Volumes, volatility and Hearst index - page 25
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
1. By rollback I mean any reversal in the process of shaping a zone segment that does not result in switching its direction. This is a restriction from above. However, I have also limited it from below, so as not to go through all the rubbish. In general, this distribution is approximate, in fact the statistics on kickbacks probably a bit wrong to collect. But is it really necessary? :)
3. no, i didn't. And can you give a reason why it should be built? I'll tell you why I should build it, but what trade ideas can I check when I look at it?
1. OK, I see, that's not what I thought. Explain the following then. During the formation of a segment, the price can roll back more than once. Which one do you take for statistics ? Or all of them ? On the chart the pullback is represented in %. In relation to what? You are working with a simple, normal tool with one parameter - the minimum value of the segment. Do you measure the percentage with respect to it or to the already formed part of the segment? Or another way ?
3. don't know. That's the ratio you're baffling me with. I didn't look in that direction somehow. Could be TP/SL valuation. Or maybe volatility valuation, instead of Hearst. After all, as we now know, that value tends towards 0.5. It seems to me that it is quite appropriate to assume that if it is consistently less than 0.5, then the market is a return market. And if it is more than 0.5, then it is a trend market. If we assume from the marginal values 0 and 1, then it turns out so.
The point is that there are two ways to find local measures. One is to transform the Hearst account methodology in such a way that it becomes sufficiently local. The second - the formation of a new value, an indicator, which has the necessary property of reflecting the nature of the market.
2. Assuming that the true value of any pullback is 23%, this is the only way we can find the level beyond which the pullback will not go, with 90% certainty. How serious this assumption is is up to you :)
1. OK, I see, that's not at all what I thought it would be. Then explain the following. During the formation of a segment, the price can roll back more than once. Which one do you take for statistics ? Or all of them ? On the chart the pullback is represented in %. In relation to what? You are working with a simple, normal tool with one parameter - the minimum value of the segment. Do you measure the percentage with respect to the bottom or to the already formed part of the segment? Or another way ?
3. don't know. That's the ratio that you puzzled me with. I didn't look in that direction somehow. Could be TP/SL valuation. Or maybe volatility valuation, instead of Hearst. After all, as we now know, that value tends towards 0.5. It seems to me that it is quite appropriate to assume that if it is consistently less than 0.5, then the market is a return market. And if it is more than 0.5, then it is a trend market. If we assume from the marginal values 0 and 1, then it turns out so.
The point is that there are two ways to find local measures. One is to transform Hearst's counting methodology in such a way that it becomes sufficiently local. The second - the formation of a new value, an indicator, which has the necessary property of reflecting the nature of the market.
1. taking everything. As we are interested in the pullback from the point of view of continuation of the movement, any of them, not only the last one, will do.
Percent - to the already formed part of the segment, of course, what's the sense to look into the future.
2. For TP/SL - do you want to work with bars? Or do you plan to work with dynamic ones? For me it always seems too rough.
For assessing market character? Well, the distribution is even less local measure than the Hurst index, imho. And what for, the average is enough.
But I really don't understand the problem with the construction, just write the values into a file, then import them into excel/matcad/matlab/... and see. I mean, I know exactly only for matlab, but it can't be that other packages don't have the function you need.
I don't think it is serious. Any pullback, at any moment can turn into a change of direction of ZZ. And then what ? Then it just doesn't fall into your stats. From that point of view it is hard to judge anything based on such statistics.
Of course in any one, just with different probabilities. That's why you need an allocation. The practical point I made above is to enter with the expectation that the move will continue.
1. Thank you, I see. As an option, I wasn't thinking about the future, but about the value of the ZZ parameter. With respect to the parameter it may also be interesting: it may draw a boundary at which a reversal with a given probability will occur. In other words, we may get an earlier identification of the reversal, with a given probability, of course. And the definition area of such a distribution would still be [0,1].
2. This was just a "bold oat sketch". :-) I'm telling you, I haven't even looked in that direction. So, ideas at a glance. And you don't need distribution to work, you just need to count the ratio. And the distribution is needed to understand what the RMS will be. If it's too big, there's no gain, and no confidence in these figures. If you remember, I calculated R, D and DD RMS in my opinion for this very purpose.
No problems with the construction. I've done this I don't remember how many times, but for other values. But I didn't do it. And you seem to have, that's why I asked, out of curiosity. :-)
Of course in any one, just with different probabilities. That's why you need an allocation. The practical point I made above is to enter with the expectation that the movement will continue.
Ahhhh, now I see what probability we're talking about. Probability of pullback = probability of continuation of movement. Yes, in this sense 0.6 is not comforting. But come on - it's Fiba ! Unless, of course, a reversal is formed. :-)
You are talking about literal similarity here, in fact patterns.
Realistically apparently we can talk about similarity not for local characteristics, but for their average values (for example see High-Low/|Open-Close| data in this thread on p. 14). However, my experience with statistics made me somewhat skeptical about the possibility to derive a trading system from them (statistics). The confidence intervals, you see, always turn out to be wrong and I am beginning to suspect a fundamental law.
I am talking about an important characteristic of fractal analysis (FA) - self-similarity. FA does not introduce the concept of "patterns", and a coefficient of 0.9, for example, does not say anything about any particular form of signal. FA essentially studies patterns that lead to self-similarity. Of course, a pattern can include elements of stochasticity and then you have to evaluate it statistically somehow.
But as I wrote before, quotes series do not practically have any similarity and the initial signal should be transformed in order to get to this similarity. I think, it will take some time.
PS: Patterns may also be useful, it is possible that it will be possible to identify the model, at least in the first approximation.
I am talking about an important characteristic of fractal analysis (FA) - self-similarity.
What a coincidence - that's what I'm talking about too :)
Ahhhh, now I see what probability we're talking about. Probability of a pullback = probability of continuation. Yeah, in that sense 0.6 is no consolation. But agree - it's Fiba! Unless, of course, a reversal is formed. :-)
Not necessarily. You're trending in on a 23% pullback. Where do you put SL? That's where you think about this allocation.
And when the pullback is 0.6, it's clearly 50/50.
What a coincidence - and I'm talking about the same thing :)
Yes, but I'm not talking about the patterns you're talking about :o)
You're talking about literal similarities here, in fact patterns.
Let's define what a pattern is? If one understands it as some kind of stable structure that also has a probability of occurrence, is it one thing or something else?
Resilience is a characteristic that needs to be proven. One of the first candidates for stable structures (i.e. patterns) are repetitive structures. The highly correlated parts of the rada you mentioned are the repetitive structures. At any rate, patterns will definitely have a high degree of correlation and thus fall under your definition of similarity.
FA, to my understanding, interprets self-similarity in a much broader way.