Trend spotting

smjones and wtb,

I can only underline what phil is explaining.
It is the very same as what I teach people in my room every day.

The way I explain it as follows:
WTB said in his first posting : ...

Meaning to have an indicator or a method to filter out the trend from the consolidation. Because we all know it is better to pick up a trade in a trend then to try to catch a trend from a consolidation.

People want to depends on indicators to tell them if there is trend or no trend or strong trend etc...
Try to open your minds now.
Price itself is THE most important indicator itself. There is no lag.
But 99% of the indicator that use a value is actualy a moving average that is taken in the calculation. Same goes for this slow stochastic that is shown in the previous posting. That means that it lags. problem is also that it needs to move between bounderies ( between 100 and 0) as so many other indicators. That has as effect if the trend is so strong that at some point the indicator will be at the extreem and there is no other way then to come back down in a up trend or to go back up in a down trend. So that means that the indicator is doing the opposite of what price is doing.
Methods who are using 1 or several MA's to identify the trend will have as effect that in a consolidation market with a bit bigger waves it will continiously show you an up trend at the top of a move and a down trend and the bottom were price just finished.
If one uses an indicator for ex. like yours or semilar indicators it will have the tendency to ALWAYS show you a trend. It will always have a bias in up or down (rarely it will say neutral bias)

To bring it back to what phil was saying higher highs or lower lows....
As WTB said: methods to filter out trending vs. non-trending markets when it comes to sniff around for a trade.
If the trend is clear (and it doesn't matter on wich time frame because one can take it from a 60min chart an other from a 5 minute chart) to "sniff around for a trade", that pattern of higher highs and lower lows will occur on every time frame.
Now important: if this happens in a clear way then there is a clear trend.
The problem is that people would like to know the bias continiously. And this is the beauty of this very effective but simple technique if at some point you don't see it in a clear way anymore because some lower high and then a higher bottom and retracing back etc...then that means there is NO TREND.
The moment that you say to yourself damm this is difficult to spot well that should be your neutral bias. Difficult to spot or you don't see it means automaticly that there is no trend. Meaning that you stay away from it.

I know it sounds to simple and that you would like to hear about some new indicator or technique. But just for fun maybe in the weekend if you have a couple of hours free. Take some charts with no indicators on it what so ever. Go back somewhere in historie and scroll forworths very slowely (with MT4 one can do tick by tick with the F12 button).
And put this simple aproach to the test and I promisse you that you will be amazed from what you see or the conclusions that you make. Don't quit after 10min Do it severall hours till you have the hang of it.
Again when it frustrates you because you do not see the trend or you find it hard to spot the trend then that means do not trade.
In the room they ask me very often igor could you help me on tracing the trend on the 5 min chart. I tell them emmiadiatly if I need to help you it is because it is not clear. If it is not clear then there is no trend.

I believe that price itself is the most important indicator in a traders tool box. But many traders seem to forget about that tool. They want to replace it by other tools that actually take the most important tool (price) as base for their calculations.
If price goes clearly up why do I need an indicator to tell me that there is a strong up trend ...if I can see that for myself.
The problem is that most traders try to spot a trade from the very beginning.
If you only take action when the trend is clear in this way higher highs lower lower lows then it is not full proof but you will see that will stay away more from false trades.
With the same technique I place also my stoplosses that in this way make a lot more sence then a fix pip stoploss.

In this price reading I develloped also a few price patterns if they occure is is better to stay away from it.
As a tennis ball pattern, were highs and lows are getting smaller and smaller (fits most of the time in a triangle) or a news bar pattern were before news price makes a little high and with news comes heavy down and then retraces back up. As long price then stays between this little high before the news and the low you better stay away from it (oposite for a reaction on news to the upisde)
A top-retracement and bottom-retracement pattern: If during the day price continously goes up. Somewhere in the US session it makes a high and then starts to retrace a bit. Very frequently you will see that price will stay for hours sometimes up to 2-3 days between that high and the bottom of that retracement (same goes for big moves to the downside). As long if price stays within that range you will see the same effect of very hard or nearly impossible to spot higher bottoms or lower tops. So stay away from it.

