This page has been archived and commenting is disabled.
Guest Post: Sketching Outlines Of Predictability
»
- 4880 reads
- Printer-friendly version
- Send to friend
- advertisements -
Contact Information:
Tips: tips [ at ] zerohedge.com
General: info [ at ] zerohedge.com
Legal: legal [ at ] zerohedge.com
Advertising: ads [ at ] zerohedge.com
Abuse/Complaints: abuse [ at ] zerohedge.com
Suggested Reading:
Make sure to read our "How To [Read/Tip Off] Zero Hedge Without Attracting The Interest Of [Human Resources/The Treasury/Black Helicopters]" Guide
It would be very wise of you to study our disclaimer, our privacy policy and our (non)policy on conflicts / full disclosure. Here's our Cookie Policy.
Ibuprofen Bitchez
Ooooh, math bitchez ;)
I rike
I was going to read this, but decided today would be the day to get my hemorrhoidectomy. I am not sure which would be more painful.
Allow me to sum up (in the interest of getting you to your appointment):
Shit happens.
Pretending you know what shit is gonna happen, ain't gonna happen.
What the hell does all that mean ??? :-)
Some of us have jobs that are dependent on predicting the future with reasonable accuracy. Taleb wrote an entertaining but perhaps inaccurate book (The Black Swan) that said no one can predict anything with any accuracy.
Kaminski's theorem (see citation) shows there are conditions under which prediction is possible.
I submitted a conjecture that, if true, offers a rule of thumb technique to determine when models are relevant, and when Taleb is right.
There is a simple counterexample to the conjecture where N=2, where the observations are not equal. Pre- and post- estimates will both have zero variance and constant non-equal estimates. I've been looking for some general conditions under which the conjecture is true.
As he also said, the market is great for amusement only. Not to make money.
What a coincidence, I was searching for those Kaminski charts just a few hours ago....
I don't think Taleb says no one can predict anything. I think he says there are more unpredictable events of significant magnitude than most "predictors" assume. In fact most self proclaimed predictors get caught in the "turkey fallacy" that everything will always be as it is in the middle of the curve.
The map is not the territory, and the more you think it is, the greater the peril.
I probably spoke too loosely about Taleb. That said, I've heard his talks/workshops outside of those books, and he is far more dismissive than you acknowledge.
Don't get me wrong, BSwan was a great read, and he helped me realize "it is better to be broadly right than to be precisely wrong."
I'm a Taleb fan, just not that stupid treasury business.
He is a very frustrated individual. With good reason.
Attn: Deflationists, short sellers, etc.
Which way are these going?
Up or down?
CRB Index
Commodity stock index
Cyclicals
India
China
Anybody caught short if we break out needs to go back to school and learn how to read charts.
LOL....
i am all long after the last one
Surely enough to make you long.
It's not that I didn't enjoy the attributes, but when looking closer I came upon the genesis of our current Fed. problems. If you look closer you will see the PDR and Policy & Politics. How befitting, let your imagination run with that one. GeeZZZ
Uuuhhhh, Keynesian economics on steriods?
Nice catch. Buuuut, http://www.youtube.com/watch?v=hQKb6lacbGs (just teasin)
How many standard deviations are those?
The weekly FTSE chart is a good tell of where we're going.
H&S top
rising wedge making up the right shoulder
volume spike and graveyard doji for the week matching left shoulder in Jan
I think we're having deflation in the stuff we don't need and inflation in the stuff we need.
http://www.marketwatch.com/charts/int-adv.chart?symb=UK:UKX&sid=123797&time=8&startdate=&enddate=&freq=2&comp=&compidx=aaaaa~0&uf=7168&ma=1&maval=50&type=4&size=1&lf=1&lf2=4&lf3=0&style=1013&mocktick=1&rand=229610621
Good comment Robo - and good call on the market for the past few weeks. You have been spot-on.
Judging from the tone of the comment threads here in the last month, there are many who are probably still short from sp1040 (or worse) and have blown themselves up trying to time the next crash.
When the next crash does happen, very few of the doomsday crowd will be left to profit from it.
Judging by the comment threads here I would venture to guess there are more who have been long gold.
http://finance.yahoo.com/echarts?s=GLD+Interactive#chart3:symbol=gld;range=6m;compare=spy;indicator=sma%2870%29+volume;charttype=line;crosshair=on;ohlcvalues=0;logscale=off;source=undefined
"there are more who have been long gold"
that would be a good guess, Tyler
stocks are for suckers - short, long, who gives a fuck?
grab some gold, pop some popcorn, enjoy the show
There is absolutely no question that the long gold/long silver meme overwhelms most threads.
