Re: Chaos theory made simple - for trading

41
That is strange Pava because Mandelbrot NEVER addressed trading and traders directly. You can look and find that it was NEVER assumed that the trader knew what he/she was doing, it was assumed that the trader did not matter and if at all, not as much as the quant, the risk and asset managers in the scheme of things.

There is NOT a single place where traders are addressed directly UNTIL NOW. Of course it could be that it was assumed that traders are not that clever infatuated as they have always been by who you call pagans and things pagan (things completely baseless) - so when math found the market no one ever gave it to them (traders) to understand the intricacies of a thing so mathematical, complex and emotion-laden (men are known to kill for money and women too). That is why I say it is strange.

Because if you have read the literature on chaos and fractal geometry and noticed as a trader, that no one (Not a single writer until now) ever mentioned trading or indeed traders (except in almost condescending passing), then you will understand why you have had indigestion reading chaos and fractal geometry before now and (unless you are not curious enough) you will also find out why you would not have same indigestion now. Now the real problem with Mandelbrot and his shared knowledge is that his target audience was quants and asset managers for whom he thought the issues of a market model was more relevant and that they had the brains and imagination (not the pagan trader) to establish this new frontier of understanding – everyone forgot that the trader who actually takes the risk in market time could be the solution – until now. The quants and asset managers simply have not been able to handle it. For example, if you are someone with the presence of mind to understand quantitative strategies – the so-called algorithms – that cause your average pagan trader to wet his/her pants, you will agree with me that, here we have very clear evidence of the catastrophe of not listening to Mandelbrot and his notion of Mandelbrotian randomness as Nassim Taleb calls it.

Algorithmic trading has now reached a point where in fact it is clear that it suffers much the same problems as do humans when the wrong math is used to read the market  (because after all Algorithmic trading is mathematical). Algorithmic trading is not more profitable than man generally, since it only relies on the brute force of machine behavior in the ability to do things on a massive scale endlessly, but not an understanding more discerning than the human trader.  So both man and machine have problems dealing with the fractality of markets. For instance, in the market maker (algorithmic) strategy, where you basically straddle a range for the spread, the extension of the learning model of Glosten-Milgrom by Das fails to resolve the issue of adverse selection.

So for all the hype algorithmic and HF trading have reached their limit and the only way out is to listen to Mandelbrot and of that I am certain they will do now because unlike the pagan trader these are open minds seeking an edge and not sulking like humans about the many times they have tried and failed. Usually, of course, these machine strategies borrow from humans and try to be more efficient (machine efficient, i.e.). But this time we are at the same starting point where in fact man has always beaten machine. But it will depend on the non-pagans amongst us using the digestible new realities of math and market discovered by traders – NOT academicians or quantitative analysts or asset managers. So it is amazing to me that the human trader (except a predictable pagan) will ignore the chance to do what he says he has always wanted to do - trade the market profitably.


