Market condition - flat or trend? Which dominates? - page 18

New comment
 
Neutron:

Here, I remembered another very, very scientific definition of a trend or flat market. It's based on a Zig-Zag (ZZ) type of pattern.

Thank you for your attention to the question. I think this method is very similar to the proposed one, as noted by Composter. I would like to compare the estimation though. But the trend/flat criteria would be fundamentally different. A good option was suggested by SK (using linear regression) But no longer want to burden a volunteer. If anyone is willing to lay out some kind of estimates, we'd be happy to accept. As for all the above, a lot of work has been done, results obtained, conclusions drawn. Everything applies to the proposed criteria. Those who read attentively will be given food for thought.

50/50 can be different. There are times when you can get a little. Did you estimate this "bit" or was it strictly less than the spread?

No in-depth evaluation was done because of the other task at hand. The "answer" coincided with the theoretical assumptions (as required to be proven), but that does not mean that no benefit can be derived from it. "Slightly" appeared in some combinations, but the stat advantage was small (57/43 on average) and all of course without spread and swaps, i.e. theoretically. Mainly EUR/USD pair was tested, others slightly different, but insignificantly. "Rip" :-) is possible, but in a different way.
 
imsgfx:
Here is the latest example (EURO, M1, 30 pips, spread 2 pips)

- break 1.5800

- BUY 1.5807

1. Check your terminal settings for deviation from the requested price (I think you have 5 pips there, which is more than likely)

2. "Live" does not always perform pip by pip. The price might not have been at this level long enough for you to open a position, the opening is usually done at the next closest to your price, see point 1.

3. The most important thing. This EA is not designed for trading. Its task is to collect statistical data on the historical period in order to evaluate the ratio of staying in the trend and flat position according to the set criteria.

 
Xadviser:

Thank you for your attention to the question. I think this method is very similar to the one proposed, as noted by Composter. I would like to compare the assessment though. But the trend/flat criteria would be fundamentally different. A good option was suggested by SK (using linear regression) But no longer want to burden a volunteer. If anyone is willing to lay out some kind of estimates, we'd be happy to accept. As for all the above, a lot of work has been done, results obtained, conclusions drawn. Everything is applicable to the proposed criteria. Those who read attentively will give food for thought.

At one time I dealt with this question, I still have the code that allows to estimate the profitability of the above TS.

On horizontal axis - a step of division H-Z, on the vertical - <Z>-2H - average return on specified instruments for 1 month. The positive return of the strategy corresponds to the trend state of the market - when it is optimal to trade on the movement. Negative - points to the flat condition and it is optimal to trade against the movement (for more details see 17 page of the topic).

From this data follows that the main market condition is uncertain (50/50) with the light prevalence of flat, which allows you to use it to reckon on an average return of 2-3 points from each entry-exit from the market. The optimum PZ for this period of historical data has H=10-20 pips.

 
Xadviser:
imsgfx:
Here is a recent example (EURO, M1, 30 pips, spread 2 pips)

- break 1.5800

- BUY 1.5807

1. Check your terminal settings for deviation from the requested price (I think you have 5 pips there, which is more than likely)

2. "Live" does not always perform pip by pip. The price might not have been at this level long enough for you to open a position, the opening is usually done at the next closest to your price, see point 1.

3. The most important thing. This EA is not designed for trading. Its task is to collect statistical data on the historical period in order to estimate the correlation of the price staying in trend and flat position according to set criteria.

The order was opened as it should (according to the EA's algorithm) on the bar following the break (the bar opened at 1.5805) with a 2 point spread (1.5807). When taking the statistics into account, the beginning of the flat of that period is taken as 1.5800 and the opening of the next bar should be taken that blurs the whole picture. The variant of how to correct such a problem is obvious and I wish you good luck in your developments.

 
imsgfx:
The order opening, as it should be (according to the Expert Advisor's algorithm), took place on the bar following the break (the bar opened at 1.5805) with a 2 points spread (1.5807). When taking the statistics into account, the beginning of the flat of that period is taken as 1.5800 and the opening of the next bar should be taken that blurs the whole picture. The variant of how to correct this problem is obvious, and I wish you good luck in your work.

Understood.

But it is not a "problem". It depends on what strategy to use. In the EA settings, you can enable "counter-trending". In that case, the example you have described would be an advantage.

Thank you.

 
Neutron:
From the above data, it follows that the underlying market condition is uncertain (50/50) with a slight predominance of flat, which allows you to count on an average return of 2-3 points from each entry-exit from the market when exploiting it. Optimal PZ has H=10-20 points.

The apparent similarity of the obtained results suggests their credibility. Although the methodology is close.

The half-long return of the strategy corresponds to the trend state of the market - when it is optimal to trade by the movement. Negative returns indicate a flat condition and it's optimal to trade against the movement.

Have you evaluated the possibility of evaluating the "strategy's profitability"?

 
Xadviser писал (а)

Have you evaluated the possibility of estimating the "strategy's profitability"?

Yes, I checked both on demo and real - the yield is exactly the same, minus the spread minus the inflation (nstability) of the optimum for H.

 

I've been doing some research and I found an interesting picture...

I think it might remind someone of something...

p.S. You have to look at each graph individually...

 
forte928 писал (а) >>

I was doing some research and discovered an interesting image...

I think it might remind someone of something...

p.S. You have to look at each graph individually...

Eugene, you are talking in riddles. Is there a certain sacred meaning or is it the style of phrasing?

Was it an observation or a study? What was its essence and what were its aims? What were the results and conclusions?

Of course it reminds, but says NOTHING :-(

What is the graph? In this drawing? The coordinate definitions are not very clear (vertical axis - grid lines).

Clear explanations would be appreciated.

Best wishes,

Sergey

 

I took four waves and added them with periodicities...1;0.618,0.5,0.382 - the result is on the screen, this constantly appears on charts and it can be observed both in forward and reverse directions, the selected form should be applied to the existing price trend and according to its correspondence to draw a conclusion about the nature of price movement... How to do this is certainly an idea but it can be imaginary...

there is mail for communication forte@inbox.ru

1...111213141516171819
New comment