The Mathematics of Moving Averaging

[tradingjournals]

I have written some maths that relate to moving averages, and I want to know some few related things:

1. What mathematical questions you may have about moving averages (MA). You can also write non- mathematical questions you have about MA s even if these may not be apparent to you as mathematical questions.

2. What insights/results/properties/etc you know about moving averages.

3. Books, sites and people who have written about the topic of moving averages in general and their mathematics in particular.

 

[jack hershey]

Quote from tradingjournals:

I have written some maths that relate to moving averaging, and I want to know few related things:

1. What mathematical questions you may have about moving averages (MA). You can also write non- mathematical questions you have about MA s even if these may not be apparent to you as mathematical questions.

2. What insights/results/properties/etc you know about moving averages.

3. Books, sites and people who have written about the topic of moving averages in general and their mathematics in particular.

How does one get the right to post values that do not occur in the raw data?To put it another way: How can a functional result not have a static value beteen inputs and even then land on a value that is not part of the data system?

How does one get the right to use data from sets that have no right to be overlapped?

What is the best way to use statistics to simulate first and second calculus type deriviatives?

Why do you believe the inventors of the two line theory used a Go/NoGo approach for generating "signals" from their added degrees of freedom?

What shift do you feel occurred in the use of stats when the PC came into general usage? Why were these factors in the "expressions" so important?

 

[tradingjournals]

Quote from jack hershey:

How does one get the right to post values that do not occur in the raw data?To put it another way: How can a functional result not have a static value beteen inputs and even then land on a value that is not part of the data system?

How does one get the right to use data from sets that have no right to be overlapped?

What is the best way to use statistics to simulate first and second calculus type deriviatives?

Why do you believe the inventors of the two line theory used a Go/NoGo approach for generating "signals" from their added degrees of freedom?

What shift do you feel occurred in the use of stats when the PC came into general usage? Why were these factors in the "expressions" so important?

Does any one understand what this post says?

 

[xelite777]

Quote from tradingjournals:

Does any one understand what this post says?

I guess after the fourth Heineken everything will be become much clearer...

 

[jack hershey]

Quote from tradingjournals:

Does any one understand what this post says?

It says big money makes 20% a year if it is lucky.

Have you ever wondered how editors let math articles slip into "Futures" or the like? Do they have professional panels who vet the proposed articles?

There is a book entitled "How to Lie With Statistics". About eleven chapters. You can probably clip any periodical and get an example for every chapter from one issue.

I kept track of writers and vendors before and after the advent of the PC. At that time I nonted how many YEARS it took these turkeys to switch from the pre PC to post PC indicator's settings. Pring took four years on the MACD. Even today, you can see people using unsynched indicators on ET.

Indicators generate "signals", when the maths are fucked up, then the "signals" are NOT in synch with taking the full offer of the market.

So answer the Q's. Or use the information provided to you for your purposes.

 

[xelite777]

Quote from jack hershey:


There is a book entitled "How to Lie With Statistics".

This book is surely the Bible of our politicians, the so called "scientific community" (Monsanto "studies" anyone?), the media and other professional liars who rely heavily on "studies" and "statistics" to sell their propaganda to the sheeple.

Quote from jack hershey:

Indicators generate "signals", when the maths are fucked up, then the "signals" are NOT in synch with taking the full offer of the market.

Not sure I quite follow your idea. What is "fu..ed up" in a simple 20 period moving average for example?

 

[jack hershey]

Quote from xelite777:

This book is surely the Bible of our politicians, the so called "scientific community" (Monsanto "studies" anyone?), the media and other professional liars that rely heavily on "studies" and "statistics" to sell their propaganda to the sheeple.




Not sure I quite follow your idea. What is "fu..ed" in a simple 20 period moving average for example?

It is represented by values that do not appear in the market data set values.

In maths it is not a good idea to change systems of numeration.

I know the unknowing and the naive do this. It is okay for them to do so since they know no better.

Platform vendors only have to get the purchaser to say yes. So that is what they do in the maths capability of thier platforms.

I probably would frighten you if I posted my expression for coloring bars red and for coloring bars black.

I know if I saw the expression you accept it would frighten me.

Bars change color, often, on an intrabar basis. There is a reason for this in the maths of a market system's operations. I would say most people do not deal with these kinds of details. We can agree that this is okay.

I can't make use of a 20EMA or an AMA. Perhaps you do.

I look at a week as a casual period for doubling on intra-day trading. The 20% bandied about by those who do stats for the finacial industry is a measure of some hours on one day for me; for them it is a year.

For me settlement is a pain in the neck since I prefer to use my profits at the same time they "take" their commissions. But the settlement is not mathematically capable in this era of mathematics and service.

Some people keep the money they extract from the market; I don't. I contribute it to those who are solving problems.

Do you know of people working on a terrific problem who do not have enough money?

 

[panzerman]

I suggest you review the work of John Ehlers. Specifically, "Rocket Science for Traders" gives a good review of moving averages. Also, look for his Super Smoother, and his recent work on Spectral Dilation.

 

[xelite777]

Quote from panzerman:

I suggest you review the work of John Ehlers. Specifically, "Rocket Science for Traders" gives a good review of moving averages. Also, look for his Super Smoother, and his recent work on Spectral Dilation.

Thanks for the info but in trading I like to keep it simple.

Personally I trade with a simple moving average (to stay with the trend) and an amazingly simple price action pattern that a 10-year old boy could spot in seconds.

That's it. Seriously.

Trading does not have to be complicated. In fact, I discovered a long time ago (the hard way) that simple systems with super easy rules are the most profitable, while the ones with 15 mathematical parameters and 72 indicators fail miserably.

One can make a nice living just trading one or two bread-and-butter price action based patterns.

 

[kut2k2]

Quote from tradingjournals:

I have written some maths that relate to moving averages, and I want to know some few related things:

1. What mathematical questions you may have about moving averages (MA). You can also write non- mathematical questions you have about MA s even if these may not be apparent to you as mathematical questions.

2. What insights/results/properties/etc you know about moving averages.

3. Books, sites and people who have written about the topic of moving averages in general and their mathematics in particular.

1 - You claim to have written something about the mathematics of moving averages.

2 - You don't disclose what that something is.

3 - You ask questions I would expect from somebody who doesn't know a bloody thing about moving averages.

I detect a homework troll.