My observation stems from the "tone" of comments targeted towards virtually any mention of an equity long bias regardless of the timeline for the trade (I am neither permabull nor permabear - merely a trader).
TD - the ZH articles are the best.
Actually, as you may have read on Zero Hedge over the last 1.5 years, the prevalent theme is how broken the stock market is. You are welcome to time every candle chart in the market. Just dont be surprised if the next one takes it down 15%, 25%, 50%... We have not checked if robo posted one of his trademark pictures of a stock jumping by a percent or two on May 6 at 2:30 pm, but it wouldn't surprise us. Just as it wouldn't surprise us if robo were to sell all his stock at the very bottom of such a move, when the market went from bidless to mini bid. Of course, robo would not disclose that.
Yes TD - I have been here since the beginning of ZH (almost) and I do agree that most markets are broken relative to the old days and the Algo-liquidity illusion presents abnormal risk to the downside.... (and I position my capital accordingly).
I never got the impression that Robo was posting charts to claim personal victory on anything. Based on your comment, you clearly think he is.
Well, since he said last week that he doesn't trade most, if not all, of the stocks he posts charts on I for one am lost as to why he continues to post these charts at all. We all know low volume ramps are rampant in today's market.
as most as he post soft porn he probably works for the SEC and despises us all for giving him work after congressmen read the posts
+1000 Tyler *appluase*
so THAT's where gordon gecko been spending his time.
He's been flat out wrong about Gold. No one has argued that the market was rallying here -- the point is will it fail the top of the trading range or break out? Robo is adamant the break out is a given -- others here are simply saying let the charts prove it don't force them. To do that you have to have confirmation which we didn't get nor have we had during this rally.
"In most of the cities and towns of this country, this Wall Street panic will have no effect."- Paul Block (President of the Block newspaper chain), editorial, November 15, 1929
"We're now in the 'middle innings' of the current economic expansion, and the next economic recession is not yet in sight.” - David Seiders, Chief Economist, National Association of Home Builders, Jan 2006
"Anyone who bought stocks in mid-1929 and held onto them saw most of his adult life pass by before getting back to even" Richard M. Salsman
When are those APPL's gonna get included in the CRB?
If you want to see an interesting end of the day spike, look at the chart for KCP on Friday. Maybe a new line of shoes?
I would like Robot Trader to specify who is going to do all the buying. Individual investors are headed to the exit doors. Sovereign wealth funds? Direct FED. purchase of the indexes? One computer to another?
I would like Robot Trader to specify who is going to do all the buying. Individual investors are headed to the exit doors. Sovereign wealth funds? Direct FED. purchase of the indexes? One computer to another?
Lovely picture, Robo.
No different than being caught long if we fail. Both positions need disciplines, stops and plans -- why is it that you won't acknowledge that anymore?
It was broadly predictable that Robo would post a picture like this.
For the last image, I would say "go down".
Once again: The economy is not a machine, it is an ecosystem. If models could be constructed like college physics lab classes there would be no need for the fancy charting and metrics. We could just apply a formula (simple or not) to predict future market movements -- like planet rotation and refraction of light. Those pesky animal spirits keep interfering.
An ecosystem is a type of machine. It is a complicated and interesting machine because component parts interact, creating complexity.
There are conditions under which even such complicated machines fall under the power of the human mind. At other times, we are trying to catch the wind. Even so, time is on our side.
You'd argue with a post, wouldn't you? Sometimes we just have to stop the madness just above the molecular level. Your determinist view is not helpful to the debate. Having said that, love ya man!
I'm trying to have a conversation here. Sorry you see things as an argument. Over and out.
Either over, awaiting further communication, or out, done; not both. It's like saying clip when magazine is meant. Just a pet peeve, carry on.
Don't go away mad...
Now you done it. Taking the arguments down to the molecular level has made the comments go all Brownian, LOL !
Random Motion is(are) my middle name(s).
Big picture small picture focus violation in 3.... 2.... 1.....
Complex does not equal complicated. I am sure you mean one or the other in this post. The relevant differences are alive in many major multi-disciplinary debates.
