DMI/ADX consistently works

See the attached chart and you know why SHORT in the
DAX Future 12-09-03 at 07.00 AM !!!!

Cheers
Jeanmichel

 

Jeanmichel,

Thanks for this morning's chart. On the second chart, bottom section, it looks like you have 2 moving averages plotted (black and yellow lines) along with the ADX/ADXR/DMI. What is the purpose of these lines when you already have your EMA's at the top?

 

Ram,


That is only for the confirmation !!!

The yellow didn't cross the dark blue line and both lines
went into the downside and that is for me "confirmation"
for a SHORT run !!!!


JM

 

Thank you for your informative posts.

My question is "trade only when ADX is near the 20 or 30 line".

Do you mean that there is something special about those particular readings, of 20 and 30?
Or,do you mean that it has to be at least 20 or 30?
Do you mean that between 20 and 30 is no good?
Would 40 for example be no good?
Thanks,

 

Hi,
I must be confused.

I have attached a screenshot of 8-25-03 in esignal along with the settings for the ADX/DMI. The confirmation you suggest in the beginning of the thread does not seem to be present.

Can you tell me what I have wrong here?

Thanks

BTW this is PST on the chart

 

There are 4 constants to consider for Wilder's system. The dmi+ and dmi- signal line constants are the same and usually have a value of 14. The adx signal line constant is usually 3. The fourth constant is the time step. These are not hard and fast rules. You can change these numbers as you like. Since dmi+ is the buy signal, I usually color that line green. Likewise I color the dmi- red because that is the sell signal.

If adx < 20, the stock is not trending.
If adx >20, the stock is either in an uptrend or a down trend.

If adx >20, the necessary and sufficient conditions are:

dmi green > dmi red, buy
dmi red > dmi green, sell

If adx < 20, stay out.

It works great for swing trading on daily charts but I have never made money with it intraday. I use two minute charts and I have something that works better than wilder's system for that.
regards.

 

If you're not scalping, try the 5m. You may be happier with it.

 

Thanks

I've seen lots of 5 minute charts where the adx looks a lot better than it does on 2 minute charts. Right now I am a scalper and I like the 2 minute chart best.

I'm looking for a system with a high probability of success and at least 30 trades a day. I used to have a system like that but it doesn't work anymore. Believe it or not it produced an average of 4 points a day. I don't think I'll every find anything like that again with technical indicators.

On my first post on ET, I explained that system. But it didn't generate much interest.

regards

 

TRADING BREAKOUTS USING THE ADX INDICATOR
by Matt Kirk

A popular ADX (Average Directional Index) trading method is the basic breakout strategy whereby a trade is executed either on the long or the short side when prices have moved out of the recent trading range. For the purposes of this article I’ll outline a channel breakout strategy however breakouts can also appear in a market that has been trading in an upward or downward sloping fashion.

There are many different ways to identify when the market has made a channel break and one of the more common is to buy a 20 day high or sell a 20 day low. The general rule is that the price bar must CLOSE on a new high or low - not just penetrate the previous 20 day high or low. Ah, but I hear you say “what about all those annoying false breakouts!”
The following method can alleviate the problems and frustrations associated with a high or low tick due to stop loss activity rather than the potential commencement of a trend. If the market travels sideways in a relatively tight range for 20 days then one can expect that volatility is generally low.

My favourite strength of trend tool is of course the Average Directional Index (more commonly known as the ADX). This little beauty will keep you out of many trades that turn out to be false breaks. This is how it works.

To start with the ADX has nothing to do with direction. It simply tells you whether the market is trending or not or, is about to commence a trend. In your charting software the ADX line is found hidden away within the indicator called the DMI or Directional Movement Indicator. The DMI will provide the direction as well as displaying the ADX line however I personally ‘toggle’ off the DMI component as I find it visually a bit messy. On a daily chart I suggest using 8 and 18 period Moving Averages (MAV) to provide the directional signal when they cross. A cross of the 8 period MAV from below and up through the 18 period MAV means that prices are now advancing or moving higher at a faster rate than they have been – the opposite is true for a sell signal.

