Free Republic
Browse · Search
News/Activism
Topics · Post Article

Skip to comments.

Limitations of computer models [system error for quants]
Financial Times ^ | 8/14/2007 | Gillian Tett and Anuj Gangahar

Posted on 08/14/2007 9:25:31 PM PDT by bruinbirdman

click here to read article


Navigation: use the links below to view more comments.
first 1-2021-23 next last

1 posted on 08/14/2007 9:25:34 PM PDT by bruinbirdman
[ Post Reply | Private Reply | View Replies]

To: bruinbirdman

25-standard deviation event” – something that only happens once every 100,000 years or more.


Yeah, right. Some $500K/year PhD trying to cover his ass.


2 posted on 08/14/2007 9:29:14 PM PDT by rbg81 (DRAIN THE SWAMP!!)
[ Post Reply | Private Reply | To 1 | View Replies]

To: bruinbirdman; ShadowAce

Lost $1.5bn? Blame the computer.


3 posted on 08/14/2007 9:31:40 PM PDT by stainlessbanner
[ Post Reply | Private Reply | To 1 | View Replies]

To: stainlessbanner
Models? Computerized trading?

What ever happened to the human brain?
Live by the sword..

4 posted on 08/14/2007 9:34:49 PM PDT by Publius6961 (MSM: Israelis are killed by rockets; Lebanese are killed by Israelis.)
[ Post Reply | Private Reply | To 3 | View Replies]

To: bruinbirdman

What’s happening is pretty simple: The stock market is in a closed feedback loop with the models. What’s that mean?

Take the stock market from its start up until today. You can make a great model that “predicts” every event in the whole thing based on past data for each point. Use that to predict the future. Won’t work long. Why? Because a lot of other big players have similar models, which in turn change the way the market operates, which eventually invalidates those models due to various types of feedback.

In other words, the 25-SD events don’t mean the market’s broken, they mean the market’s changed due to the feedback effect.

In fact, it’s mathematically provable that you can NEVER build a complete model that will work out into the future for this type of system because once you do (or rather get close to doing so) and it becomes part of the system, the feedback will force an out-of-model behaviour.

(Maybe not well stated, but I’m tired — it’s Miller time.)


5 posted on 08/14/2007 9:38:05 PM PDT by piytar
[ Post Reply | Private Reply | To 1 | View Replies]

To: bruinbirdman

Computer models are only as good as the data they’re based on. GIGO (garbage in, garbage out).

So what do these bright young bucks do? Leave the computers on auto-pilot while they summer in the Hamptons?


6 posted on 08/14/2007 9:38:10 PM PDT by LibFreeOrDie (L'Chaim!)
[ Post Reply | Private Reply | To 1 | View Replies]

To: Publius6961
Ripe for hackers I would imagine.

Well, every 100,000 years or so.

“People say these are one-in-a-100,000-years events but they seem to happen every year...”

Kinda Biblical. "100,000 years is as 365 days"

This stuff is over my head...

7 posted on 08/14/2007 9:38:45 PM PDT by Syncro
[ Post Reply | Private Reply | To 4 | View Replies]

To: piytar

I’d say you’ve pretty-much nailed it. Well done.


8 posted on 08/14/2007 9:40:10 PM PDT by Petronski (imwithfred.com)
[ Post Reply | Private Reply | To 5 | View Replies]

To: piytar
Last para should have been:

In fact, it’s mathematically provable that you can NEVER build a complete model that will work infinitely far (or even really far) out into the future for this type of system because once you do (or rather get close to doing so) and it becomes part of the system, the feedback will eventually force an out-of-model behaviour.

9 posted on 08/14/2007 9:41:21 PM PDT by piytar
[ Post Reply | Private Reply | To 5 | View Replies]

To: bruinbirdman

Breaks down to computers can’t anticipate real life.


10 posted on 08/14/2007 9:41:54 PM PDT by TASMANIANRED (Taz Struck By Lightning Faces Battery Charge)
[ Post Reply | Private Reply | To 1 | View Replies]

To: Petronski

Thanks...