Human choices are to a certain extent predictable. There is great insight in this topic available from marketing, such as campaign responses. The fact that advertising prices can be predicted and justified is based on historical data. So, to say that modeling markets is difficult just because there is human choice overestimates the unpredictability of humans. Meanwhile, even systems with simple rules quickly turn into highly complex structures, as elegantly described by fractals. It is the ability to aggregate states in a nested model which is crucial. For example, Newton's law F=M*a holds even though there are a huge number of atoms and electrons in different states in the matter. Alas, the article is devoid of any discussion of complexity management.
Having said that, it is fairly simple to predict what will happen in a system which is entirely centralized and debt-based. The function from the past to the future is the compounding rate of interest. This function works well to describe the change of state for every debtor or creditor, hence is universally applicable. Thus there is a very high degree of predictability in debt-based systems as there are built-in time depenencies. In a system of truly "free" economic agent, meaning "free" cash flows without any commitments to pay debt, there is less predictability, as there is more room to change decisions as cash flows are not locked up, i.e there are fewer contiuous function mappings across the two spaces. However when cash flows are locked in, the system in fact becomes more predictable. Ultimately, it will be the lender who will suck in all the cash and hold all assets, as continuing defaults will continue to cause assets to gravitate inwards, towards the lender of last resort, who is also the receiver of collateral of last resort.
I would agree that if-then logic does not work well, but the reason is that complex systems should rather be described as networks. In the case of a debt system, it is a predictable network which is emanating from a few central nodes, the central banks, and then spreads out in a tree form (acyclic, directed lending-contract graph where the directionality of the arcs signifies the direction of the collateral pledge) where the arcs in the network map past to future state via debt repayment terms. Using such hybrid modeling techniques, both discrete and continuous, along with carefully delined aggregation, it becomes very easy to predict the economic system.
I'm not a believer in complexity management for adaptive systems when the component are ultimate survivors like people. The I-Ching is right, you try to manage things and it only makes poepel more cunning and perverse to your ends. Not sure how you can say possibly say that it is very easy to predict the economic system. I mean... dude.
I don't know what this means. Please elaborate on this.
UPDATE: I get it. Excellent 2nd paragraph. You mean it is predictable once the system reaches some critical mass and undergoes condensation.
Yes, consolidation increases predictability considerably, but also note that when the tendency of consolidation is an intrinsic feature of the system it is predictable even before it consolidates.
The full sentence: Using such hybrid modeling techniques, both discrete and continuous, along with carefully delined aggregation, it becomes very easy to predict the economic system. (Full quote please).
"Easy" is on the precondition that hybrid models which are both quantitative and discrete, but this is rarely the case. Banging one's head against a stone wall has never been easy. Either it is all about continuous measures (M1, M3, national debt, budget surplus, number of unemployed) or a few discrete actors (Fed, Government, Private Sector) chosen seemingly haphazardly without detailed descriptions of their interrelationships. With such intellectual starting points no wonder there is confusion.
One of my points was that the intrinsic degree of complexity is not really relevant to its predictability. This was the reason why I used F=MA as an example. Put differently, just claiming that something is complex does nothing to make it unpredictable per se. The key word here is choosing a level of aggregation with a high degree of predictability which is relevant to the decision being made.
As for your survivor example, a survior is even more easy to predict compared to someone who has surplus resources. People with surplus resources (intelligence, money, authority etc) create far more unpredictability in an economic system than survivors. Freedom in terms of allocation decisions without any surplus beyond getting the basics has little systemic unpredictability tied to it. A survivor needs food, shelter and basic safety. This is why, in (hyper)inflation, it is very easy to predict where spending goes: necessities and commodities. Not just by historical macro data but by looking at individuals. It is not arrogant to say for sure that individuals need to eat, that is a fact. It is easy to predict.
Economically one could put incoming cash flow and a store of wealth at the base of the free agent's hierarchy of needs. The debt based system constantly attempts to establish inward links by having free agents pledge their capital as loan collateral. This strikes at the basic needs of economic security. The counter-response to this is naturally to cut off the collateral dependency and seek stores of wealth that are not linked to the credit system (e.g. commodities, preferably non-perishable).
This is why the Federal Reserve System is one of the most intellectually brilliant economic systems ever devised. It has both a network structure to it (a network of banks, interrelated through clearly defined relationships between institutions, e.g. dividend payments, authority etc, see Federal Rerve Act) and it links into commercial banks and it establishes incoming cash flows through interest rate payments, i.e. a quantitative function that creates time-dimension predictability (a higher certainty of state transitions).
