AutoPilot Trader V3.2: New Features, Backtest Results, and What’s Next
V3.2 dropped on April 1, 2026. And no, that's not a joke.
This wasn't a ground-up rebuild. The core entry logic is identical to V3.1 - the same Two Hour Trader framework I've been trading for over a decade. What V3.2 delivers is smarter trade management and the most requested feature we've ever had: dollar-based risk control that actually adapts instead of just slamming the door shut.
If you're an existing APT user, some of this will look familiar - you've been watching the changelog evolve in real time. If you're evaluating AutoPilot Trader for the first time, this article gives you the full picture of where the system stands right now, what the backtest shows from January 2025 through April 2026, and what forward performance looked like in Q1 2026 during one of the choppier market stretches we've seen in a while.
Let's go through it systematically.
What's New in V3.2
Risk Ceiling with Dynamic Contract Scaling
This is the headline feature, and it's a meaningful improvement over what we had before.
The old V3.1 Maximum Stop Filter was binary: if a trade's stop distance exceeded a set threshold, the trade was skipped entirely. That worked as a blunt instrument, but it left a lot of valid setups on the floor. A trade with a 12-point stop on NQ might be filtered out even when trading 1 contract would have been perfectly acceptable from a dollar-risk standpoint.
V3.2 replaces that with a Risk Ceiling that scales intelligently:
Set a maximum dollar risk per trade in the settings
If the setup's full-size stop exceeds that ceiling, the system sizes down: 3 contracts to 2, then 2 to 1
If even 1 contract exceeds the ceiling, the trade is skipped
This is off by default. Enable it in Settings under Risk Ceiling
One important clarification before you set this and forget it: this is planned risk based on your stop placement, not a hard intraday stop. APT uses bar-close stops, which means the actual realized loss on a given trade can occasionally exceed the ceiling if price moves sharply through your stop level within a bar. Build in a 15-20% cushion when setting your ceiling - don't put it at your maximum acceptable dollar loss.
This is the first time APT offers risk management that responds to the setup rather than just accepting or rejecting it. For traders scaling from micros to full-size, or running APT through a prop firm evaluation with strict drawdown rules, this feature alone justifies the update.
PT Refresh Tighten-Only
This one is subtle but important for understanding how the system behaves in expanding volatility conditions.
In V3.1, if ATR expanded significantly while a trade was open, the profit target levels could drift farther from price mid-trade. That sounds counterintuitive - and it is. Profit targets should tighten as price moves toward them, not widen.
V3.2 locks this with a tighten-only constraint. Once a profit target level is set, it can only move closer to price, never farther. The system uses a math.min() comparison between the locked level and any updated calculation, keeping the tighter of the two. Clean fix.
Trail Ratchet (V3.2.3)
The trailing stop now only tightens. It cannot loosen once it's been set at a given level.
This sounds obvious - trailing stops should trail in one direction. But in earlier builds, under certain conditions, the trail level could briefly reset to a looser position as new bar data came in. The V3.2.3 fix anchors the trail using a math.max() comparison for longs (math.min() for shorts), so the trail level is always the most protective value seen up to that point.
What Did NOT Change
This matters as much as what did change:
Core entry logic is identical to V3.1. Same Two Hour Trader framework. Same signal generation.
Alert JSON structure is unchanged. Existing members do not need to recreate their alerts when updating to V3.2.3.
Trade management flow is the same: PT1 hit moves stop to breakeven, PT2 hit sets a hard floor, then the trail activates.
V3.2 is trade management and risk control improvements layered on top of a framework that was already working. We didn't touch the engine because the engine doesn't need touching.
V3.2 Backtest Results: January 2025 to April 9, 2026
The full V3 backtest analysis from last year covered a different testing window. Here's where V3.2 stands on the updated dataset.
Combined Portfolio: 3 NQ + 6 YM (2-Way Mode)
Combined P&L: $279,425
Total Trades: 1,187
Weighted Win Rate: ~72%
The 3 NQ + 6 YM weighting is intentional. One NQ contract is roughly equivalent to two YM contracts in notional value, so this portfolio structure creates balanced exposure across both instruments rather than overweighting the larger contract.
