Why ADX > 20 Is a Lie + Easy Trading Method

[daedalus 10]

Recently i've made somewhat of a paradigm shift in how I trade (and simplified things about 10x's over) but I wanted to share a small observation I have made concerning "trending" markets and ADX readings and how i'm trading. This example is on the Forex markets (as I made the original post for a FX site, but I use this to trade the e-mini's on a 1-5min and tick charts these days)

 

First off, let me preface this by saying if you are a pure trend trader looking for MA cross overs or something into a bigger trend then this might not help you. For the rest of us going short and long based off S/R, trendlines, or Fibs then this will be relevant to you!

 

Ok, so we are all told from various software vendors, trading forums, trading websites, educational vendors that the GOOD trades are in trending environments correct? We are told to wait for the ADX to be > 20 to take a trade? Because then we know the market is trending right? I'm here to argue that this SIMPLY IS NOT TRUE. In fact - its the opposite.

 

I argue that the prime time to look for entries is in fact when the ADX < 20 and by definition the market is NOT in a trend. Think about it like this: If you are getting into a move when the ADX > 20 you are by definition jumping into a position after price has ALREADY MOVED A GOOD DISTANCE IN ONE DIRECTION. Thus, its probability of that trend stopping and a reversion to its mathematical mean is much much more likely which means a move against you is much much more likely as well.

 

Now on the other hand, by getting into a trade when the ADX < 20 we are saying that we want to be in the move BEFORE ITS OCCURS and let us make profit while the ADX is increasing and the TREND is increasing.

 

Here are some examples from Friday on the EURUSD 1 minute chart. The indicator on the chart shows nothing other than a GREEN BAR if the ADX < 20 and a RED BAR if the ADX > 20. i.e. we only are looking for positions when there are green bars.

 

First thing to note - IF you trade like this it will be some of the easiest most carefree trading you've ever done. Why? Because ADX filters 70-90% of the price action on most days. And you sit and wait without a care in the world. I have the indicator set up to alert me when the ADX drops below 20 so I get a message saying pay attention and then we look for trades. This will DRASTICALLY reduce trade amounts, increase efficiency, reduce commissions, and eliminate the possibility of overtrading a market. Lets face it folks - most of us aren't good enough to be 100% on every swing 100% of the day. It gets us into trouble and we all know it. So why fight it? Especially if you can filter 70-90% of the day and only focus on tiny points on the chart where the probability for a trend developing is in your favor?

 

I apologize for the size of these charts but nothing peeve's me more than someone posting a chart example thats too small to see. You should have no issues here.

 

These examples will be nothing more than trendline breaks snapped to a Pivot High/Low reading on the charts. Entries on breaks of the trendlines with the ADX < 20. NOTHING ELSE. You decide the management and targets but it would be pretty hard to lose money here.

 

 

You are jumping in at the momentum shift BEFORE THE TREND OCCURS.

 

Now take a look at the entries where the ADX > 20 and the pundits would like to tell us its a GREAT TIME TO LOOK FOR TRADES. Almost everything loses.

 

 

Coincidence? That's for you to decide. I don't think it is coincidence. In fact this is exactly how I trade from here. ADX and trendlines. Call it a day and make money. Only we can complicate the hell out of it "looking for the trend". The truth is the trend is relative, and thus WE DEFINE IT. And if that's the case, everyone else has a difference definition and no one can be right or wrong. So why bother? Swing your middle finger up at the "trend" and forget about chasing it. Its a fruitless effort and a lie, and one that hasn't made too many of us rich.

 

Think its a fluke? Go look at it yourself.

 

ADX and trendlines. The only tools you need to make yourself rich.

 

Cheers!

[daedalus 10]

Easy trading so far today...

 

I got out early at the profit target my management rules said to but you can see... easy, and effective.

 

 

I'll post more trades from today and we progress! Happy Trading!

[waveslider 10]

I think ADX is better as an exit tool, though I don't use it...

[ianf0ster 10]

.....

Ok, so we are all told from various software vendors, trading forums, trading websites, educational vendors that the GOOD trades are in trending environments correct? We are told to wait for the ADX to be > 20 to take a trade? Because then we know the market is trending right? .....

