Rate of price change, how to calculate - page 11
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
:)
Oh, man... I don't have enough time
So: Werner Heisenberg Quantum Mechanics and Kant's Philosophy (1930-1932), chapter from Part and Whole. In this chapter... I won't, though. Let's just read it. It is not that long. And very (from my point of view) fascinating.
// There at the link all this book, I hope will be wishing to rummage in other chapters, and maybe read the whole.
// There's another book in there too: "Physics and Philosophy" - it really impressed the hell out of me. I have them both in paper form (in one collection). I poke around in it sometimes when I'm in the mood...
Then ... I hope to continue trying to measure/sort the measurable market facts ... in an effort to satisfy our own needs, of course... well, what else can we do? ... ;)
I have many books -- Heisenberg, Weil, Poincaré... (in paper form). Yes, these books are very rare nowadays.
For my part, I recommend reading the profound works of Henri Poincaré, "Science and Hypothesis", "The Value of Science", "Science and Method", and "Last Thoughts", collected in Henri Poincaré's book "On Science".
The last 4 pages have raised my spiritual level :-))
The conundrum of not being able to divide the journey price by the time in the journey worries me.
:)
that's a load of crap... I don't have enough time
The previous speaker's excellent speech will give us an extra powerful impetus to accelerate our search for new methods of calculating the rate of price change!!!
Right. All reactions are within expectations. :)
I won't make excuses anyway, don't even ask... :)) I'll just explain: sometime something like that should have been put on the forum all the same. Maybe it was necessary a long time ago, maybe it's high time now (well, it seemed to me so), or maybe it's too late... After all, "Nada Fedya, nada! You can argue until you're blue in the face or just laugh at the size and style... BUT. In terms of content, it's mostly correct. Well, one must eventually recognize that causality-determinism in this world is leaky in many ways and that at the limit this leakiness is not at all reduced to our subjective uninformedness. It's broader, deeper and more fundamental than that.
;)
I have many books -- Heisenberg, Weil, Poincaré... (in paper form). Yes nowadays these books are very rare.
For my part, I recommend reading Henri Poincaré's profound works "Science and Hypothesis", "The Value of Science", "Science and Method", "Last Thoughts", collected in Henri Poincaré's book "On Science".
The last 4 pages have raised my spiritual level :-))
The conundrum of the impossibility of dividing the journey price by the time in the journey worries me.
Who has it easy now? :) On the other hand it is possible (take away and divide, and that's it), but they say it's of little use.
// In 1917 they also tried "take away and divide" ........
:)Just in the topics on mutual information it was found that the lion's share of this dependence is the daily dependence of volatility, from which it seems no one has learned how to make much profit. Among such topics there was one in which I participated (I cannot find it in my search, I have completely forgotten, maybe it was something similar and not mutual information). There I kind of suggested a way to filter out this component and see if anything remains after that at all? But then no one did.
P.S. I said I couldn't find it and I found it :), there it is.Read that thread in the evening. (Just before you shat out that tribunal piece on the previous page that you giggled at.)
I think I got a little bit into the subject. The formulas are not fully understood, but the train of thought traced. In particular, yes, rationing on the daily volatility has not been done and the residual has not been measured.
If I was smart I'd get the hang of it, but I feel I missed something in primary school.)
- I agree with the percentages. Distance is measured in relative units.- Each minimum acceptable price change is a new member of the time series.
There's the hard part, the price. If you ask me, the concept of price depends a lot on your goals)
(a) Suppose you are a market maker, work on ECN (that's easier for example) and are very much interested in market orders
Take T&S for some fixed/non-fixed period (FIR/BIX filters) and calculate a geometric weighted average. Here's your speed = +- X% per trade of minimum volume
(b) Suppose you are an arbitrator and also work on the exchange, but are interested in limiters of slack market makers.
Here is more complicated. First, you are interested in bid prices (market history) and not deals. Secondly, you have at least two prices: for bid and for ask and you should even synchronize them.
