Technical Analysis book recommendations

[Indoril]

Hi all,

I am interested in the content of the CMT certification textbooks however I'm not really concerned about getting the certification itself. Unfortunately the textbooks are rather expensive so I was wondering if anyone had any recommendations of other books of similar content?
https://www.mta.org/cmt/chartered-market-technician/
https://www.efficientlearning.com/cmt/

Regards,
Ty

 

[Modest]

http://thepatternsite.com/

 

[Boggo]

When you get a basic understanding of technical analysis then read this at least twice.
It will have a positive influence on your way of thinking and applying analysis.

Just my

 

[Indoril]

Thanks Boggo. Actually I've been looking into quantitative/algorithmic systems so this perfect.

 

[Boggo]

Thanks Boggo. Actually I've been looking into quantitative/algorithmic systems so this perfect.

Don't forget Howard Bandy's books too, he is a significant contributor to this site and you will find a lot of useful info from him if you use the search thingy on here.

 

[howardbandy]

Don't forget Howard Bandy's books too, he is a significant contributor to this site and you will find a lot of useful info from him if you use the search thingy on here.

Thanks for the plug.

Begin with the free stuff. Watch the videos I have posted to YouTube:
http://www.blueowlpress.com/video-presentations

Understand the concept of "stationarity." In order for patterns, rules, and profits to continue in the future as they have in the historical data you will be mining as you work through system development, the future must resemble the past. That is, the distribution of what is important to your trading must be stationary.

Understand the concept of "risk normalization." Given two or more alternative uses of funds, use the techniques of risk normalization to be able to compare them, choose among them, and decide which (if any) should be traded.

My mantra: Trade frequently, trade accurately, hold a short period, avoid serious losses.

I recommend trading a single issue long/flat. Create as many systems that trade a single issue long/flat or short/flat (but this is harder) as you wish. Evaluate their performance daily. Trade the current best one. One. Creating a portfolio gives the illusion of safety through diversification, but guarantees that some of the funds are being used suboptimally.

The field of technical analysis / quantitative analysis is changing very rapidly. I feel traditional technical analysis based on visual examination of charts, drawing trend lines, looking for traditional patterns such as flags, support, resistance, divergence, etc is no long state-of-the-art and no longer worth learning. If you have the math aptitude for it, consider going straight to machine learning. Read the thread: "Getting started in machine learning for trading."

Separate the trading system from the trading management system. The trading system evaluates the price data in search of patterns that precede profitable trades and issues signals to buy or sell. Importantly, it has no position sizing component. The trading management system evaluates recent trades, determines the health of the system, estimates risk, estimates profit potential, computes maximum safe position size.

There are no challenger tournaments, no handicaps, no do-overs. From your first trade through your last, your competition is the best traders in the world -- Goldman Sachs, David Shaw, James Simons.

The single best book is "Thinking, Fast and Slow," by Daniel Kahneman. Dr. Kahneman explains how easily we fool ourselves. My recommended reading list, as of mid-2016, is:
http://www.blueowlpress.com/wp-content/uploads/2016/08/FT-Bibliography-Appendix-D.pdf

Search through ASF postings for more of my musings and rants.

Best regards, Howard

 

[Indoril]

Thanks for the plug.

Begin with the free stuff. Watch the videos I have posted to YouTube:
http://www.blueowlpress.com/video-presentations

Understand the concept of "stationarity." In order for patterns, rules, and profits to continue in the future as they have in the historical data you will be mining as you work through system development, the future must resemble the past. That is, the distribution of what is important to your trading must be stationary.

Understand the concept of "risk normalization." Given two or more alternative uses of funds, use the techniques of risk normalization to be able to compare them, choose among them, and decide which (if any) should be traded.

My mantra: Trade frequently, trade accurately, hold a short period, avoid serious losses.

I recommend trading a single issue long/flat. Create as many systems that trade a single issue long/flat or short/flat (but this is harder) as you wish. Evaluate their performance daily. Trade the current best one. One. Creating a portfolio gives the illusion of safety through diversification, but guarantees that some of the funds are being used suboptimally.

The field of technical analysis / quantitative analysis is changing very rapidly. I feel traditional technical analysis based on visual examination of charts, drawing trend lines, looking for traditional patterns such as flags, support, resistance, divergence, etc is no long state-of-the-art and no longer worth learning. If you have the math aptitude for it, consider going straight to machine learning. Read the thread: "Getting started in machine learning for trading."

Separate the trading system from the trading management system. The trading system evaluates the price data in search of patterns that precede profitable trades and issues signals to buy or sell. Importantly, it has no position sizing component. The trading management system evaluates recent trades, determines the health of the system, estimates risk, estimates profit potential, computes maximum safe position size.

There are no challenger tournaments, no handicaps, no do-overs. From your first trade through your last, your competition is the best traders in the world -- Goldman Sachs, David Shaw, James Simons.

The single best book is "Thinking, Fast and Slow," by Daniel Kahneman. Dr. Kahneman explains how easily we fool ourselves. My recommended reading list, as of mid-2016, is:
http://www.blueowlpress.com/wp-content/uploads/2016/08/FT-Bibliography-Appendix-D.pdf

Search through ASF postings for more of my musings and rants.

Best regards, Howard

Thank you so much for your detailed response. I'm taking it all in.
I have certainly been more interested in quantitative trading over traditional technical analysis but I guess I was more looking into the maths of how certain indicators work and why, in hopes it would help in developing a quantitative system. Basically, I haven't been able to find many good sources on how to start learning quantitative trading outside of the very basics. Having said that, I have started this course on machine learning at Udacity, Machine Learning for Trading, which so far has been very good.