....... Think about it like this: If you are getting into a move when the ADX > 20 you are by definition jumping into a position after price has ALREADY MOVED A GOOD DISTANCE IN ONE DIRECTION. Thus, its probability of that trend stopping and a reversion to its mathematical mean is much much more likely which means a move against you is much much more likely as well.

 

Now on the other hand, by getting into a trade when the ADX < 20 we are saying that we want to be in the move BEFORE ITS OCCURS and let us make profit while the ADX is increasing and the TREND is increasing.

 

Here are some examples from Friday on the EURUSD 1 minute chart. .......

 

Hi Daedalus,

I have no wish to offend you. perhaps you have more experience than me, but if I assume that you are not tyring to deliberately mislead people, I can only conclude that you do not understand either the ADX or what your strategy is based upon.

 

You have a point about the ADX getting you in late when a trend changes,

but you didn't wait for a trend change confirmation from the ADX on your chart. All your exampels of failing trades when the ADX is > 20 are counter-trend trades. Thus it should be obvious that the stonger the trend, the less likely they are to work out - so they give better results if the ADX is <20 when they are taken.

I did not see any examples of you using the ADX to get in with the trend (e.g. on the break of a bull/bear flag or a pullback).

 

Unless you know that you are doing it, and know the risks involved, it is deadly to one's account to trade an indicator in a way that goes against the theory behind the indicator.

 

I am not an expert, but I have learnt not to do that.

 

Ian

[daedalus 10]

The method outlined is a breakout method. I understand that the moves were counter trend which is why they failed. The point of it was that lots of people try and use trendlines and can never seem to figure out why some patterns that are advancing H, HH can snap and make a profitable trade, and others that appear to reverse in their face.

 

I just wanted to show an easy way to qualify trendline trades, nothing more!

 

Cheers!

[trade-samarai 10]

Hi daedalus:

 

I understand your thinking on this, having chased ADX trades. It really is crucial to get in early. Your ADX-Mod idea seems to me to have merit. I would add that a Bollinger Band-Keltner squeeze can confirm your theory. Thanks for your insight. The "ADX Modified" seems easy enough to code, even I could do it. It's the first time that I've heard this idea.

[snowbird 10]

I also believe this method has merit. Many of the trades will be out of consolidations, allowing a small stop loss and greater risk/reward potential. I've heard some traders look for trades within consolidations or when ADX drops to a very low # for this very reason. The tricky part of course being not knowing in advance which way price will break out of consolidation. The trendline break entry you show appears to let price action confirm the direction out of consolidation.

 

snowbird

[daedalus 10]

Thanks for the kind comments folks. I trade the e-mini's and just wanted to share with you the four trades i've taken over the past two days in the ES (including this last trade about 10 minutes ago).

 

Today:

 

Yesterday:

 

My results were as follows: Par, +87.50, +125.00, +112.50. The first trade went +2 in favor but my target was slightly higher at +2.5. I don't know how you all want to manage them, etc...

 

But I wil say this for the few things I've found to help me out.

 

1. I use a Keltner channel to skip trades that "break trendlines" but area already up or down at a band. Keeps me from getting in too late and catching trades that are going to retrace or reverse without giving me a chance to get to par.

 

2. The highest probability trades are the wedge setups (ie like the trade from today, H, HH, HL, LH, failure for entry. If the trade is an advancing setup, where you are taking a trendline break of a swing that is going H, HH, LH, HH, failure for entry the odds are less in your favor. I would recommend looking for some kind of divergence to further filter these entries.

 

The end result is much less trades. But you know what? I'm perfectly fine with only taking 1 - 4 good trades a day and leaving it at that. Keeps my commissions low and my trade efficiency high.

 

Since I've switched to this method over the past weeks while i've still been tweaking the method (adding keltner rules, and wedge vs. advancing attributes) i've traded basically like crap and am still making consistent money.

 

Just keep things simple and you'll do just fine!

[Tams 20]

put a 50+ MA on it...

when the ADX breaks, it will most likely break in the direction of a medium term MA.