11 posted on 08/14/2007 9:42:05 PM PDT by piytar
[ Post Reply | Private Reply | To 8 | View Replies]

To: bruinbirdman; All
... something that only happens once every 100,000 years or more.

But happened real quick-like. As they say, back to the drawing board.

12 posted on 08/14/2007 9:47:10 PM PDT by dighton
[ Post Reply | Private Reply | To 1 | View Replies]

To: piytar

Further compounding the model problem is that while it is a closed feed back loop, it is an open system and influenced by external factors. The inputs of today are not going to be the same inputs as tomorrow.

I’m not totally against computer model ... but business types and often computer types don’t understand the limitations of the systems like math types.... and there is the first step to a failure.


13 posted on 08/14/2007 9:49:33 PM PDT by taxcontrol
[ Post Reply | Private Reply | To 5 | View Replies]

To: piytar

What you said reminded me of Goedel.


14 posted on 08/14/2007 9:52:00 PM PDT by Moonman62 (The issue of whether cheap labor makes America great should have been settled by the Civil War.)
[ Post Reply | Private Reply | To 5 | View Replies]

To: piytar

Your explanation could be the reason “why” the article speaks to neural networks, whihc would seem to me to mean that all the models are interconnected and marching to the same song.


15 posted on 08/14/2007 9:58:55 PM PDT by padre35 (Conservative in Exile.)
[ Post Reply | Private Reply | To 5 | View Replies]

To: bruinbirdman

“The dog ate my homework” will now be “the computer ate my billion.”


16 posted on 08/14/2007 10:07:05 PM PDT by TrueKnightGalahad (Your feeble skills are no match for the power of the Viking Kitties!)
[ Post Reply | Private Reply | To 1 | View Replies]

To: piytar
Looks like that 25D got sucked into the norm again. Deja vu? The confusers are so fast, they probably don't know what a 25D is. It is part of the norm in nano seconds.

"Medallion, has had an annual return of 30 per cent since 1988 – is suffering badly from recent movements."

30%/yr for 19 years leaves quite a profit to take a hit on.

So, it takes my house 5 years to double in value, free money. I am not sad when the market price falls 12% in 3 months. yitbos

17 posted on 08/14/2007 10:09:23 PM PDT by bruinbirdman ("Those who control language control minds." -- Ayn Rand)
[ Post Reply | Private Reply | To 5 | View Replies]

To: piytar

Why of course. The stock market, after the original sales of the stock, is a game of trying to outbet everybody else in real time. Earlier buyers hope to make money off of later buyers (because, of course, they expect to sell sometime), the winners make money off the losers.


18 posted on 08/14/2007 10:14:23 PM PDT by HiTech RedNeck
[ Post Reply | Private Reply | To 5 | View Replies]

To: bruinbirdman

“It tends to be more consistent, since machines – unlike people – never get tired.”

Wrong. Computers don’t have emotions and their trading is strictly based on Technical Analysis. The normal trader will allow emotions in and sell when he should be buying or vice versa. A good trader sets the rules and should be like a machine in carrying them out.


19 posted on 08/14/2007 10:19:30 PM PDT by jwh_Denver (http://www.youtube.com/watch?v=1k08yxu57NA&NR=1)
[ Post Reply | Private Reply | To 1 | View Replies]

To: bruinbirdman

Self referencing systems are generally chaotic - nearly anything can happen at any given moment. (”Godel, Escher & Bach”). I guess the rich madmen/institutions using emancipated computers for trading are like nukes with random timers in the center of our economy. We had better slow them down! Let humans and computers trade in entirely different markets, with no cross overs. I know, cannot happen, but it would be interesting to see what the outcome would be.


20 posted on 08/14/2007 10:23:08 PM PDT by GregoryFul (how'd that get there?)
[ Post Reply | Private Reply | To 1 | View Replies]


Navigation: use the links below to view more comments.
first 1-2021-23 next last

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.

Free Republic
Browse · Search
News/Activism
Topics · Post Article

FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson