Blog

• Flip Signals: Early Evidence of a Low-Quality Trade Regime
• Adding Random Forest to Machine A: Dual-Model GTO Signal Comparison
• Building Trade Observations Academy: From Market Data to Structured Learning
  How we built an automated pipeline that transforms market observations into structured lessons and quizzes to improve trader decision-making.
• AI Exit Warnings: From Narrative to Actionable Signal
  Converting narrative AI Analysis into a simple, actionable signal.
• Closing the Loop: Scoring AI Market Analysis with Real Trades
  How I linked AI market analysis to real executed trades, scored the outcomes, and wrote the results back into the analysis record.
• Trade Observations Becomes Part of the System Architecture
  The research site is no longer just a blog. It now reads directly from the trading system's central database to curate and publish AI market analysis.
• When a Trading System Outgrows a Single Market
  Planning the expansion of a systematic trading architecture from one asset class to another.
• Parallel Systems vs Shared Infrastructure
  Why adding a new asset class should usually begin with a parallel trading stack instead of a shared one.
• Model Design Across Multiple Symbols and Asset Classes
  How modeling approaches must evolve when systematic trading expands beyond a single instrument.
• Designing Data Pipelines for a New Asset Class
  How market structure differences shape data engineering when expanding a systematic trading framework.
• Stop Guessing, Start Observing: Using Trade Data to Measure Market Climate
  How I began using trade-level MAE/MFE data and SQL to measure market climate instead of relying on intuition.
• A Trade Within a Trade
  Designing robust re-entry logic with acceptance + follow-through.
• Stop Guessing: When Different Models Agree on the Same Stop Level
  Comparing Random Forest and Gradient Boosting stop models using MAE-to-date and recovery probability.
• Two Sides of Risk: Using MAE and MFE to Govern Stops
• Building Machine B: From Research Model to Production System
  Turning a trailing stop model into a live risk engine with retraining, safeguards, and fallback rules.
• Building a Machine-Learned Trailing Stop Engine
  A practitioner’s journal on engineering systematic exits using Machine Learning.
• Why Machine-Learning Trailing Stops Rarely Work (and How to Fix Them)
  Backtest evidence from an ES/MES two-machine architecture: ML trailing stops often add zero alpha. Here’s why—and how to redesign exit research.
• RTH vs ETH: The Data Distribution Mismatch That Broke My Stop Model
  Why training on overnight futures data sabotaged a model I only use during RTH.
• The Futures Rollover Trap: Why Your ML Model Breaks Every Quarter
  Futures contracts roll. Your model quietly degrades unless you engineer around it.
• Random Forests for Trailing Stops: Labels Without Lookahead Bias
  How I defined 'good' and 'bad' trailing stop decisions without cheating with future data.
• Why Trailing Stops Are Harder Than Entries
  Entries get the attention. Exits shape the equity curve.
• I Had Rules for Entries and Feelings for Exits
  Why trailing stops became the core problem I decided to systematize.
• When the Second Brain Hesitates: RFStopManager, Stale State, and a Stop That Didn’t Move
  A real failure-mode walkthrough where Machine B published tighter trailing stops, yet RFStopManager didn’t record applying them—why apply_status became SKIP/not_tightening, and how to harden the loop.
• Fragility and Failure Modes: Where the Second Brain Can Break
  A candid map of the weaknesses in my distributed trading architecture—and the roadmap for hardening it.
• The Operational Loop: Supervising a Second Brain in Real Time
  What 'good state' looks like during a trading session, how I monitor a distributed trading system, and how I avoid the two classic failure modes of automation.
• A Live Trade Through the Second Brain: From Signal to Exit
  A real short trade traced through my multi-machine trading system—showing how models, databases, and execution logic coordinate in real time.
• The System Behind the Second Brain: Machines, Models, and a Database as the Nerve Center
  A high-level tour of the components in my trading system—two Linux machines, two Windows machines, and a database that turns separate tools into a coordinated decision pipeline.
• Trading With a Second Brain: How ChatGPT Changed My Decision Process
  My trading bottleneck wasn’t information—it was unstructured thinking under pressure.
• Random Forest Trailing Stops: From Training to Live Stop Decisions
  How my trailing-stop engine uses a Random Forest as a risk gate to choose the tightest valid stop for current conditions.