Understanding the Contract Scaling Numbers
This is where a lot of first-time readers get tripped up, so I want to be precise.
1 contract (Trail Only mode): No partial exits. The entire position runs to the trailing stop. This produces a fundamentally different return profile - not just a scaled-down version of the 3-contract results. Win rate drops to approximately 45% because the system never books PT1 or PT2 profits along the way. Fewer total trades as well, since some setups only trigger when partial exit logic is active. If you're evaluating 1-contract performance and expecting it to mirror 3-contract results at 1/3 the size, you'll draw the wrong conclusions.
2 contracts (PT1 + Trail): PT1 exit + trailing stop. No PT2. Win rate improves over 1-contract because partial profit locking is in play, but you're giving up the PT2 lock-in that the 3-contract setup captures.
3 contracts (1/1/1 split): PT1 exit, PT2 exit, trailing stop. This is the designed base unit. The backtest numbers referenced throughout this article use 3-contract sizing.
4+ contracts: Scale linearly from the 3-contract base. Multiples of 3 are recommended to maintain the 1/1/1 split cleanly.
Micro contracts (MNQ/MYM): Divide full-size results by 10 for comparable dollar figures. For best signal quality, chart the full-size contract (NQ or YM) and set your Execution Contract to the micro symbol.
If you want to model what these results look like at your specific account size and contract count, the V3.2 Risk Calculator handles all of this. It has explicit Full/Micro toggles, exact backtest data for 1/2/3 contracts, linear scaling for 4+, safety ratings by drawdown percentage, and a smart recommendation engine that suggests contract count based on your account size. Worth spending 10 minutes with before you finalize your setup.
Q1 2026 Forward Performance
Backtest numbers are useful. Live results are what matter.
Q1 2026 on the 3 NQ + 6 YM portfolio:
January: Flat (-$485)
February: +$6,000
March: +$11,000 (76% win rate)
Q1 Total: approximately +$17,000
January was genuinely slow. Low volatility, choppy sessions, the kind of market that grinds manual traders into bad habits. APT went flat - which, given the conditions, I'd call a good result. Not giving back money in a difficult environment is part of what the system is designed to do.
March is the one worth understanding in detail. The month opened with a $12,000 drawdown. That's not comfortable to sit through, and I won't pretend otherwise. But the system required zero intervention - no manual overrides, no adjustments, no shutting it off out of frustration. It came back and closed the month up $11,000 after rallying roughly $25,000 from the drawdown trough.
That's the behavioral pattern that separates systematic from discretionary trading. A manual trader almost certainly flips the system off at the bottom of a drawdown and misses the recovery. The system doesn't have that option - and that's a feature, not a limitation.
"I’m now eligible for a payout from Topstep today, using APT. This is one of the best investments I have ever made!” - Jacob Cable
If you want the full Q1 walkthrough with chart visuals, I did a complete YouTube review: Q1 2026 performance + V3.2 feature walkthrough.
Where We Are in the Development Roadmap
There's a question I get regularly: how much more optimization is possible before you've just curve-fit the backtest?
Good question. Here's the honest answer.
Over the past several months, I tested a significant number of potential improvements to signal quality: higher timeframe alignment filters, time-of-day exclusion windows, adaptive volatility scaling, and several others. Every single one was rejected. Not because I stopped looking, but because the data showed they didn't improve risk-adjusted returns in out-of-sample testing. Some looked good on the backtest and fell apart when I walked it forward. Others reduced drawdown but killed the win rate.
What that testing revealed is that the single-strategy edge in V3.2 is essentially maximized. The Two Hour Trader framework is extracting close to its theoretical maximum from NQ and YM within the current structure.
The next meaningful improvement comes from a different angle: adding a second uncorrelated strategy. We're targeting the Spring setup - a Wyckoff-based entry pattern that behaves differently from the Two Hour Trader framework and has historically performed well when the current system is in a sideways grind. Target launch is end of April 2026. When two genuinely uncorrelated strategies run together, the portfolio-level Sharpe ratio improves even if neither strategy individually gets better. That's the goal.