Re: Chaos theory made simple - for trading

42
Krelian99 wrote:You are there where many were before. They wanted to slay the markets with pure math and failed. They did everything to make Turtles and HFT failing in the end. In homogeneous systems you can slay it with math only, no question. All the very nice approaches when you have no direct physical connection of input and output. But ask yourself, is the market homogeneous? Yes, the market repeats patterns all the time and that at certain places.
There is no conspiracy anywhere and there has never been. Man has always sought to overcome by the knowledge of things. There is nothing in present life that has not evolved in similar steps - modern medicine, flying (combat and commercial), politics, etc. The contest is always between stark ignorance and knowledge. Knowledge is not always intellectually accessible to everyone and innovative applications of knowledge follow a normal diffusion curve so I would argue that you want to know a lot more than you do as of now to reach any valid conclusions. Let me also give you a small shock as a warning about reaching conclusions too quickly. THE MULTIFRACTAL TRADING MODEL - has NEVER been traded by anyone alive or dead. Including me. Why? First we just found and confirmed it (we started talking only in the last 6 weeks) and second, it needs highly specialized and expensive machinery in order to trade it. The first of such has just been designed and will be setup for manufacture, delivery and commission in the months ahead. So I do not know where you have seen the Multifractal model traded as to freely speak of it as if it is something that exists. Now if you do not know whether or not a multifractal model exists as of now, how could you say so much about it? The only reason is that you are discussing something completely different from what I am talking about. But be that as it may, i.e. you are not on same page as me - certain things are fixed and cannot be varied by belief. The mathematics of iterates is firm and clear in the aperiodic instances that define markets - you cannot have "price patterns" period. This comes from the behavior of the numbers that form the shapes you see. A secondary school kid will vouch for that. My point is lets not leave science (where it is established and can be easily referenced) and insist on what pava describes as pagan. It is empirical so lets grapple with the meaning of the implication that price patterns are an illusion of the mind, the argument is over. The other thing that does not appear clear to you (in terms of the topic) is the term homogeneous. If you took time to understand the math even on a cursory level - then you will know why it is called Multifractal and not Monofractal analysis. The mathematics captures to the dot all the nuances of the market place in real-time. That is why it is a bomb. But in fact not everyone (as we conceive of it now - though that could change in future) can afford the equipment, gain the learning and the practice to be the super trader it will make an average trader, and therefore few will be able to join in at the inception of this market turn in trading (however, knowing about it is important to be ready to follow). Now why am I so sure I will prevail? Because I am one of the first to understand this power well enough to invest in the equipment (having spent years learning the science). In fact I designed it. Now once that is going for us we will make returns - the returns will separate truth from falsehood and imitation will guarantee the rest. So it must NEVER be an argument but a possibility until proven or otherwise. And if proven? The Law. Very simple.

Re: Chaos theory made simple - for trading

43
Strange? So many mathematicians working for banks and institutions and noone had the idea to apply chaos theory to the market? You are the very first who comes around the corner with that? Do you really think that. Many groups did before, that's for sure, but you find not much (or they call it totally different so it is intended that you can't find anything - it wouldn't be the first time). Either they failed or they got so rich and never shared a word. But I wonder why London banks which trade FX recruit engineers right away from universities to trade the markets. And why they are high-profitable without Mandelbrot? How do they read the market? Do you think the banks let the engineers loose the money and they don't want to maximize their wins? 'Trading is not easy, but simple.' said your friend Bill Williams. But you don't need chaos theory or Mandelbrot to read the market. My scalping style is mainly influenced by Fuzzy Logic, not easy but simple, and super predicting
I am a pagan and pagan trader through and through. I love science and religions, but I'm always bring it into question and take the useful stuff for me and I don't feel like a Polynesian either. Be open for other approaches. :)

Re: Chaos theory made simple - for trading

44
Krelian99 wrote:Strange? So many mathematicians working for banks and institutions and noone had the idea to apply chaos theory to the market? You are the very first who comes around the corner with that? Do you really think that. Many groups did before, that's for sure, but you find not much (or they call it totally different so it is intended that you can't find anything - it wouldn't be the first time). Either they failed or they got so rich and never shared a word. But I wonder why London banks which trade FX recruit engineers right away from universities to trade the markets. And why they are high-profitable without Mandelbrot? How do they read the market? Do you think the banks let the engineers loose the money and they don't want to maximize their wins? 'Trading is not easy, but simple.' said your friend Bill Williams. But you don't need chaos theory or Mandelbrot to read the market. My scalping style is mainly influenced by Fuzzy Logic, not easy but simple, and super predicting