The combination of a strong discrete centralized network with a quantitative function which describes the state transformation from the past to the future is just sheer brilliance. To me this implies that banks in general and central banks in particular will soon own all assets of the world (they will flow inwards along the links and stop there). Same goes for inflation. One issuer, millions of holders. Thus, the flow of purchasing power goes to the center (governments, Fed) from the pheriphery, from pheripheral debt holders to central debt holders, etc.
So the predictability of the system is in my opinion high. It explains the fall of the dollar, the deprivation of of lower/middle/most of upper class' purchasing power, decline in pension funds, escalation and centralization of debt, aversion of deflation and many other things.
I hope you don't think I am arguing here. I'm not. Just respectfully engaging you.
Those pesky humans are predictable only when they play by the rules and don't burn the playhouse down. Some of the behavior I saw in Hungary and Poland in the 80s made this plain, at least to me. They would screw up production assembly lines not because of negligence, but rather on purpose, just to stick it to the man.
I think this is just definition differences. Topolgical complexity implies somethign close to the sensitive dependence on initial conditions--"chaos stuff". Chaotic systems are inherently unpredictable. I admit they are under some conditions controllable, though.
I suspect your definition of complexity doesn't imply chaos.
No problem. What I'm trying to focus on is the behavioral mechanics of the system, not a single term. It doesn't matter whether we use different terms, to have a 'gut feeling' of the system behavior (to internalize it) is priority #1.
People's sense of dignity and self-respect - individual soverignty i.e. freedom from government/debt - are part of the pattern of the economic (fractal) system. However these forces are more suited to be represented as non-continuous functions, e.g. when somebody says "enough is enough". It's not like the production line burns a little bit every day, rather it's all in one day. Taleb would call it a black swan event. So sure this potential transition capability exists, however it tends to require extreme preconditions to stimulate (human suffering/suddenly revealing lots of unknown information), yet it has systemic predictability (we may not know when, but we know what leads to what). It would temporarily appear chaotic, but I would rather call it large-scale rare systemic state transition. It's not a very practical description so let's just call it chaos then...
That assumes that the governments will honor the contracts made with those banks.
IMO that is not very likely.
They will either break them to stay in power, or the current representatives will be replaced, and the old contracts will then be declared null and void.
Some things are simply unenforcable.
Wouldn't spend to much time pondering that statement as it's nonsense. A question for you, though, JM: why only 2 buckets? If you chose 2 arbitrarily, how would you determine if n buckets were enough? What's in between or prior could be different and be what also happens tomorrow, or what happens to morrow could be unprecedented.
Stationarity does imply there are many possible variations. I advise 2 buckets for economic and practical reasons. Data is expensive, and you need to economize on its use. Practically, if you use, say, an ARMA fitting, it is easier (but not optimal) to handle the comparison with only two buckets. This applies even if you Monte Carlo, I should think. So I guess you could say this simple procedure also economizes on time.
SORRY I DIDN'T SEE YOUR REAL POINT EARLIER... there were a lot of comments going up.
There is always a risk of the unprecedented blowing up a model. So even given stationarity, one truly can't assume that it will hold 4ever. Quant models are a part of making sense of events because they are a way to objectively organize data. There are clear rules guiding how data is condensed into knowledge. I won't rely on any single model 100%, but the conjecture offers a possible route to the conditions under which I won't count on them at all.
The real issue is epistemological: how are you going to make sense of the world, or even "know" anything.
Deduction fails because the world fits definitional categoriesonly occasionally.
Induction fails because ultimately that outlier blows up the argument.
Abduction is the only rational alternative. Conclusions from a quant model are ideal inputs into abductive reasoning. Not the end-all be-all.
Agree/Disagree?
JM, are you familiar with "Why Markets Crash" by Didier Sornette? He fits a log periodic function to price data. This seems to work very well in not only predicting crises, but also their aftermaths.
I would also be interested in whether you are familiar with John Ehler's work. His MESA software, based on the Burg algorithm, is used to discern the dominant cycle in a price series. For this purpose, it has distinct advantages over Fourier methods. Once he has the dominant cycle, he subtracts away that to arrive at an instantaneous trendline. Thus, his methods enable one to distinguish between when the market is cycling and when it is trending, with little or no delay. Though this is not prediction per se of when a market will top or bottom, it can be back tested and has been found to be reliable on out of sample data. The predictive aspect is whether this software will be able to accurately distinguish future trending vs cycling in a price series; if the answer is yes, then your invertability criterion would appear to be satisfied, if I understand correctly what you are discussing.
