What kind of statistic tool is helpful?

You know what I object to is when you open your mouth and in fact you don't know what the fuck you are talking about

Here is a definitive text that talks about the ways in which a trader can use statistics.

"The Mathematics of Technical Analysis" by Clifford Sherry
Published by Wiley"

I mentioned this about a dozen times, but you have your head so far up your ass, that you may not have heard it.


Steve
 
Quote

"statistics is invaluable in analyzing strategy performance; what else is there? But in actual strategy itself? "


Actually there is quite a lot. Statistics are used to characterize the underlying data not only for stationarity, but for independence, randomness, serial and auto-corellation and persistence.

Skilled researchers do this work prior to testing strategy performance. You dont seem to understand this. Unless the underlying data series exhibits specific characteristics, any trading strategy that you test will product spurious results. This is why traders start to use a strategy, then find that suddenly it isn't working. They didn't do the prep work.

And finally, this is why most traders usually don't make money with statistics. First you have to know what you are talking about. Its like learning to use any tool. Before you can use a wrench, you have to learn which end to hold in your hand.

Thanks
Steve
 
You're better off applying Fuzzy Logic to trading instead of Statistics, its just as applicable and it sounds like you know what you're talking about.

The funny thing is, you never see a guy with a PH.D in Statistics making big money in the markets. You would think they would have an edge on everyone.

Not the case, the market is random. No one can predict the way the market will move 100% of the time.
 
Quote from kut2k2:

Now you know why every statistician isn't obscenely wealthy from applying his or her vast knowledge of statistics, both basic and ultra-esoteric, to the financial markets.
More...
My sentiment exactly. Think Hound of Baskerville.

Quote from kut2k2:

Some traders may use stats in trading strategy but most of those apps are humorous at best. Bollinger bands are the clown cars of TA. :D
More...
From what I have seen and read in what passes for trading literature, I would have to agree. And you are absolutely correct about the bands: the premise is not theoretically valid. If those things actually work for anyone, it has nothing to do with the proper use of statistics.
 
Quote from steve46:

You know what I object to is when you open your mouth and in fact you don't know what the fuck you are talking about

Here is a definitive text that talks about the ways in which a trader can use statistics.

"The Mathematics of Technical Analysis" by Clifford Sherry
Published by Wiley"

I mentioned this about a dozen times, but you have your head so far up your ass, that you may not have heard it.Steve
More...
Obviously you've confused me with someone who gives a damn about whatever you've babbled in previous threads. How utterly weird.

Given your experience with having your head up your ass, you still may not get the point of the following Amazon review of your Bible:

"It is more like a draft than a serious book"
February 21, 2003
Reviewer: A reader
"Books which teach trade-methodology never give you any real result. They will show you some successful examples but never tell you what is the real chance of their systems. So you can learn different kinds of trading systems from those books but you can not do any real comparison for them. I have bought the Sherry's book with the hope that I can learn some analysis for real results. I was dissapointed. It is not difficult to understand the math in the book if you have college degree. However, there is not sufficient discussion on connecting probability/statistic theory to financial market."


I repeat: Why isn't every statistician filthy rich from using his/her tools on the markets? For that matter, why isn't every econometrician? For the intellectually honest, the answer is obvious. :)
 
Well lets be clear about this. First I DON'T have a Phd in statistics.

I do have a very strong background in math. Its just my opinion, but I don't think you have to know everything about statistics to make money. You just have to know how to use the tools that work in this business.

The text I referenced provides a good start for beginners who have a background in math. To read and understand it, you need to know math out to advanced algebra. Also It would be helpful to have a nodding acquaintance with differential equations.

Fuzzy Logic? I understand the term, but I am not competent to comment. I am sorry. I will look into it.

Thanks for your comment
Steve
 
As I understand it the basis of the options business is statistics, eg, the probability of an option expiring with any worth determines its premium.
 
Quote from steve46:

How amazing:

When the ignorance gets to be just too amusing, I figure I will step in and say something.

Statistical tools come in all varieties. The best thing you can do for your career, is to obtain a good education in statistics.

There are two broad classes of statistical tools that you should know, and learn to use. Parametric statistics are used to analyze data arranged in discrete intervals or ratios. These tools are used to describe parameters such as "mean" and "standard deviation".

Non-parametric statistics are used wherever data has flexible parameters, and no inference can be made about the distribution of the result. One example is the "Chi-Square" Test. Other non-parametric tools include:

Spearman's rank correllation
Binomial test
Mann-Whitney
Median Test

To the extent that you learn to appreciate and use these tools, you will have an advantage over your competitors. I always suggest that traders who are serious at least take the first year of statistics at your local Community College.

Good luck
Steve
More...
Steve,

Mr. Fong is looking for precision and you offer him nonparametric statistics? If those tools can improve his timing and initial protective stop placement, then either I really forgot what I learned about statistics in school about 20 years ago (quite possible), or you have a marvelous sense of humor. And if you think that a first year statistics course at a community college will give a trader an edge of any kind, then I will have to surmise that it is indeed your sense of humor.

If you are a consistently profitable trader, and you genuinely attribute your success to the use of nonparametric statistics, then my hat is off to you. You must be in very rare company indeed.
 
Continue to work on your sense of humor. What I am offering is the concept that you need more than a class of statistical tools to make a living. You need skills and an understanding of how to use these methods.

I dont think I can say it in many more ways.

You can talk abou precision all you want. I have found a way to do this that offers me an advantage. I make money almost every day in a difficult market. I am sure that is because of the work I have done and continue to do using the tools in the book I referred to.

As an example, I have an old Corvette. I know how to work on it, but I prefer that my mechanic does the work. Why? because he has more skills with the tools than I do.

I hope people get the message, but then again who am I to try to protect you folks.

Steve
 
You must log in or register to reply here.
Back
Top