ZigZags Shepherds - page 11

 
Uladzimir Izerski:

... The deep frustration with ZZ is that it was originally built on a flawed algorithm.

Excuse me, what exactly is meant by the above?

"Volumes as such we don't see. Full stop." - You can agree with that. But we all see the total numbers of purchases and sales in a given timeframe. And this is already interesting.

 
Andrei:

Open and Close are random variables within an interval, they make no sense initially. Only the extremes make sense...

Where do you think High and Low are? Not within the interval? What if we take a larger interval?
 
aleger:

Excuse me, what exactly is meant by the above?

"Volumes as such we do not see. Full stop." - You can agree with that. But we all see the total number of purchases and sales in a given timeframe. And that's already interesting.

In the lag of course. It's good on its own.))

 
Uladzimir Izerski:

In lagging, of course. Good in itself.)))

The reason for the "lag" is the delayed visualisation of the current trend and it may well be "corrected"!

 
sibirqk:

Perhaps it's useful to state common knowledge once again - repeating the commonplace truths won't make things worse.

There are several ways of displaying price in the market - renko, renji, ekeioboost, kagi, etc. They are usually divided into time-independent and time-independent, but in my opinion this is not a sensible breakdown. It is better, imho, to classify according to the constancy of any characteristic. In this case, the usual price representation can be called equi-temporal, i.e. every bar always has the same amount of time, while the opening/closing price difference and Hg/Lw difference may vary greatly.

Renko is equi-opn/cls, i.e. the open/close price difference is always the same modulo and equal to the renko size, while the difference between Hg and Lw is different, but in the range of one to two renko sizes. The amount of time in the renko bar is also different.

Rencos are equi-Hg/Lw, i.e. the difference between the highest and lowest point of a bar is always constant, the price difference of an opn/cls bar can be anything within the size of the reng. The amount of time in the bar randge is also different.

If we consider that the tick representation of price is the most natural, then a chart of rendezvous bars is essentially a tick chart with the size of a tick per rendezvous. For example, if there are tick quotes with the size of a five-digit tick, and it is necessary to programmatically build a four-digit one - find min/max as soon as their difference exceeds 10, a new four-digit tick chart is formed, i.e., a four-digit chart with the size of 1 is formed.

Equi-bars, as understood from the name, are bars in which the number of ticks is the same, all other characteristics associated with the price are different for each bar. In my opinion - the most meaningless representation of price. The price flow, if we consider it from a multitude of providers, always has a width of a few spreads. This width is not constant over time and widens and narrowed. Every brokerage company broadcasts the prices to its clients terminals using the program - let's call it a price filter. You have to be very simple not to understand that a significant part of brokerage companies profit comes from the quotes filter, and therefore its logics is not simple. And the amount of generated ticks is determined by this logic. I.e. equivolume bars will have different appearance in different brokerage companies and moreover they can have different appearance even in different accounts inside one brokerage company. Coming back to the topic of the branch - statistical characteristics of zig zags will be different for different brokerage companies due to quotes filters working, and within the same brokerage company they will be constantly "floating" because of the nonlinear internal logic of the quotes filter. It is clear that these differences and changes will be small, but statistical advantage of Pastuhov, as I understand it, is also small, so it is only theoretically possible to earn on this, imho.

Bravo! It is impossible to describe the existing ways of representing the price more clearly and succinctly. You would have made an excellent teacher. Although I don't know, maybe you are a teacher).

 
aleger:

The cause of the "lag" is a delay in visualising the current trend and it may well be "corrected"!

I have had the lag issue resolved a long time ago.

Nice to see you on this forum as a researcher.

P.S Read my post and was surprised to find. How ambiguous these two sentences can be perceived)).

One has to get it right. Respect to you as a researcher.

 
Evgeniy Chumakov:
And where do you think the High and Low are? Not within the interval? And if we take the interval higher.

If you take the interval above, of course they will be inside, but there will be High and Low...

 
Uladzimir Izerski:

The deep frustration with ZZ is that it was originally built on a flawed algorithm.

There are many different algorithms that can be used to build a partition... There is unlikely to be a single best ZZ...

 
Alexander_K2:

To be honest, I haven't researched ZigZags...

But, if it can be argued that the market is dominated by the Laplace distribution (or rather a double geometric distribution), then the sum of the numbers with the Laplace distribution is what gives the xi-squared, which tends to be normal as K increases.

By the way, I've read a lot of posts on ZigZags - many people use themonly for their ability to work with extremums HIGH and LOW, believing that they can "hit" the best entry point. Right?

That's not exactly the point, Pastukhov uses zigzags in his work not for technical analysis. Based on the breakdown by extremum he derives indicators of market fractality, i.e. the same thing as Hurst, at H = 2 the phenomenon of a non arbitrage market, at H>2 the market is trendy, at H < 2 the market is in a flat state. Besides, currency pairs are divided according to this feature, for example EURUSD is more trendy, than flat, so accordingly the strategy. This is an opportunity to earn on "hard tails". But unfortunately the costs in the form of commission and spread nullify this opportunity over a long period of time. Also zigzags are a way to move away from the usual averaging, which suffers from the effects of Slutsky-Yule.
 
Andrei:

If you take the interval above, of course they will be inside, but there will be their High and Low...

High and Low is just a measure of price in a given time interval. What are we talking about? I'm thinking deeply)).