The markets are Chaos...and Order

[Gringotts Bank]

I take no offense- did you see how I avoided the question? Let me ask you one. Given that I have said that the larger the data set the more accurate youcan become, do you think equities has the required characteristics I am looking for?

Given that there were only eight trades a week ( and there would be plenty more)That level of return just tells me that it's relevant.

Cheers

Sir o

By the way you've worded the question, I'm assuming the answer is 'no'. I don't understand chaos theory, and don't want to make any effort to do so, so I'm just asking questions from the end-user point of view (if you end up selling your method, that is). So let's say it's a forex or futures system that's highly scalable and can return 40%pa. Now what? Apply for a huge margin loan and trade it for huge profits? I'm happy for you, but how do I get some of the action?!! And what about that equity curve!??

 

[Sir Osisofliver]

By the way you've worded the question, I'm assuming the answer is 'no'. I don't understand chaos theory, and don't want to make any effort to do so, so I'm just asking questions from the end-user point of view (if you end up selling your method, that is). So let's say it's a forex or futures system that's highly scalable and can return 40%pa. Now what? Apply for a huge margin loan and trade it for huge profits? I'm happy for you, but how do I get some of the action?!! And what about that equity curve!??

Gringotts,

Cool your jets buddy. I appreciate that you're approaching this from and end-user point of view and being a sceptic. It gives me insight into the kind of things I may need to provide if I were to sell this as a product of some kind. However, IF (and it's a big if) I share the methodology it will be some time away before I can produce a polished product. What you're talking about as an end user would require that I hire some talent that can turn what is currently a whole bunch of numbers that makes my computer groan with effort into a visually orientated indicator.

There is also the issue of training. With the variables involved - it is nowhere near as simple as changing the sample period, altering the std deviation etc etc - that a normal indicator has. Then there is the issue of phase shifts occuring (where the length of the component waveform changes - most commonly at the end of a cycle). If I preset an indicator, it would not remain valid forever. The people who do not understand what is happening would thow their hands up in digust and say it doesn't work or that it is giving false signals.

Anyway I had a think and I'm posting a couple of equity curves - but I've taken out the X&Y axis.

Cheers

Sir O

 

[Sir Osisofliver]

@ Motorway

I really want to engage you in some discussion in the thread. You've raised some excellent material that is worthy of discussion. I need some time to do it properly however but wanted to thank-you for your contributions.

@ Everyone else

Now that you have the broad brush stroke version - should I leave it there or try and continue to build when I have time?

Cheers

Sir O

 

[Gringotts Bank]

Oh come onnnnn! That chart means nothing, and you know it.

What is this thread but a tease? What's the point?

You have people interested. Fine. Now what? Where's this going?

If your system is 5 years off going commercial, tell us and I'll stop posting.

If it's about thrashing out ideas about chaos theory, get on with it. 95% of people reading won't understand it.

 

[Sir Osisofliver]

Oh come onnnnn! That chart means nothing, and you know it.

What is this thread but a tease? What's the point?

You have people interested. Fine. Now what? Where's this going?

If your system is 5 years off going commercial, tell us and I'll stop posting.

If it's about thrashing out ideas about chaos theory, get on with it. 95% of people reading won't understand it.

You didn't get any meaning from that? Not even that you can visually see that the chaos model equity curve has a degree of unpredictability taken out of it? That's why I posted it after all.

Gringotts - as I said before for me to produce this as a course/book/ system package with special indicators (or whatever) will take a considerable period of time. I would need to devote myself to this in full-time development (If I choose to do it at all). The purpose of this thread was to find out (where people with a range of experience in trading) have difficulty with the material. Not simply give away all my hard work on this which I would consider to be my Intellectual Property.

actually I wanted to use the forum members for a sounding board to see where I need to tighten up the text or explain things better before I use it somewhere else.

I get it - you don't want to understand it - you just want to verify that it works and then use it to improve your own trading.

I'm sorry if I wasn't clear about my aim here - but I have to ask myself that if I write this up and the majority of people are in the same boat as you - what's the point of publishing it? I'd be better off in making a system of some kind and selling that and keeping the IP to myself.

