Serving the Quantitative Finance Community

 
User avatar
GA
Topic Author
Posts: 0
Joined: August 6th, 2001, 9:11 pm

Quant Finance with Complexity

August 6th, 2001, 9:39 pm

Hi list - I have been reading some work from Santa Fe's Doyne Farmer on modeling the financial market using very simple physical characteristics, referring to his agent-based model. Using simple simulations, he was able to reproduce certain statistical properties of financil time series. It helps me to appreciate quant finance beyond pdf and stochastic calculus. Unfortunately, it seems to me that so many advanced quant finance books never mention current research in complexity field applying to finance. Searching Amazon.com's database, a thorough technical book on complexity in the context of finance a rare find. Any suggestions and pointers to read on complexity in finance? - Ga
 
User avatar
Paul
Posts: 6604
Joined: July 20th, 2001, 3:28 pm

Quant Finance with Complexity

August 9th, 2001, 7:03 am

I don't know of any text books but there are plenty of finance papers out there. Also take a look at one of the books we review (Olsen et al). Although there is no direct modelling in that, it has some useful pointers.Ten years ago Guy Kember and I got very excited about modelling market microstructure. You know, different classes of economic agents interacting leading to interesting dynamics and chaotic behavior that mimics randomness. Our conclusions were that it was pretty pointless, and ultimately very boring. It wasn't difficult to get nice dynamics but the modeliing was totally arbitrary (insensitivy to initial data!!!). I've known quite a few people who claim to have found the perfect predictive model based on some of these ideas. The story always goes like this:1. Person gets dirty financial data set.2. They reckon they can use the techniques they've used previously (modelling height of the Nile, for example) to model their financial data.3. They find a model that predicts FX rates a few minutes in advance.4. They get lots of publicity or seek research funding or both.5. When tested 4 real it falls apart.6. They go back to modelling the height of the Nile. I've yet to read the autobiography "How Market Microstructure Modelling Made Me My First Million." But that's just cynical ol' me!P
 
User avatar
Aaron
Posts: 4
Joined: July 23rd, 2001, 3:46 pm

Quant Finance with Complexity

August 9th, 2001, 2:15 pm

I was introduced to the Black-Scholes model in 1980 and told that it was the same equation as classical heat transfer in physics. I got a brilliant idea. Real financial markets work in discrete units (ticks and trades), so does quantum physics. Why not read up on quantum heat transfer and come up with a major breakthrough.I learned a little quantum physics, a fair amount of topology and something about heat transfer. Nothing about option pricing. I repeated the experience, more or less, investigating Mandlebrot's "sheep and wolves" paradigm for crashes.My point is that clever ideas and elegant mathematics rarely combine to produce profit. Fun and education, yes. But progress in finance has been made by people who started with a problem, not with a tool or a vision of how the world should be. Finance is real, real people, real money. Progress in finance contributes to human happiness by giving individuals more control over their economic lives. People who show up to make a quick buck by using a cleverer equation to extract money from less-sophisticated people are not contributing to anything worthwhile, and lose far more often than win. Over the years I have become less and less tolerant of people who push radical financial ideas but have no taste for nor experience in finance. They're like people who will explain how the entire world should be governed, but can't get along with anyone they actually know.Doyne Farmer in particular burned through a lot of Swiss money buying fancy computer equipment. He published some interesting work and apparently had a lot of fun. I read his stuff for pleasure and education. He would probably be a good guy to know. But the FX markets ignored him.
 