So this technique is doing exactly what WTB said: filter out trending vs. non-trending markets.
But I can understand that severall people are not helped with this technique because it is not applicable in mecanical trading systems or one can not define them als rules for an EA or in an indicator.
I know for many people it is not "sexy" enough but used as a discretionary trader it is a charm (again when it frustrates there is no trend)
I never found an indicator that can beat this technique in reliability.

regards...iGoR

PS. If you want to do a bit of home work for that technique to make it visualy a little bit more appealing then you can place a heiken-ashi bar and a HULL MA 15 on it (doesn't matter wich time frame, from 5min to weekly)

FXigor you have not looked at my math. This method of determining a trend is only based on price action. Please actually look at the math. I do not use lagging indicators.The very chart I posted shows no trend.. Please read my posts.

Your post is not based on my statements or the info I posted.

My method of determining trend is directly drawn from price, not indicators and certainly not MAs or any other lagging indicator.

SMjones I have indeed not looked in to it. I will do...
But my posting is not against your indicator. It is only a contribution that I wanted to make on the question that WTB asked in his first posting (techniques on finding trending versus not trending).
So to be clear I do not think or say that your indicator is not good.

And felt also a bit that phils technique (as one of mine) was not appreciated to the fullest.

friendly regards..iGoR

PS. To have no misunderstanding : I do not think or say that your indicator is not good. As long as I did not look in to it I believe you if you say it is a good (reliable) indicator.

I have the highest regard for Dow theory which is what Phil posted.

Nice discussion. I have to go get some sleep now. So I will catch up later. thanks much.

I am the proud owner of a 1938 2nd edition copy of the Dow Theory. It's a smelly old book with brittle paper but packed with wisdom. It is dedicated To Business Men who, perplexed by present day uncertainties, seek a reliable method of anticipating forthcoming changes in business activity.

On a slightly related note, I know a lot of traders that believe in head and shoulders patterns but fail to realize that the left side of the pattern is simply a higher high and and higher low and that the right shoulder is a violation of that previous high and low thereby signalling a reversal in trend. It's just the opposite for a bullish pattern.

Well, the most important part of any trend discussion is always to bear in mind that trend is 1000% realtive to time. Direction cannot exist without time.

One mans trend is another mans range. How you determine that trend really doesn't matter as long as your within a degree of accuracy. Obviously an MA would need to be matched to time and given a degree of tolerance.

On a volatile pair trend can be as elusive as the lochness monster. First thoughts would be to increase the time frame out until you see an obvious direction. But does that give you an accurate picture of trend in YOUR timeframe. Not always. Bottoms and tops can be hard to discern and harder even to trade in.

If you can honestly ask yourself why you even need to know what direction the trend is in and answer that question coherently, then you probably have no problems discerning it in the first place.

Absolutely, which is exactly the reason as of why one of my SMA conditions is to have the SMAs hugging price as it retraces after every swing of the trend. If price is not hugging the SMAs, then I am looking at the trend in the wrong time frame.

Have a look at the screenshot below: it's the same currency pair throughout the same time span, but one is the 4H chart and the other one the hourly chart. The SMAs help me to identify which time frame this trend is better suited for.

FXiGoR,

Thank you for your amazing post. Honestly, I am tempted to call it the best if not one of the best posts here at the factory.

So many people think in black and white and forget about the grayscale.

"I think this will go up..."(hesitant) -- Newbie #1
"I think this will go down..."(hesitant) -- Newbie #2

"I think this will not go..."(hesitant, but not losing money) -- Experienced Trader

And we can argue the point of the "best indicator" to a bloody pulp, but the reality is that it may take a hammer in the head to wake me up, but for you it could be an anvil.

Allow me to make my point a little clearer. It could take you just price to determine market conditions, but for me, I might use my MA's. Is there really a "best" way to do anything?

Answer: Yes, your way(Individually).

i dont know about you guys, but i have found that the ADX is the greatest trend spotter of all. im not big on indicators, cause they basically tell you the same thing as the price does, but ADX is different. study the formula that welles wilder put together and it becomes clear why - the DMI (which ADX is built upon) is ingenious!

i use ADX when trading discretionally and with quant system. every quant method i have relies somehow on the ADX, to me its like salt to a chef!

oh, another point about trends!

ever asked yourself what causes a trend?? behind every chart there are real people making trades, some short term, some long term. for the technical guys that is sometimes hard to accept. they want to think the price is doing this or that because it hit some fib line or is overbought etc. but it really comes down to to the human side of things. you know that commercial from Dow chemical... "the human element"?

so what causes a long term up trend? well, long term buyers of course! likewise, shorterm sellers are the cause of a short term down trend. thinking of the market in human terms is a big advantage i believe, it really gets you focused on what is actually moving the market and can help you predict where its going too!

this is why im such a fan of watching price action during big news reports. you can get some real insight into what everyone is thinking by how they react to the news. for instance, if you are unsure who is buying the market lately, you can somewhat decipher who it is and what their intention is by seeing how they react to positive or negative news. and then during the following week, you can use this information to help you answer questions like "is the buying over with?", because you know who the buyers are!

great thread btw!