As I understand your piece, the invertibility criterion is met perfectly by absolute physical laws, so the question of invertibility has more to do with the degree of predictability. For instance non-laminar fluid flow is inherently unpredictable.
Aside from the question of predictability, I've been reading John Casti's "Reality Rules" and in one of his discussion questions, he raises the topic of the Kondratieff wave, which, if it is a real phenomenon and "not just an artifact of the economist's imagination" should provide long term predictions. Casti raises a fascinating question, can we account for the wave in one of the following three ways?
a) "The wave is a stable limit cycle arising from the Hopf bifurcation mechanism:
b) "The wave is one of the periodic orbits emerging during the period-doubling bifurcation of a logistic-type map."
c) "The wave is a quasi-periodic orbit on an n-torus."
The answer to this question would seem to precede the question of predictability, and could probably be generalized to most price series that involve waves. I would assume that picking the a, b, or c would make a huge difference in determining how the Kaminsky theorem would apply.
Very familiar with Didier Sornette's work. To his credit, he put his research program on the line with his market calls. This is not to abuse him, but his call on a China equity meltdown in September of last year was completely wrong. I respectfully think at least some of his log periodic model fits the so data well because much of the fit is derived from hindsight. What I think is powerful is his ability to conceptualize cascade effects in the financial setting.
You understand well what I am saying. Just about any quantitative model falls under the topological gaze, but the models I use are of the wavelet/decomposition type. They probably are yours too, as you see the connections so well.
You know as well as I that wave decompositions are tricky (and somewhat proprietary). Reliable sampling from a dynamical system for attractor reconstruction requires a voluminous amount of high-frequency data to be robust. Is your reconstruction good enough to make conclusions about Kondratiev waves? Quasi-periodicity or more complicated? I could go on and on.
I think that there is a powerful place for quant models in that they enable one to say "this is XYZ", as opposed to "this is like XYZ". It raises the quality and standard of the model used, and the quality of the decisions possible.
Escapeclaws:
I spent some time trying to calibrate that log peridic model of his:
A + E(t) Osc(t) = A - B (tc - t)α / SQRT[1 + { (tc - t)/Δt }2α] [1 + C COS[ w Log(tc-t) + (Δw/2α) Log(1+{ (tc - t)/Δt }2α) ]].
How do you fit w and tc?
Thus a centralized, debt-based system is a gradient vector field with a point attractor (sink), to look at it from a dynamical point of view. Could you explain what are the two spaces between which there are continuous mappings? I'm trying to reconcile that with the vector field point of view in which the vectors just measure the speed and direction of the flow toward the central bank.
May I ask, what is your line of work...? The only reason I ask is I have no fucking idea what you just said... Graphic Design/Print Broker is my gig' sorry to sound so stupid.
Yes, that is a very succinct way to put it!
The two spaces would be the net worth of any pair of debtor and creditor from present time and future time. There are many factors but the important one to focus on is the net flow of interest rate payments over time.
I think of the system as composed of flows. We could characterize debt as a flow vector. The basic formula for compound interest is FV=PV(1+i)^n which would be the space mapping function. The 'strength' of the vector could be said to be the right-hand side factor, (1+i)^n. This would imply that when interest rates are low, velocity of money as induced by rates payments ultimately slows down as "debt saturation" occurs (free cash flow locks up and return yields reach asymptotic minimums). The field is conceived by nesting. The interest rate spread between savings and borrowing is a gravitational force towards banks. Dividend payment and profits in turn allow bank owners to grow the business. This creates a feedback mechanism (there is a cyclical short cut in the field). While not entirely sink-oriented (there's also interbank lending) the simplified equation makes sense at end-point with one-way interest payments (lender pays bank). For a bank one ought to sum all FV from all PV over each debt contract (which may have different i and n) to arrive at net flows.
One conclusion would thus be that all banks that lend at interest must be non-profit and modify the interest rate spread by paying back excesses savers to make the net profit 0. Otherwise banks will gradually increase economic inequality.
There are many signs that this is actually happening in the system right now (increasing gini coefficient, increasing financial sector % of GDP, the bonus symptom, nationalization of lenders).
If You Knock over a Domino in a line of Dominos... that are equaly spaced and within reach of each other while falling...
When You Knock over one.. You can assume that the others will fall.. unless there is an intervention of some kind.