User avatar
gammashark
Posts: 0
Joined: August 10th, 2001, 12:34 am

Quant Finance with Complexity

August 10th, 2001, 12:44 am

I am not sure this is still the case, but I thought the Prediction company (the commerical off-shoot of the complexity research at Santa Fe) was very successful. If not very successful, then at least successful enough to continue to have UBSWarburg buying their signalsm and a bit of their company.The following is from the Prediction company website: www.predict.com News and Press Releases2/2001 - Still Expanding!Prediction Company has multiple open positions. See the Employment pages. 3/24/2000 - UBS AG Purchases 24.9% of Prediction Company LLCCompanies also extend exclusive relationshipSANTA FE, New Mexico and STAMFORD, Conn., March 24, 2000 - Warburg Dillon Read and Prediction Company of Santa Fe, New Mexico, today announced the purchase by UBS AG of a 24.9% minority interest in Prediction Company LLC.UBS and Prediction Company have also agreed to extend their contractual relationship, under which Prediction Company LLC is developing technology that is deployed under an exclusive license to UBS AG and its Warburg Dillon Read investment banking division.Prediction Company is a privately held company based in Santa Fe, New Mexico, that was founded in 1991 by Doyne Farmer, Norman Packard, and James McGill to build models for predicting financial markets. The technology developed for this purpose includes substantial research and data infrastructure, along with technology for high-speed trade execution in real time.Norman Packard, President and CEO of Prediction Company LLC stated, "We welcome this investment by UBS. The investment, as well as the continuation of our relationship, will enable our company to realize the potential of its technology in a broad array of markets, with access provided by Warburg Dillon Read's strong global platform."David Zyer, Managing Director of Warburg Dillon Read's Prediction related business area, noted that "This investment further solidifies Warburg Dillon Read’s access to the world-class technology and risk management methods being developed and refined by Prediction’s research scientists and systems engineers. This exclusive relationship also enhances WDR's expertise in the deployment of sophisticated data infrastructure and electronic execution technologies."The relationship between Prediction Company and Warburg Dillon Read began in 1992 as a relationship between Prediction Company and O'Connor and Associates, which was integrated into Swiss Bank Corporation in 1995. In 1998, SBC merged with Union Bank of Switzerland to form UBS AG.Warburg Dillon Read is the investment banking division of UBS AG. In the United States, Warburg Dillon Read’s securities activities are conducted through Warburg Dillon Read LLC, a U.S.-registered broker-dealer. Warburg Dillon Read was created on June 29, 1998, when Swiss Bank Corporation and Union Bank of Switzerland merged to form UBS AG, one of the largest banking groups globally, with more than $1 trillion under management. While Warburg Dillon Read's origins and culture are European, the orientation is truly global, with a strong presence in the Americas and Asia Pacific. Warburg Dillon Read is a market leader in risk management, financial structuring, fixed income issuance and trading, foreign exchange dealing and derivatives, and maintains leading positions in corporate finance and equities throughout Europe, the Americas and Asia Pacific. ###Enquiries:Jim Galvin: +1 212-821-5503, of Warburg Dillon ReadNorman H. Packard 505-984-3123, of Prediction Company Text
 
User avatar
Hamilton
Posts: 1
Joined: July 23rd, 2001, 6:25 pm

Quant Finance with Complexity

August 10th, 2001, 1:50 am

That press release says a 24% minority interest but nowhere does it mention what they paid for it.Perhaps nothing?Cynical ol' me.
 
User avatar
Hamilton
Posts: 1
Joined: July 23rd, 2001, 6:25 pm

Quant Finance with Complexity

August 10th, 2001, 1:52 am

I read Mandelbrot's book on Complexity, and, although he has a very interesting writing style and some interesting theoretical observations, found it useless for actually doing anything applied.<sigh>
 
User avatar
Lennart
Posts: 4
Joined: August 15th, 2001, 11:26 pm

Quant Finance with Complexity

August 17th, 2001, 1:57 pm

I might be really far off topic here... (All I know is what I read in an article by Mandelbrot in Scientific American some time ago - "A random walk..." something I think). But aren't these ideas somehow related to stochastic volatility? Perhaps with Poisson jumps... Perhaps not of much practical interest, but theory is also fun...
 
User avatar
Paul
Posts: 6604
Joined: July 20th, 2001, 3:28 pm

Quant Finance with Complexity

August 17th, 2001, 2:04 pm

The chaos/dynamical systems idea is that completely deterministic systems (i.e. ones where the equations have no randomness in them) give results that look random. So a chaotic system could ending up looking just like what you've mentioned, stoch vol with jumps, say. The trick is to finding the correct deterministic system for your problem. And, of course, in finance that is going to be very, very complicated. Even then the problem with chaos is that you may only have predictability for a very short period of time before the 'sensitivity to initial conditions' kicks in. So the big question is how useful is the idea in practice?P
 
User avatar
gammashark
Posts: 0
Joined: August 10th, 2001, 12:34 am

Quant Finance with Complexity

August 17th, 2001, 3:22 pm

Paul,I am sure that is why everyone looks at FX series to start - a frenetic market (cash and 3m anyway) where even a very short term advantage could be turned to profit... at the right shop of course. That is my (limited, I admit) understanding of what the Prediction company does: it provides signals to UBS, the majority of which are FX. One hears rumours about eurodollars being subject to similar sorts of analysis (inevitably at a 'Japanese Bank'), but not having laid eyes on one yet, I remain doubtful.A couple of years ago, a Sunday Times article in Britain highlighted a company called Neuralytics, run by Frank Dunn, who claimed they had very good short term predictions of the S&P 500. His partner was a slightly mad, reclusive South African mathematician. I haven't heard much out of them since, but perhaps other forum members have had chance to encounter them?
 
