|
|
|
Dude
SFN Die Hard
USA
6891 Posts |
 Posted - 12/26/2007 : 08:40:45
|
Looks like we may have the beginnings of Asimov's Psychohistory happening for real in the lab.
The University of Arizona is working on it now.
Pretty interesting stuff, especially if you are a fan of Asimov's foundation novels.
|
Ignorance is preferable to error; and he is less remote from the truth who believes nothing, than he who believes what is wrong.
-- Thomas Jefferson
"god :: the last refuge of a man with no answers and no argument." - G. Carlin
Hope, n.
The handmaiden of desperation; the opiate of despair; the illegible signpost on the road to perdition. ~~ da filth |
|
|
|
bngbuck
SFN Addict
USA
2437 Posts |
Posted - 12/26/2007 : 11:28:07 [Permalink]
|
Whooeeee.....! Put that ASTRAP sucker to work on devining the NYSE after the algorithm matures, publish the results in Murdoch's new propaganda organ - the WSJ - and what happens to the US economy? It would mean financial nirvana or disaster, depending how fast you could get in or get out of the market. And then program enough sophistication into the damn thing to be able to predict the effect (resulting from its own prediction) upon that very prediction, and take appropriate action to prevent that predicted outcome, and what do you have? Did Issac tinker with that concept in Foundation? (Haven't read it, but think I will, now)
This is really fascinating stuff, Dude! Takes me to speculating about using a Wellesian Time Machine to go into the future and visit my grandchildren, then return back way to my youth and abort my own children at conception (the Catholic view of birth control), which would create a prior-to-the-future situation in which the potential for my grandchildren did not exist, even though they did exist when I started the time-travelling and went to visit them in the then-future!
Literally hundreds of these kill-your-own-grandfather time paradoxes exist, which at least casts some philosophical doubt upon the possibility of time travel. But when one contemplates the consequences of extremely accurate future event prediction however, it becomes a whole different horse race! Literally! Not paradoxical, because here we are dealing with statistical probability, not the 100% fact of absolute knowledge of the future that the paradoxes assume.(They also assume that the future is predetermined, which is very probably untrue!)
Like 90% of the speed of light, there is no existential or mathematically logical reason that extremely accurate statistical prediction - that with a standard deviation or sigma approaching zero - cannot be possible.
But if the folks in Tucson get really good at extending and improving their algorithm, scenarios like the one I described in the first paragraph could develop, and the possibilties are fascinating! Think about it! |
 |
|
|
Dave W.
Info Junkie
USA
26020 Posts |
Posted - 12/26/2007 : 19:12:13 [Permalink]
|
If we can take action to prevent the predicted economic outcome of being able to accurately predict stock values, then the model (algorithm) used to make the predictions was fatally flawed, either due to incompetence or incompleteness. In other words, if the predicted outcome fails to come about, then sigma is necessarily larger than intended.
What would an accurate stock-market model show? I think it would show some interesting peaks and valleys as the model's inventors get rich using it while keeping knowledge of it private, followed by a few days (or more) of level prices when the SEC shuts the markets down because they've caught wind of the model (due to the all-too-human, hedonistic and avaricious trading practices of the inventors).
Then, we'd see one of two things. Either the model itself (and all notes, scribblings and knowledge) is destroyed or quarantined, and things go back to normal, or public trading ends entirely because nobody can sell a stock which buyers know will go down in value, and nobody can buy a stock that would rise because nobody would sell them. The model itself would tell us which would happen, either by continuing past the date of the SEC's involvement or not.
A third possibility is that the inventors of the model are inhumanly wise and never spill the beans or make trades of which the SEC would take notice. In this case, the inventors get slowly rich (and sometimes get poorer to maintain the illusion) and nobody's ever the wiser. For all reasonable intents and purposes, there is no way to tell whether (for example) Warren Buffet already has this prediction algorithm (and superhuman restraint) or not. |
- Dave W. (Private Msg, EMail)
Evidently, I rock!
Why not question something for a change?