 

[Modest]

I feel traditional technical analysis based on visual examination of charts, drawing trend lines, looking for traditional patterns such as flags, support, resistance, divergence, etc is no long state-of-the-art and no longer worth learning. If you have the math aptitude for it, consider going straight to machine learning. Read the thread: "Getting started in machine learning for trading."

Translated: Technical Analysis doesn't work but my method works so buy my book!

 

[rb250660]

Check out the books, podcasts and videos of these two guys:
- Kevin Davey
- Howard Bandy

Everything else is recycled rubbish that doesn't work anymore and not worth your money. For ideas see quality blogs, publications, etc., take them and add your own touch. Generally, whatever everyone else is doing, do the opposite or do something completely unheard of. Make your own indicators and unique ways of describing market action to make trading decisions from. As stated above, head and shoulders, flags, divergence, Elliot Wave (don't get me started on this crap), etc. is all a waste of your time. Also books on psychology, journaling your trades, etc. is BS. Psychology should be largely dealt with during the development phase. If you want to get into quantitative systems my recommendation is to get AmiBroker Professional, good data (don't ever be a tight a** with data) and do focused study on Kevin's and Howard's materials. Probably a $1000 initial outlay and keep it to that.

But this is just my opinion.

Good luck.

 

[Trembling Hand]

Psychology should be largely dealt with during the development phase.

One line that sums up my 8700 post. love it.

 

[howardbandy]

Translated: Technical Analysis doesn't work but my method works so buy my book!

What I said was chart analysis and traditional indicators do not work as well as they did as recently as a few years ago. I gave links to several hours of videos and other free material that will help explain what I meant and give techniques to analyze risk and profit potential. No need to buy anything. After watching and reading the free material, the reader will be able to decide which direction to continue. Is that approach not reasonable? What would be better?

 

[howardbandy]

One line that sums up my 8700 post. love it.

I agree. Psychology can be quantified, encapsulated into the objective function, and used to direct development. During live trading it continues to measure system health, determine maximum safe position size, estimate risk, and estimate future profit potential.

 

[howardbandy]

If you want to get into quantitative systems my recommendation is to get AmiBroker Professional, good data (don't ever be a tight a** with data) and do focused study on Kevin's and Howard's materials. Probably a $1000 initial outlay and keep it to that.

If you want to use a traditional trading system development platform, AmiBroker and a subscription to Norgate Premium Data are a great combination. $1000 will easily fund the first year and will leave plenty of change for some reference material. Even less expensive if you start with Yahoo or Google free data. Add Norgate when you want higher quality data.

If you want to go the machine learning route, I recommend Python for the development platform and Quandl for data. Python will be free forever. Quandl will be free until you decide you want higher quality curated data, then subscribe to their premium service at about the same price as Norgate.

Traditional platforms are limited to one model type -- decision trees. Machine learning expands the number of models to several dozen, including decision tree.

 

[Indoril]

I currently have a subscription to another platform with data but after the 12 months is up I was planning on going down the Amibroker/Norgate route. I've been trialling it and AFL is really nice to use in comparison to this other platform I've been using.

I had considered just using Yahoo data but I read somewhere that their ASX prices only go to 2 decimal places and so not accurate enough.

 

[Quant]

Translated: Technical Analysis doesn't work but my method works so buy my book!

What I said was chart analysis and traditional indicators do not work as well as they did as recently as a few years ago. I gave links to several hours of videos and other free material that will help explain what I meant and give techniques to analyze risk and profit potential. No need to buy anything. After watching and reading the free material, the reader will be able to decide which direction to continue. Is that approach not reasonable? What would be better?

Ignore the haters Howard , i appreciate the work you do , thanks for your input

I have a couple of Howards books and one the he very kindly posted to me 4 free after a very generous offer here on ASF , Howards videos and books have singularly been the most helpful things from an individual ive had in my 18 years of trading journey and i wish i went down this road a decade ago .. Kudos to Mr Bandy , we are lucky to have him here

 

[Quant]

Depends on your objectives but this worth a read for many fwiw

 

[tech/a]

Translated: Technical Analysis doesn't work but my method works so buy my book!

I cant believe you said this?
It shows immaturity, ignorance, lack of respect ( Howard's more than earnt it! ) and While Howard pushes his barrow
He gives volumes of information you wouldn't get free from anyone I know.

If you have no interest or in put
Say nothing

 

[gartley]

Very little has changed and the same patterns are present in todays charts compared to 100 years ago!! They maybe not be as frequent and even some newer ones have evolved but they are still there. What TA really boils down to is finding an approach that suits your trading style and personality.

 

[jjbinks]

Depends on your objectives but this worth a read for many fwiw



View attachment 70459

is there evidence that this book is good

 

[howardbandy]

It is so very easy for us fool ourselves. Begin by reading Kahneman's book, "Thinking, Fast and Slow."

If you will be selecting your trades using subjective interpretation of chart patterns, traditional books might be helpful.

If you will be using rule-based models, define rules precisely, test everything rigorously using sound modeling techniques, and validate using virgin data not used during development.

The field of trading system development is changing with astonishing rapidity. The overwhelming majority of counterparties to your trades will be machine learning artificial intelligence based. From your first trade on, you will be in the deep water and strong current with Goldman Sachs, David Shaw, and the like.

Best, Howard