From theory to practice - page 1272
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
There was no significant economic news in May, but summer starts in June.
and that's where the spikes come in.
Let's see what Volchansky has to say after that.
So he's probably not trading during the news.
Here's the thing...
I decided to look at optimal indicator periods for different months.
the result is the following, for the last 12 months:
710
150
400
60
830
320
70
330
80
140
560
300
is there any relation between these numbers? autocorrelation? Autocorrelation here is negative, -0.38. Anyway, this parameter changes very sharply, no smooth parameter change.
"for any random graph you can find an indicator period, on which it will show a good profit".
Is this data on minutes? You are averaging a period = 1 hour. Try working with ticks, with a sample size averaging 1 hour.
I've long wanted to ask, what does Hearst have to do with forex?
None, no matter what anyone says. The main problem is the stationary Gaussian distribution of this parameter. It shows trend/float with a lag when the main events have already passed.
However, I would like some Person to do a comparative analysis of the performance of 3 indicators of decay: Hurst, autocorrelation, some other one to choose from.
And with sample volumes in 60 minute increments: 60, 120, 180,....
This is an assignment.
Don't expect me to personally do it. I do not need it.
It shows a trend/float with a lag when the main events have already passed.
That's what I wrote. We are analyzing the last period of n bars, so the analysis will always be lagged, because we are analyzing the history that has already passed.
there is nothing that allows us to see whether the market is trending or flat.
There's nothing that allows us to look directly at what's trending or flat at the moment, except to see what happens on smaller timeframes, assuming that the market is self-similar and the same processes occur on all timeframes.
that's what I wrote about. we're analyzing the last period of n bars. therefore the analysis will always be lagged, because we're analyzing the history that has already passed.
there is nothing that allows us to see whether the market is trending or flat.
There's nothing that allows us to look directly at what's trending or flat at the moment, except to see what happens on smaller timeframes, assuming that the market is self-similar and the same processes occur on all timeframes.
Correct, but not quite.
The central tendency measure (average) - can and should be lagging, because what matters to us is how far price has gone over a certain period of time.
A divergence indicator - no, it has to work here and now. Instantaneous speed as an example.
However, most indicators are built on averaging over the sample volume. So what to do? The answer: look for periods in the market structure, some kind of bearers. They do exist and I use some of them in my TS. And everything works properly in these periods.
Therefore, I would be interested to see a summary table of Hearst values, autocorrelation, etc. on different periods - to make sure that independent research has resulted in the same periodic market structure that I have, on which the breakdown indicators show correct values with at least 80% probability.
I have long wanted to ask, what does Hirst have to do with forex?
exactly the same as "econophysics" i.e. none.
here's the thing...
I decided to look at the optimal indicator periods for different months.
The result is as follows, for the last 12 months:
710
150
400
60
830
320
70
330
80
140
560
300
Is there any correlation between these numbers? autocorrelation? autocorrelation here is negative, -0.38. In short, this parameter changes very sharply, there is no smooth change in the parameter.
"for any random graph you can find an indicator period, on which it will show a good profit".
These figures are just the result of a self-adjusting process in the market. You can see it well on history. Abrupt changes This is the result of news portions.
Any significant price change in one instrument will pull a change in the others to smooth out the imbalance.
Is this data on minutes? You are averaging a period = 1 hour. Try working with ticks, with a sample size averaging 1 hour.