Visit Dave's Psoriasis Info, too. |
|
|
|
pleco
SFN Addict
USA
2998 Posts |
Posted - 12/26/2007 : 20:34:38 [Permalink]
|
| Maybe they watched Pi - Faith in Chaos too many times. |
by Filthy
The neo-con methane machine will soon be running at full fart. |
|
|
|
|
Dude
SFN Die Hard
USA
6891 Posts |
Posted - 12/26/2007 : 21:04:38 [Permalink]
|
This thing, as with Asimov's plot device, is intended to look at the potential actions of large groups of people. It isn't going to tell us jack about the stock market (except what you can divine on your own as the consequence of those actions it does predict).
If it ever actually churns out an accurate prediction I will be suprised. Just thought it was pretty cool that someone thinks they may be able to do it.
|
Ignorance is preferable to error; and he is less remote from the truth who believes nothing, than he who believes what is wrong.
-- Thomas Jefferson
"god :: the last refuge of a man with no answers and no argument." - G. Carlin
Hope, n.
The handmaiden of desperation; the opiate of despair; the illegible signpost on the road to perdition. ~~ da filth |
|
|
|
|
bngbuck
SFN Addict
USA
2437 Posts |
Posted - 12/26/2007 : 22:15:45 [Permalink]
|
Dave.....
Assume the program works about as outlined - tracking, categorizing, collating and comparing million of pieces of investment-related information and processing this data until until predictive patterns begin to emerge, then summarizing them into a specific investment strategy with an extraordinarly high probability of success. If information about it's own existence and success probability in predicting the market was kept from it by it's selfish creators, it would have no data that could be used to assess its own predictive effect upon the market until the substantial trading that the creators would be doing (a hell of a lot of trading in certain favored stocks that would send their prices soaring) had an effect on the algorithm. That effect would have to be one that even further enhanced the creators' portfolio - ever higher prices for those stocks. (I would think that they would want this lead time to make as much as possible before the inevitable disclosure.)
Contrarians would quickly reduce these values, because that's the way the market works, but that new information would quickly be inputted and result in an ongoing constant reprogramming that would still acrue to the creators' benefit. Unhindered, this would shortly lead to astronomical pricing, all kinds of stock splits, and a roaring, assymetrical bull market. The Federal Reserve would go nuts, and demand immediate federal action to control the market after raising the prime to usurial rates. Only exposure of the system would stop the process, but draconian measures might be taken by the Fed and FCC to temper the funnelling of hundreds of billions into a few hands.
The owner-creator-users of the system would have to very cautiously use it or lose it, exactly like Ed Thorp's blackjack system back in the '60's. As the article shows, Thorp hasn't done too bad in the stock market, either; but nothing to rival Warren Buffett.
Therefore,I think you are right, in the supposition that ultimately the SEC and the Fed would take note of some funny business. As to whether the authority exists for the SEC to shut down the markets I guess it does predicated on the actions that were taken several years ago to inhibit computerized auto-trading.
But investor stasis would almost have to be the final outcome if everyone had access to the predictions, because every dollar won on the market is a dollar lost by someone else. I was totally joking when I suggested that owners of such a predictive systen would put the results on the front page of the WSJ. Of course they would get very rich, very quick, but after they were outed, the programming would quickly be made illegal, because of it's potential for creating economic chaos.
As to Buffett, I don't think he could keep a secret like that any better than the folks at Los Alamos kept their deadly secret in the forties during the development of the first atomic weapon. Too many people involved in the creation of such a system! Once it was known to be successful, thousands would find a way to develop and use similar systems, even though they might be illegal. This would probably result in a damping effect, correct assymetry, and probably result in a depressed market!
It sounds like the only way a really huge fortune could be amassed from the use of such a program, would be if it was kept super secret, applied judiciously and with careful balance, and considerable diversification employed. But, except for the secrecy, that's pretty much the way smart investors work, even without a computerized crystal ball! |
| Edited by - bngbuck on 12/27/2007 09:17:30 |
|
|
|
Dave W.
Info Junkie
USA
26020 Posts |
Posted - 12/26/2007 : 23:51:30 [Permalink]
|
Yeah, bngbuck, that's what I was saying. 
By the way, the SEC has the power to suspend trading on individual companies (for up to ten days), so I don't see any reason why it can't just suspend all of them. At once. Googling, I find a hint of a bill that would have given the SEC the power to suspend all trading on all domestic markets for 24 hours in the event of some vaguely defined catastrophe, but can't find anything about whether it became law or not.
Hey, who decided to shut down trading for four days after 9/11? Or was that just Wall Street? |
- Dave W. (Private Msg, EMail)
Evidently, I rock!
