THE IDEA EXCHANGE - page 12
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
1. one shows the High[] position relative to the EMA, the second shows the Low position relative to the EMA, so I think we should introduce thresholds (significance levels). Compare Delta not with zero, but with these thresholds.
2. You have a small error in your code. The variables cx and lx cannot equal 50, the condition must be cx<50 and lx<50. There is an overrun of the array.
We can do it with thresholds too - I tried it, but it's an extra parameter for optimization, while it has little effect. As for the error, it does not seem to affect it: while (cx<51), while (lx<51)
... ... ...
However, IMHO, the testing is unremarkable. You feel more confident if, for example, you've been optimizing from May to September,
but you're testing from January to December. That is, there is both a run and a run out.
The trend-detector showed good results (including out-of-sample profits) even in the simplest Expert Advisor - the crossing of two MAs.
And also in its pure form - when entering - when the Delta crosses the zero line. And trailing, of course, was present there.
Things went much better when I implemented independent long and short entries. Thus the drawdown is compensated and the number of trades increases. For long and short positions I've provided my own trend detectors. Even GBPCHF is able to make profit!
Things went much better when I made independent long
and short entries. The drawdown is compensated, and the number of
trades increases. For long and short positions I provided my own
trend detectors. Even GBPCHF is able to make profit even with this
!
If I remember correctly, the entry parameters should be optimized separately for buy and sell.
If I remember correctly, you have to optimise the inputs separately for buy and sell.
The scale on the graph is too crude. Four decimal places is not enough! Four decimal places are not enough! We need to add a fifth. In this case, the MA values themselves are simply rounded to four decimal places. For example, the MA values on the chart vary (approximately) from -0.00015 to -0.00025, while calculations show only 0.0002!
I have shown this point with yellow arrows on the chart.
Dear people ! Please, advise me what I have to do to increase the sensitivity of the scale in the indices by 1 decimal place.
My Ruslan: "How do you do it? I have a good impression of the performance of the tr-detector. But during the test I had some unclear moments. I noticed that sometimes there seems to be a signal on the chart. A trade is not opened. I started to analyze it (with my usual scrupulousness - modestly speaking!). Include a comment. The following misunderstanding was found.
The scale on the graph is too crude. Four decimal places is not enough! Four decimal places are not enough! We need to add a fifth. HOWEVER, the MA values themselves are simply rounded to four digits. For example, the MA values on the chart vary (approximately) from -0.00015 to -0.00025, while calculations show only 0.0002!
I have shown this point with yellow arrows on the chart.
Dear people ! Please, advise me what I have to do to increase the sensitivity of the scale in the indices by 1 decimal place.
In the init section of the indicator we just need to write IndicatorDigits(Digits+1); We will get one more digit. If we add +2, we get two more digits.