February 16, 2026
The market opened. Sim101 was live. PCS was running.
NinjaTrader's Sim101 isn't a backtest — it processes real market data in real time through the actual market feed, accounting for real bid-ask spreads and slippage. Fills in Sim101 are as close to live fills as you get without real money on the line.
Ten months of development. 5.5 years of walk-forward OOS testing. Sharpe 4.29 in the historical record. Now we find out if it works in the present.
PCS fired its first trades.
What the MFE/MAE Data Revealed
The first thing I pulled after the opening session wasn't P&L. It was MFE/MAE data — Maximum Favorable Excursion and Maximum Adverse Excursion for each trade.
MFE tells you how far in-the-money a trade moved before it closed. MAE tells you how far it moved against you before it resolved. Together, they tell you whether your entries are right and whether your exits are capturing the available edge.
✓ Entries: Right
Average MFE per trade. The market is moving in the predicted direction by nearly $62 before the trade closes. That's regime detection working in real time.
⚠ Exits: Constrained
The system is giving back a meaningful portion of that $61.88. Exits are leaving money on the table. A specific, addressable problem.
This is actually good news.
The system isn't wrong about direction. It's premature about timing the close. That's a solvable problem. And we have a solution in development.
The Bobby Trade Manager
The next major development is a trade management module called Bobby.
Bobby's job: dynamic exit management using the same real-time regime classification that drives entries. Instead of fixed targets and stops, Bobby will read coherence state as the trade develops, extend targets when the regime suggests continuation, and tighten exits when coherence breaks down mid-trade.
The MFE data confirms the opportunity. If the average trade reaches $61.88 in the right direction, smarter exits should recover a meaningful portion of the give-back. Even 15–20% recovery has significant impact at multi-contract scale.
Entries are working. Now we optimize what we do with them.
The Path to $500/Week
Let's talk about the math.
At one contract, PCS's historical performance projects to approximately $1,895 per year. That's the OOS walk-forward number — real, unoptimistic, and net of the noise that comes with a $5/point instrument.
At five contracts: $9,475 per year. That's $182/week.
$182/week is not $500/week. But it's the baseline — before Bobby's exit optimization, before any parameter refinement from live data, before scaling beyond five contracts.
None of this is guaranteed. That's the point of the live phase.
What Live Trading Will Prove (or Disprove)
The backtest is a closed question. The walk-forward validation is as rigorous as I know how to make it. But only live data can answer the real questions:
Is the edge real-time detectable?
Regime classifiers work on historical data. Live markets have latency and feed anomalies that backtests don't capture.
Does slippage eat the edge?
MES is liquid, but at 10 contracts we'll find out what that actually costs.
Is the $459 max drawdown realistic?
That number over 5.5 years is extraordinary. Live trading will either confirm it or reveal the scenario historical data missed.
Does Bobby actually help?
The MFE data says the opportunity is there. Execution will determine whether we capture it.
I expect surprises. The question isn't whether anything unexpected will happen — it's whether the system's framework is robust enough to handle what we haven't anticipated. The architecture is designed for regime change, not regime assumption.
Follow the Journey
This is the beginning of the live chapter, not the end of the story.
We'll publish live performance data, trade logs, MFE/MAE updates, and Bobby development as they happen. Real numbers. No cherry-picking. If PCS has a bad month, we report it and explain it.
Ten months to Sharpe 4.29. Now we find out what that actually looks like in the real world.
The system is running.
This is Part 4 of a 4-part series. Start with Part 1: "The 10-Month Gamble" for the full story from the beginning.