Market etiquette or good manners in a minefield - page 76
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
HideYourRichess, you are confusing two values - H-volatility and the H split step and thus confusing a person! I entered two different designations for a reason.
H-volatility is the average of H segments on a series. This is why there are two strategies. When segments average more than 2H and when they average less. It's simple as that. I have a suspicion that by H-volatility you mean something else than what Pastukhov describes.
My understanding is that H-volatility is the average (or RMS) of the lengths of all distances from local extrema, to the points of the cagi-partition made with H step.
No. Wrong.
I introduced dimensionless volatility (unlike Pastukhov): Hvol, is a dimensionless value, defined as the average length (projection onto the ordinate axis) between Kagi-building counts (red circles) referring to the step H. For a trend market (market parameter H-volatility>2) and H- for a counter-trend market(Hvol<2). Pastukhov himself introduced a dimensional value in his work, and as HideYourRichess correctly says above: H-volatility is the average of H segments on a series.
Both definitions are essentially the same. It's just that I'm used to using dimensionless values in my calculations.
That is, Hvol is a characteristic of the arbitrariness of the market at a step of division H (in points), which you personally selected for some reason. Pastukhov in his dissertation does not pay attention to the choice of this parameter. He only states the fact and doesn't consider the market as a function of this parameter. It seems to me that this is a key point and it is not by chance that it is quietly "neglected". Indeed, the TS is based on the principle of an iron strategy reversal and proves that this behaviour in the market is statistically significant in terms of profitability and there is no word about how to select the "best" Н and whether it is worth looking for it again, having chosen it once...
Thank you, things are clearing up a little bit. And here's my "take two". Take a look, please:
It looks more like a kagi to me now. In any case, the algorithm you mentioned has been fully followed. Just in case, I'm attaching the listing in 11 format.
I introduced dimensionless volatility (unlike Pastukhov): Hvol, it is a dimensionless value defined as the average length (projection on the ordinate axis) between Kagi-building counts (red circles) related to the segmentation step H. For a trend market (market parameter H-volatility>2) and H- for a counter-trend market (Hvol<2). Pastukhov himself introduced a dimensional value in his work, and as HideYourRichess correctly says above: H-volatility is the average of H segments on a series.
Well, why so. Pastukhov uses 2H - this is the dimensionless value, because it doesn't matter what size H is, the main thing is that there should be 2.
Thank you, things are clearing up a little bit. And here's my "take two". Take a look, please:
It looks more like a kagi to me now. In any case, the algorithm you've voiced is fully respected. Just in case, I'm attaching the listing in 11 format.
It's a perfect match:
The red line is mine, the squares are yours. Congratulations!
paralocus, I didn't get the parameter s=3*10^-3 in your code... Anyway, I put it equal to 1, then your parameter m became equal to the step of division H in points.
It doesn't look like a kagi.
Perhaps you don't have a good understanding of what Kagi-building looks like.
Why would you do that. Pastukhov uses 2H - it is a dimensionless value, because it doesn't matter what size H is, the main thing is that there should be 2.
This is something new... unusual!
Thus, H is the dimension of the vertical axis of the price chart, i.e., points, and corresponds to the value of points that the price should retreat from the extremum to fix the Kagi oscillation. Hence, the Pastuhov volatility 2H is [pips], in other words, it is an average value of Kagi Leverage and it is measured in pips.
Correct me if I'm wrong. I think, HideYourRichess, you are getting carried away.
I think I've gone over it!
The value of 2H characterizes the average size between extrema of the original series (kotir), not the entry/exit points shown in the figure above.
To be honest, I no longer remember what Pastukhov calls a Kagi formation and what is called an auxiliary series - a series consisting of kotir extrema or entry/exit points. In any case, HideYourRichess, if your "Kagi is not like that" has as its nature this fact, I apologize for my harsh statement to you.
At the end of the day, to form trade orders, we are only interested in entry/exit points. That's why I meant by Kagi-building this particular BP.
"Let's turn to the classics" (c) Here's an excerpt from my dissertation.
On page 15 it states that kagi H-building is black and white circles. You make up your own categories and attribute them to Pastukhov. You can't do that. And the 2H theorems refer specifically to black and white circles. Don't call kagi something else.
With such unexpected insights on your part, I'm afraid I'm never going to wait for the magic kagi algorithm, which is no more complicated than renko.
And, 2H has no dimension because it's just a 2 or so. There are no points there because it is impossible to compare H-waves in points, which means there is always a dimensionless conversion. However, this is a matter of taste.
"Let's turn to the classics" (c) Here's an excerpt from the dissertation.
...
Here, I don't remember exactly, but if I'm not mistaken, the statistical advantage (specifically for "forex rows"), for kagi turns out to be almost negligible. It can only be recouped by a very, very long presence in the market, and given a non-uniform MM, it's virtually impossible. Or am I wrong?
paralocus, in your code I didn't understand what parameter s=3*10^-3... Anyway, I set it to 1, and then your parameter m corresponds to the step of division H in points.
One night I had an epiphany: suddenly I understood that brokerage companies don't trade currency! They trade spreads for instruments! Hence I got a very simple conclusion that for the kotier distribution function (on minutes) the projection of maximum onto the X-axis will be equal to the spread, which proved to be true:
And since this is the case, it means that the most correct step of quotient partitioning should be a multiple of the spread. The spread on the pound at my DC is 3 pips, i.e. it is 3*10^-4. That's where the s parameter comes from.