MACD is the simplest thing.

Posted By black Saturday, January 22, 2011
Add to Favorites0
Author Message
black
 Posted Saturday, January 22, 2011
Supreme Being

Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)

Group: Forum Members
Last Active: Sunday, July 13, 2014
Posts: 520, Visits: 1,004
A MACD is simple. Yet, it is a good indicator. It is the difference between a smaller exponential moving average and a larger exponential moving average. For example, the EMA 12 minus the EMA 26. That is a MACD.

The MACD signal is a simple moving average of the MACD itself. The standard is 9.

The OsMA that shows in the form of a histograph on a MACD chart is simply the difference between the MACD and its signal. For example, MACD 12/26 minus Signal 9.

The formula for the EMA (Exponential Moving Average) is (2/(Periods+1)*(Price-PreviousEMA))+PreviousEMA.

The point of this post is that we can make and test the MACD of just about any indicator using the above information. Some of these might give good results.

Has anybody tested a the MACD of a lot of different indicators?

Smile
rezist
 Posted Monday, January 24, 2011
Supreme Being

Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)Supreme Being - (9,415 reputation)

Group: Forum Members
Last Active: Monday, August 15, 2011
Posts: 63, Visits: 164
Hi, black

I thought about that... but my opinion is that all good working indicators we have now were invented by good mathematicians, and I think they knew what they did, and the one, who invented MACD tried it on other indicators Smile
So I think such variations probably can give good results from time top time, but will be losing on long distances.

Though that's imho Smile and I'll be glad to hear that I'm wrong and someone suggests some good new indicator Smile
black
 Posted Monday, January 24, 2011
Supreme Being

Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)

Group: Forum Members
Last Active: Sunday, July 13, 2014
Posts: 520, Visits: 1,004
rezist (1/24/2011)
Hi, black

I thought about that... but my opinion is that all good working indicators we have now were invented by good mathematicians, and I think they knew what they did, and the one, who invented MACD tried it on other indicators Smile
So I think such variations probably can give good results from time top time, but will be losing on long distances.

Though that's imho Smile and I'll be glad to hear that I'm wrong and someone suggests some good new indicator Smile


Good thinking, rezist.

How well do the indicators work? The ones that the math gurus figured out. They work quite well... some of the time. But are there any that work well all of the time?

There has to be some kind of technique or indicator combination that works well all the time. Some method to make profits if you don't have a large starting investment amount.

Most of the Forex "advisors" who have websites and sell Forex products wouldn't be selling if their products worked well enough that they could use them. They would have long since retired wealthy using their own products.

So, what works? Obviously the standard indicators don't work well enough.

Smile
LouManChew
 Posted Sunday, February 06, 2011
Supreme Being

Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)

Group: Forum Members
Last Active: Friday, February 18, 2011
Posts: 8, Visits: 72
An excellent book that is based on cold hard scientific testing is 'the encyclopedia of trading strategies' by Katz and McCormick. It was determined that it was possible to make a profitable model with the MACD based on divergences combined with limit entries. For some reason however, RSI divergences performed very poorly. It was also determined that obvious market behaviors such as clear tradable cycles are traded away before most traders can capitalize on them. The lesson: Anything too theoretically appealing or obvious will tend not to work. I spent years testing and building models based on MVAs such as the MACD and have yet to find something that is constantly reliable. This book was the sword of Democles regarding my use of standard indicators in my trading. I think indicators that come with trading packages are used by so many traders that fail is an indication that using them in their standard way is a sure way to see an account go down to zero.
black
 Posted Sunday, February 06, 2011
Supreme Being

Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)

Group: Forum Members
Last Active: Sunday, July 13, 2014
Posts: 520, Visits: 1,004
I love your handle, LouManChew. And I agree with you most of the way. But here is where I move away.

Theories of probability suggest that the 2nd Law of Thermodynamics is truly a law... but only in most cases. What I am looking for - as are we all - is the 1 in a billion method for conquering the Forex that probability suggests must/might exist.

I tried using Linear Regressions instead of EMAs with the MACD, but that was to "tight" of a MACD. It reacted too fast and gave many false signals. What I haven't tried is 2 Linear regression / EMA averages to form the MACD. I could also try a Linear Regression as the first indicator and the EMA as the second... or vice versa.

Lots of stuff not tried. Much of it garbage. Some of it better. Nothing perfect... not yet anyway.

Thanks.

Smile
LouManChew
 Posted Wednesday, February 16, 2011
Supreme Being

Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)

Group: Forum Members
Last Active: Friday, February 18, 2011
Posts: 8, Visits: 72
Oh I'm still looking too BigGrin
I think if someone wants to try to get better results with standard package indicators, even with most custom ones, one needs to try thinking outside of the box by using statistics. I always wanted to mess around with martingale like you do in many of your systems black. I remember an idea that came to me but that I never actually tested years back. I was thinking that if a simple system like a MVA cross gave really bad win%, all I had to do was to wait for the system to make 3,4,5,6 trades resulting in a loss and then start to take signals. The chances of getting a winning signal would be much greater and since most trends happen after a consolidation the win might be bigger than usual. If you also combine a martingale and doubling your bet once you start taking signals I'm guessing the results could be interesting. I am too poor a programmer to test that but I decided to get involved and to start learning. Let me know if you think you can write it I would really be curious to see the results.w00t
black
 Posted Wednesday, February 16, 2011
Supreme Being

Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)Supreme Being - (81,025 reputation)

Group: Forum Members
Last Active: Sunday, July 13, 2014
Posts: 520, Visits: 1,004
Lots of things not tried. I have been working on a funny twist to the Martingale. Let's say you sold with a limit of 20. Rather than a Stop, if the trade went the wrong direction, at -20 pips you would sell again... but this time 2 lots... same 20 Limit. If this trade went the wrong direction as well, you sold again... this time 4 lots. When you finally won a trade, you would close all the Sell positions. The winning trade would make you a profit. The trade next in line would be zero. All the others would be a loss. Of course, there would be the spread to consider. But if you knew your market, you could keep this at a minimum and make up on the winning primary trades.

Smile
LouManChew
 Posted Wednesday, February 16, 2011
Supreme Being

Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)Supreme Being - (1,390 reputation)

Group: Forum Members
Last Active: Friday, February 18, 2011
Posts: 8, Visits: 72
Very interesting indeed. It could probably be useful to use time of day depending on what gives the signal. I'm guessing a system like this would do better in non trending markets, unless you do it on the pullbacks of course.
mwitamath
 Posted Sunday, February 27, 2011
Supreme Being

Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)Supreme Being - (4,069 reputation)

Group: Forum Members
Last Active: Monday, August 27, 2012
Posts: 22, Visits: 135
Hi, I think you'll find something better on small time frames up to 4 hours using 5 34 5 settings for macd!! by my side it helps me a lot!!

Similar Topics

Expand / Collapse

Reading This Topic

Expand / Collapse

Back To Top