Prediction Company Performance

[Maggette]

I was asked about their performance and didn't find anything on the web. Any information about their performance available?

Thanks

[FDAXHunter]

No, because they got folded into UBS a long time ago. I doubt that even they have a "performance" in the sense of a track record that's easily reconcilable.

[Maggette]

Yeah , I know. But since they got their own web page and describe themself as localy managed and I heard they get advertised by the UBS as a product I thought there might be some information floating around. I'd be happy with "information" on "rumor" level;)....don't need a real track record. I am just curious.



Maybe they are just a marketing instrument for the UBS (hey guys, look how quant we are...).



Thanks anyway

[rod]

I wonder if I am the only one who finds it very intriguing that the only sources of information on the Prediction Company are articles on Wired and the New Yorker:



Cracking Wall Street (1994)



Black Box (1999) [pdf]



plus Thomas Bass' distasteful book, The Predictors. The PC founders are in the "complex and adaptive systems" cabal, which is full of intellectual charlatans. They're also on edge.org, which seems to be little more than a massive circle-jerk. I hate belittling other people's work, but PC looks more like a marketing outfit than a quant trading house. The fact that the founders, Farmer & Packard, left PC long ago is somewhat revealing, I'd say.

[Maggette]

I have no idea. I like Doyne Farmer and his research papers. But I am no practitioner and I doubt my own qualification to judge if his stuff has any practical value (beyond entertaining me).



Also Farmer has always been quite open that even though he has Dynamic System Theory/ Chaos Theory backround, his work at prediction company was pure statistics, IT and some machine learning.



Still I won't reject the concept of PC beeing a marketing outfit (or maybe some blue sky research).



Some of the PC veterans ae still in finance I guess....I think the pipeline buy side trading stuff is a spin off from the Bios Group...(quite ridicolous web presence though: http://www.pipelinetrading.com/).



Maybe all a bunch of clowns. Maybe not. I don't know.

[cowpoke]

Some years ago somebody at UBS gave me a copy of The Predictors. (A great read for anybody who enjoys superficial stereotypes. To be fair, it's not all fiction.) The UBS push on me was all marketing - looky, looky, we have been in the quant biz for years. I asked about investing in the "Prediciton Fund" and was told there is no such fund available for investment. I've heard from a few sources (including some UBS folks, and maybe the book) that UBS was more interested in the trading platform technology than the quant work. I've vaguely known a couple of folks that have gone in and out of PC, and it would surprise me if they were involved with running a fund.



@rod: If you are suggesting that Farmer is an "intellectual charlatan", I think that is unfair. Posterity will decide whether any of his work has lasting value, but charlatan suggests a motive which I doubt exists. The edge.org link pops up Steven Pinker, hardly a charlatan.

[rod]

For the record, I did not say that Doyne Farmer is an "intellectual charlatan". I like some of the papers he wrote in the 1980s while at Los Alamos, and also some of his more recent papers at the Santa Fe Institute. But I am an academic, so I can't assess their usefulness in the "real world", so to say.



Unfortunately, there's a lot of pseudo-science in the field of complex systems, which tends to diminish the original, quality work some have done in the area. Farmer is praised quite a bit at edge.org, an interesting site that gathers high-caliber thinkers such as Gell-man, Dyson, Dawkins, Pinker, and... Taleb!?! How can one take their words seriously when they lump together true geniuses and intellectual lightweights?



From the little I read about PC, I got the impression that Farmer left Los Alamos due to the lack of funding as the Cold War came to an end, he founded PC in order to make some money in order to finance his own projects, worked at PC for almost 10 years, allegedly made some cashish, then went back to complex systems research (this time at SFI instead of LANL). I would have more respect for Farmer if he hadn't allowed Thomas Bass to write pop-sci books about him, though.

[jslade]

Hey Rod: I agree with you that a lot of the edge dudes are pop science charlatans, and the whole thing is a nerd circle jerk. I'd include Dawkins and possibly Pinker in this category -the things they put into books as "science" are unconscionable. Memes? "Selfish" genes? Gimme a break!



I think Farmer is a respectable scientist, but he does have a big streak of Barnum about him, in part because of his old college bong chum, Clarence Bass (who is an atrocious writer) keeps writing books about how awesome Farmer is. He's somewhat modeled his career after Ed Thorp, which is admirable, but a hard act to follow. Apparently his Roulette project wasn't so good, and I'm guessing PC was much less successful than Princeton Newport. Since they were a support group for a Derivatives team, yeah: it's probably more of a trading platform rather than some kind of awesome alpha finder.



Santa Fe did produce some neat stuff in the 80s/90s, but a lot of it seems to have been dead ends. KNN was the main tool that seems to have come of it.

[hedgeQuant]

@ Maggette



In what context did you need the information you were looking for? I might be able to help you to some extent.

[Maggette]

Simple curiosity. No business or academic need. Any info yould be appreciated....but if not, it is absolutely no drama Smiley

Thanks



Edit: I decided to like PC for the simple fqct that they mentioned as donators for the R project.