The action of knocking over the Domino has a re-action of the last Domino in the line falling (along with all the others along the way).
I think thats the best way to say what he is saying with A,B,C,D, and so on.
I liked the read, even if I did lose myself a few times!
An interesting exercise would be to work with topological predictability in terms of non-linear functions or even Riemannian branch cuts and sheets. Non-linearity could be simply the squaring function which has two pre-images requiring a “different” space for the entire solution.
The power of the topological approach to predictablility is that it applies to any homeomorphism. But even topological dynamics theorems are expanding beyond these limitations to non-invertible functions. See Akin's book on topological dynamics.
You are thinking through to special cases, which is needed and the relevant thing to do.
I requested Akin's "The general topology of dynamical systems" through the Interlibrary Loan (ILL or Iliad) System. When some of my other requests clear I may ask for 1.) Dynamics of topologically generic homeomorphisms, 2.) Simplicial dynamical systems, 3.) The topological dynamics of Ellis actions, 4.) Recurrence in topological dynamics : Furstenberg families and Ellis actions, in that order. Any favorites in Differential Topology? Thanks for the reference - Cheers
Nice addition to the night table reading...
Unless someone can magically solve the three-body problem, I don't see prediction from price series being possible to any degree of usefulness. (That would be an n-body problem, where 'n' is the number of participants, carbon based or not.)
Reference:
http://en.wikipedia.org/wiki/Three-body_problem
I take issue.
Is prediction possible to an arbitrary tolerance? No.
Is prediction possible to some degree of usefulness? Yes.
Every decision is based on prediction in a loose sense. I don't discount luck as a determinate of success. But I don't think luck is everything. If there were no predictions made (again: in the loosest sense), then we would all just be a bunch of irrational zombies lumbering through an empty life.
The three-body problem has already been solved (Karl F. Sundman). Yet, time series are somewhat "myopic" perspectives on the overall economy. There will most likely be patterns to be found (if fitting to enough models) but one ought to be humble about the bounds of the pattern-induced knowledge derived from a single quantitative metric e.g. a single stock or option price.
If you had all the information of all elementary particles in the world, you would probably be able to predict it. But in order to setup the machine one would need to replicate the current universe's state. In order to do so there would have to be some form of measurement, alas, such measurement would affect the subject (Schrodinger's cat). And even if possible, it would take the setup of another parallel and independent universe, and such an enormous Turing machine is way too expensive to set up for mere mortals. We still lack an understanding of the essence of matter and the behavior of N bodies where N>3, so let's just live happily and randomly...
That is, to put it bluntly, a very naive statement.
It was the general scientific consensus 150 years ago. In 1850 then we had the observations of Faraday, and the increasingly more complete explanations of that category of observations by Planck, Heisenberg, Einstein, et al.
Quantum effects are thought to be fundamentally random and are thought to be fundamentally unobservable (observing them changes the end result). There's plenty of quantum effects in the macro world as well: analog TV no signal picture, newer random number generators in CPUs, and, of course, the human brain.
I think we're saying the same thing here...
It is a very big "if", and it's not even remotely possible, both due to the impossibility to replicate the state and second due to the inability to even make an independent observation that doesn't change the state of the subject (hence the example of Schrodinger's cat).
Let's see what comes out of the CERN accelerator.
I do not mean to get carried away on a rant. Your wiki reference was interesting. In today's world of computers, numerical methods are sufficient for the practical solutions of problems, including the 3-body problem - and I do not mean "Ménage à trois." Aircraft and aerodynamics is an example where there are working solutions in place. Another example is interplanetary space travel or air-to-air missile countermeasures. Since all of these subjects, including finance, use the mathematical language for solutions, an advance in one area of computation can affect others.
Even though n-body problems are thought not to have analytical solutions, your wiki reference states that they do: "N-body problems have a global analytical solution in form of a convergent power series, as it was proven by Sundman for n=3 and by Wang for n>3. However the convergence of the Sundman and Wang series is so slow that they are useless for practical purposes, and that is why, for the time being, approximate solution by numerical analysis in the form of numerical integration (or, for some cases, classical trigonometric series approximations) is unavoidable."
