how do u determine if something is statistically significant?

>Nice to see some stat talk... "Taguchi Loss Function", it's been years since I've heard that mentioned... sometimes I miss stats
(I used to be a quality engineer/statistician for Ford Motor Co. etc years ago)..

Great :). I was more humble in that function just a spc engineer in a food industry and then in a pharma industry where I was also software engineer and I finally switched to software industry as I wanted to be independant and create my own business.
Did you see Deming ? I never saw him, I just met a great friend of him who told me the story of Quality in Japan and then America and how upset Deming was that America was so short-term sighted and how Japan would decline because the new Japenese managers were formed in American Business Schools and came back with the same mentality. It was before Japan collapse when I was told that so that I was very skeptical finally Japan has collapsed.

>fwiw, I look at the market in terms of the inverse of our goal in a production environment, where we look at keeping processes in control, improving mfg process capability (Cpk) and reducing variation..

To say the least you're right the market is not really in a state of statistical control :). According to my model it is even the opposite: it maximises the variability that's how it realises "efficiency" (... for him but for economy this is not really efficiency ...).

>in the markets, I'm looking for a consolidation that then shows a process of "going out of control", eg buyers or sellers clearly taking over in the time/sales and 1-minute candlestick chart, moving the price outside of a normal curve for a breakout/breakdown..
As for me I don't use stochastic approach - but my deterministic model for which I will nevertheless use spc for improving my automate trader precision - except for rough estimation of extreme zones and I would rather use Tchebycheff than normal law.

>I used to publish on QFD and other quality deployment tools, here's an article mention from one of my ASQ articles back in 1992 (see the first article link under "Articles"):
http://qfdcapture.com/references.htm

Huh you're a QFD guru then :D

>There's plenty of statistical methods application to stock chart trading and the data points, I'm thankful for all the great corporate training + experience opportunities I was given to understand data patterns early in my career.

When I was at engineering school, I used to hate statistics, it was so boring I sometimes didn't even go to the lesson. It is really when I enter real world industry and met the guy above that my eyes opened, not immediately because my brain was too "mathematically" oriented because of school, I was always focused on maths and each time I was said, it's not important, it's just calculation details you will find in any book or a software can do it, what's important is the true philosophy behind : After some experiences now I understand why "mathematicality" is not really important but "physicallity" :D.


Quote from Ken_DTU:

Nice to see some stat talk... "Taguchi Loss Function", it's been years since I've heard that mentioned... sometimes I miss stats (I used to be a quality engineer/statistician for Ford Motor Co. etc years ago)..

fwiw, I look at the market in terms of the inverse of our goal in a production environment, where we look at keeping processes in control, improving mfg process capability (Cpk) and reducing variation..

in the markets, I'm looking for a consolidation that then shows a process of "going out of control", eg buyers or sellers clearly taking over in the time/sales and 1-minute candlestick chart, moving the price outside of a normal curve for a breakout/breakdown..

having the bkgrd of working in production data processes for years, it's helped, since the charts and data patterns that we used to help improve mfg process control, can be used w/modifications for loss of control breakout points.

I used to publish on QFD and other quality deployment tools, here's an article mention from one of my ASQ articles back in 1992 (see the first article link under "Articles"):

http://qfdcapture.com/references.htm

There's plenty of statistical methods application to stock chart trading and the data points, I'm thankful for all the great corporate training + experience opportunities I was given to understand data patterns early in my career.

ken
More...
 
Quote from Gordon Gekko:

here is why i'm asking this...

when i backtest something, the less i go back in time, the better the results are, however, they still aren't bad.

but how far back do i need to go? don't i want to know what the market is doing NOW? the market was different 8 years ago, why should i include that in my tests?

what i want to do is walk forward. test only on a fixed number of days, every so often. however, do i need a certain number of trades for the tests to become meaningful? i want to use the least amount of trades, but still statistically significant.

thx
More...

Example. One of the stocks I trade i had a 73% win record for one month. When I review all trades for the last year in this stock, the average settles down to about 53-54%. A bit more realistic.

The program Prosizer can take ALL your trades, scramble the order 500 times and come up with a realistic probability of sucess, max drawdown, positive expectancy, equity curve, and tell you how many shares, etc to trade.

http://unicorn.us.com/trading/prosizer.html
 
yes re Deming, I was able to see him once, when I was at Ford... he always surrounded himself with a bevy of young women who were his assistants... rather gruff and broad-picture re his comments, a nice experience though, to have seen him the one time...

my master's thesis was on masaki imai's "kaizen' book, excellent...

interesting re looking at variation... explaining it, is a key to the market, eg characterizing sources of variation, then capitalizing on them... example when one sector is stronger than the others, eg SOX semis up +0.8% at 9:50 am and all the other sectors less than .4%, I'll be looking for plays more in semis than other sectors, etc..

true re stats are completely boring when studied in book fashion.. but when you're out there on a live manufacturing production line that produces millions of $$ of product monthly and see the use of realtime data to minimize line errors/downtime and improve yields etc... it's very interesting... a lot of applied math to improve "how things get made" ..


ken
 
Deming with his great age with so many young women :D

I don't trade stocks except for 3 months when I realised that they were not worth compared to futures at least for daytrading.

