Market condition - flat or trend? Which dominates? - page 11
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Therefore, for a (more accurate in my opinion) assessment, the logic for calculating the TFS needs to be slightly modified. (Your comments and suggestions are accepted)
The value of segments in points will not change relative to the previous variant, while the time value (number of bars in the flat and trend) will considerably change, see Fig. 1.
Please note an "interesting" segment. At transition of one trend to another there was no flat section. Formally it could have been attached to the preceding trend, as in the diagram above, see post of Xadviser 04.04.2008 16:21
But to estimate the time factor of price being in the flat range, it would be logical to attribute it to the flat. It also does not contradict the principle of the price chart construction. After all, we never know the moment of price transition from one state to another until the specified (accepted by us for estimation) conditions are fulfilled. That is why I propose to carry out the estimation with such criteria. We can see from the figure that flat segments have increased (in time), while trend segments have decreased. This change will have a significant impact on the time correlation of the TFS.
Yes, perhaps I should have looked at the details of the calculation before picking on it :). I just don't have this as my main topic at the moment.
I understand, but with the new changes, I'd like an answer to this question: Was it you who counted the time plot relationship? And what would the relationship look like with the TFS relationship in points?
And another untested thought. On your graph of TFS percentage change you can determine in which time interval (for the pair in question) the extremums of the obtained graph are located. It clearly shows some "threshold" levels. If we connect them with the diagram of the pair and see "what happens there", maybe some conclusions can be made. The article 'Pulse-based Market Diagnostics' and this phrase in it came as a reminder: "The three-month market (January - March 2008) is becoming regular but neutral-noise. This could mean that the initial months' trends are beginning to be superseded by the opposite - the market is turning around."
Yes, perhaps I should have seen the details of the calculation before picking on it :). I just don't have this as my main topic right now.
I understand, but due to the new changes, I'd like an answer to this question: Was it you who counted the time plot relationship? And what would the relationship look like with TFS ratios in points?
And another untested thought. On your graph of TFS percentage change you can determine in which time interval (for the pair in question) the extremums of the obtained graph are located. It clearly shows some "threshold" levels. If we connect them with the diagram of the pair and see "what happens there", maybe some conclusions can be made. The article 'Pulse-based Market Diagnostics' and this phrase in it came as a reminder: "The three-month market (January - March 2008) is becoming regular but neutral-noise. This could mean that the initial months' trends are beginning to be superseded by opposing trends - the market is turning around."
The great advantage of Andrei's script is the ease of making minor modifications :). The ratio picture for points (synchronized) looks like this
The horizontal axis is time (seconds). A modification of the script is attached. Time is written in MT format - to make it easier to apply the data to a larger timeframe.
As for threshold levels, to be honest, I don't see them - not the right statistics for such statements.
As for threshold levels, frankly, I don't see them - not the right statistics for such statements.
A big advantage of Andrei's script is the ease of making minor modifications :)
Your version doesn't really hit the percentages on the bars
Some of the tools are just begging to be put to work. Their statistics are a joy to behold... see reports
It would be nice to show the maximum flat sequence in segments and most importantly in pips (wish)
It would be great to derive a maximum flat sequence in segments and most importantly in points (wish)
The relationship picture for the points (synchronised) looks like this
The horizontal axis is time (seconds). A modification of the script is attached. The time is written in MT format - to make it easier to apply the data to a larger timeframe.
By Sunday evening, sum up everything that needs to be done, and I'll come and do it. I was beginning to forget ;)
I'll do my best. Thoughts are coming up as I go along, I'm trying to keep track of them so I won't forget them myself.
Still interested?