From theory to practice - page 693

 
Alexander_K:

For the oligophrenics, let me repeat:

- if the number of ticks in the sliding window will satisfy the Poisson distribution, then we have a quasi-stationary process, which the Ornstein-Uhlenbeck model requires.

Provided the window >=24 hours has, on average, a negative correlation coefficient on all pairs, then by calculating the sliding window size we will 100% come out with a process returning to the mean.

See, kiddies, it turns out the sliding window = 24 hours is not enough. And my +18% for the month is still nothing to say - just lucky.

I suspect that the coveted "trend/fly" key is not there.... Well, there isn't one at all - ghost and fairy tale.

The most important key is the size of the temporary sliding window. But, about that - shhhhhhhh..... Not a word to anyone.

You have to decide what kind of trend you're looking for. How many pips? There are trends on almost every timeframe. There are on the monthly TF and on M1. What is the scale? For example, what we consider a trend on the minutes is a "false swing" on the hourly timeframe.

 
Alexander_K:

For the oligophrenics, let me repeat:

- if the number of ticks in the sliding window will satisfy the Poisson distribution, then we have a quasi-stationary process, which the Ornstein-Uhlenbeck model requires.

Provided the window >=24 hours has, on average, a negative correlation coefficient on all pairs, then by calculating the sliding window size we will 100% come out with a process returning to the mean.

See, kiddies, it turns out the sliding window = 24 hours is not enough. And my +18% for the month is still nothing to say - just lucky.

I suspect that the coveted "trend/fly" key is not there.... Well, there isn't one at all - ghost and fairy tale.

The most important key is the size of the temporary sliding window. But, about that - shhhhhhhh..... Not a word to anyone.

Too much sarcasm, but thanks for the info, accessible.

 
Baruban:

Too much sarcasm, but thanks for the information, accessible.

That's how I'm used to communicating here - as a defence against trolls. But, everything I write is serious research.

 
If the London session opens, the entire previous intraday story with the statistical calculations can mostly be thrown in the trash. The nature and number of participants changes dramatically. The frequency of trades and total lots bid and ask changes. And the positioning of collective price targets. If currency intervention on the instrument has started, the entire previous history with its statistical and probability parameters can be thrown under the tank. GBPJPY trading is most often managed and conducted mostly from "round numbers". Exchange rate movements - look like random wandering, but they are managed on the other side. The birth of a stable trend out of a flat is 70% man-made.
 
Oleg Papkov:
If the London session opens, the entire previous intraday story with the statistical calculations can mostly be thrown into the trash. The nature and number of participants changes dramatically. The frequency of trades and total lots bid and ask changes. And the location of collective price targets. If currency intervention on the instrument has started, the entire previous history with its statistical and probability parameters can be thrown under the tank. GBPJPY trading is most often managed and conducted mostly from "round numbers". Exchange rate movements - look like random wandering, but they are managed on the other side. The birth of a stable trend out of a flat is a 70% man-made process.

+1

BigBen trading contradicts the theories, but conforms to practice... So decide which one is wrong: theory, BigBen or practice...

 
Alexander_K:

For the oligophrenics, let me repeat:

- if the number of ticks in the sliding window will satisfy the Poisson distribution, then we have a quasi-stationary process, which the Ornstein-Uhlenbeck model requires.

Provided the window >=24 hours has, on average, a negative correlation coefficient on all pairs, then by calculating the sliding window size we will 100% come out with a process returning to the mean.

See, kiddies, it turns out the sliding window = 24 hours is not enough. And my +18% for the month is still nothing to say - just lucky.

I suspect that the coveted "trend/fly" key is not there.... Well, there isn't one at all - ghost and fairy tale.

The most important key is the size of the temporary sliding window. But, about that - shhhhhhhh..... Not a word to anyone.

I open the page and the first thing that catches my eye is "I'll say it again for the oligophrenics", it hasn't been this hard yet))))))))

It's just fatigue...

 
Novaja:

I open the page and the first thing that catches my eye is "I'll say it again for oligophrenics", it hasn't been this hard yet))))))))

It's just tired...

No, a lot of people here aren't really interested in the substance of the issue. They are interested, as M.M. Zhvanetsky said, in the personality of the person arguing. "An argument without facts, an argument with a shift to the personality of the guy. "What can a bald man say to us?! Let him grow his hair and then we'll listen to him, of course!!!" Bravo MM. Oh, and Alexander set himself up. He promised the grail. Sometimes the promise of money works as well as the money itself. Now he owes everyone. :))

 
Oleg Papkov:

No, a lot of people aren't really interested in the substance of the question. They are interested, as M.M. Zhvanetsky said, in the personality of the person arguing. "An argument without facts, an argument with a shift to the personality of the guy. "What can a bald man say to us?! Let him grow his hair and then we'll listen to him, of course!!!" Bravo MM. Yes, and Alexander has set himself up. He promised the grail. Sometimes the promise of money works as well as the money itself. Now he owes everyone. :))

I don't keep my promises.

I'm sitting here analyzing: was my +18% for the month a fluke or not?

I have left aside my desperate attempt to find a trend/flat key. In my opinion, the key is the size of the sliding time window. Could be wrong... Checking....

 
Alexander_K:

I am not giving up on my promises.

I'm sitting here analyzing: is my +18% for the month a fluke or not?

I have left aside my desperate attempt to find the trend/float key. In my opinion, the key is the size of the sliding time window. Could be wrong... Checking....

wrong

 
Олег avtomat:

wrong


I think I agree, the size of the window has nothing to do with it.