ADX Alternative?

[carlosmckenna1]

Hey you guys,
Just wondering if I could get some help?

I am working on a system for trading forex; part of the system incorporates the ADX 14 day for identifying trends. However I find this to be a little slow and tried other periods on the ADX i.e. 5 and 9 day. But these produce too many whipsaws.

Can any of you suggest another trend identifying mechanism that would improve on the lagging delay of the ADX.

Would be grateful for any suggestions.

Thanks Carlos.
P.S I have tried various MAs and again either lags to much or too many whipsaws
.

 

[Rhody Trader]

You've identified the short-coming of trend indicators. There's not a great deal to be done about it, though dealing in multiple time frames may help work around the problem.

 

[Black Swan]

adx by itself? Utter waste of time, most folk can't interpet in anyhow...as for advice..dyor...

 

[tomorton]

Hi carlos - The greatest teachers of trading put huge emphasis on positions in relation to the trend but in reality there is no such thing as THE trend. There is an infinite number of trends occurring at any one time, and excactly 50% are up and 50% down.

The trend that you should be interested in is the trend that equates to the horizons of your TA abilities and risk acceptance and, therefore, your preferred holding period. Not much use looking at 4 hourly bar chart for trends if your holding period is in the months: not much use looking at a 200 day MA for trend if your holding period is 48 hours.

Back-test your strategies: this will tell you the periods of validity of these positions based on these signals and the average time after which you had better (or might as well) get out. Let the market tell you the trend to seek to identify. Don't try to make the market's behaviour fit into your preferred trading style.

(Stones - Gimme Shelter)

 

[tomorton]

Oh yes - To identify trend.

I look at 3 EMAs in relation to price and each other, ADX, stochastic and RSI, but most importantly, swing highs and swing lows of price EOD. I do sometimes trade off price:the shortest MA: the others are interesting as corroboration but I would never recommend initiating a position based on ADX, RSI or stochastic reaching an extreme.

 

[carlosmckenna1]

Thanks for the advice.

Sorry what is dyor please?

 

[tomorton]

Forgot - I also look at my own Weekly Trend Strength Indicator, WTSI%.
This is simple stuff - plot weekly HLC bars with 50EMA (daily) overlaid. Count H's, L's and C's above MA and convert to %age. In a strong uptrend, all bars would be wholly above the 50EMA, so 100%. Select a time period for the chart that accounts for volatile periods and gives reasonable confidence for likely holding period: I suggest period should be at least 10 x your average holding period.

 

[tomorton]

Come on Swannie, its just lying there.....

 

[bevok]

Yes I would suggest looking into dyor, it really is the key to the problem

Seriously I think a trend filter is an excellent addition to most systems. I like a concept of having an entry mechanism, which is then filtered by a longer term trend identifier to chop off lower probability entries. This is simply because I believe the longer term trend tends to persist, giving you better rewards when you enter in that direction. ADX can be a good identifier of this (telling you whether to be long or short) but you can equally use the slope of a 20MA on the next timeframe up, last MACD crossover etc. You could also use a fundamental view i.e. Gold is going to go up because of investors belief in it as a antidote to risk, and the belief in the story about the decline of the dollar, hyperinflation etc etc so for now I'm only going to take entries from my technical setup (breakout, MA cross or whatever) that are long. This would only work in a daily or above system - lower timeframes are too divorced from such considerations.

 

[carlosmckenna1]

Thanks for all the advice... its given me a few ideas to look into... I have 'done my own research' and will continue to do so...
Thanks again..
P.S.
BSFOYAT