lets define a trend..quantitatively - page 2

 
luenbo:
never define,but follow.
 yes but how you know its a trend? maybe its a correction within a major trend in the opposite direction. And because of using smaller time frames you see it as a new trend. 
 
TalonTrader:
Thought I'd say something else.  95% of traders lose.  And most of them are trying to catch the trend.  HMMMM.    What does that tell you?   The market spends more than 50% of the time in consolidations/range conditions.  And it is at these times that traders get chopped up the most.  So perhaps the better question is "how do I determine range vs. trend conditions?"  and "Can my trading style or EA adapt to both conditions?"  ADX works well for this, but in manual trading, I just keep my eye on the support and resistance levels and momentum, watching for turns.  -TalonTrader

its an interesting comment. I heard many times that the market does range xx% of the time, but never done my own research, how you know its 50%? and again for what time period..

 

 " So perhaps the better question is "how do I determine range vs. trend conditions?  ..its king of the same i think, because if you know its a trend , then its not ranging. 

 
rob:

The last chart was going to be the 15min but I could not see the point as there nothing more to see other than the price rising and a little distortment along the way, the point I'm making, maybe to quantify the trend we may have to start with a greatly higher time frame otherwise we can only see chaos on the smaller time frames, in the instances above we are clear there is a trend however this is after the event but strip out the 1&4hr chart, could you deduce a trend that would last ten days, possibly the approach would be from a position management view, as the price moves into profit, you move the stop-loss to the open position, you could aslo reduce the position size as the market "notches" up, just a thought.

This concept of higher time-frames I read about, "bird watching in lion country" no publicity intended, in the book there was a great story of how 3 blind men described what they were touching, it was strange that they were all touching an "elephant" but each one had a different description of what this was, they could only interpret what they thought they could "see", a chart and all information, to each individual is viewed very  differently 

thanks for the comment and the charts. I actually do agree with you that to quantify the trend you should start with a greatly higher time frame. The problem of this approach is when the trends starts to revert, you will necessarily take a loss especially if using an ea. Some kind of time restriction may be usefull.
 
losing its necesary evil in forex so dont worry about losing when the trend is changing. worry more about winning more than losing...but then again u will have a problem because if u d know when to exit maybe you d see the end of the trend

my advice, don t think about making an ea without having exact strategy for entry, exit, partial profit what ever...
 

When you say time restriction I can only see part of a trend being traded, you would have to some kind of parameter broken by the price, ultimately the price would dictate direction, also, the price could retrace many points but still keep the trend in tact, ref, the first chart posted, the price spiked then retraced but carried on in the desired direction, hence the use of a trailing stop, it would have triggered and the best part of the trend lost.

It has always been said that you don't want  to find exact "tops" and "bottoms" but the bit in the middle, if you are trading 1lot you could reduce into the trend as it moves, this way on reversal you will loose a few pips but the lion share of the trend would have been had.

Are you a programmer O/S I think some thing as complex as what you ask could only be done by you, the programmer, possibly paid for, but trial and error would need lots of tweaking, maybe too much to pay for, I hope you achieve your goal though.

Documentation on MQL5: Standard Constants, Enumerations and Structures / Trade Constants / Trade Orders in DOM
Documentation on MQL5: Standard Constants, Enumerations and Structures / Trade Constants / Trade Orders in DOM
  • www.mql5.com
Standard Constants, Enumerations and Structures / Trade Constants / Trade Orders in DOM - Documentation on MQL5
 
onlysolo:

its an interesting comment. I heard many times that the market does range xx% of the time, but never done my own research, how you know its 50%? and again for what time period..

 

 " So perhaps the better question is "how do I determine range vs. trend conditions?  ..its king of the same i think, because if you know its a trend , then its not ranging. 

50% is just my best guess after manually trading for 7 years on stocks, options, futures, and now FX. 

Yes, if it is trending, then it is not ranging.   Many traders just try to stay out of the chop, while I thrive on it using a fast chart and a small target.  (Well, that is my goal anyway).

Some say picking tops and bottoms is a fool's errand.  I disagree 100%.  Picking the trend reversal is my forte.  How do I do it?  For a short, a combination of the 1 min x 50 sma, and the break of the most recent bounce low point.  I wait until the 1x50 is flat or falling and then make sure that weakness is proven by taking out that low.  The stops get run at that point so the support is weakened.  THEN I wait for a pop back into resistance.  Then I short.  I have about a 70% success rate with this method.  (again, just a guess)  This method is for extended trends.  And, no, I am not perfect at following my own advice!

 

TT, wouldn't disagree with you on your last comment, if you have a way of picking tops and bottoms then go for it, I suppose it's all relative, experience prevails on such trading methods,

I have included a pic below, is this near to what you mean, I have plotted the 50SMA and from the example, if you ran the trade there was a possible 70+ pips to be made on this particular trade.

 

 
rob:

TT, wouldn't disagree with you on your last comment, if you have a way of picking tops and bottoms then go for it, I suppose it's all relative, experience prevails on such trading methods,

I have included a pic below, is this near to what you mean, I have plotted the 50SMA and from the example, if you ran the trade there was a possible 70+ pips to be made on this particular trade.

 

Yes, the 50 sma is great.  Although with my scalping mentality, I would have been out at the first big support level.  I have a very hard time holding on for 70 pips, LOL.  I have been bit so many times.  I like to take the base hits and quietly win the game... (well, at least that is what I try to do).


 
onlysolo:
 yes but how you know its a trend? maybe its a correction within a major trend in the opposite direction. And because of using smaller time frames you see it as a new trend. 

https://www.mql5.com/en/articles/136

Here is a great article on trend determination.

Several Ways of Finding a Trend in MQL5
Several Ways of Finding a Trend in MQL5
  • 2010.09.10
  • Dmitriy Skub
  • www.mql5.com
Any trader would give a lot for opportunity to accurately detect a trend at any given time. Perhaps, this is the Holy Grail that everyone is looking for. In this article we will consider several ways to detect a trend. To be more precise - how to program several classical ways to detect a trend by means of MQL5.
Reason: