Urban Jaekle

for curve-fitting. This book describes the nitty-gritty stuff most authors gloss over (stop loss sizing, trailing stops, profit target selection, etc). This book is a highly practical approach to engineering employed in trading system development. Readers with a background in physics or engineering will feel at home. The authors start with a very simple moving average crossover system, and then slowly add complexity. At each step, the authors perform a sensitivity analysis to both a) verify that the added complexity is worthwhile, b) choose optimal (higher return) parameters, and c) most importantly, choose parameters that are robust to the ever-changing complexities of the markets. Finally, using the completed system, the authors perform a robustness test over many instruments and time frames. If you are looking for a book to copy a system out of and make millions, this is not for you. If you are at least somewhat experienced, and know that trading system development is more than just buy and sell rules, this book should be very helpful. Do not buy this book for the automated system it describes - buy this book for the robust system development process it presents. My only gripe is that there is no discussion regarding backtesting vs forward testing. There is an excellent chapter describing walkforward optimization. But, the authors stop just short, and do not compare the out-of-sample backtest results with the results from forward testing."