Quick opinion: MA Cross strategies

Are you for or against versions of MA Cross strategies?

For!
2
7%
For -- for certain market conditions.
2
7%
For -- so long as it isn't just a simple x/x cross.
3
10%
For -- so long as the MA cross is a part of some other, broader strategy.
15
52%
All of the above.
3
10%
Against!
3
10%
REALLY against -- they NEVER work out over the long term.
1
3%
 
Total votes: 29

User avatar
ironrick
Trader
Posts: 85
Joined: Wed Nov 16, 2011 7:04 pm
Location: North Idaho Boy, USA

Quick opinion: MA Cross strategies

Post by ironrick »

Just wanted to get a quick opinion of what everyone thought about MA Cross Strategies OVERALL.

I keep seeing some people like them and some people really despise them and say they never work. Very polarizing discussions.

What say you? For or against. Please take the poll.

If you have a bit more time:

Why do you hold your views? Your personal trading experience, money management, some fundamental flaw?

Under what circumstances would you/ could you see yourself using an MA Cross strategy?

Am I asking the wrong questions?

Thanks for the input!

Cheers,
Rick
"A problem well put is half solved."
garyfritz

Re: Quick opinion: MA Cross strategies

Post by garyfritz »

Simple answer: MA crosses tend to work well in trending markets, i.e. markets that move for a long way in one direction. They tend to work crappy in ranging markets -- unless the market moves happen to "sync" with the natural "resonant frequency" of the MA cross. Then they work great.

More complicated answer: (warning, math concepts ahead :D)

An MA cross is essentially an oscillator: MA(fast) - MA(slow). If you look at the math of it, MA cross systems are bandpass filters. They pass midrange frequencies while attenuating high and low frequencies.

What do I mean by frequencies? Fast, noisy moves are high freqs. Long, slow trends are low freqs. You can treat the market like a signal waveform and use digital signal analysis on it.

Positive MAs filter out high freqs -- they're called a "low-pass filter." That's why a MA smooths the price signal -- it got rid of all the high-frequency noise. Faster (shorter) MAs have a higher cutoff than slower (longer) MAs. MA(fast) - MA(slow) basically says "cut off the high freqs above the fast MA, and subtract the lower frequencies." Here's a crude illustration:

Image

The "fast" and "slow" values determine the high and low range of the bandpass.

For the übergeeks in the crowd (who me??), here's an illustration. The blue line is a sine wave of gradually-increasing cycle length (a few cycles are marked on the chart). The red line is SMA(blue,20) - SMA(blue,40). Notice how the red line peaks in value in the 13-17-bar cycle range. That's the effect of the bandpass filter allowing more of the signal through in the bandpass range. (Unfortunately at lower frequencies [longer cycles] the red line inverts and gets stronger, but that's just because an SMA cross is a lousy bandpass filter.)

Image

Now imagine feeding the MA cross a price signal instead of a sine wave. It's going to pass the frequencies in the 13-to-17-bar cycle range, and attenuate higher and lower frequencies. You're left with an oscillator that picks out 13-to-17-bar cycles and hands you that as a trading signal.

So MA crosses filter out the trend (low freq) and the noise (high freq). The result is a nice medium-speed oscillator.

**BUT**

The bandpass is tuned to a SPECIFIC FREQUENCY RANGE determined by "fast" and "slow." That's why the image above peaks in the 13-17 bar range. If the market conditions you want to trade (say, swings that go from low to high [half a cycle] in about 7-9 bars) fit into that frequency range, you're golden. The MA cross will produce perfect signals and all is wonderful.

But if your desired market conditions DON'T fall into that range, they get filtered OUT. You're left with what IS in the bandpass range, which might not be tradeable using your methods. Other market conditions (like the longer cycles at the right of the chart) get distorted.

So class, to summarize today's lecture: MA crosses work great. Except when they don't. Which is usually.
User avatar
gaheitman
Trader
Posts: 655
Joined: Tue Nov 15, 2011 10:55 pm
Location: Richmond, VA, US

Re: Quick opinion: MA Cross strategies

Post by gaheitman »

garyfritz wrote:Simple answer: MA crosses tend to work well in trending markets, i.e. markets that move for a long way in one direction. They tend to work crappy in ranging markets -- unless the market moves happen to "sync" with the natural "resonant frequency" of the MA cross. Then they work great.

More complicated answer: (warning, math concepts ahead :D)
So class, to summarize today's lecture: MA crosses work great. Except when they don't. Which is usually.

Wow, wonderful post! Thank you.

George
User avatar
ironrick
Trader
Posts: 85
Joined: Wed Nov 16, 2011 7:04 pm
Location: North Idaho Boy, USA

Re: Quick opinion: MA Cross strategies

Post by ironrick »

garyfritz wrote:Simple answer: MA crosses tend to work well in trending markets, i.e. markets that move for a long way in one direction. They tend to work crappy in ranging markets -- unless the market moves happen to "sync" with the natural "resonant frequency" of the MA cross. Then they work great.