 

 

may I recommend this MA with extension:

http://www.traderslaboratory.com/forums/f46/moving-average-with-extension-5642.html

Edited May 7, 2009 by Tams

[snowbird 10]

50 MA should definately help.... and also keep you out of some bad trades (pass on signals or be very carefull if the 50 MA is flat/directionless)

 

snowbird

[thalestrader 14]

These examples will be nothing more than trendline breaks snapped to a Pivot High/Low reading on the charts. Entries on breaks of the trendlines with the ADX < 20. NOTHING ELSE ... In fact this is exactly how I trade from here. ADX and trendlines. Call it a day and make money. Only we can complicate the hell out of it "looking for the trend".

 

Hi there,

 

Interesting post - I have a few questions:

 

1) Are you still trading this way?

 

2) If yes, have you tried just using trendlines and dropping the adx altogether? What I am getting at here is do you think that the adx reading has anything to do with the the price movement subsequent to the trendline breaks?

 

3) In my opinion, trading properly drawn trendline breaks is an excellent trading strategy. From the charts you posted, it seems that you are not consistant in the manner in which you select your pivots on which to anchor your trendlines. So my question here is how do you determine which pivots you use to draw your trendlines.

 

Thank you,

 

Thales

[ckait 10]

Are you aware that the adx only told you what happened before and not what will happen in the future? Do you use support and resistance in your trading? By what means do you determine what will be support and resistance? If not the road to riches will elude you longer than your account will last,(in most cases anyway).There is no easy trading method. If there was everybody would do it and it would no longer work. The markets work on pure supply and demand. If you can figure out some kind of an edge to guage a short term imbalance between the supply and demand then you would be able to make an informed trading decision. Keep looking but look inside the price and yourself not the list of canned indicators and holy grails.good luck

[Tams 20]

Are you aware that the adx only told you what happened before and not what will happen in the future?....

 

 

 

Most indicators do that -- tells you what happened before.

That doesn't mean it provides no useful data.

 

 

Just like the speedometer in your car, it is telling you what happened before.

When you are speeding toward a tight curve... what happened before can tell you what is going to happen soon.

 

;-)

Edited June 6, 2009 by Tams

[Frank 10]

ADX measures past directional movement but has a kind of funky aspect to the calculation...

 

ADX(14) doesn't look back over the past 14 bars and compute -- it instead uses the raw reading from the last bar and weights it 13/14 --- then takes the current bar DMI and weights it 1/14.. For this reason, going from low to high ADX (or vice versa) is actually a very slow indicator and not all that useful, imo. I think it is much better to understand what makes ADX go up and down and think about what will have to happen to get ADX to move in the future.... ie, one big bar with strong volume could kick-start it if ADX was low --- or a pattern of higher highs and higher lows would do it if done over a 14+ bar sequence.

 

I find that you should simply look at the bars and see if the high or lows of the bars are being consistently taken out of not. higher highs and higher lows will make ADX go up. lower lows with lower highs will make ADX go up. but another way ADX can go up is if you simply going up slowly but there is little movement below the bar lows (or above the bar highs).

 

If you study the calculation from the inventors book (Welles Wilder) -- you can understand why a given price bar is additive or subtracting from the latest ADX reading. This means you have to know what makes +DMI and -DMI go up or down --- and, if the current ADX reading is high --- is it high due to high or low net DMI number.

 

You can then watch the price bars in real-time and pretty much figure out a more 'real-time ADX' mental calculation and what type of action will reverse that action. I like using trailing high-low channels in conjunction with watching the ADX indicator in the lower pane.

 

Welles Wilder writes that he chose 14 as his ADX figure because that was basically 1/2 of a month and he was using daily charts. Well, I have found that if you take 1/2 of the 14 figure --- and run a high low 7-bar channel off that, you can anticipate what action is consistent with high or low ADX and what is not.

 

One other thing, if ADX is at a low level, it won't take that much strong movement to jumpstart ADX cause there is already tight compression and ADX will naturally expand from that low level. Thus, ADX is somewhat mean-reverting to the 20-22 level.

 

Here is an example from Friday -- 5-min ES chart. Note that when a 7-bar high low channel reverses, ADX always drops. When a channel sustains, ADX expands. This obviously isn't rocket science but understanding the nature of the calculation really can help better understand how to use ADX.