I am a pagan and pagan trader through and through. I love science and religions, but I'm always bring it into question and take the useful stuff for me and I don't feel like a Polynesian either. Be open for other approaches.
Like I said my friend it is NEVER an argument it is a possibility until proven wrong (or faulted mathematically) and after that law. You feel that it is beyond me to know? I am not human enough? Or maybe you imagine I have not been to the same universities with multiple degrees to my name (letters we call them)? Or it is not clear to you what a quant is and how they work? Why are you so unsure of your own abilities and knowledge? Do you think quants are made of matter superior to what you and I are made of? Of course you cannot and do not dare to mention a single error in mathematical logic in all of what I am bold enough to state to the world but even with that serious lapse you feel competent enough to declare your faith in nothing in particular? And oppose everything I am saying without even the most basic justification? For instance what makes anything I am saying or have said disputable? Is there any one single thing in the math and logic that you can fault? So if you cannot fault a thing - where does your impetus come from - objectivity or emotion? What kind of emotion? Fear, anger, envy, insecurity, what? It is so easy to see what you struggle with something - yet this is beyond me and you and certainly we wait for a future in which our determined actions (you and I) have a definite roll to play. Relax it cannot be changed now it must be confronted for truth (but not by feeble questions such as you ask of the presentation). That is why we tell kids not to lie - because invariably you must live that lie. It will soon come to pass.



BTW: You really love Bill Williams do you not? I have never read him, do not know what he says and not sure it will matter - simply from his implementation of chaos in indicators. His implementations tell me to ignore his ideas as not fully formed to lead (plus I wanted to hear from chaologists directly). That is why we have had to put effort at all into understanding what Mandelbrot was on about (if your good friend Bill had resolved it we would have just used his ideas). But it is not just you having problems with the solution. Some quants are raving mad too and like you they cannot fault the math; so all we get is feeble protests but nothing more. But some other quants are engaged and asking rational questions. I was hoping with human traders we could even share more than one would with the competition. But may be I am wrong.

Re: Chaos theory made simple - for trading

45
No, Bill Williams is a trader, his view on the market isn't wrong, but far from complete. So he tells no BS, but actually some interesting stuff. He isn't a mathematician and so he uses some indicators. His A/D line is ok, the rest is so-so. He misuses the term Chaos Theory, that's why you don't like him or make fun of him. So, it's a running gag. But he tried something and misunderstood fractals with local pivot points. This happens in the best families.

It isn't my task to fault math. I just say money makers test all the time whether there are still unwanted traders in the market and they do fake moves all the time to get rid off them or suck other traders into the market. In a homogeneous system with physical boundaries it is makeable, you have a matrix with e.g. 20 orders (when the physical laws between input and output are missed) and calculate your estimation error and prediction error. Very nice, as long as the system doesn't change with time. FX is time-variant and inhomogeneous without clear boundaries. So you need a frequency- and time-variant approach and yes, here I'm out, at least for the moment. And as I said I'm skeptical due to the logical reasons I mentioned in my post above. And yes, I know how quants works ;-)


Re: Chaos theory made simple - for trading

46
BTW: They are upset about similar things as you - stuff like patterns. But I assure you the patterns fractal geometry speaks about are statistical patterns random series, anti-persistent series, and persistent series. These mathematical patterns exist as mathematical indicators - they allow us judgment. That is why we are saying that to get that edge everyone seeks you must understand shapes for the behaviors of the numbers that form them. These patterns (the mathematical ones) are not expressed as price patterns in the common understanding of the word - especially by technical analysis. These patterns of fractal geometry are visible to the eye but making sense of them for reading the market requires a special understanding of fractal behavior, and the specific implications of the fractal patterns seen. Also the patterns mix up in any given flow (and hence the chaology plus a special console to read the matrix). Of course my only question would be know what your objection would be to knowing the details of this question? Also algorithms are in decline in terms of their rate of employment (growth). That has to do with them reaching a threshold which I say is easily overcome if the writers listen to Mandelbrot

.