User avatar
Chukchi
Posts: 0
Joined: December 15th, 2001, 3:43 am

Quant Finance with Complexity

March 11th, 2002, 11:17 pm

Here is a 'new' Almanac: Beyond Equilibrium and EfficiencyEdited by J. Doyne Farmer and John GeanakoplosOxford University Press, September 2002, ISBN: 0195150945 Table of ContentsPreface, J. Doyne Farmer and John GeanakoplosMoney and Goldstone Modes, Per Bak, Simon F. Norrelykke, and Martin SchubikPower Laws in Economics and Finance: Some Ideas from Physics, Jean-Phillip BouchaudScaling in Financial Prices: I. Tails and Dependence, Benoit MandelbrotScaling in Financial Prices: II. Multifractals and the Star Equation, Benoit MandelbrotMultifractal Returns and Hierarchical Portfolio Theory, J.-F. Muzy, D. Sornett, J. Delour, and A. ArneodoFinancial Markets as Nonlinear Adaptive Evolutionary Systems, Cars HommesFrom Minority Games to Real Markets, Damien Challet, A. Chessa, M. Marsili, and Y-C. ZhangTowards Evolutoinary Game Models of Financial Markets, Daniel FriedmanStatistical Mechanics of Asset Markets with Private Information, Johannes Berg, Matteo Marsili, Aldo Rustichini, and Riccardo ZecchinaOn a Universal Mechanism for Long-Range Volatility Correlations, Jean-Phillipe Bouchaud, Irene Giardina, and Marc MzardEmpirical Properties of Asset Returns: Sylized Facts and Statistical Issues, Rama ContWhat Good is a Volatility Model?, Robert Engle and Andrew J. PattonCorrelated Adaptation of Agents in a Simple Market: A Statistical Physics Perspective, Juan Garrahan, E. Moro, and D. SherringtonPrice Fluctuations, Market Activity, and Trading Volume, Vasiliki Plerou, Parameswaran Gopikrishnan, Xavier Gabaix, Lus A. Nunes Amaral, and H. Eugene StanleyHigh-Frequency Cross-Correlation in a Set of Stocks, Giovanni BonannoA Builders Guide to Agent-Based Financial Markets, Blake LeBaronIndex
 
User avatar
Omar
Posts: 1
Joined: August 27th, 2001, 12:17 pm

Quant Finance with Complexity

March 12th, 2002, 12:25 am

Beyond Equilibrium and EfficiencyHave you read anything in this book that is 'solid'?"Money and Goldstone Modes"I really detest it when physicists talk like that.
 
User avatar
Chukchi
Posts: 0
Joined: December 15th, 2001, 3:43 am

Quant Finance with Complexity

March 12th, 2002, 12:50 am

Here it isCondensed Matter, abstract http://xxx.lanl.gov/abs/cond-mat/0009287Money and Goldstone modesAuthors: Per Bak, Simon F. Norrelykke, Martin ShubikComments: 7 pages, 3 figuresSubj-class: Statistical Mechanics; Adaptation and Self-Organizing SystemsWhy is ``worthless'' fiat money generally accepted as payment for goods and services? In equilibrium theory, the value of money is generally not determined: the number of equations is one less than the number of unknowns, so only relative prices are determined. In the language of mathematics, the equations are ``homogeneous of order one''. Using the language of physics, this represents a continuous ``Goldstone'' symmetry. However, the continuous symmetry is often broken by the dynamics of the system, thus fixing the value of the otherwise undetermined variable. In economics, the value of money is a strategic variable which each agent must determine at each transaction by estimating the effect of future interactions with other agents. This idea is illustrated by a simple network model of monopolistic vendors and buyers, with bounded rationality. We submit that dynamical, spontaneous symmetry breaking is the fundamental principle for fixing the value of money. Perhaps the continuous symmetry representing the lack of restoring force is also the fundamental reason for large fluctuations in stock markets.
 
User avatar
Omar
Posts: 1
Joined: August 27th, 2001, 12:17 pm

Quant Finance with Complexity

March 12th, 2002, 4:09 am

Money and Goldstone modesThe title is a lot sexier than the contents. That's what puts me off.
 
User avatar
pjakubenas
Posts: 0
Joined: March 26th, 2002, 10:14 am

Quant Finance with Complexity

March 28th, 2002, 6:51 am


Money and Goldstone modes
Authors: Per Bak, Simon F. Norrelykke, Martin Shubik
Comments: 7 pages, 3 figures
Subj-class: Statistical Mechanics; Adaptation and Self-Organizing Systems

I've read it several times, couldn't do heads or tails of it.
Can anybody explain what it was supposed to mean?

Pjakubenas


 
User avatar
amacrdec
Posts: 0
Joined: June 10th, 2002, 6:16 pm

Quant Finance with Complexity

June 10th, 2002, 9:34 pm

Hello,Just read through some of the messages posted on this site and I would like introduce myself as the "slightly mad" mathematicial from South Africa that has patanted the non-linear neural networks that can predict price movements to a very high degree of accuracy. If you need to get a hold of me please feel free to do so at [email protected]. Andranik Macrdechian