Since the convergent power series solution is slow, there is work here for mathematicians and scientists to speed things up. How about hyper-frequency trading? Too bad that myopic Program Managers (PM) are in charge and holding us back. Managers and end-users will need a better education or more indentured statisticians, mathematicians and physicists to counter-act management ineptitude. I wonder which way this will trend? Probably with more greedy managers with their poorly educated, powerful customers and fewer indentured scientists. Remember that the boss is not in charge because he is smart, but due to experience and guile. Who really believes that the duplicitous argument that banks need money to retain smart people? Ever meet a scientist more passionate about money than the study of Nature? In the mean time keep your head down and do not get shot. Sucking-up is the best way forward.
In summary - I believe in complexity arising from simple rules. Stephen Wolfram has done the most work on this, and he has a brilliant TED talk that outlines his main points.
http://www.ted.com/talks/stephen_wolfram_computing_a_theory_of_everythin...
The problem, however, is being able to reduce the behavior backwards towards these simple rules. It may take all the time in the knowable universe to do so. That is why I see price series, which reflects the vagaries of human emotion and greed in addition to a myriad of other inputs such a gargantuan task to solve.
It may well be that someone has, but I bet they'd take that secret to their gold-cast coffin.
As complex and sophisticated folks most certainly the markets, being the data set of humanity, can only be explained in an extremely complex and sophisticated way.
Or not...
Your post is much too advanced for this fella but it does tickle me to consider the brillance on the other end of my 'deterministic' trades.
Best of luck!
Globalization presumes sustained economic growth. Otherwise, the process loses its economic benefits and political support.
Paul Samuelson
If you have always believed that everyone should play by the same rules and be judged by the same standards, that would have gotten you labeled a radical 60 years ago, a liberal 30 years ago and a racist today.
Thomas Sowell
'If you have always believed that everyone should play by the same rules and be judged by the same standards' ...
Question is SS what will such a person be labeled as in the future?
Da times they are a'changing.
http://anonymousmonetarist.blogspot.com/2010/09/united-states-of-america...
I wish you could speak in "laymens terms" one time and throw down the gauntlet ....
Your one of my all time favorites but some of your stuff,,, Voooom' right over my head ..
You seem to know the interworking of the banking industry, is it really that scary ???
Update: Slouching Towards Gomorrah: Modern Liberalism and American Decline
A must read .... Thanks for the link, it had been a few days since my last visit.
I suspect that tries to argue against positive discrimination for minorities, without saying it out fair and square?
That statement brushes over the small detail that minorities are being discriminated against actively, due to a historic dynamics that are very hard to change - and hence they are being judged differently already.
It also brushes over the detail that generations of discrimination left their mark on minorities which makes them handicapped in the marketplace straight away. Such effects are hard to change as well: unemployment is way higher and general wealth is still way below that of comparable caucasians.
If you accept that view then racial positive discrimination can be thought of as a counter-force against that existing force of discrimination - i.e. it tries to restore (or at least, counter-balance) that implicit discrimination that happens in the USA every day.
If you dont accept that view then you claim that racial discrimination has been eradicated and does not exist in the USA anymore and that babies born into minorities do not start with a handicap straight away. Are you making that claim?
(Note, I am making no statement about the implementational details and general wisdom of current positive discrimination practices/laws in the USA.)
Here's a great passage from Taleb's book:
However, if you believe in free will you can't truly believe in social science and economic projection. You cannot predict how people will act. Except, of course, if there is a trick, and that trick is the cord on which neoclassical economics is suspended. You simply assume that individuals will be rational in the future and thus act predictably. There is a strong link between rationality, predictability, and mathematical tractability. A rational individual will perform a unique set of actions in specified circumstances. There is one and only one answer to the question of how "rational" people satisfying their best interests would act. Rational actors must be coherent: they cannot prefer apples to oranges, oranges to pears, then pears to apples. If they did, then it would be difficult to generalize their behavior. It would also be difficult to project their behavior in time.