@ Everyone - Given you've only seen the broad stuff - Is it simply far too complex to try and sell and sell a "this is how it works - now go apply it to your own trading"?

Cheers

Sir O

 

[Gringotts Bank]

You're right, there is no point in publishing it for the mass trader/investor, because they either won't understand it, or even if they have the capacity to understand it, are not interested in investing the required time.

You could publish it in a journal of finance, but what's the point in being an academic? No point at all.

Ward systems (Neuroshell, Chaos Hunter etc) have the right approach. Their software developers have taken extremely complex tasks and made them so that any doofus can use them. Their support and background literature is easy to understand and apply. You'd pay a fortune over many many, years to get someone to do that for you properly, the way they have. For those that have the maths background, Ward systems products will allow virtually unlimited scope. The internet is littered with trading software that is just total bullsh1t.... completely unusable. You need to be very skillful at communicating the written word, and I don't think you have that skill (again, no offense). I know you understand it, but imparting that to others is a whole new ball game.

So the way I see it you have two options -
1. Sell the whole thing to a hedge fund, offering to go on staff and help the IT guys develop it into something usable.
2. Trade it yourself.

 

[peter2]

Sir O: The thread started out with the potential to be both challenging and thought provoking. I knew that I would have to read it carefully to grasp some of the broad concepts in a field that I know little. I am disappointed that you decided not to take us on a journey of discovery where new ideas and other ways of interpreting the market data would surprise me as we journeyed forth.

I would describe your trading as just another pattern trading strategy. You have identified a pattern that repeats enough to profit by its existence. Very simplistic I know but the thread didn't explore the concepts and their application fully for me to appreciate the extent of your IP.

I would ask you to consider resuming the journey rather than be distracted by the pot of gold at the end of the rainbow.

 

[Gringotts Bank]

I've just realized that's the answer, O.

Buy Chaos Hunter software, $1500. Input your fundamental and technical data there. Will give you everything you need, and more.

For this little bit of priceless advice, I'd like you to PM me the inputs. I've just saved you tens of thouasands of dollars and years of system development.

 

[Sir Osisofliver]

Sir O: The thread started out with the potential to be both challenging and thought provoking. I knew that I would have to read it carefully to grasp some of the broad concepts in a field that I know little. I am disappointed that you decided not to take us on a journey of discovery where new ideas and other ways of interpreting the market data would surprise me as we journeyed forth.

I would describe your trading as just another pattern trading strategy. You have identified a pattern that repeats enough to profit by its existence. Very simplistic I know but the thread didn't explore the concepts and their application fully for me to appreciate the extent of your IP.

I would ask you to consider resuming the journey rather than be distracted by the pot of gold at the end of the rainbow.

So that's one person who wants me to continue...and I still want to talk to Motorway...anyone else want me to keep plugging away?

Cheers

Sir O

 

[motorway]

I've just realized that's the answer, O.

Buy Chaos Hunter software, $1500. Input your fundamental and technical data there. Will give you everything you need, and more.

For this little bit of priceless advice, I'd like you to PM me the inputs. I've just saved you tens of thouasands of dollars and years of system development.

Chaos Hunter might be OK.. But the secret is to be found in asking the right questions , and in the correct syntax .

A spread sheet could do this...

The secret is in to do with what is the nature of X that is correlated to the Ys

Ys are all the things you can think of
( The market properties you are interested in )

The italics are because it isn't a real secret

Motorway


This====> Does not fill me with confidence

ChaosHunter requires that you enter text files of historical or example data from spreadsheets or data feeds. Then you choose arithmetic and other math functions that you want ChaosHunter to use, and it produces numeric formulas that you can read, understand, utilize or even sell outside of ChaosHunter. ChaosHunter functions available include neural network and chaotic functions.

Buy/sell signals generated by ChaosHunter formulas can be transferred to many popular trading platforms, which enables you to trade your models with a number of brokerages. We have made ChaosHunter model transfer easier with ready-made interfaces to the following platforms:

 

[RandR]

So that's one person who wants me to continue...and I still want to talk to Motorway...anyone else want me to keep plugging away?

