Which EMAs work best for catching trends?

[ArmyDoc]

I am planning on backtesting a system similar to the Cowabunga system tomorrow, but using longer time frames, (determine trend with weekly, trade daily, and time entrance and exit with 3 hr).

I plan to determine the trend using the EMA crossover method and I was hoping to get some advice here on which EMAs would work best for trading daily charts?

I know Cowabunga uses 5 and 10, but some brief experimentation with 10 and 20 on EUR/USD and USD/CHF (only for the most recent 7 months) indicates that perhaps this combo would be better in that there are less false trend changes, (the kind that force you out of a profitable trend trade).

For example, using 10 and 20 EMAs, the EUR/USD rally began in Mid March and has not had any cross overs since, (only close calls and perhaps a touch, but no confirmed cross over). This would have resulted in 1500+ pip profit and counting. The 5 and 10 EMA did have crosses that would have caused one to exit profitable long positions several times as well as some false long signals, (I will only trade the weekly trend, so no shorting the EUR/USD for now).

A brief check of USD/CHF shows the same thing, 10 and 20 would have kept you in a short trade since late April, for a 1000+ pip profit.

For my system to work, it is rather important that one not prematurely exit such multi-month, 1000+ pip trades, since the SLs you need to use are enormous, (200-250) and limit you to 1.5:1 leverage at best, (assuming you risk 2% of capital/trade, if you risk 3% then leverage can be increased to maybe 2-2.5:1)

Any suggestions? Perhaps other combos I have not considered?

[ThePhoenix]

Ichimoku is extremely good in determining longer term trends. It is actually designed specifically for this.

It also has several methods in place of placing a SL, that will let you ride the larger trend and if the SL is hit the trend is considererd invalitdated.

Go to Ichimoku Discussion Forum: Forex, Stock, Commodities, Futures, Bonds to learn about ichimoku. thats the forum look for the link for the actual teaching about it.

[xtraction]

I am planning on backtesting a system similar to the Cowabunga system tomorrow, but using longer time frames, (determine trend with weekly, trade daily, and time entrance and exit with 3 hr).

I plan to determine the trend using the EMA crossover method and I was hoping to get some advice here on which EMAs would work best for trading daily charts?

I know Cowabunga uses 5 and 10, but some brief experimentation with 10 and 20 on EUR/USD and USD/CHF (only for the most recent 7 months) indicates that perhaps this combo would be better in that there are less false trend changes, (the kind that force you out of a profitable trend trade).

For example, using 10 and 20 EMAs, the EUR/USD rally began in Mid March and has not had any cross overs since, (only close calls and perhaps a touch, but no confirmed cross over). This would have resulted in 1500+ pip profit and counting. The 5 and 10 EMA did have crosses that would have caused one to exit profitable long positions several times as well as some false long signals, (I will only trade the weekly trend, so no shorting the EUR/USD for now).

A brief check of USD/CHF shows the same thing, 10 and 20 would have kept you in a short trade since late April, for a 1000+ pip profit.

For my system to work, it is rather important that one not prematurely exit such multi-month, 1000+ pip trades, since the SLs you need to use are enormous, (200-250) and limit you to 1.5:1 leverage at best, (assuming you risk 2% of capital/trade, if you risk 3% then leverage can be increased to maybe 2-2.5:1)

Any suggestions? Perhaps other combos I have not considered?

Instead of working with moving averages, which are known to lag, work with the weekly open price, the daily open price and last week’s high/low prices. These don’t lag and traders react to them every day, much more so than moving averages. Apply some common sense and you will realize that traders have been chasing moving averages for decades. Don’t you think that by now if there was anything of value there, it would have been discovered? Consider all of the institutions with the best minds and equipment have been working on cracking this nut and you would have to conclude it’s a waste of time.

[pipso_facto]

I am planning on backtesting a system similar to the Cowabunga system tomorrow, but using longer time frames, (determine trend with weekly, trade daily, and time entrance and exit with 3 hr).

I plan to determine the trend using the EMA crossover method and I was hoping to get some advice here on which EMAs would work best for trading daily charts?

Moving average crosses tend to work best on much longer averaging periods than 10/20 days.

[pipso_facto]

For example, using 10 and 20 EMAs, the EUR/USD rally began in Mid March and has not had any cross overs since, (only close calls and perhaps a touch, but no confirmed cross over). This would have resulted in 1500+ pip profit and counting. The 5 and 10 EMA did have crosses that would have caused one to exit profitable long positions several times as well as some false long signals, (I will only trade the weekly trend, so no shorting the EUR/USD for now).

A brief check of USD/CHF shows the same thing, 10 and 20 would have kept you in a short trade since late April, for a 1000+ pip profit.

Be careful to cherry-pick examples of when your system would’ve worked in the past. Throw a good measure of objectivity in your back-testing.

… the EUR/USD and USD/CHF are highly corelated

[ArmyDoc]

@ThePhoenix

Ichimoku is extremely good in determining longer term trends. It is actually designed specifically for this.It also has several methods in place of placing a SL, that will let you ride the larger trend and if the SL is hit the trend is considererd invalitdated.

Go to Ichimoku Discussion Forum: Forex, Stock, Commodities, Futures, Bonds to learn about ichimoku. thats the forum look for the link for the actual teaching about it.

ThePheonix, thanks so much for opening my eyes to the power of Ichimoku.

I have only begun my preliminary research into this but let me so you an example of how good Ichimoku is at finding trends.

In my new daily swing trade system I have been back testing, I shorted USD/JPY on 8/31/08 @108.332 and exited on 1/04/09 at 103.102. This was using 10 and 20 EMA crossover to spot trend reversals.

Had I used the Chikou span to plot my entry, I would have shorted 2 weeks earlier at 109.691 and had I used the Kumo cloud with chikou span confirmation to exit, I would have held on another month and exited @ 91.812.
I would have made an extra 257 pips using Ichimoku.

A better example is my worst trade so far, a false “buy” signal for the GBP/USD on 12/15/08. That trade was made right after closing my best trade, a shorting of the cable that netted me 2100 pips in profit and the long position stopped out within a day.

BUT had I used candles closing beneath or above the Kumo with Chikou beneath or above the price to confirm as my entry point, I would have been much better off.

I would have entered at the same time and price, but instead of prematurely exiting on Dec 15 and then incorrecly trading long the Cable, I would have hung on until Mar 21.

This would have netted me an extra 1100 pips in profit, (total profit 3200 pips) AND prevented my falling for the short term fake out that cost me a 250 pip stop loss.

I will dig more into how to use the Ichimoku, especially how to set stop losses, but as I head off to the gym I am inspired to new heights!

I don’t think I will sleep much, as I’ll be researching and backtesting long into the night!

Thanks Pheonix, along with ForexBrokerRatings, you are one of my biggest heroes, (he for showing me the value of long term trading, and you for pointing me to the most powerful trend spotting tool I have ever seen!).