In orthodox economics, rationality became a straitjacket. Platonified economists ignored the fact that people might prefer to do something other than maximize their economic interests. This led to mathematical techniques such as "maximization," or "optimization," on which Paul Samuelson built much of his work. Optimization consists in finding the mathematically optimal policy that an economic agent could pursue. For instance, what is the "optimal" quantity you should allocate to stocks? It involves complicated mathematics and thus raises a barrier to entry by non-mathematically trained scholars. I would not be the first to say that this optimization set back social science by reducing it from the intellectual and reflexive discipline that it was becoming to an attempt at an "exact science." By "exact science," I mean a second-rate engineering problem for those who want to pretend that they are in the physics department - so-called physics envy. In other words, an intellectual fraud. - Nassim Taleb, The Black Swan: The Impact of the Highly Improbable, 2007
I think what is being missed here is the role of innovation. Even in the notion of "debt", i.e., credit with purchasing power, the supply has very much been shaped over time by innovation. Now, in a broader economic sense, it is the unexpected, PROFITABLE new way of organizing relationships that is key to growth and rising productivity. As I like to point out, in a competitive economy people are busy learning the most efficient ways to cooperate. Are there patterns of change thereof? I do believe so, but to consider them is to consider the force of creative transformation begotten by mind over matter over time....no simple scatter-plot diagram if God can help it, lest spontaneity not be spontaneous! In other words, free will necessitates uncertainty and may be the fundamental source of change within the system.
Here's another thought. At any given moment is there a "better future"? If so, relative to what or whom? What is the source of the "rational choice" that allows the future superior state to realized? The bottom line is that we are communicating here in the midst of a SUBJECTIVE REALM by definition. Our choices matter with respect to future states. Hence, the power of choice, free will, is the key "determinant" of change over time....and when are we not experiencing change? Never while here.
So those seeking "objective truth" and certainty in this realm are wasting their time as far as I can tell. This is not the place for that, although the effort has been very creative even if the product is effectively false....a useful illusion if you will.
I'm going to pick on Taleb here a bit, not you. This is only because your quote is opportune.
Taleb makes it easy for many people to settle an argument they don't understand by copying him and just mock it. Somehow when it comes from Taleb's pen this is considered profundity. It is not.
I find the quote distasteful because it anchors the notions of rationality, predictability, and tractability within some economic theory. This is so off-base that it is just laughable.
There are circumscribed conditions where models have relevance in decision making. This three page write-up tries to establish simple criteria for those conditions. May be right, and may be wrong.
Taleb is right to point out that seeking to box in the true "rationality" behind economic change is a fallacious equation at the outset. Unpredictability is the essence of the source of innovation that can not be modeled, and all should be thankful for this since otherwise "free will" could not be. It's a nightmare for mathematicians and statisticians, but who cares!
and the best scam is to give the illusion of choice... shell game anyone?
that is what our economy is in now.. so yes, many things ARE predictable and frontrunned, etc etc by TPTB
even the BP oil spill is being magic tricked away
Taleb is interesting, not especially for the profundity of his thought, but because of the audience he addresses in such a manner. IMHO, not for nothing that economics is called the dismal science. However, I've always been concerned that his insights have been dumbed down for his audience and as a result more interesting thought has been laid aside.
Quantitative financial analysis is very interesting but the degree of mathematical sophistication required to follow your post is, for most readers sorely lacking (enjoyable to gloss over none the less).
Taleb uses a vivid metaphor, the black swan, to drive home his point that certainty is elusive and doomed to fail when most needed, in market tails.
The metaphor takes him far from mathematics and allows for an immediate grasp of what he is talking about on an intuitive level while also allowing for an endless stream of divagation and misunderstanding of what he is talking about.
That however also describes the topos of the market.
For you intelligent quant guys and mathematicians, and for Taleb as well, perhaps begin with more Aristotle and end with more Beckett. For Aristotle, a well formed narrative has a beginning, a middle and an end. Yet Aristotle goes to great pains in the Physics to question forms of causality and chance and his analyses are not as neat and tidy as many of his commetators have promulgated. And for Beckett? Facsinating question on reflexive systems gone awry and in many ways a good intro to Aristotle. And markets. Maybe then it's time for math class.
Quant guys and mathematicians aren't always as intelligent, much less wise, as you give credit. Wisdom = humility, which too few possess in this world, self included.
I would add Shlomo's Ecclesiastes to Aristotle's Politics and Beckett's Godot. There is no better book for understanding decision under uncertainty. I still haven't grasped everything in it.
Good stuff.
As well. And if you have the time not wasting away in idle discourse: Molloy, Malone Dies and The Unnameable ...
I'm not a tech guy. But I do follow them. We banging the 1130 a few times and we need some serious juice to punch thru. Keep in mind we moved 7-10 pc in just over two weeks. Not an easy thing to do. Bulls think it's a couple days then bam 1150+
We've been quite choppy and range bound. No bull market acts like this.
Macro data has been poor, we will likely get the pullback this week.
My pull-back to look for will be when the TBTF bullion boyz crap all over the COMEX.
I need some more bullion.
Good luck with Gandhi.