selection
Continuous refinement across eleven years of tested data surfaced one specific cohort of signals that consistently lost money. That cohort is gone. V4 does not take those trades. The strategy is smaller by design, and that is a deliberate improvement.
Entry Mechanics
V3.3 entered on the close of the signal bar. On a large signal bar, that meant chasing the move by the time you got filled. V4 marks the signal bar and then waits for a partial retrace before entering. The stop stays the same. PT1 stays the same. When price continues to target, the improved entry price becomes captured edge. It is a small mechanical change with a measurable impact across the record.
The Runner Now Has a Target
V3.3's final runner was open-ended with a trail. In practice it left a lot of trades dependent on market conditions you cannot control. V4 gives the runner a true target alongside the active trail. Price still trails, but if it reaches the final target first, the trade locks profit there. This is a more disciplined structure.
PT2 Stretches on Explosive Moves
This is the change I want to explain most carefully, because it will occasionally look like a mistake in real time. When there is an explosive move immediately after entry, PT1 stays static and PT2 extends further out than it would have in V3.3. Here is what that means on a specific type of trade: price reaches where PT2 used to sit, misses the new extended target, and then trails out at breakeven after PT1 fills. You watch price hit V3.3's old target, and the trade ends at breakeven. In that moment, the extended PT2 looks like a bad decision.
Across 11 years of tested data, stretching PT2 on those moves yielded more profit than the fixed PT2 approach. The math is in the record. But the math does not make the individual trade feel better. I am telling you this up front because if you run V4 long enough, you will see that trade play out, and I want you to understand why it is there.
Two Builds
Inside the Two Hour Trader settings, you now select Core or Shield. These are not two separate strategies. They are one strategy at two different risk settings, with 75% shared entries and a 0.76 correlation. Do not think of them as additive.
Core runs full signal frequency and is built for the conditions where the strategy has its strongest edge. When conditions favor APT, Core captures more of that upside.
Shield adds two additional filters on top of Core. It takes fewer trades, is more patient on entries, and its historical drawdowns are substantially smaller. It earns less in favorable conditions by design. It was built with prop accounts and smaller accounts in mind, where controlling drawdown matters more than maximizing upside in a good month.
My recommendation: if you are new to automation, sensitive to drawdown, trading a small account, or running a prop account, run Shield.
The honest version of the tradeoff is this. Core does best when conditions are good, Shield does best when conditions are poor, and nobody knows in advance which environment they are walking into. That is the entire reason both builds ship.
The Simulated Performance Record
HYPOTHETICAL PERFORMANCE DISCLOSURE: All figures below are simulated backtested results, not live trading results. Past simulated performance is not necessarily indicative of future results. CFTC Rule 4.41 and NFA Rule 2-29 require disclosure that hypothetical results have inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading and may under- or over-compensate for factors such as liquidity. No representation is being made that any account will or is likely to achieve results similar to those shown. V4 has no live forward record and is currently running in shadow.
The figures below are backtested on NQ at 3 contracts, net of $4.50 round-turn commission and modeled slippage, over the window 2024-01-01 through 2026-07-31.
APT V4 Core
Basis: NQ 3 contracts, net of commission and slippage, 2024 through July 2026.
Total over the window: $53,347
By year: 2024 -$1,993 | 2025 +$38,243 | 2026 (Jan through Jul) +$17,098
Max peak-to-trough drawdown: $10,724
1,394 positions across 263 trading days | Win rate at 3 lots: 42.5%
Best month: April 2025 at +$17,721 | Worst month: September 2025 at -$4,587
16 of 31 months positive
APT V4 Shield
Basis: NQ 3 contracts, net of commission and slippage, 2024 through July 2026.
Total over the window: $48,212
By year: 2024 +$7,287 | 2025 +$21,700 | 2026 (Jan through Jul) +$19,225
Max peak-to-trough drawdown: $5,043
897 positions across 171 trading days | Win rate at 3 lots: 39.5%
Best month: March 2025 at +$13,551 | Worst month: September 2025 at -$3,738
20 of 31 months positive
A note on win rates: both builds show a much lower win rate at 1 lot (Core 29.4%, Shield 28.5%) than at 3 lots. This is a property of position sizing, not a flaw in the strategy. At 1 lot there is no PT1 peel and no runner structure, which changes how trades resolve. The 3-lot numbers are the appropriate reference for anyone running the standard contract configuration.
How to Compare V3.3 and V4 (the YTD Strategy Reports)
The release video (https://youtu.be/kOxI04cym8o) shows TradingView strategy reports for a side-by-side view. These are on a different basis from the performance page figures above: NQ 3 contracts, $100,000 account, gross of commission, year-to-date 2026-01-01 through 2026-08-31. V3.3 shows roughly $9,000 at a profit factor of 1.04 for that period. V4 Core shows roughly $26,000, approximately 26% on a $100,000 account, at a profit factor of 1.21. V4 Shield shows roughly $39,000 at a profit factor of 1.57. These are gross-of-commission strategy snapshots; the performance page figures above are the auditable net-of-cost record covering a different window. The two sets cannot be compared directly.
Where to Audit Everything
The performance page is at powertrading.group/apt-performance. It has equity curves for both builds, month-by-month P&L from 2024 forward, and full CSV downloads (daily and trade-by-trade). A risk calculator on the page lets you enter your account size, your maximum tolerable drawdown, your account type, your build, and your contract count. It tells you whether the sizing fits, including a micro option.
A backtest you cannot audit is a marketing asset. Every trade is downloadable. V3.3 is marked retired on the page.
Access and Pricing
AutoPilot Trader is capped at 250 members. Roughly 30 seats remain as of today. If you are already a member, V4 is live in your Two Hour Trader settings now.
If you are not yet a member, code OPINICUS is 15% off. A separate strategy called Spring is also coming in the next few weeks, with its own video, its own article, and its own performance record. Watch for it.
Disclosure: All performance figures in this article are hypothetical simulated results produced by backtesting on historical data. V4.0 AutoPilot Trader has no live forward trading record. Hypothetical performance results have many inherent limitations and no representation is being made that any account will achieve results similar to those shown. The development window for V4 Core and Shield rules (2025-03-01 through 2026-07-31) overlaps with the published backtest window; results from that period are in-sample and should be weighted accordingly. Only out-of-sample forward performance will provide meaningful evidence of sustained edge. This content is for educational purposes only and does not constitute financial advice or a solicitation to trade futures. Futures trading involves substantial risk of loss.
