FR H-Volatility - page 26

New comment
 
Mathemat:

Prival, for you: the redhead is gold.


Here gentlemen, thank you so clearly.
 

Here, compare the autocorrelation function built for bars containing equal number of ticks n=1...100 (red line), and equal number of minutes n=1...100 min.

EUR/JPY tick history data 2007

 

OK Candid. I think this is a true disaster for people, as the volume is high, and people usually look at bars by astronomical time. But I'm not from here, I'm from Sirius; and I was taught on Sirius that true market time is operational time. So I don't see much difference between what appears to people to be a disaster and a marketplace in normal, non-catastrophic time. After all, to me a long hairpin at NFP of 150 pips on the oira, which went up in 15 minutes, is equivalent to three regular bars on the hourlies, which are not too different from the adjacent ones :)

So which is still the correct view - the terrestrial view or the Syrians' view?

 
Mathemat:
OK Candid. I think this is a true disaster for people, as the volume is high, and people usually look at bars by astronomical time. But I'm not from here, I'm from Sirius; and I was taught on Sirius that true market time is operational time. So I don't see much difference between what appears to people to be a disaster and a marketplace in normal, non-catastrophic time. After all, to me a long hairpin at NFP of 150 pips on the oira, which went up in 15 minutes, is equivalent to three normal bars on the hourlies, which are not too different from the adjacent ones :)
Got it, Siriusians don't need money from the market :)
 
Prival:
Neutron:

Thought it had long been an open secret - the advantage of operating time...

One step further, and the benefits of Renko time will be revealed to the select few. This is when a time quantum of the market is taken as the time of price change by the value H.


A little easier for the military :-) please, both on the X-axis and the Y-axis. What is the "time quantum of the market". Please help me, I spend half a day digging, sometimes to understand the meaning of some phrases. Like volatility is simple - it's volatility. Like in the anecdote sometimes it turns out that a soldier asks you what an orange is? He replies, "Do you know what a tram is? No, he says I do not. Answer. An orange does not look like a tram at all.

To make it easier to understand, here there are several different representations of this price trajectory.

http://vtsystem.narod.ru/market.htm
http://vtsystem.narod.ru/market2.htm
http://vtsystem.narod.ru/market3.htm

If we are talking about what I called Renko time, information about this division is contained in the third link (Fig.4). But I didn't find anything about the operating time in your links. There are similar constructions (in price-volume coordinates), but it's not quite the same. You have to dance from the tick history, while the author is referring to minutes.

P.S. On the chart above is the correlation coefficient between adjacent bars drawn from minutes and from ticks, where is the greater pattern?

 
Neutron:

Here, compare the autocorrelation function built for bars containing equal number of ticks n=1...100 (red line), and equal number of minutes n=1...100 min.


I do not quite understand what this diagram has to do with adequacy of the concept under discussion.
 
If by concept we mean transition to operating time (OT), and by adequacy we mean consistency between expected results of such transition and those actually obtained, then I do not see the reason for your misunderstanding. Looking at the figure, there is a noticeably greater relationship between the price increments in the OM coordinates. This, in turn, allows for more efficient exploitation of some market analysis tools.
 
Neutron: You have to dance from the tick story and the author is tied to the minutes.

That would be nice, but we have what we do not have: a deep ticking history we do not have. The history from ratedata.gaincapital.com does not count, as it is not my DC's history.

I intend to build equivolume bars exactly by minutes (or even by 5 minutes) - and not by a fixed number of these bars in an equivolume bar, but approximately by a fixed volume.

Example: on the 5-year history of the euro, in each 4-hour tick - about 1000 ticks (I don't remember exactly, but the order of the figure is such). Ok, let's pass to 5-minute bars and collect every equivolume bar corresponding to 4hours (we will draw it on H4) on condition that there should be 5-minute bars so that their total volume is close to 1000. In reality it may be 1050 or 950, but it cannot be helped. Hopefully a market portrayed like this will contain far fewer disasters.

 
Mathemat:
Neutron: You have to dance to the tick story, but the author is tied to minutes.

That would be nice, but we have what we don't have: a deep tick story. The history from ratedata.gaincapital.com does not count because it is not my brokerage company's history.

I intend to build equivolume bars exactly by minutes (or even by 5 minutes) - and not by a fixed number of these bars in an equivolume bar, but approximately by a fixed volume. Example: on the 5-year history of the euras in each 4-hour bar - about 1000 ticks (I don't remember exactly, but the order of the figure is such). Ok, let's pass to 5-minute bars and collect every equivolume bar corresponding to 4hours (we will draw it on H4) on the condition that there should be 5-minute bars so that their total volume is close to 1000. In reality it may be 1050 or 950, but it cannot be helped. I hope that the market that is depicted like that will contain fewer disasters.


For that, the number of ticks per unit of time would need to be constant on average. What happens if it increases slightly over time?
 
Yes, Vinin, I agree, there is no such consistency (you can see it on the daily data - the volume varies from 4 to 13 thousand). So we can try to normalise the volume within the equivolume bar by some average of the actual volume over a longer period.
1...192021222324252627282930313233...42
New comment