 

I agree with the original poster. Once ADX hits a high or extreme level, the move is 'mature' --- though may still be a continuation trade for a smaller profit target.

 

Edited June 6, 2009 by Frank

[momentom 10]

Firstly, this has to be one of the best posts I've seen on TL. Thanks Frank.

 

What follows from this is that the result of correctly following Frank's idea is that you would have the best trade location = tight stop + best r/r.

 

Taking early aggresive entries more than pays for itself. One of the reasons people lose or get faked out is the less than optimal trade location of their entry. You really need the trade to go your way immediately and not take heat.

[sunilrohira 10]

Hello,

 

Does anyone have the .ELD or code for ADX Modified? Thanks in advance.

[trade-samarai 10]

//ADX-Mod (Format type as point and weight from 3 to 5 as desired to look as pictured.)

 

inputs:

Length( 14 ) ;

inputs: UpColor(green), DnColor(Red);

variables:

ADXValue( 0 ), Color (0) ;

 

ADXValue = ADX( Length ) ;

 

if ADXValue < 20 then

color = UpColor

else

color = DnColor;

 

plot1(0,"ADX-Mod",color);

[Frank 10]

Lets take a look at todays S&P chart. I like the 5-min chart for watching ADX as its not TOO low a timeframe but certainly not too high either.

 

So the bigger idea is the high-low price channel. When the channel forms a V, ADX drops. When the channel forms an /\ (inverted V), ADX also drops....

 

Once the channel begins to change directions, you can watch the highs and lows of the bars for help. Note how price attempted down but no lower high pattern formed. Eventually, the channel formed a V and ADX continued to stay low. Just as the high end of the channel began up --- again, the lows could not hold --- so ADX would be expected to stay low. But then note what happens next, the highs of the bars begin to consistently make lower highs as the low end of the channel is taken out.

 

So there are 2 equally important things going on here -- there is the channel and there are the individual price bars. Then you also have the consideration of where the absolute level of ADX is.... if ADX is very low already, it will begin to expand as the channel begins to extend. If ADX is already high, it matters where price came from --- if the channel was down but highs are getting taken out consistently, then ADX will drop because the channel is not extending.

 

To some extent, you can think of the individual price bars as the lower timeframe and the channel as the higher timeframe -- a good trade needs to incorporate both timeframes.

 

First look at just the chart without the ADX indicator in the lower pane and try to visualize what is going on with the indicator (ADX is slow indicator --- it lags a bit -- so this is a good exercise):

 

[Frank 10]

Here is the same chart with the ADX indicator in the lower pane.

 

At Point A, the channel starts lower. But as it does this, the next 5-min bar is green and the low is not taken out. You can see that this green price bar is not going to help ADX get going as we know ADX moves only when a pattern of lower highs and lower lows ensues. The next bar makes a higher high and higher low to the last bar. Obviously, this bar is going to begin to erase the effect of the attempt lower.

 

Then the lower end of the channel turns up. Then the high end of the channel turns up --- so the overall channel has now formed a 'V'... The V pattern makes ADX go down. (If ADX is already very low, ADX may just stay at very low level).

 

Next the upper end of channel turns up at Point B. Thus you begin to again watch highs and lows. Once again, the price bars cannot get going and this leads to another reversal.

 

At Point C, you can see how something new happens --- the highs of the bars are not being exceeded while the lows are. The lower timeframe is signalling sell pressure. ADX eventually responds as the channel turns down while lower highs and lower lows continue.

 

[thalestrader 14]

Lets take a look at todays S&P chart. I like the 5-min chart ... First look at just the chart without the ADX indicator in the lower pane and try to visualize what is going on with the indicator (ADX is slow indicator --- it lags a bit -- so this is a good exercise

 

Here is the same chart with the ADX indicator in the lower pane ... ADX eventually responds as the channel turns down while lower highs and lower lows continue.

 

 

Hi Frank,

 

Thank you for sharing you expertise concerning the ADX.

 

I have a question, or rather a group of questions: How do you trade based upon the information you are focusing upon? Where did you enter? Where do you place your stop(s)? When do you get out?

