How to Apply Math to Technical Analysis?

[suby 10]

Hey guys,

 

I was wondering if any of you guys apply math to your technical analysis when trading?

 

(i.e. Determining probabilities of patterns completing using Price and Volume...)

 

If so how can you do this?

 

Don't know if this makes any sense to technical analysis but I was wondering if any of you guys out there have converted your technical analysis in trading into a science....

[suby 10]

not one person on here uses stats in there trading....?

[steve46 12]

In order to USE statistics one has to have knowledge of how statistics work.....and to do that takes at least a couple of years of study....my impression is that the general population here (read "Predictor's" comments about statistics)....has very little background and worse yet the "knowledge" they THINK they have is full of half truths and distortions....

 

If what you want is good information you will need to invest in yourself and get a solid background...go to your local College or University and take the 100 & 200 level courses of "Probability & Statistics"....if you pass those you will have a basis for applying what you learned to trading...if you have a background in general math, you might just go buy the course texts and go through at your own speed.

 

Good luck to you

[TradeRunner 10]

Hey Steve I think you once recommend in this book in your first TL thread:

 

Amazon.com: The Mathematics of Technical Analysis: Applying Statistics to Trading Stocks, Options and Futures (9780595012077): Clifford Sherry: Books

[steve46 12]

  TradeRunner said:

Hey Steve I think you once recommend in this book in your first TL thread:

 

Amazon.com: The Mathematics of Technical Analysis: Applying Statistics to Trading Stocks, Options and Futures (9780595012077): Clifford Sherry: Books

 

Yes, Cliff Sherry is good author, however if a person doesn't have at least a little bit of basic background, they will find that they don't really know how to use the information..

 

Just one man's opinion, but I think eventually you have to learn the basics

 

A couple of recommendations

 

Statistics for Dummies (Text)

or

Khan Academy (Internet)

 

Good luck folks

[SIUYA 11]

  suby said:

Hey guys,

 

I was wondering if any of you guys apply math to your technical analysis when trading?

 

(i.e. Determining probabilities of patterns completing using Price and Volume...)

 

If so how can you do this?

 

Don't know if this makes any sense to technical analysis but I was wondering if any of you guys out there have converted your technical analysis in trading into a science....

 

A lot of the maths might be handy after the technical analysis - to determine its validity, modify money management behaviour or check out the actual entries and efficiencies of a system......otherwise, you dont really need a lot of maths to apply simple TA. (my preference)

 

Thats not to say that you cant also approach it from a purely maths point of view and do it without a chart, and use maths to optimise parts of a strategy......I suggest it will all depend on your love of maths.

 

(check out people like ralph vince, kelly formula, optimal f, Ed Thorp, Larry Williams - there is plenty of info and debates out there - otherwise start as others suggest - at the beginning - remember there is the off said "lies, lies, and damn statistics" a lot has to do with how its presented, what it represents and then if you can follow it.)

[suby 10]

  steve46 said:

In order to USE statistics one has to have knowledge of how statistics work.....and to do that takes at least a couple of years of study....my impression is that the general population here (read "Predictor's" comments about statistics)....has very little background and worse yet the "knowledge" they THINK they have is full of half truths and distortions....

 

If what you want is good information you will need to invest in yourself and get a solid background...go to your local College or University and take the 100 & 200 level courses of "Probability & Statistics"....if you pass those you will have a basis for applying what you learned to trading...if you have a background in general math, you might just go buy the course texts and go through at your own speed.

 

Good luck to you

 

Hey Steve,

 

Thank you for your reply.

 

I don't have an undergrad in stats but I have taken stats classes in my undergraduate degree. To what level of stats do you suggest having? Are there any books do you recommend?

[suby 10]

  steve46 said:

Yes, Cliff Sherry is good author, however if a person doesn't have at least a little bit of basic background, they will find that they don't really know how to use the information..

 

Just one man's opinion, but I think eventually you have to learn the basics

 

A couple of recommendations

 

Statistics for Dummies (Text)

or

Khan Academy (Internet)

 

Good luck folks

 

Please disregard the comment above

 

I was wondering what is your opinion on Cliff Sherrys book?

 

I have it and have read the first few chapters. Do you apply some of the knowledge in this book to your trading?

[steve46 12]

  suby said:

Please disregard the comment above

 

I was wondering what is your opinion on Cliff Sherrys book?

 

I have it and have read the first few chapters. Do you apply some of the knowledge in this book to your trading?

 

First, the more you know, the better.....in all things...the problem is that you have to invest quite a bit of time to get enough knowledge to make a significant difference...I would estimate that taking the first year of probability & statistics is a minimum for most serious professionals...You also need to know how to use an Excel Spreadsheet, how to import data and to analyze it using the available menu options...

 

Sherry's book is an excellent example of what I was saying...He goes through many difference aspects of statistical analysis...if you have some background you can follow what he is doing and make sense of it...if on the other hand you don't have an adequate background it will probably not make much sense to you...You probably won't like this but it is just like learning math after not having done it for a while...if for example you took beggining

algebra in high school or college and then did nothing with it for a while, you cannot expect to simply move ahead to intermediate algebra without doing some review...do you see what I mean...? Sorry but in this case there aren't many "shortcuts"....