Stats and anything at school could be interesting if they have coped with real industry problems. I was never doing the (obligatory) maths exercices during my engineering preparation because I found them useless whereas I like to cope with more big problems so that I've done all the big problems not the small ones: the teacher suspected once that I wasn't doing the exercices and waited that I showed him that I've done them, I have to faint to look for them until he was fed up to wait :). In automation it was the same thing, I once get a very bad note at the final exam because I just refuse to learn the syntax of the langage by heart - I don't see the interest - whereas at the project I got the best note because it copes with (nearly) real problem: the automation of a whole factory. Happily the project counted as twice as the exam so it saved me :). I got the best note not because I worked for it but because I was interested by the subject and so do my best for that. The note is a consequence not a target. I'm pretty sure that in general many children that are apparently bad pupils have some motivations that could be exploited to turn them into good pupils. But schools today are more interested by quantity than by quality: children have to learn more and more materials ... and then just have to forget them all the more later so that it is really useless. They should rather be teached how to learn by themselves by discovery process which is one of the primary motivation of a human at least when he is a child because when one comes adult some curiosity is lost (but some stay childs all their life and so curiosity stay with them : I think it's my case :) ).


Quote from Ken_DTU:

yes re Deming, I was able to see him once, when I was at Ford... he always surrounded himself with a bevy of young women who were his assistants... rather gruff and broad-picture re his comments, a nice experience though, to have seen him the one time...

my master's thesis was on masaki imai's "kaizen' book, excellent...

interesting re looking at variation... explaining it, is a key to the market, eg characterizing sources of variation, then capitalizing on them... example when one sector is stronger than the others, eg SOX semis up +0.8% at 9:50 am and all the other sectors less than .4%, I'll be looking for plays more in semis than other sectors, etc..

true re stats are completely boring when studied in book fashion.. but when you're out there on a live manufacturing production line that produces millions of $$ of product monthly and see the use of realtime data to minimize line errors/downtime and improve yields etc... it's very interesting... a lot of applied math to improve "how things get made" ..


ken
More...
 
Finally this rejoins Deming's thought about psychology of motivation (see below) and this of course is also applicable to trading: what is your intrinsic motivation to come there ? Mine is that it is a business and that I've found an edge to cope with this business. I've never played lotto or put a cent on a casino's table, if I haven't had any edge in stock market I wouldn't care coming because my motivation is not gambling. If my motivation was gambling then I would rather choose the real casino : with no edge it's more joy and less costfull. I had a grandmother that was very fond of casinos, happily my genes didn't inherit from that :D.

http://www.elitetrader.com/vb/showthread.php?s=&postid=394133#post394133
Deming's thought:

A Knowledge of Psychology includes a knowledge that people are different from one another and knowledge of how to use these differences to optimize everybody's abilities and inclinations. It includes the concepts of intrinsic and extrinsic motivation and the phenomenon of over justification.

People are born with intrinsic motivation that is often destroyed by various practices at school and work. Grades cause students to work for grades, or a reward from parents for grades, rather than to work for the purpose of learning. Rewards at work such a merit pay cause people to work for rewards rather than for job satisfaction and to find meaning in their work and lives. Some extrinsic motivation helps develop an individual's self-esteem, but over emphasis on extrinsic motivation eventually destroys an individual's intrinsic motivation and leads to detrimental effects on self esteem. Work and life eventually have no meaning.

Quote from harrytrader:

Stats and anything at school could be interesting if they have coped with real industry problems. I was never doing the (obligatory) maths exercices during my engineering preparation because I found them useless whereas I like to cope with more big problems so that I've done all the big problems not the small ones: the teacher suspected once that I wasn't doing the exercices and waited that I showed him that I've done them, I have to faint to look for them until he was fed up to wait :). In automation it was the same thing, I once get a very bad note at the final exam because I just refuse to learn the syntax of the langage by heart - I don't see the interest - whereas at the project I got the best note because it copes with (nearly) real problem: the automation of a whole factory. Happily the project counted as twice as the exam so it saved me :). I got the best note not because I worked for it but because I was interested by the subject and so do my best for that. The note is a consequence not a target. I'm pretty sure that in general many children that are apparently bad pupils have some motivations that could be exploited to turn them into good pupils. But schools today are more interested by quantity than by quality: children have to learn more and more materials ... and then just have to forget them all the more later so that it is really useless. They should rather be teached how to learn by themselves by discovery process which is one of the primary motivation of a human at least when he is a child because when one comes adult some curiosity is lost (but some stay childs all their life and so curiosity stay with them : I think it's my case :) ).
More...
 
Quote from harrytrader:


This implies that with only 30 trades you won't get much about consistency of mean if you decide to take a sample of a size of 30 since you have only one SINGLE SAMPLE of 30. It would be better in fact to reduce the number thirty for getting more samples than the contrary (in quality control one often uses MULTIPLE samples of only 4 or 5 items). Secondly the premisce is important: INDEPENDANCY.
More...

How do you tell if the trades come from different samples? How can you tell that the different market you look at are independent? I think most markets aren't independent.

Can you recommend some books on statistics? I think as long as I can't tell an edge from random I can't tell if I'm winning or losing.
 
You must log in or register to reply here.
Back
Top