The difference between a MA and a non lag MA?

Hi folks


I pretty much make my livelyhood through MAs. Recently i had a look at a strategy that uses non lag MA, and i thought i might incorporate it into my existing strategy, if there is an advantage in doing so.

The MAs i use for moving S/R is 20, 39, 69, 200 on M1 apply to close, exponential, the MA 200 being the strongest, something you can count on in 90% chance of halting price when things are moving a bit. These are quite generic, and tells me something about almost all pairs/gold/oil or whatever. It would seem they get pretty physical/respected at M30 and H4 too. So it really seem to have to do with the human subconsious, or its just that central banks have these MAs and respect them/look at them. Not that i understand why they bother with M1, therefore im not sure that they do.

I use 4 other MAs for levels/channels to contain price, calculated for 4 diffrent strenghts of PA pushing its waves through the market. There probably is no point in me going into this deeper then this, because there is a lot to say about it.

These MAs that i use, is not non lag MAs, but i am wondering if there is something i am missing because im not using non lag MAs. Ive had a look around for info in non lag MAs, but i havent yet found what i was looking for.
So my wish is to work on some things together.


I use levels on some of my existing MAs as i said, but i cant seem to find this function in the non lag MA v.7.1.1. I would want to be able to make visual levels at certain distances from the MA, lets say 8, 13 or 23 pips away, just as we can do with the MT 4 stock MA. How can that be achived with the non lag MA? Maybe it is possible and i just missed something...

What is the diffrence between a MA and a non lag MA mathematicaly? Obviously the non lag MA follows the price more closely, but how is this calculated?

I think you're missing the fact that all MA's are lagging.

a 240MA on a 1M chart would be the 4hour average, 60MA would be the 1-hour average.

on a 5min chart, the 1hour would be the 12MA, and the 4H would be the 48MA

No im not missing that. But i would like to know what the diffrence is between these two types of MAs.

difference between the types of MA's? like SMA, EMA, WMA, that kind of thing? the difference is just how they're weighted or calculated.

That difference

the difference is in the formula of how they are calculated. I don't have an answer for the question sorry dude, I'm not into MA's like that. I'm sure you can google that and get an answer.

I was just confused by the whole "non-lag MA" thing you're talking about...

http://en.wikipedia.org/wiki/Arnaud_...Moving_Average

If you still want to know, check this out

Unfortunately, there is nothing called a non-lag MA till now. There are only attempts to reduce the lag of the MA by using different weighting methods in the MA formula.

The SMA (Simple Moving Average) averages the closing prices of the mentioned periods with equal weights

The WMA (Weighted Moving Average) averages the closing prices of the mentioned periods with decreasing weights proportionally, with the highest weight being given to the recent data and then weights decrease uniformly as we go back

The EMA (Exponential Moving Average) uses a smoothing factor. The smoothing factor is calculated by 2 / 1+N where N is the desired periods. This ensures that the weighting is exponential instead of uniform across the desired periods and this means that as you go back to previous periods, the weighting of the data decreases exponentially.

The DEMA (Double Exponential Moving Average) attempts to further reduce the lag of EMAs and uses the following formula:

DEMA = 2xEMA - EMA(EMA)


The TEMA (Triple Exponential Moving Average) again attempts to reduce the lag of EMAs further and uses the following formula:

TEMA = (3xEMA - 3xEMA(EMA)) + EMA(EMA(EMA))


Other types of MAs are available as well, none of which though is non-lagging.

Just a quick reminder for you. When you reduce the lag of a MA, you are also increasing the potential false signals. Since reducing the lag makes the MA more sensitive to price movements, it will definitely produce more false signals as prices move sideways, so having a reduced lag MA is not always a good thing.

If you want to have a look at something which is non-lag, you can try PRICE, it should be non-lag.


Best Regards,

Nader

Right on!

signal analysis concepts by john ehlers

http://www.docstoc.com/docs/9289613/...By-John-Ehlers

It gives a good explanation what and where about lag in MA.

all MA's are lagging. there is nothing like a non-lag MA or indicator.

search for nonlagema. it's an ema that uses the formula from the link i provided.

i think from my basic understanding is that normal ma is static period and doesnt take current cycle period into consideration. but a non-lag ma take current cycle period and adjust period accordingly. i suggest John Ehler sinewave or polynomial projections come to mind, but thats just my two cents, could be completely wrong.

Thanks for the explanation Narafa

Found this ALMA download (Free)

http://www.arnaudlegoux.com/?p=1

I installed it and will be using it in my scalping system tomorrow.
Thanks for the info.

You are more than welcome. Hope it helped...

Regards,

Nader

Check out the Hull Moving Average.
It is about as close to a non-lag MA as you will get
Tonkar

The Hull is the best I have seen.

DEMA, AMA, Hull, Tillson, Jurik, ...... they all attempt to reduce lag and overshoot, without compromising too much on smoothness.

However, the biggest enemy is not lag, it is randomness. For those who want a simple proof, it's this: the nearer price approximates a completely random walk, the nearer EVERY system approaches to delivering zero expectancy. Or in other words, if MAs are your signal generator, then unless they somehow align themselves with genuine market inefficiencies, they're not going to be any more effective than randomly generated signals.

Disagree with this. The nearer price approximates a completely random walk (in the gaussian, IID sense) the better chance you have at a non-zero expectancy. In moving averages and actual market behavior, it's the fat tails and non IID (serial correlation) behavior that are killers.