How to determine the TREND?

[oem7110]

On fig 1, up trend is established, way to go ...
On fig 2, market starts to go down on the next day, do we still consider up trend at this moment?
On fig 3, market keeps dropping on the next day, do we still consider up trend at this moment?
On fig 4, market break below previous peak, on the next day, do we consider down trend at this moment?

[Chris67]

depends what method of trend identification you are using surely ?
in all 4 examples the last point is higher than the start - so in all 4 instances the market direction is up ? in all 4 examples they could be 10 minutes of 2 years of charts in which case that makes a huge difference to ?
Its a bit of a tough one to answer really
C

[Trend Rider]

OEM7110 - I agree with Chris67. It all depends on the timeframe and perspective you are looking at. I like to start out with a long term chart 5-10 years at least to see the overall perspective - the sum of the total so to speak. I then move to an intermediate timeframe 1-2 years to see what the latest market actions have been and how the overall trend looks. I then move to a short term timefame such as a 3-6 month timeframe to put it all in perspective. Anything shorter for me is a Swing Type or Momentum Trade and I will look at shorter timeframes to gain the proper perspective for those particular types of trades.

I think you propose a good set of questions that most traders - new or in-experienced have a hard time identifying.

[Robwynge]

Stepping back a bit, the issue you raise points to an unsolvable trading dilemma, which is that there is simply no 100% sure way to know if the pull back in your uptrend is just a short-term pull back prior to trend resumption, or whether it is the end of the trend (there may be ways to handicap the odds, but still...). Therefore, it is up to us to define the types of market movements we want our system to capture.

In your figures, obviously the sooner we declare the trend over and sell to take profits, the more likely we are to cut off what could become a big winner. Therefore, the question is whether we have the type of system that generates enough small winners to pay for our losers (we likely need a high win rate). If not, then cutting off small winners that might have become significant will almost certainly harm our system, perhaps fatally.