Why not question something for a change?
Visit Dave's Psoriasis Info, too. |
|
|
|
bngbuck
SFN Addict
USA
2437 Posts |
Posted - 12/27/2007 : 00:17:20 [Permalink]
|
Good God Dude!, what is the stock market other than the potential actions and reactions of large groups of people? Every work day of every week, hundreds of billions of dollars are transferred back and forth between and amongst large (dozens of millions) groups of people! Every single transaction is predicated upon more or less educated guesses as to the potential actions of large numbers of shareholders - the market as a whole - and their actions and reactions to the condition that they perceive the market to be in that day! I have played this game for years, Dave, and that is the essence of it!
I agree as to whether the work being done in Tucson is leading in this direction or not, as being highly problematical today; but Dave and I were merely indulging a little SF speculation, whether or not Asimov would have agreed! After all, he was just a master storyteller - no one of his puny talent can hope to reach the depths of the bathysphere of bullshit that Dave and I roll around in in! |
|
|
|
bngbuck
SFN Addict
USA
2437 Posts |
Posted - 12/27/2007 : 00:39:23 [Permalink]
|
Dave.....
| Hey, who decided to shut down trading for four days after 9/11? Or was that just Wall Street? |
Yeah, that was Wall Street, all right! And it was the Department of Transportation that shut down all the airlines for several days!
And it was the Congress of the United States that gave a cackling, crowing, little rooster cock a permanent hard-on together with the authority to start a debilitating war for Irag's oil - because of their collective fear of offending their constituencies! If there is any threat to their cozy 1 mil a year+ in salary, benefits and perks, they will gladly suck any thing in range that asks for sucking. Even today, with our "Democratic majority"! Yes, and Hillary's lips are pursed as sweetly as Tom DeLay's used to be! Sometime ask me how I really feel! |
|
|
|
dv82matt
SFN Regular
760 Posts |
Posted - 12/27/2007 : 05:56:57 [Permalink]
|
Originally posted by Dave W.
If we can take action to prevent the predicted economic outcome of being able to accurately predict stock values, then the model (algorithm) used to make the predictions was fatally flawed, either due to incompetence or incompleteness. In other words, if the predicted outcome fails to come about, then sigma is necessarily larger than intended. |
There would be an implicit assumption that the model would only accurately describe the outcome when the acting agents do not themselves have knowledge of the model (or where they don't make use of it for some other reason). Knowledge of the model would tend to reduce the usefulness of the predictions but that wouldn't mean the predictions were invalid just that they weren't as applicable to the new situation that developed as a direct consequence.
|
|
|
|
Dave W.
Info Junkie
USA
26020 Posts |
Posted - 12/27/2007 : 07:23:05 [Permalink]
|
Originally posted by dv82matt
Originally posted by Dave W.
If we can take action to prevent the predicted economic outcome of being able to accurately predict stock values, then the model (algorithm) used to make the predictions was fatally flawed, either due to incompetence or incompleteness. In other words, if the predicted outcome fails to come about, then sigma is necessarily larger than intended. |
There would be an implicit assumption that the model would only accurately describe the outcome when the acting agents do not themselves have knowledge of the model (or where they don't make use of it for some other reason). |
Well, that would be covered by the incompleteness I spoke of, but it would also mean that the predictions have largish errors. If I have a model, and buy a stock that the model says will soar, and the model's prediction is 1% low because it didn't take my action into account, then how is that not going to ripple through to other related stocks or even become compounded over time? |
- Dave W. (Private Msg, EMail)
Evidently, I rock!
Why not question something for a change?
Visit Dave's Psoriasis Info, too. |
|
|
|
Dude
SFN Die Hard
USA
6891 Posts |
Posted - 12/27/2007 : 08:45:14 [Permalink]
|
bng said:
| Good God Dude!, what is the stock market other than the potential actions and reactions of large groups of people? |
You are applying to much specificity to what this thing could be capable of.
It might, for example, be able to predict that people will (in large numbers) invest in the market.
It would not, for example, be able to tell you what exact stocks they would buy.
Also, the value of any given stock (as you well know) can be effected by a thousand things that have nothing to do with large groups of people. A small group of insider traders can cause fluctuations. The weather can alter crop futures. And so on and so on.
|
Ignorance is preferable to error; and he is less remote from the truth who believes nothing, than he who believes what is wrong.