In the Thinktank, we've been discussing the Spring setup extensively as part of the broader market structure curriculum - if you want context on how that entry pattern works before the APT version launches, our community is the place to be. Members who understand the manual version of a trade make better decisions about when and how to run the automated version.
"My first ever 5-figure profit day came from APT. APT paid my rent today!” - Dez
How to Use the V3.2 Risk Calculator
One of the consistent points of confusion for new APT users is figuring out what contract size makes sense for their account. The Risk Calculator is designed to answer that question before you fund a live account.
Navigate to the calculator here. Key inputs:
Instrument selection: NQ only, YM only, or the combined NQ + YM portfolio
Full or Micro toggle: Explicit toggle, not a decimal input
Contract count: Pulls exact backtest data for 1, 2, and 3 contracts; linear scaling for 4+
Account size: Generates safety rating based on projected drawdown as a percentage of account equity
The safety ratings are:
Conservative: Maximum drawdown at or below 15% of account
Moderate: Maximum drawdown at or below 25%
Aggressive: Maximum drawdown at or below 40%
Dangerous: Maximum drawdown above 40%
The calculator also has a smart recommendation feature: input your account size and it suggests a contract count, automatically switching to micros if your account is too small for full-size contracts at a safe drawdown level. Combined portfolio view uses a 75% drawdown correlation estimate between NQ and YM positions.
Spend time here before you ask me what size to run. The data is in the calculator.
Frequently Asked Questions
The Risk Ceiling is showing different contract sizes than I expected. Why?
The Risk Ceiling calculates dollar risk using the Execution Contract's tick value, not the charted symbol. If you're charting NQ but executing on MNQ, the risk calculation is based on MNQ's $2/point value. This is the correct behavior after the V3.2 instrument detection fix.
Is 1-contract mode just a smaller version of 3-contract mode?
No, and this is important. 1-contract mode runs Trail Only - no partial exits at PT1 or PT2. The win rate is approximately 45% in this mode versus ~72% in the 3-contract portfolio. These are fundamentally different trade management profiles. Don't project 3-contract percentages onto 1-contract sizing.
Can I run APT through a prop firm evaluation?
Yes. The NQ Long-Only mode in particular has been used for prop firm challenges. The Risk Ceiling feature in V3.2 gives you additional control over per-trade exposure, which is useful when you're managing against a hard daily loss limit. See how a previous APT build passed a $50K prop firm evaluation.
What's the minimum account size?
APT works with accounts as small as $10,000 using micro contracts, or 150k prop firm accounts. Use the Risk Calculator to find the right contract size for your specific account.
When does the Spring setup launch?
Targeting end of April 2026. I'll announce it to the existing APT community first, with a full walkthrough on the Thinktank platform before it goes live.
The Bottom Line on V3.2
The entry logic hasn't changed because it didn't need to. The Two Hour Trader framework has been generating consistent results for over a decade, and V3.2 backtest data covering January 2025 through April 2026 confirms the system continues to perform - $279,425 combined P&L across 1,187 trades at a ~72% weighted win rate.
What V3.2 adds is the infrastructure that makes this system more usable across a wider range of account sizes and risk tolerances. Dollar-based risk management that scales intelligently. Visual improvements that make trade review cleaner. Bug fixes that were annoying in V3.1 and are no longer a factor.
Q1 2026 wasn't a gift from the market gods. January was flat, March opened with a significant drawdown. The system handled both without intervention and closed the quarter up roughly $17,000. That's the result I care about.
"I've managed to double my port since joining the group and have finally found my stride." - Martin Chavez
If you're ready to take a serious look at what your account setup could look like, start with the Risk Calculator. If the numbers make sense, the AutoPilot Trader product page has everything you need to move forward, including the white glove setup call where we configure your entire system live on Zoom - average session is about 35 minutes.
V3.2.3 is live. The Spring setup is coming. The roadmap is clear.
Trading futures involves substantial risk of loss and is not suitable for all investors. Past performance is not indicative of future results. Backtest results represent historical simulation with specific parameters and do not guarantee future performance.