Cheers

Sir O

Ive been enjoying it ! Ive been enjoying the in depth reasoning and discussion of the theory ....

But I think you need to work on coming across 'in laymans terms' a touch more.

Am I on the right path with the following ? ...

1. Your identifying macro trends and wave patterns in the economic cycle
2. Your looking for indications/confirmation of the current point in the cycle

3. Your identifying the results of these macro patterns at previous points on specific equities etc.
4. Your looking for indications that the equities etc. will conform to previously identified patterns in the economic cycle.

Am I way off and need to re read the thread or fairly close ?

 

[Gringotts Bank]

Chaos Hunter might be OK.. But the secret is to be found in asking the right questions , and in the correct syntax .

A spread sheet could do this...

The secret is in to do with what is the nature of X that is correlated to the Ys

Ys are all the things you can think of
( The market properties you are interested in )

The italics are because it isn't a real secret

Motorway


This====> Does not fill me with confidence

According to Ward, the secret, (if there is one), is in choosing ther right inputs. CH works out the syntax for you. That's it's job. The manual explains how it works. It's all there. O isn't going to share anything of value. It's called grandstanding. Big freakin deal!

 

[MacP]

So that's one person who wants me to continue...and I still want to talk to Motorway...anyone else want me to keep plugging away?

Cheers

Sir O

Yes, please do.

 

[KurwaJegoMac]

According to Ward, the secret, (if there is one), is in choosing ther right inputs. CH works out the syntax for you. That's it's job. The manual explains how it works. It's all there. O isn't going to share anything of value. It's called grandstanding. Big freakin deal!

Sir Osisofliver quite clearly stated, in the very first paragraph of his first post, that the purpose of this thread was to serve as a sounding board, to bounce ideas and for comments on how to tighten up his text. So dont talk rubbish that his intentions have not been made clear.

You asked (selfishly and rudely) for the end user specifications and Sir Osisofliver denied your request. Then you proceed to whinge like a spoiled child because you didnt get your way. If you have no desire to adhere to the original premise of the thread then just go away and stop with the childish antics.

Sir Osisofliver, please continue with your original few posts. I agree with your original statement that it's worth understanding the foundations first. Coming from a Chemical Engineering background I've Looked at chatoic systems in fluid, reactions, etc. your original posts align with what i know, but i never thought to apply it to investing due to what i perceived to be too many variables present. Without repeatable observations i dont see how it can be applied to markets - i dont have the investing experience yet but keen to read more about your epiphany.

 

[Sir Osisofliver]

All right so I've had a re-think. Instead of leading out and trying to prove that the market is chaotic (which I've said is a- not a new concept and b- probably where we might get bogged down) I've decided to shift the order around. I am simply going to post some references for you to read or not read - your choice and then I want to demonstrate the principles of how we can take advantage of the Macro expression of chaotic principles on the economic cycle, and then discuss the implications and the things that are important to assist understanding.

Below are some links (not an exhaustive list) of useful infomration. - This is academia guys and girls - some of this stuff is major headache inducing.

1) Probabilistic causation http://en.wikipedia.org/wiki/Probabilistic_causation - I know it's just wiki - but this will give you a starting point. It's only us nerds that go read the cited reference materials.

2) Analysing Chaos Models - http://www.newton.ac.uk/programmes/HOP/seminars/070310001.pdf - This one is bitchingly math heavy.

3) http://necsi.edu/projects/baranger/cce.pdf

4) http://iic.wiki.fgv.br/file/view/CH...+APPLICATION,+AND+MANAGERIALIMPLICATIONS.pdf

5) http://sprott.physics.wisc.edu/chaostsa/

6) http://gwydir.demon.co.uk/jo/numbers/interest/golden.htm - Irrational number sequences

Below I've reposted some info from my newbie thread - I'm going to expand on it a bit - but not today.

Economic Cycles and how to spot them


Our market is cyclical. CYCLICAL. As is every market. Do not delude yourself that somehow this current market is different. Do not listen to the crowing of a broker who spouts rubbish about supercycles and how the emergence of China will create a cycle that will last for twenty years. (I heard this at the tail end of the last cycle) Do not listen to those that say "We are different". We are the same as everyone else because the things that drive us are the same. Therefore the market goes through periods of boom and bust. NEVER THINK OTHERWISE.