-- Thomas Jefferson
"god :: the last refuge of a man with no answers and no argument." - G. Carlin
Hope, n.
The handmaiden of desperation; the opiate of despair; the illegible signpost on the road to perdition. ~~ da filth |
|
|
|
|
dv82matt
SFN Regular
760 Posts |
Posted - 12/27/2007 : 08:55:18 [Permalink]
|
Originally posted by Dave W.
Well, that would be covered by the incompleteness I spoke of, |
All models are incomplete in some sense. They are limited to describing a particular set of phenomena. We may wish to generalize the model or feel that it should not be so limited for asthetic or practical reasons but all models are limited or "incomplete" in what they can predict.
| but it would also mean that the predictions have largish errors. If I have a model, and buy a stock that the model says will soar, and the model's prediction is 1% low because it didn't take my action into account, then how is that not going to ripple through to other related stocks or even become compounded over time? |
If Scenario A where no one uses the model to guide stock trading is what the model is designed to predict then its inability to make equally accurate predictions in Scenario B where the model is used is not a failure of the model but an improper application.
That said if such a model did exist it could probably be generalized fairly easily to arbitrary precision (the limit being the precision of the original model) simply by running it iteratively. |
|
|
|
Dr. Mabuse
Septic Fiend
Sweden
9687 Posts |
Posted - 12/27/2007 : 12:56:00 [Permalink]
|
The Psychohistory as told by Isaac Asimov depended on the participants of history not being aware of active participation by the Guardians of the Future.
The Foundation series of novels by Isaac Asimov has, to me, been the most important sci-fi literature I've read. |
Dr. Mabuse - "When the going gets tough, the tough get Duct-tape..."
Dr. Mabuse whisper.mp3
"Equivocation is not just a job, for a creationist it's a way of life..." Dr. Mabuse
Support American Troops in Iraq:
Send them unarmed civilians for target practice..
Collateralmurder. |
|
|
|
BigPapaSmurf
SFN Die Hard
3192 Posts |
Posted - 12/27/2007 : 13:10:40 [Permalink]
|
| PKD RULZ! [/runs away] |
"...things I have neither seen nor experienced nor heard tell of from anybody else; things, what is more, that do not in fact exist and could not ever exist at all. So my readers must not believe a word I say." -Lucian on his book True History
"...They accept such things on faith alone, without any evidence. So if a fraudulent and cunning person who knows how to take advantage of a situation comes among them, he can make himself rich in a short time." -Lucian critical of early Christians c.166 AD From his book, De Morte Peregrini |
|
|
|
bngbuck
SFN Addict
USA
2437 Posts |
Posted - 12/27/2007 : 20:18:20 [Permalink]
|
Dude.....
Methinks you overreach a bit in this comment:You are applying to much specificity to what this thing could be capable of.
It might, for example, be able to predict that people will (in large numbers) invest in the market.
It would not, for example, be able to tell you what exact stocks they would buy.
Also, the value of any given stock (as you well know) can be effected by a thousand things that have nothing to do with large groups of people. A small group of insider traders can cause fluctuations. The weather can alter crop futures. And so on and so on. |
The stock market, being essentially a closed system, should be subject to Boolean analysis according to complexity theory. Take your piki from the wiki! Until the recent emergence of the computer world into the terabyte era, no computing system on the planet had the computational power to handle a project of this size. The conventional wisdom on the street had always been the random walk hypothesis. Now, it appears, the random walk may turn into a primrose path! This is an abstract of a restricted article on predictive methodology that gives you a feel for the confidentiality surrounding this subject.
The work of Ed Thorp, the guy that developed the famous blackjack algorithm back in the late 60's, appears to be currently directed to the stock market.
He has an investment firm and is definitely picking stocks with a computer program. Rumors have been circulating for some time in investment publications concerning this guy's prowess (he is worth several hundred million, and in just a few years), so I decided to become a client. I'm floating a few thousand in a straight investment account using his picks, gonna give it about six months and, we will see! There are hundreds of picking-program gurus lurking in the alleys behind Broad and Wall streets, but this is the first one I've run into in thirty years of investing that seems to have real technical chops. We will see!
|
|
|
 |
|
|
|
Psychohistory for real?