I found this appropriate here since Merlin brought it up and answered. I was typing this out in my own words, but decided to revisit the actual source itself and quote him(why butcher something done right ).

This conversation came out of talk about support and resistance.


the words of Darkstar

"So what is a trend exactly? Technically it is any sustained unidirectional price movement. As should be apparent by now, for a trend to develop there needs to be a greater volume of demand to buy (sell) then there is to sell (buy) across multiple price points.
Trends develop and are sustained for many reasons, but the most relevant for our purposes is the fact that they have a tendency to perpetuate their own continuation. That is, movement in a given direction will generally entice others to initiate orderflow that will help exacerbate the move. This is of course the direct result of the common trader’s propensity to implement trend following systems. You can visit any random trading system thread on ForexFactory to confirm this propensity.
Operating from the premise that a sustained unidirectional price movement (trend) is the result of significant unidirectional orderflow, and that this unidirectional orderflow has a tendency to increase in volume as more and more trend following traders discover a justification for initiating trades, how is it that trends are ever reversed? If you have been following along, it should be obvious that the only permissible way to halt the advancement would be to offer supply in greater quantity then is being demanded."

I like the psychology element in that. Like Merlin said, people in the end. But it can also be strategic reasons, too, I guess?

I.e. with a primary target other than the FX investment itself, but it as only a means to another end?

Anyhow, the "how can a trend turn around then" question is most intriguing if one does adhere to the domino logic of herd mentality, i.e. once a breakout to one side is made, everyone jumps on board the gravy train to get their share...

I guess the biggest male sheep has to blurt out "uncle'" at some point(or simply has enough profit or needs to reinvest or something) and drop dead so hugely, causing so much "damage", that all others suddenly go "uhm, that 50 pips ago was the price I originally woulda closed out at...oh darnit..." and then wait until that is met again on a retrace....THEN ditch out, too, causing first the big plunge for a first downwave, then a retrace, then as price is near or met again the next downwave, which in turn causes the next folks to realize "darn, it's too late to even get out at the retrace, crap, and it isn't stopping, best get out while we still can" and the third, strongest wave downwards starts...and snowballs the other way, in turn....

Something like that, I guess.


I still rather like trendlines best. All MAs I use (no I'm not telling) are wavey versions of trendlines, really. They coincide so much, it's scary.


Also trendlines in effect do nothing but say higher highs, lower lows graphically but give you a spot where to actually trade from with em.


Hm, now what was my point again...oh yea, add 1 to my post count ;p

Moving on...gotta look at ADX again sometime, but last time it was just 3 very squiggly lines giving me a lot of headache...but I remember back in my indicator days or RSI and whatnot, some other guy really loved them, too...

Mh..still though, if everything's calculated off price, then surely we too can at some point feel how price is feeling if the info is already in front of us...just gotta tune in with the market....ohm....

Amen.

Reading this, I am SO happy I am long past that phase

At some point you realize that simple is best (half the PF will probably go "whaaaaaaaaaaaat did HE just say THAT?" right now) as Mike will ascertain

You should tell them "wtf are you looking for a trend on the 5m for" instead if you ask me, but to each his own hehehehehehe ;D

(I know everyone keeps going a trend is depending on where you look for it, but to me that's hogwash. There's ONE main trend (and that is anti-USD right now imho, and I mean USDX by this) and then there's current direction - in trend or corrective against it, but there are not 50 trends on 50 timeframes. There's only one trend and there's only one current direction. Thinking differently will put you into hells kitchen and annihilate chances of being capable of realizing what a retrace and a retrace entry is and why they are truly salvation)

Si, va bene.


This is something so simple nobody starting is capable of understanding this. I wasn't, either, I think. Not sure, been a while.

Which is possible.


Really nice post there iGor

ADXDMI doesn't have to be 3 squiggly lines. Here's what it looks like on my chart:

The formula for ADX as listed here.

From Welles Wilder

ADX (Average Directional Movement)
--------------------------------------------------------------------------------
The Formula . . . . . .