Historically our market is rising roughly 80% of the time and during this time we are all gods. It's easy to make money when the market is rising. It is however the 20% of time that tends to kill people. This is where the cowboys and those that have let the excitement and emotion of the bull run lead them into disaster as we approach the inevitable correction. Every cycle there will be someone who thinks that the conditions that we see now are going to stay around, (or that the correction won't be bad). This is where the Storm financials, the Opus Primes, the Tricoms, lose money and destroy wealth for their clients. They do NOT understand the principle of the economic cycle and how to use this to their advantage.

So the biggest question is How do you tell when everything is about to go pear shaped? What are the signals? What do I look for to know when it's time to liquidate or protect my share market assets?

Economic Clock
View attachment 42839View attachment 42840

If you do a search for Economic Clock you'll get images like those I have attached above (hopefully). The Economic clock dates back to the 1930's where it was proposed as a forecasting methodology by English Economists and released in the London Evening Standard as a series of articles. The basic theory is that certain events which are:

a) Easily recognizable; and
b) highly correlated;

will allow you to ascertain at which point of the broader economic cycle we currently sit. Theoretically if you know where we are in the cycle, you know which asset class in a broad sense is the correct one to invest in for an optimal investment. Each event is related to the previous event. Each previous event causes the following event to occur. It's Cause and Effect, but in a cycle of events that roll around and around in a never ending cycle.

You'll note some differences in the two diagrams I attached, even though the basic principals of one event following another event in a series remains intact. This is where the theory gets subjective and it's subjective because we are dealing with a mechanism that is attempting to take a complex aggregation of factors and deliver them to you in a simple easy to understand diagram. The clock does not tell good time. The clock is an imperfect and generic indicator - but is still a valuable tool, particularly for analysing the market over the longer term.

The clock is a simple representation of a complex aggregation and there are differences between the interpretation of the theory because each cycle is an aggregate of several different cycles, with each of these individual cycles having their own characteristics in terms of length, midpoint and amplitude (depth or height). Have a look at the wave below to see what I mean.

View attachment 42842

Imagine that the wave represents that of the interest rate cycle. This wave will be highly correlated yet different to the wave representing the currency cycle. It's only when these two waves move in combination with each other that the effects will be felt. In terms of the broader economy there are 14 different cycles that are correlated to a greater or lesser extent. It's only when all these cycles move in a negative aggregate direction that our economy as a whole moves into a corrective pattern. Where academics and economists differ is where this cycle of events starts and finishes. Where exactly on the circle do you point to an arbitrary position and say "this is the start of the cycle". It's a cycle...it never ends..hence the differences in the two diagrams and the differences between academics.

I like to conceptualise the 12 o'clock position as being the peak of the share market cycle, and the six o'clock position as the start of the new cycle. From 12 to 6 the market is falling, and other asset classes are more appropriate to invest in.

Note that because it is an aggregate of a bunch of individual cycles, each time the cycle rolls around, it looks different. Some cycles are 5 years in length, some cycles have been 14 years in length with the long-term average of the cycle 7.8 years.

The important thing to remember is that the sequence of events around the outside of the clock is always consistent. Yes they look different each time but the major events (or the signposts if you will) stay the same every cycle. The sequence is consistent.

Now I'm going to add a layer to make the concept a little bit more complex for you. Global economies are correlated. They synchronise their clock's, but they aren't perfectly synchronised. Larger economies tend to lead and smaller economies tend to lag. Australia makes up 1.6% of world economic activity. So we are a very small player. This is GOOD FOR US. Other economies lead the way into and out of booms and busts, meaning that we have a higher level of predictability in our market that the larger economies do NOT. We get an indication of what is to come for us, by looking at larger economies. This is the reason why I frequently rant at Australian Managed Funds because of their behaviour during the GFC. They could see what was about to happen in our economy and they did nothing to protect the assets in their control.

