From theory to practice - page 1276

New comment
 
Uladzimir Izerski:

So what was the advice? Don't be too long now that you've started.

If he understood it, he would have told me a long time ago.
 
Uladzimir Izerski:

So what was the advice? Don't be too long now that you've started.

Yeah, I didn't write about advice. Something was glimpsed, and that's what I thought by association. What he said exactly, I don't remember, and it doesn't matter.
And you know what I do.
 
Yuriy Asaulenko:
Yeah, I didn't write about advice. Something was glimpsed, and it was already suggested by association. But what he said exactly, I don't remember, and it doesn't matter.
And you know what I do.

I know what you do, you just forgot to answer, how does a dandelion know there's wind on Earth?!

 
Aлександр Антошкин:

I know what you're doing, you just forgot to answer, how does a dandelion know there's wind on Earth?!

+++

Lucky.))

 
Aлександр Антошкин:

I know what you're doing, you just forgot to answer, how does a dandelion know there's wind on Earth?!

The interesting thing is, it doesn't know. But without the wind, there would be no dandelion.
 
Alexander_K:

A breakdown indicator - no, it has to work here and now. Instantaneous speed as an example.

it often happens that price has travelled a lot in a second bar, i.e. moved at high speed, but that does not mean that it is trending now. such big second bars appear from time to time.

I think that there is no indicator that can show here and now (on the second bar or on the one-minute bar) what phase the market is in now.

If there were such an indicator, the market could be decomposed into states

When the market is in a trend - switch on the trend EA, when it's in flat - switch on the flat EA.

There is no instantaneous indicator, we will have to watch the last time interval for several bars.

 
multiplicator:

There is no instantaneous indicator, you will always have to look at some recent time span over several bars.

As it were, I'll add.
The market is definitely a random process. The market is not SB, although SB is a very good market model, and in many ways it is a very similar process. The market, like SB, has no derivative. No matter how much one can thin it, it does not help - we obtain the same SB with different characteristics and no derivative. More precisely, the rate is undefined, on any interval. There is no indicator, not only instantaneous, there is no indicator at all.
 

Alexander_K:

So it would be interesting for me to look at a summary table of Hearst values, autocorrelation, etc. on different periods - to make sure that independent studies have resulted in the same periodic market structure as I have, where the deviation indicators show correct values with at least 80% probability.

I used to check it once. But I did not look at the whole period, but at the period of deviation, in the figure - red.

autocorrelation was calculated for this piece of the chart. and then, depending on the indicator value, I made a decision whether to enter or not to enter the trade. this piece of the chart is a trend or a flat.
(I also tried something similar to Hurst.)
it changed the system results.

did your results change much after introducing such an indicator?

 
Yuriy Asaulenko:
I'll kind of add to that.
The market is certainly a random process. The market is not SB, although SB is a very good market model, and in many ways they are very similar processes. The market, like SB, has no derivative. No matter how much one can thin it, it does not help - we obtain the same SB with different characteristics and no derivative. More precisely, the rate is undefined, on any interval. There is no indicator, not only instantaneous, there is no indicator at all.

I disagree. There is such an indicator - otherwiseDmitriy Skub would not have kept his modified Hearst method of calculation in the greatest secret.

 
multiplicator:


have your results changed much since the introduction of such an indicator?

Sure. But, I'll be 100% sure after 3 months of real-life practice. Let's see.

1...126912701271127212731274127512761277127812791280128112821283...1981
New comment