 

Respectfully,

 

Thales

[Tams 20]

Here is the same chart with the ADX indicator in the lower pane....

 

 

 

Thanks for the illustrations and explanations ...

 

An indicator is only good/useful if you understand its operations.

This is good description.

[thalestrader 14]

I thought Tams was Happy ... what happened?

[Tams 20]

I thought Tams was Happy ... what happened?

 

 

I am obliviously happy. ;-)

[Frank 10]

Hi Thales,

 

Actually, I am on vacation this week and was just looking at the chart after a day at the beach --- I will be the first to tell you that after the fact analysis is always easy compared to real-time trading. I am not saying this is so obvious and that I would have caught an entry even if I had been focused today as I may not have -- but one thing I can say is that I would have been 'looking' on short-side given this price action.

 

In general though, with a low adx situation I would be looking at 2 types of entries ----

 

1) enter on a sell stop and look for near immediate continuation. you want to see price continue through your entry price for a bit so you have some room -- if don't get continuation, you are often better off exiting out for small loss and re-evaluating. if it does go in your favor after your stop, you should try to stay with it by giving it some room room but not too much as you do not expect a 5-min high to be exceeded by much. in that situation, one idea is to default to a stop that is 3 ticks above the 5-min bar high. if the ADX coil is breaking, that stop should not get hit.

 

2) the other entry is to try to to catch a 'lower high' -- ie, the channel may need to fill out a little more b/c sometimes the main channel has turned but there is some testing action back up that holds the channel-high but still violates some 5-min highs. in this situation you want to see price test up and then find a price and go with it. sometimes I will start with smaller position and add only when I see selling pressure resume. again, defaulting to a stop 3 ticks above a 5-min bar.

 

the best thing about catching ADX moving from low ADX to high ADX is that you when you catch it, you generally get a 'direct move' --- where your stop will never be threatened. then you can scalp out of a piece and just maintain a tight stop and now in a no-lose situation.

 

today, it looks like the only way to enter was to use Method 1 or get caught chasing. sometimes there will also be other types of things you notice that tip you off --- like if NQ or DAX break hard first. or if you see a large NYSE ticks reading and can catch first retracement back to +300. because I wasn't watching all this in real-time, tough to say.

 

one note of caution -- try not to be trading both sides of a coil. ie, if you had a short bias today, you could wait for the action that goes with that bias --- sometimes the coil gets sloppy (wide and loose) and this is very difficult situation. today, the range was very narrow so this doesn't apply --- but sometimes you get a wider coil and you have to be careful not to be going long at high side of a trading range and short at the low end of the range. overtrading a range can be account killer.

Edited June 17, 2009 by Frank

[thalestrader 14]

Actually, I am on vacation this week and was just looking at the chart after a day at the beach --- I will be the first to tell you that after the fact analysis is always easy compared to real-time trading. I am not saying this is so obvious and that I would have caught an entry even if I had been focused today as I may not have -- but one thing I can say is that I would have been 'looking' on short-side given this price action.

 

Hi Frank,

 

I agree completely about after the fact chart reading. I did not trade the ES today (I am primarily a stock trader with the occassional futures trade thrown in for the sport of it). But I do watch the ES, and I have included today's 5 minute ES with short entries (blue lines with arrows) and exit (thicker red hashed line) marked.

 

This is what I meant in my original question. Where would you have entered today, had you been trading, and had you traded it as you would like to think you would have given what you se on the chart tonight. For example, first sell stop would have been 919, second short entry at 916.50, and 3rd entry at 915. Exit would have been trailing stop at 910.50 (again, I grant that this is a fair bit of "monday morning quarterbacking").

 

I try not to go into the day with a bias one way or the other, but just let price go where its wants to go. Once that 919.25 gave way I did start looking for short trades more than long trades, and once 916.75 (the prior low) gave way, I could then be said to have had a "short bias."

 

I am very curious as to how you make entry and exit decisions based upon the indicators you are looking at. From your response, I suspect you are relying more on price itself and less on those indicators than one might presume from your detailed descriptions of how you use them.

 

Thank you,

 

Thales