So the clock is a general tool and it is the events on the outside of the clock that determine where you are in terms of the clock. I'm going to focus on the events at the start (bottom) and end (peak) of the share market cycle, rather than go through a full cycle, because I think that will be the limit of the time I can give and then open it up to questions.

 

[Sir Osisofliver]

Start of the Share Market Cycle

What is the trigger for the start of the new share cycle? As we approach the six o'clock position we've already gone through the optimal buying window for both Residential Property and Fixed interest. Interest rates, which started high, have been lowered by the Reserve Bank under their charter to stimulate the economy. This time around the guvmint also handed out great wads of cash in the form of stimulus packages to...you guessed it, stimulate the economy from the depths of recession.

The technical definition of a recession is two consecutive quarters of negative GDP. So this is one of the signals that you will use to determine your position on the clock...are we in a technical recesssion. (Note that the release of these figures is three months after the event so the lag can catch you if you are not aware of it.)

The other main indicator of the start of the new cycle is what is happening in the unemployment cycle. (See the image attached below).



When we look at the chart above you can see how in a general sense it follows a sine wave pattern...IE it's cyclical. You'll also hopefully note the correlation that exists between the peak of that cycle and the bottom of the share market that occured in March 2009. Yes there is a lag effect..it's not a perfect correlation as the unemployment cycle lags the start of the share market cycle, but the peak of the unemployment cycle is significant evidence that we are at the beginning of the new share cycle.

Let me explain what is happening and what drives our economy out of the correction. The guvmint draws back a vast amount of our overseas reserves whith which to employ capital expediture projects (and give free money away making Kevin Rudd everyone's best friend). Reserve Bank has lowered interest rates, making it cheaper and easier for business to do business. HOUSEHOLDERS however at this point in the cycle, are counting every penny. Businesses have been laying off staff. They have to if they want to survive in the corrective environment. The peak of unemployment however is when the company starts re-hiring. All the existing employees breath a sigh of relief because they now feel more secure in their jobs. All those expenses that they have been putting off to put some money away just in case now get paid. The car gets a much needed service, the kids get new school shoes, hubby takes the wife out to dinner etc etc etc and it's this activity and increase in consumer confidence that has a significant effect in driving us out of the recession.

Because these expenses build up and all occur within a very short period of time, this leads share prices to have a bounce at the immediate bottom, (the tipping point if you will) around people being fired and people being hired. See the picture of the All Ords below and note the rapid rise in share prices that occured after the tipping point. Now look at the Unemployment cycle picture. During that initial very profitable rise in share prices, the unemployment cycle only started to improve at the end of that run. Thats the correlated yet lagging effect I was talking about. So your trigger for the start of the new bull market run, is when the unemployment cycle reaches peak, NOT, when it starts to improve.



This is golden time for long-term and savvy investors. Yields are HUGE on some really great shares that have had the **** kicked out of them. Banks with near 10% yield? I'll have me some of that thanks. So Yields (particularly Aggregate yield or EPS) can be used as an indicator to tell whether you are near the bottom of the market. This is an expression of the cyclical market. Brokers are yammering on about great yield and technicals as if they are the be all and end of of analysis. Stochastic this, MACD that.

When the Reserve Bank starts to increase interest rates again (guess where we are in terms of the interest rate cycle), you can be assured that the beginning of the new market has occurred. If you look at one of the clock diagrams in the previous post...you'll see rising share prices as one of the indicators I was talking about. Hey look another cycle

A bit later and we start to see commodity prices begin to improve, because the stockpiles that were created when the market corrected, have now been eaten into by the increase in demand. Queue another indicator of improvement and the movement of the clock away from the 6 - 7 o'clock position. Good Grief yet another cycle within the broader cycle.

It's at this time that the Reserve Bank starts to rebuild our overseas reserves, and will continue to raise rates in order to get good purchasing against oversea's currencies. Heaven's yet another cycle within the broader cycle.

It's around this time (say about 8-9 o'clock) that Banks and lending institutions start to free up the credit market. They turn the faucet off in a BIG way as we approach the recession because they can't justify lending money to any tom, dick and harry because heads have rolled because of the stupidity they had when lending to unemployed people to buy houses. (Ok this wasn't us...it was the U.S. ) and look another cycle.

