Trading Systems for Software Engineers

Book Architecture

Part I - Foundations (written in full below)

1. Why Engineers Lose Money
2. The Market as a System
3. The API Surface: Instruments, Orders, Execution
4. Expectancy: The Only Equation That Matters
5. Your First System

Part II - Building

1. The Data Layer - survivorship bias, corporate actions, timestamps and timezones, tick vs. bar aggregation, why two vendors disagree about the same candle
2. Backtesting as a Test Harness - why it's an integration test against a mock that lies to you, and how to make it lie less
3. Overfitting - degrees of freedom as a budget, plateaus vs. peaks, the multiple-comparisons problem
4. Validation - in-sample/out-of-sample discipline, walk-forward analysis, Monte Carlo, deflated performance metrics
5. Position Sizing and Risk - fractional risk, Kelly and why fractional Kelly, correlation and portfolio heat, drawdown budgets

Part III - Categories of Edge

1. Trend Following - the oldest surviving edge, why it works, why it's miserable to trade
2. Mean Reversion - pairs, stat arb, and the difference between "cheap" and "falling"
3. Breakout and Volatility - regime dependence, volatility as an asset rather than a filter
4. Structural and Carry Trades - risk premia you get paid to absorb, and the tail that eventually arrives
5. Microstructure - what the professionals actually do, and an honest account of why you can't compete there

Part IV - Operating

1. Backtest → Paper → Live - the three transitions, each of which will surprise you
2. Production Engineering - idempotency, reconciliation, state recovery, kill switches, observability, the 3 a.m. page
3. Psychology as a Systems Problem - treating your own behaviour as an unreliable component to be designed around
4. Records, Tax, and Jurisdiction - why this is a chapter and not a footnote
5. Knowing When to Stop - killing a strategy, and killing the project