When you calculate the price, you should take into account:
- - Quotes are relative
- - volumes are inversely proportional to execution probability
I gave some general thoughts about the algorithm here(c) Suppose you are a long term investor
Then you don't bother with all this nonsense and just trade normally, for example with a wristwatch (arithmetic average of relative values) and measure the speed in pips per candle.
P.S. Although arbitrage and market makers is one and the same, only with different signs, a common definition of price can not give / pick up.
It's OK. What we have is what we have. Let me summarise your "predicament" as best I can. At the same time, I'll summarize a little, so that not only the "price" can be frustrated in a productive way.
Thus:
1. The definition of notions may (may even have to) depend on the purposes (intentions, needs) of the defining person. That is, it depends on the observer.
2. In case of an intention to measure something, both the object of measurement (in terms of choice of object), and the measuring tool (choice of tool) depend on the aims of measurement (or more exactly, on the aims of the measuring subject).
I would like to add something else of my own accord:
3. On the other hand, after a choice of a measurable object, a method of measurement and a "ruler", the initiative kind of goes already to "the measurable object in the context of a measurement situation". Or more exactly to the interaction "object to be measured - measuring instrument", which in the output gives us long-awaited figures - "information about ...".
In other words: The measuring object, the measuring instrument and the measuring subject are inter-acting, interdependent. A subject defines (assigns) a measurand and a measuring instrument, and these in their turn generate in interaction with each other a number (a set of numbers), a quantitative characteristic of the measurand.
That's all for now. I mean about difficulties. :)
As for the speed of price change... what's the problem with it? The problem is with the targets. And with speed almost none. Speed = distance / time interval. Distance S = Price(t) - Price(t - Δt); If you consider it a relative distance, then divide it by, say, their geometric mean. Thus, we obtain Sotn = (Price(t) - Price(t - Δt)) / sqrt(Price(t) * Price(t - Δt)); of course we are tempted to take an arithmetic mean instead of geometric one, especially taking into account that they differ very little at short distances. Well, if we agree not to forget about this shamanistic substitution, we can do it. Then (conditionally) we will take that Sotn = 2 *(Price(t) - Price(t - Δt)) / (Price(t) + Price(t - Δt)). Ah yes, the unit of measure. Well, after such normalization the price has become relative, it is no longer measured in concrete gulden-trunks, because they have decreased in the process. It now remains to divide this dimensionless shit by the time interval to get the desired speed. And let's define (for greater specificity) the units of measurement. Let's say percent per second. Then: Votn = Sotn / Δt = 2 * 100% * (Price(t) - Price(t - Δt)) / ((Price(t) + Price(t - Δt)) * Δt) per cent/sec.
Well somehow so, in a general way. Which "price" or what period of time and in what units to take - let it depends on concrete purposes of measurement. It used to depend on it too... if anything. :)
That would have been finished here but for one subtlety: the entire profit of this construction is destroyed by a simple observation that being measured by the above method, this very velocity behaves more or less decently only "at a long distance", and at a short distance it runs wild, and the more precisely (instantaneously?) we are trying to measure it, the faster it fucking flailing about, acting more like a random value trying to confuse us than an intelligent analogue of an inertial physical process, ...... Or to put it "cleverly" - undifferentiated in time bullshit, this very "price", oh undifferentiated... at point [0]. And to measure velocity on a segment is always welcome - see formulas above.
I'll also add, perhaps, the following: do not get upset or try to upset me by pointing out (fairly) that I have not said anything new here. I did not mean to say "new". I just wanted to arrange the "old" in a way that made it perhaps more understandable. That's how it happens sometimes, when you look at the same thing in several ways. Anyway, as one pleasant acquaintance on a neighbouring forum said: "I'm sorry if I've crushed someone's dreams )".
Delightful.
Soaring.
I wish I could add fractals, patterns, levels, channels and multitudes.
But judging by the amount of work involved, I can't wait.
Pluck.