 

If you know someone with a good solid background it may speed up your learning process if you can ask them questions. This is why the Khan Academy is a good idea for many students. You can go at your own pace and learn quite a bit.....

[Tams 20]

[Tams 20]

if you did not study stats in uni,

i would not bother starting now.

getting a passing knowledge is ok, but

applying stats requires deep understandings,

and deep understanding in stats do not come overnight.

 

there are many ways to skin a cat,

some trade with high maths,

some trade with basic arithmetic,

some even trade without a chart.

 

many ppl go through life without academic stats

they lived just fine

so did you so far

 

be happy with what tools you have

and learn to apply them,

there are money awaiting.

[addchild 10]

The purpose of applied statistics is to answer questions, If you don't know what it is your looking for there is nothing to find.

 

Once you have an idea of what it is that you're looking for, and what questions you are trying to answer, you can venture down the rabbit hole.

 

UC Berkeley Webcasts | Video and Podcasts:

 

Statistics 110: Probability | Harvard Video Course

 

 

Khan Academy

[ForexTraderX 10]

If your looking to determine how price patterns and such play out in terms of their statistical edge... i strongly suggest you google bulkowski. He's written several books on what your asking about, and he has a great free website (not sure the URL..google him, it'll come up)

 

for example, a bearish engulfing candle on a daily chart in any given equity in the stock market has I believe a 78% chance of accurately forecasting a 5% drop in the stocks price (if I remember correctly anyway... I know for a fact it's over 68%)

 

How did he determine this? He found over 30,000 examples in over 500 different stocks taken over a 10+ year period of time, and he has software he developed that specifcally screed for this criteria. This was the result.

 

Not exactly sure what your looking for, but if you want to see statstical validity in your chart patterns, bulkowski probably has what your looking for.

[SIUYA 11]

http://thepatternsite.com/setups.html

 

what you were looking for

[suby 10]

  SIUYA said:

http://thepatternsite.com/setups.html

 

what you were looking for

 

Thank you very much for this!

[suby 10]

  ForexTraderX said:

If your looking to determine how price patterns and such play out in terms of their statistical edge... i strongly suggest you google bulkowski. He's written several books on what your asking about, and he has a great free website (not sure the URL..google him, it'll come up)

 

for example, a bearish engulfing candle on a daily chart in any given equity in the stock market has I believe a 78% chance of accurately forecasting a 5% drop in the stocks price (if I remember correctly anyway... I know for a fact it's over 68%)

 

How did he determine this? He found over 30,000 examples in over 500 different stocks taken over a 10+ year period of time, and he has software he developed that specifcally screed for this criteria. This was the result.

 

Not exactly sure what your looking for, but if you want to see statstical validity in your chart patterns, bulkowski probably has what your looking for.

 

Thank you very much for sharing this with me, this is exactly what I was looking for!

 

Do you trade according to Bulkowskis methods?

[suby 10]

  addchild said:

The purpose of applied statistics is to answer questions, If you don't know what it is your looking for there is nothing to find.

 

Once you have an idea of what it is that you're looking for, and what questions you are trying to answer, you can venture down the rabbit hole.

 

UC Berkeley Webcasts | Video and Podcasts:

 

Statistics 110: Probability | Harvard Video Course

 

 

Khan Academy

 

Thank you for this ADDChild!

[ForexTraderX 10]

  suby said:

Thank you very much for sharing this with me, this is exactly what I was looking for!

 

Do you trade according to Bulkowskis methods?

 

Heh... ya, I kinda figured this is what you were looking for. I remember once upon a time having this same question on some long gone chat room, and shortly after found bulkowski's books, then later, his website.

 

Do I trade according to bulkowski's methods. Interesting question... No, I can't say that I do, but this does not take anything away from his work!

 

You see, I'm a day trader primarily. I may hold a trade for 2 or 3 days on occassion, but the vast majority of my trades resolve within 24 hours of entry.

 

Bulkowski does not day trade to my knowledge. So it goes nearly without saying that we obviously have 2 very different styles of trading.

 

That being said, I use bulkowski concepts heavily to help me determine my daily bias.

 

For example, If I see a bearish engulfing candle on a daily chart of the EUR/USD, or GOOG, or BAC...etc... I know that there is about a 79% chance that the following day will close lower than the bearish engulfing day.

 

With this in mind, I will ONLY take short trades. If this aligns with the overall daily trend, even better.

 

This is how I use bulkowski's concepts. I don't trade using his method exactly, but if a bulkowski pattern has a high probability of determining the following daily candlestick (like the bearish engulfing candlestick pattern)... then I will only trade in the direction of the likely resolution for that following day. I would never take any day trades long if I saw a bearish engulfing candle on a daily chart in just about any market, I would only look for places to get short.

 

Hope this helps...