Mid Cycle

Mid cycle we've reached the old high of the market. Straight away you should be looking at the All Ords and saying OK we are not yet at 9 o'clock. we are between six and nine, but probably after seven because we've had that initial rise in share prices, and an improvement in commodity prices, so we're somewhere between seven and eight. It's this point where the clock tends to slow down. We can sit at this point for a long while whilst we wait for the Synchronisation in Global Economies to occur to lead us to those ecstatic peaks.

End of the cycle (Peak of the market)
Banks are giving credit away like's it's water with a use-by date. Commodities have razor thin stockpiles and prices are jumping like a cockroach on a hot barbeque. Share price yields are looking THIN and brokers are saying...Look its all about the fundamentals.


Okay - now read the above and think about chaos, what is happening?

Cheers

Sir

 

[Danneman]

Sir Osis, I'm following your threads with great interest.

This is highly interesting, especially since I watched a BBC documentary about the mathematics of chaos the other week(pretty light entertainment, mind you). But even these doco's tend to make you think.
The idea was that nothing is as linear as us humans like it to be, or want to believe. And once you push a system (be it economic or climate), it eventually reaches tipping points where the previous rules (i.e the economic clock) don't apply anymore.

Obviously, one would think we would be a such a tipping point in the economic system. It's tempting to think that, being a natural bear, once all highly indebted countries fail like dominoes, no economic rules, but chaos will apply.

It's actually rather depressing..

thanks for your kind donation of knowledge though..

Daniel

 

[Punta]

I don't agree that markets are chaotic. A basic property of a chaotic system is one where a tiny change in the current state of the system leads to large changes in the future state of the system. E.g. the weather, where the proverbial butterfly flapping it's wings could reasonably affect the formation/path of a hurricane some time later.

I doubt that the markets behave like this. It seems unlikely that individual buy/sells affect the prices at some later date, even for the individual securities involved, let alone the entire market.

The broad market moves (bubbles and crashes) are probably more of a forced response to existing legislation, or innate human nature, or well understood cycles, rather than chaotic.

 

[OldFart]

Insane.....no wonder the market is all screwed up.....if banks can get away with fraud, why not wallstreet and the newsmedia as well?

Bank Debt Is Near Record Low

http://www.cnbc.com/id/45466176

and here's why they have no debt:


Secret Fed Loans Helped Banks Net $13B


http://www.bloomberg.com/news/2011-...congress-gave-banks-13-billion-in-income.html

 

[Sir Osisofliver]

Apologies for the absence from this thread guys and girls, as I indicated at the beginning I would post in this thread when I had time. Unfortunately my control and Chaos modelling account with live trades were with MF Global, which has blown what free time I have out the water and put the kibosh on the monetarised trading model.

I do intend to get back into this material but my time is extremely limited and I'm not prepared to give up sleep to write about this...but I wanted to pop in to address punta's comment below...

I don't agree that markets are chaotic. A basic property of a chaotic system is one where a tiny change in the current state of the system leads to large changes in the future state of the system. E.g. the weather, where the proverbial butterfly flapping it's wings could reasonably affect the formation/path of a hurricane some time later.

This statement is correct but somewhat misleading. The weather as you've used in your example is not based within a natural system which is also a growth model. Chaos expressed within a natural system which is based upon a growth model would be something like a population curve, or plant growth. Usually there is a constraint on the growth, that creates a replicating pattern of expansion and contraction, but these are still natural patterns that exist within chaotic principles. EG Introduce a new species into an environment such as a cane toad in Queensland and watch what happens. A minor change to the initial conditions IE three more individuals released, may create a vastly different growth rate...an expression of chaos within a growth pattern. Can you see how this would apply to the market?

I doubt that the markets behave like this. It seems unlikely that individual buy/sells affect the prices at some later date, even for the individual securities involved, let alone the entire market.

The broad market moves (bubbles and crashes) are probably more of a forced response to existing legislation, or innate human nature, or well understood cycles, rather than chaotic.