The secret here however, is when to take the trade? I find that you can stack the odds in your favour when the ADX has moved up for two periods between 20 and 30 on the Index. The ADX line can come from below 20 on the first move up but the second period move higher, must be above 20. A falling ADX line or a sideways ADX mean that the market is not trending or perhaps has completed the previous trend. Moves below 20 are generally false moves and should be ignored. OK then, the ADX has moved up two days or bars (if you’re intra-day trading) as per the above guidelines and the MAV’s have crossed to give a buy signal. You must then BUY on the open of the next bar and place your initial stop half way between the range that the market had just broken out from. The size of the previous range will, of course, determine the dollar value of the stop loss so it will be the size of your account that determines whether you should do the trade or not. (You should generally risk only 5% of your capital in any one trade). The opposite is true for a sell signal. As the market moves in your favour, trail the stop up using your own money management techniques to lock in profit. There are plenty of different ways to do this and perhaps should be left to be dealt with in another article. Feel free to email me your address (matt.kirk@tricom.com.au) and I’ll post out a chart showing the perfect setup which can be used on stockmarkets, futures and foreign exchange trading.

This article first appeared in Your Trading Edge Magazine Sept/Oct 2000

Matt Kirk has been writing freelance articles under the pseudonym Phil Dew since 1999. He is represented by Big Picture Management and is an Associate Director of Tricom Futures Services. He can be contacted at matt.kirk@tricom.com.au

This article is copyright Matt G Kirk, Sydney 1998-2000. No part may be reproduced or distributed without the express permission of Matt G Kirk.


Cheers,
Jeanmichel

 

I always like to test ideas like this when I have some free time. The author of the article that Jeanmichel posted stated that a rising ADX between 20 and 40 "stacks the odds" in your favor for a trade in the direction of the 8/18 moving average cross. So, the best way to see if he is right is to put it to the test.

From the article it sounded like he was saying to trade a MA cross that occurs when the ADX has risen for the second consecutive bar. I went looking for these occurances on a 5 minute chart of the ES and only found 5 over a 60 day period. Its very rare that you get a MA cross at the precise moment the ADX is rising. So for the sake of having a better sample (and a tradable system) I entered the trade in the direction of the MAs at the time of the ADX move. So, if ADX is between 20 and 40, has moved up for two bars, and the 8 EMA is > the 18 EMA, we buy...vice versa for sells.

For the exit I did not want to get overly complicated. The author mentioned using the halfway point of the previous range. How do you define the previous range? He doesn't give any rules like this, so for simplicity I first looked at exiting when the moving averages crossed back over. The results with this were terrible due to the lag of the moving averages. So, in an attempt to exit closer to the end of the move I chose to exit when the ADX conditions were no longer valid. So if the ADX declined or went over 40, we exit the trade. I did not use any stops, as the ADX rolls pretty quick when the trade turns bad. Plus, I have found that if a system is bad without complicated stop rules, then it will probably be worse once those are added to the mix. I also set it to exit on the close if still in a position at the end of the day.

The results are not that good. Over the last 60 days, it is up $550 trading 1 lots. With 216 trades, that doesn't cover commissions. Also, the biggest winner is $450, so most of the gain is in one trade, which for some reason I never get when I trade a system live. So from these preliminary results, the odds aren't really stacked at all.

Now I know that my interpretation of the method is loose. Since the author did not give specifics, I had to improvise a bit. But I kept it simple, and if it doesn't work with simple code, it is certain to fail with lots of extra restrictions, with the added danger of curve fitting. My entries look very similar to the ones that Jeanmichel has posted (although this sytem is exiting pretty quick when ADX changes). Since JM has not really specified how he exits, then I can't compare this with his method. But I think the general idea is the same. Certainly it will catch nice trend moves, but it also gets chopped a lot, giving back the gains. Perhaps JM will post more specifics of his trade rules, particularly with regard to exits.

I am attaching the system report, a chart of the last few days with some trades plotted, and the TradeStation code for anyone that wants to play around with it. If you improve on it or have something to add, then please post it so that others can learn from it.