Evidence Based Technical Analysis – David Aronson (Review)

Last Updated on June 19, 2022 by Quantified Trading

I have finished rereading David Aronson’s book Evidence- Based Technical Analysis, Applying the Scientific Method and Statistical Inference to Trading Signals, a book I bought back in November 2007. This is no easy read and a bit technical, but worthwhile its price tag. The book is very good for those who have no background in statistics. It’s a great book to get to know statistics without having to learn the mathematics behind the theory.

Who is David Aronson?

David Aronson got interested in technical analysis as early as the 1950s. He later went on to write technical memos for Merrill Lynch and later advised Tudor Investment Corporation in 1990.

He has wide experience in writing trading systems and how to apply artificial intelligence to trading. Aronson applied data mining to traditional computerized trading strategies.

What is the scientific method?

The book is divided into two parts. The first part examines how you can apply the Scientific Method to test your strategies. Aronson writes how you can apply the scientific method and statistical tests to determine the significance of your trading signals:

  1. Observation
  2. Hypotheses
  3. Prediction
  4. Verification
  5. Conclusion

The scientific method is the only rational way to extract useful knowledge from market data and the only rational approach for determining which TA methods have predictive power. He calls this evidence-based technical analysis (EBTA).

The first 7 chapters are all about statistics and how to use that in testing/trading. It gives a good explanation of statistical inference, frequency distributions, standard deviation, confidence intervals, data mining, probabilities, and p-value. Especially the chapter about data mining is interesting. Aronson is good at explaining in a simple manner so all these technical concepts are easy to understand. He also explains how it’s possible to derive sampling distribution on back-tested results: the Bootstrap and Monte Carlo simulation. The author argues that these two methods are the most important test you can perform on your sample to identify the degree of randomness.

The second part of the book consists of tests of 6402 trading rules.

 

 

Similar Posts

  • Hello,
    I also read this book and I only regret that my english is not so good to understand all contents. Nonetheless I also recommend this book.

    I am going to buy his second book – ” Statistically Sound Machine Learning for Algorithmic Trading of Financial Instruments: Developing Predictive-Model-Based Trading Systems Using TSSB”

    TSSB is free software. Now I am working only on excel so it will be nice to take a look for other possibilities. I have tried to learn R or Python but it is difficult in self-study mode.

    P.S.
    Have you ever tried Model Risk add-ons in excel? It is something like Solver with much more functions (optimalization is very simple) and you can also make Monte Carlo analysis. You can grab trial version for 15 days.
    Lukasz

      • Oddmund, thanks for Y post. Because I am new on your website, could you remind what kind of software do Y use in backtesing and trading?

        thanks

          • Hi,

            do you use API to trade? I know you have someone that write the codes for you but my question is how do you set the orders every day? Do you need someone to write the execution codes every time you build a new strategy?

  • Hi,

    No, I use Excl for execution. I have a simple API code there, and I can just change my formula in Excel. No need to change code for every strategy 🙂