Trading Systems: A New Approach to System Development and Portfolio Optimisation

Author: Emilio Tomasini and Urban Jaekle | Categories: Trading Systems, Algorithmic Trading, Quantitative Finance, Backtesting

Executive Summary

"Trading Systems" by Emilio Tomasini and Urban Jaekle, published in 2009 by Harriman House, is a rigorous, quantitative guide to developing, testing, optimizing, and evaluating mechanical trading systems. The book takes a scientific approach to system development, emphasizing statistical robustness over curve-fitting and providing frameworks for determining whether a system's historical performance is likely to persist in live trading.

The authors address the complete lifecycle of trading system development: from initial design and programming through backtesting, optimization, walk-forward analysis, and portfolio construction. Unlike many trading books that present cherry-picked results, Tomasini and Jaekle focus heavily on the pitfalls of overfitting and the statistical methods needed to distinguish genuine alpha from noise.

Core Thesis & Arguments

The authors argue that successful mechanical trading requires a disciplined, scientific methodology that guards against the human tendency to see patterns where none exist. Their central thesis is that most trading systems fail not because of poor strategy design but because of inadequate testing methodology -- specifically overfitting to historical data. They advocate for walk-forward optimization, Monte Carlo simulation, and portfolio diversification across systems and markets as the primary tools for building robust, long-lasting trading strategies.

Chapter-by-Chapter Analysis

Chapter 1: What Is a Trading System?

Defines mechanical trading systems, explains why systematic approaches outperform discretionary trading for most people, and introduces the scientific framework for system evaluation.

Chapter 2: Design, Test, Optimisation and Evaluation

The core methodology chapter covering the complete workflow: designing testable hypotheses, choosing appropriate time frames, backtesting with proper data handling, optimizing without overfitting, and evaluating results statistically.

Chapter 3: Trading System Examples

Presents several complete trading systems with full code, backtested results, and analysis. Includes trend-following, mean-reversion, and breakout systems across multiple markets.

Chapter 4: Walk-Forward Analysis

Detailed treatment of walk-forward optimization, the gold standard for testing system robustness. Explains how to divide data into in-sample and out-of-sample periods and why this prevents overfitting.

Chapter 5: Portfolio Optimisation

Extends single-system analysis to multi-system, multi-market portfolios. Covers correlation analysis, capital allocation, and the benefits of diversification across strategies and instruments.

Key Concepts & Frameworks

• Walk-Forward Analysis: The methodology of optimizing on in-sample data and testing on out-of-sample data in rolling windows.
• Overfitting/Curve-Fitting: The danger of tailoring a system too closely to historical data, destroying future predictive power.
• Monte Carlo Simulation: Randomization techniques for assessing the probability distribution of system outcomes.
• Robustness Testing: Varying parameters around optimal values to ensure performance is not dependent on precise settings.
• Portfolio of Systems: Combining uncorrelated systems to smooth equity curves and reduce drawdowns.

Practical Trading Applications

1. Always use walk-forward analysis rather than simple in-sample optimization to validate trading systems.
2. Test system robustness by varying parameters and ensuring performance degrades gracefully, not catastrophically.
3. Use Monte Carlo simulation to estimate realistic ranges of future performance and maximum drawdowns.
4. Build portfolios of uncorrelated systems trading different markets to reduce overall portfolio risk.
5. Be deeply skeptical of any system with many parameters -- simplicity correlates with robustness.

Critical Assessment

Strengths: Rigorous statistical methodology. Strong emphasis on avoiding overfitting. Practical code examples. Portfolio-level thinking is rare in trading system books.

Weaknesses: Requires programming skills and quantitative comfort. Some technical sections are dense. Market-specific examples may need updating for current market structures.

Best for: Quantitative traders and system developers who want a scientifically rigorous framework for building and validating mechanical trading strategies.

Key Quotes

"The biggest enemy of the system trader is overfitting. If you do not rigorously guard against it, you are guaranteed to fail."

"A robust system should work across a range of parameters, not just the optimal ones."

"Walk-forward analysis is the closest thing we have to a crystal ball for trading systems."

Conclusion & Recommendation

"Trading Systems" is an essential reference for anyone serious about developing mechanical trading strategies. Its emphasis on statistical rigor, overfitting avoidance, and portfolio-level thinking elevates it above the many superficial system-trading books on the market. While the technical demands are significant, the methodology presented here will save system developers from the expensive mistake of deploying overfitted strategies with real capital.