ADXi = [(ADX(i-1) * (n - 1)) + DXi] / n
where n = Smoothing Period
DX = [ 100 * ABS( (+DI) - (-DI) ) ] / ( (+DI) + (-DI) )


Directional Indicators
+DI : Positive Directional Indicator Line
-DI : Negative Directional Indicator Line
--------------------------------------------------------------------------------
The Formula . . . . . .

+DI = 100 * wSum(+DMI, n) / wSum(TR, n)
-DI = 100 * wSum(-DMI, n) / wSum(TR, n)
where n = DI Interval
wSum = Welles Summation (Sum = Value + Sum.1* (n - 1) / n)
The DMI is calculated as follows ...
HiDiff = HI - HI.1 (current high - previous high)
LoDiff = LO.1 - LO (previous low - current low)
if (HiDiff < 0 and LoDiff < 0) or (HiDiff = LoDiff) then
+DMI = 0
- DMI = 0
if (HiDiff > LoDiff) then
+DMI = HiDiff
-DMI = 0
if (HiDiff < LoDiff) then
+DMI = 0
-DMI = LoDiff


True Range (Average True Range)
--------------------------------------------------------------------------------
The Formula . . . . . .

MA(TR, Type, Period)
TR = MAX{
ABS(HI - LO),
ABS(HI - CL.1),
ABS(CL.1 - LO)
}

As jacko put it once -- give a chart to a five year old. He'll tell you what the trend is in about five seconds.

And it works on all timeframes too.

Boy... I couldn't agree more with this statement. You hit it right on the head. We traders have such a bad problem with overcomplicating things.

I dont think it's overcomplicating things but rather wrapping up price with a neccessary "blanket". Let me give you an analogy: price might be a beatufil canvas: very coloful, symetric, etc - but if we give it a nicely polished wooden frame it might easily boost its beauty!

What I am trying to say is that simply saying "higher highs and higher lows" doesnt cut it. We need to "frame" that up a little bit in order to "measure" the trend as it unfolds. Trading is difficult enough already as it is, which is why trading a naked trend just because it's making higher high and higher lows doesnt seem very wise to me. Every trader uses different tools to "measure" the trend -I use SMAs, smsjones uses his own mathematical formula, Merlin uses ADX, Jacko uses round numbers and 50% fib, etc- but regardless of the tool, you need SOMETHING to base your judgement upon that trend you have at hands.

Naked "higher highs and higher lows" is not enough in my opinion.

And that's true. As the markets have changed over the years, as have the ability to define the trend. When Charles Dow developed the theory there was no global information age, no CNBC, no Bloomberg, no news source to base instant reaction upon. With that, you often had to wait hours to days for economic data and possibly even longer for the markets reaction.

So higher highs/lower highs would have been more prevelant in a less volatile environment. Now, with data streaming within seconds and the market reacting practically immediately, the trend lines can become blurred and uncertain; especially when dealing with shorter time frames outside of the weekly and monthly; which are the shortest time frames that Dow had applied the theory to himself. The theory needs adjusting when reverting from macro to micro application.

Basic knowledge of what establishes a trend is invaluable to a trader regardless of market. How that trend is interpreted in the varying time frames will always be a matter of debate.

The importance of the direction of the trend will also vary from trader to trader allowing differing views as to the direction and possible turning points.

If your system or method doesn't allow anti-trend positions, then the percieved direction of the trend should be based on your own personal analysis.

Hi WTB,

Sorry but saying that this is not very wise is the same as saying to someone who is going from the ground floor to the 5th floor with a staircase that this is not very wise.

And saying that you need SOMETHING is the same as saying if you want to go to the 5th floor you need an elevator or whatever other solution...but not on foot.

I know that taking the staircase can be tiring so can be looking to your chart to track the bottoms and tops. But I know that both are very effective.

I don't know what the dow theory is about or how he explained it or to watch witch certain time frames or what the volatility was in his days.
But I know that price is price and that when a trend is clear up it will translate itself in each bottom will be higher then the previous and the opposite for down trend and that no indicator will show me more then what I see from the price itself. And when I don't see it because of to confusing that this is the moment that I better stay away from it because there is no clear trend.

I will never say that looking to a (laging) trend indicator is not very wise nore someone who is taking an elevator.

You openend this topic to ask each of us what our tips or tricks are. Wich we gladely share. So I don't understand that you find some of the techniques not very wise ;-)

friendly regards...iGoR