Re: Chaos theory made simple - for trading

47
Krelian99 wrote:Strange? So many mathematicians working for banks and institutions and noone had the idea to apply chaos theory to the market? You are the very first who comes around the corner with that? Do you really think that. Many groups did before, that's for sure, but you find not much (or they call it totally different so it is intended that you can't find anything - it wouldn't be the first time). Either they failed or they got so rich and never shared a word. But I wonder why London banks which trade FX recruit engineers right away from universities to trade the markets. And why they are high-profitable without Mandelbrot? How do they read the market? Do you think the banks let the engineers loose the money and they don't want to maximize their wins? 'Trading is not easy, but simple.' said your friend Bill Williams. But you don't need chaos theory or Mandelbrot to read the market. My scalping style is mainly influenced by Fuzzy Logic, not easy but simple, and super predicting

I am a pagan and pagan trader through and through. I love science and religions, but I'm always bring it into question and take the useful stuff for me and I don't feel like a Polynesian either. Be open for other approaches.
Banks operate a complex of trading assets and many such assets have nothing directly to do with spot markets. We know that outfits like proprietary trading (in their basket of functions) are the least celebrated for performance. But then they might also earn income from their so-called trading operations through a broker type service, and have retail and institutional traders as clients, etc. So their scope and interests are too wide to allow for the not too specific issues you raise about performance. What can be said at most (in terms of the profitability of their prop operations) is that prop trading or other near effects in their operations is not reported in a transparent manner (and therefore not enough to make some of the assumptions you make on face value).

Re: Chaos theory made simple - for trading

48
Hi,

I have been reading the discussion. I happened to have a paper by Mandelbrot "A Multifractal Walk down wall street" (attached) which could help to illuminate. Mandelbrot addresses the idea of simulating prices using multifractals. In the paper there is diagram of charts, page 4, "pick a fake", where we are asked to guess which of the charts 4 to 8 is from a real market data, and which is simulated by multifractals.
Have fun!

Re: Chaos theory made simple - for trading

49
Teduh wrote:Hi,

I have been reading the discussion. I happened to have a paper by Mandelbrot "A Multifractal Walk down wall street" (attached) which could help to illuminate. Mandelbrot addresses the idea of simulating prices using multifractals. In the paper there is diagram of charts, page 4, "pick a fake", where we are asked to guess which of the charts 4 to 8 is from a real market data, and which is simulated by multifractals.
Have fun!
Hi bro that is really nice of you to share. I will take a look myself and hope others will. Mandelbrot's book The (Mis) Behavior of markets actually contains the same thing. Notice that in the paper and in the book Benoit Mandelbrot showed how the Iterated Function Systems (IFS) formalism of fractal geometry, may be applied to the market and what a fractal generator is in terms of a model that simulates the market. Our context is different in that we do not speak to a model that simulates or projects the market but one that allows the reading and trading of markets in real-time using actual price charts. As such, the need was to find a fractal footprint in situ that defines the overall structure of markets in ways anyone can reasonably conclude to be deterministic of market price. That is what I did and felt it would be really nice to share with fellow traders and show them how this changes everything for us as traders. Mandelbrot has the correct understanding of the market as well as the mathematics to allow us model and trade it. But until now his paper and book have been lying fallow (useless even to the quants and asset managers he was addressing) because in fact no one in the entire world (funny really) could relate (in a simple and direct way) his thinking in simulating the market to actually trading the market using the charts as they are - because of course the charts do not show fractal behavior directly as price bars hide too much information (fractalities). We are now showcasing that ability and asking traders to get interested enough to understand the implications (at least).
These users thanked the author Darkdoji for the post:
global

Re: Chaos theory made simple - for trading

50
Thank you Darkdoji for your reply.

Although I like the insights from  fractals and chaos as a mathematician, I haven't yet seen anything concrete for practical trading.

I respect Williams, and his "fractals" have some usefulness, but they have nothing to do fractals in Mathematics.

IFS are good for generating fractal objects (e g ferns, etc) but I doubt if the market is deterministic. By the way, the paper I mentioned before does not mention IFS.

Prof Didier Sornette is an expert on multifractals, I believe he is working on using multifractals to detect forthcoming market bubbles. This could be very useful.

Peters with his Fractal Market hypothesis (FMH) sees the market more like a stochastic process (not deterministic), but where the probability distributions are not normal. In FMH outliers are "normal".

These are my views.

Best regards
These users thanked the author Teduh for the post:
Black Sheep


Who is online

Users browsing this forum: No registered users and 17 guests