So class, to summarize today's lecture: MA crosses work great. Except when they don't. Which is usually.

Excellent help! Thanks, George!! I have a feeling your post may get linked to quite a bit!

Thank you all for your input - keep it up! :)

Cheers,
Rick
"A problem well put is half solved."
User avatar
ironrick
Trader
Posts: 85
Joined: Wed Nov 16, 2011 7:04 pm
Location: North Idaho Boy, USA

Re: Quick opinion: MA Cross strategies

Post by ironrick »

Thanks again, Everyone, for the feedback -- keep it up, the poll is still open and only takes seconds to complete.

You can find my reason for wanting some feedback and a new trading strategy here: http://www.stevehopwoodforex.com/phpBB3 ... f=12&t=212

Cheers!
Rick
"A problem well put is half solved."
User avatar
ironrick
Trader
Posts: 85
Joined: Wed Nov 16, 2011 7:04 pm
Location: North Idaho Boy, USA

Re: Quick opinion: MA Cross strategies

Post by ironrick »

More excellent insight on this topic from Khalid. Emphasis mine. (quoted from here: http://www.stevehopwoodforex.com/phpBB3 ... 2&start=10 )
Khalaad wrote: The first thing to do [in a MA Cross strategy] is to define risk.

My experience is... that moving average crosses and stops do not go together very well. Hence I would define risk as average loss for individual pairs you wish to trade measured through back tests over a sufficiently large sample of trades.

Khalid
Very well said. And a good rule rule of thumb to remember!

Thanks, K!

Rick
"A problem well put is half solved."
User avatar
Gamma_gallus
Trader
Posts: 118
Joined: Wed Nov 16, 2011 2:03 pm
Location: France

Re: Quick opinion: MA Cross strategies

Post by Gamma_gallus »

Hi Rick & all,

Still based on MA cross, I found something interesting from below site. I'm trying to see if there's something we can get out of it. Thought of sharing it with you guys, because i think it has some potential, especially on the H4 (but still need to investigate the entry & exit points)

http://forex-strategies-revealed.com/scalping/ema-bands

So from it, I have added 2 SMA 100 & 200 to see the bigger picture.
I'm trying now to see (visual backtest) if we can add some filters using the SMA 100 and/or 200 to decide wether we go long or short (maybe also using the ATR with a EMA), like for ex. this one

- if price is above 100 and/or 200 SMA then buy only
- in between : no action
- below short only

I've just started the analysis, so need to further train my eyes on those charts ..

Gamma
You do not have the required permissions to view the files attached to this post.
The_Snowman
Trader
Posts: 383
Joined: Sun Nov 27, 2011 8:24 am
Location: Australia

Re: Quick opinion: MA Cross strategies

Post by The_Snowman »

MA systems do work; every Indicator or EA developed works SOME of the time, the problem is they don't work ALL of the time.

There are volumes of books on the subject as well as pages of internet, but that wasn't the question was it?

The question was whether, in our opinion, we thought they worked, or knew they worked I suppose, the ultimate goal of the poll could be to find someone with something successful to share ;)

I ended up here by searching for a thread about "Triple Screen Trading" by Dr. Alexander Elder - I didn't find any in the Manual Trading forum and no mention of him here in General Discussion either.

This method of trading uses MACD as the tool to determine trend or trend changes on the higher time frames. He has recently written a book called "Two Roads Diverged: Trading Divergences (Trading with Dr Elder)"

Recently I have been following a great thread (yes, they do exist despite the bias :o ) on FF called "Third Screen Strategy" - I would like to see something similar here on SHF.

Has anyone seen a thread on this topic, or suggestion where I might best open one? I would think under the aforementioned Manual Trading would be best as that's what it is.

Thanks for any reply or suggestions, sorry for going off topic sort of anyway...

8-)
AnotherBrian

Re: Quick opinion: MA Cross strategies

Post by AnotherBrian »

garyfritz wrote: An MA cross is essentially an oscillator: MA(fast) - MA(slow). If you look at the math of it, MA cross systems are bandpass filters. They pass midrange frequencies while attenuating high and low frequencies.
Great post! I just had a flash back of grade 12 electronics! I still remember all that bandpass and frequency stuff.
effluvium

Re: Quick opinion: MA Cross strategies

Post by effluvium »

AnotherBrian wrote:
garyfritz wrote: An MA cross is essentially an oscillator: MA(fast) - MA(slow). If you look at the math of it, MA cross systems are bandpass filters. They pass midrange frequencies while attenuating high and low frequencies.
Great post! I just had a flash back of grade 12 electronics! I still remember all that bandpass and frequency stuff.
What I have found that If you attach TDI indicator to any chart which you attach 2 EMAs (1 EMA 3period Shift 1 + 1 EMA 6 periods shift 1) you'll find that many of TDI crosses correspond to the crosses of the 2 EMAs
Post Reply

Return to “General Forex Discussion”