Moving Averages-simple or exponential

Quote from Walther:
Neil,
What setting do you use for QQQ ? 4-5 good signals a day is great.
Thanks,
Walter
More...

Walter,
Its still work in progress really and I never said 4-5 GOOD signals a day!! lol Actually it has a low percent win ratio so far.. less than 50%.. but average gain over loss is good.. small losses mostly.

Its a ma crossover system ( tempered by the cci indicator) which is always in.. but a delayed start and closing always before end of day.. have been testing it real time to compare against back test results over last six months. still needs a lot of tweaking but nicely promising..

keep in touch if you want to know how it progresses, nearly end of feb and is ok so far altho its horrible trading conditions

Neil
 
I don't trade with MA but I like to look charts with theses bigs MAs in Fibo ratios :)

<IMG SRC=http://www.boursorama.com/cgi-local/graph?type=0&pcent=&mo=0&code=%24INDU&symbole=%24INDU&bid=3301&tc=4&cl=&pe=0&dd=-1&df=-1&st=3&is=2&i1=3&i2=5&i3=0&mm1=83&mm2=126&mm3=233&zo=&comp=0&indiceComp=1&codeComp=&symboleComp=&choix_bourseComp=pays=33&bidc=&duree=&>



<IMG SRC=http://www.boursorama.com/cgi-local/graph?type=0&pcent=&mo=0&code=350000&symbole=%24INDU&bid=3301&tc=4&cl=&pe=0&dd=-1&df=-1&st=3&is=2&i1=3&i2=5&i3=0&mm1=610&mm2=987&mm3=1597&zo=&comp=0&indiceComp=1&codeComp=&symboleComp=&choix_bourseComp=pays=33&bidc=&duree>

Quote from TG:

Have changed my moving averages from one to other lots of times, mainly based on comments from trading sites I respect. Just curious if others have gone through similar experience and what results have been. My own tend to favor exponential, but not a decisive edge. My trading tends to be intraday with a few swings.
More...
 
Quote from Neil:



Walter,
Its still work in progress really and I never said 4-5 GOOD signals a day!! lol

Its a ma crossover system ( tempered by the cci indicator) which is always in.. but a delayed start and closing always before end of day.. ..

keep in touch if you want to know how it progresses, nearly end of feb and is ok so far altho its horrible trading conditions

Neil
More...

Funny thing is that I use something similar . I try to use cci as a setup and MA's as a confirmation signal . Only 1-2 signals a day on average.
Walter
 
I use only 2 basic periods. 50 day and 200/250 day (or 40, 50 week). I run both SMA and EMA, knowing that if one is going to act as S/R it could be either this time. I look to see which the market has been "respecting" during this move and expect that one will continue to work. I don't believe that between SMA and EMA one is better... just one happens to be working better during this phase.:D
 
Quote from Walther:



Funny thing is that I use something similar . I try to use cci as a setup and MA's as a confirmation signal . Only 1-2 signals a day on average.
Walter
More...

What is even more funny is that I have also been looking into a similar system with a CCI and MA, exactly the way you seem to be using them. But this is still a work in progress as I am too busy working on other systems as well, those that rely mostly on the price action.
 
Quote from Walther:
Funny thing is that I use something similar . I try to use cci as a setup and MA's as a confirmation signal . Only 1-2 signals a day on average.
More...

Thats interesting Walter and also Wally.. we seem to be on the same basic track here. The trick is to get enough signals.. but not too many.. are you holding overnight or only day trading your signals?

For me so far I have found best results to come from a late-ish start.. so the ma's can settle after any gaps etc.. and never holding overnight, altho the jury is still out on both those factors.

I am using esignal feed running Investor RT by linnsoft as my charting... is good for backtesting and organising strategy signals for actual trading.. what you guys using?

Might start a thread one day about this cci/ma strategy.. since there seem to be a few of us working on it..

Neil
 
The EMA is a filter smoothing technique used in traditional statistics as one of simple prediction model for time series. In traditional statitiscs an alpha coefficient is used and must be chosen by optimisation technique whereas in TA alpha is fixed by a ratio formula.

On the contrary MA is a bad filter since by conception it is delayed and also autocorrelated (so the fake of believing into correlation when looking at an MA when there is in fact none) so MA is not used as filter but rather as gravity center for anticipating mean reversion. So MA is more used for support whereas EMA is more used as filter (ie as less biased as possible) for example as center line of a bollinger band. But since EMA in TA used a fixed ratio formula instead of being adapted it is also retarded so it is a bastard between a filter and a MA.

As for general remark you can use any MA, EMA, CCI, RSI on any time frame etc. since they are rather poor indicators of mean reversion what is important is then money management and "feeling". You can't have more with that by design. That's why searching the Holy Grail with stochastic indicators is a lost of time. The Holy Grail is elsewhere :D

In conclusion take 2 indicators a slow and a fast and stick with them so as to get the feeling and don't bother changing. And be courageous with all the false signals if you are not disciplined enough well :)

Quote from zentrader:

A big drawback of the SMA is how it is affected by price jumps. The SMA's memory is only limited to the lookback period. If there was a huge move 20 days ago, when you calculate your 20 period SMA tomorrow there will be a big jump in the SMA as this observation is removed from the sample. On the other hand, the EMA will not show such a jump as all previous data is weighted in the calculation.

It is a bit hard to explain, the best thing to do would be to create an artificial time series, put a huge spike in the middle of it, and look at the differences between the SMA and EMA.

For this reason I think the EMA is superior, but in reality it probably does not make much difference.
More...
 
Quote from toby400:
Fitzy and company over at "Talkstoxx" use the CCI and ma's as a trading stratagy. Lots of chart examples and commentary...all free.
Hope this helps you.
More...

Thanks Toby, yes I know talkstoxx site.. well worth looking at as you say. I did spend a bit of time watching woodie and others work the cci on paltalk but wandered off after a while since that was trading faster mostly than I am interested in.. 5 trades a day average max for me.. a quick 2pts on the es is fine but you lose that in seconds too sometimes..

I know others there also trade swing and longer term day trades.. might re-visit the threads again anyway..

Anyone know what happened to woodies cci room on platalk? Seems they must have moved on elsewhere.. no sign of them now.. hundreds of traders vanished into thin air!

Neil
 
There are some interesting issues on this discussion about Moving averages : in order to get a full overview of all MA variations including SMA, WMA, EMA, DEMA, TEMA, T3, VIDYA, and AMA a la Kaufman or Jurik, I recommend this 5 pages pdf (written by M. Jurik and named "Evolution of Moving Averages") which summarized nicely all their main characteristics :

http://www.jurikres.com/down/ma_evolv.pdf

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