Evidence-Based Technical Analysis

[darthtrader 10]

Well, this is easily the strangest book in my entire library. I had bought this book almost immediately when it came out. I tried to read a few chapters the first day I got it and it basically sat on the shelf since that day. My intial opinion was that it was complete and utter garbage having nothing at all to do with the title. Having recently started to get into data mining I ran across mention of it in a thread, opened up it tonight and completely ripped through half of it in awe.

 

The problem with this book is the title really should be in quotes as the author is essentially proposing a new form of technical analysis based on data mining and statistical inference as opposed to showing what traditional chart patterns or what not have positive expectancy. A more honest title would be "data mining and statistical inference of financial data for dummies".

As far as trading wisdom there really is none, if you embrace the concepts of this book it basically is setting out on a path of much learning and down a very different road from traditional TA. What I found quite interesting is that it brought together alot of things I found in this area from different sources in one book. Data mining traps and what to watch out for, the importance of judging your results against a random system....In all though it really is more of a philosophy book with a stats primer.

 

If you trade purely discretionary this book will be utterly worthless. Even for alot of systems trading it may be totally worthless. If your open to the idea of setting off on a new course though of basically leveraging what the computer is good at and what your brain is good it this may be worth several times its weight in gold but an incredible amount of work lays ahead if you embrace the concept of "EBTA".

 

I'm pretty much sold on the concept.

Edited April 23, 2008 by darthtrader

[Northern boy 10]

hey Darthtrader, could you give an example used in the book?

[darthtrader2.0 10]

Well, I'm not really sure what you would want as far as an example. It basically gives a primer in bias/heuristics, frequentist statistical analysis, monte carlo vs the bootstrap as far as generating a random system to judge your system against..Its all theoretic though and only scratches the surface..your not really going to be able to do anything with the stuff in this book without alot of heavy lifting on the ideas presented. Its really just a bridge between something like John Murphy's books and the quant world.

[Do Or Die 10]

This is one of my favorite books.

 

If you are a discretionary trader stop reading after chapter 1 (or 2, cant recall) where the author talks about bias in marking chart patterns.

 

If you are interested in automated trading or automating your trading it can deliver you solid concepts. That's why I liked it. I also have some seminars by same author on data mining and backtesting traps/mistakes which helped me much.

[clmacdougall 11]

Finished the book a few months ago. It takes some time to digest but I agree with you. It's a good book.

[Beebers 10]

Some chapters of this book are required reading for the CMT certification (from the MTA).

 

5. Hypothesis Tests and Confidence Intervals

6. Data-Mining Bias: the Fool’s Gold and Objective TA

7. Theories of Nonrandom Price Motion

 

 

It also is a good primer on statistics, and proper (and improper) backtesting written in an easy understandable style (for the non-statistician).

[doubletop11 10]

The ideas that have been brought up so far are very intriguing. Being a trader that looks as much as the big levels as the small. I would like to get your view of what this book says on how human nature in the marketplace and the natural laws of supply and demand?

[Beebers 10]

I am wayyyyy not as articulate as the writer of that book. You probably can get some more info via a first look feature on amazon.com

[clmacdougall 11]

The ideas that have been brought up so far are very intriguing. Being a trader that looks as much as the big levels as the small. I would like to get your view of what this book says on how human nature in the marketplace and the natural laws of supply and demand?

 

This book does not make any attempts to make a study of the things you've mentioned. It only leads you as to how you should go about studying the markets in a statistically relevant fashion.

The title of the book is a little misleading. There are no technical analysis tools mentioned, just a leading of how to properly analyze the validity of your proposed method to see if it might be successful or not.

[doubletop11 10]

The title of the book is a little misleading. There are no technical analysis tools mentioned, just a leading of how to properly analyze the validity of your proposed method to see if it might be successful or not.

 

From what I have read on the free amazon preview it seems to talk about how to really look at a strategy or trading system from an empirical standpoint. I totally agree with that in concept! In my trading endeavors, I have made a semi-automated system that I believe doesn't rely on faith, magic numbers, expert interpretation, or lagging indicators with divergence or only correct after the fact. Rather it calculates and draw these zones the same way every single time. The interesting thing is that as I continue down this road of pure automation I'm wondering if this is a good book that will shed light into good theory of markets for current day market dynamics and how automation should look at the last 10 years of the new "electronic trading era" that we are in or if it still generalizes the topic so much that it's ambitious enough to always be correct. IE: psychic telling you general stuff that could be true of anyone. I noted that it said it looked at 25 years of buy/sells in the S&P. That's all well and good but if the underlying conditions of the marketplace have changed, "electronic trading era", would not half whatever ideas it assumes at the beginning of that time be inconsistent with what we are really seeing on the charts. Frankly, while the S&P/ ES is major market. I personally think that its dynamics are shewed from the norm of almost every other market globally. I don't wish to get on my soapbox about the dangers and traps of the ES.

I have not read the book and would like to know a little more before I buy it off amazon, please enlighten me.

sincerely thank you!

[clmacdougall 11]

This book teaches you how to identify problems with your data mining techniques. It also teaches you how to apply a Monte Carlo or bootstrapping technique to your statistical sampling process.

I'm not sure if that's what you're looking for or not. It does not uncover any secret indicators. It simply teaches you how to analyze your own methodology and decide upon it's statistical reliability to be profitable or not.