SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
MCarreira
Topic Author
Posts: 1724
Joined: July 14th, 2002, 3:00 am

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 8:35 am

FT - Gillian TettQuote...What really damaged the financial system in recent years was not so much “maths” or “economics”; instead the crucial problem was bad maths (and economics) that was used and abused....
 
User avatar
daveangel
Posts: 17031
Joined: October 20th, 2003, 4:05 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 8:42 am

gillian tett is getting to be quite boring ... such little knowledge taken to such great lengths.
knowledge comes, wisdom lingers
 
User avatar
Paul
Posts: 10791
Joined: July 20th, 2001, 3:28 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 9:33 am

I find the FT generally bad at risk, derivatives, credit, etc.Interesting phrase used by Gillian Tett, maybe she'd like to read "The use, misuse and abuse of mathematics in finance" Royal Society Science into the Next Millennium: Young scientists give their visions of the future Phil Trans 358 63, published in 2000, by one P Wilmott. This is the paper with the great line "It is clear that a major rethink is desperately required if the world is to avoid a mathematician-led market meltdown."I was invited to last weekend's inaugural meeting of the Institute for New Economic Thinking, funded by Soros, at King’s, Cambridge that Tett mentions. My 'takeaways' from this event were not as optimistic as hers: There wasn’t much that seemed to be exactly new There was more than a whiff of arrogance about many of the speakers. Very little in the way of self doubt, they really do still believe all this stuff One of the speakers clearly thought he’d ‘solved’ all of econometrics. I hope no one believes him otherwise I see this leading to another mess down the line Physics envy is still prevalent. The biggest laugh was when one speaker said how physicists just had to model atoms and forces, and that economics was much harder. Well, yes, if you are aiming for the perfect model. But that’s just plain silly. To me the most important point that should have come out of this conference, but didn’t, was to acknowledge the limitations of economics, and, having made that acceptance, to decide what on earth economics can do that is useful and productive Nor was there any talk of robustness of modeling. Robustness is fundamental to the whole purpose of economics in my view Some of the speakers were pro maths and some con. On a positive note, a couple of speakers did comment that economists do only know a small subset of mathematical methods. A point I’ve made about quants as well Observing so many economists at close hand I have to say that I wouldn’t lend any of them my money. They are all far too keen on defaulting as the solution to any number of problems. And they like inflation a lotP
Last edited by Paul on April 15th, 2010, 10:00 pm, edited 1 time in total.
 
User avatar
rmax
Posts: 6080
Joined: December 8th, 2005, 9:31 am

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 10:43 am

QuoteOriginally posted by: Paul..... Observing so many economists at close hand I have to say that I wouldn’t lend any of them my money. They are all far too keen on defaulting as the solution to any number of problems. And they like inflation a lotPThis is a very worrying point.
 
User avatar
Paul
Posts: 10791
Joined: July 20th, 2001, 3:28 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 10:58 am

Yes, they really perked up during this topic of conversation. It was as if all this was beautifully theoretical, and wouldn't it be absolutely fascinating to watch while this country or that defaults. The ivory-tower thing. P
 
User avatar
exneratunrisk
Posts: 3559
Joined: April 20th, 2004, 12:25 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 11:15 am

Who is a mathematician?One who understands what can be described in the language of mathematics and what not?One who understands what a model is - in relation to the mathematical object level?One who understands that models and their algorithmic representations need to be organized and validated orthogonally?One who understands that axiomatic and algorithmic mathematics is not the same?One who has met many special functions "personally", from Sine to the Meijer G, and can manipulate them virtuously?One who says: mathematics without computers is worthless?One who says: it is all numbers?One who corrects Wittgenstein: the world is everything that has a probability to happen?One who can construct models from petabytes of data?One who says: mathematics is only data compression?What of that is relevant to finance?Where are the traps?What does robustness mean in this context?
 
User avatar
daveangel
Posts: 17031
Joined: October 20th, 2003, 4:05 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 11:26 am

QuoteOriginally posted by: PaulYes, they really perked up during this topic of conversation. It was as if all this was beautifully theoretical, and wouldn't it be absolutely fascinating to watch while this country or that defaults. The ivory-tower thing. Pif you are lending anyone money be prepared for default.There is nothing wrong with defaulting - it clears problems, brings new specialist investors to a field, clears balance sheets, resets risk premia. companies in default have greater room to manoeuvre especially if there are trade unions involved.
knowledge comes, wisdom lingers
 
User avatar
Paul
Posts: 10791
Joined: July 20th, 2001, 3:28 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 11:39 am

...especially when it's mathematically optimal P
 
User avatar
Paul
Posts: 10791
Joined: July 20th, 2001, 3:28 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 11:41 am

The high point of the INET conference was during a lunch. Dominique Strauss-Kahn, Managing Director of the International Monetary Fund, gave a speech which was nicely interrupted by protestors.P
 
User avatar
frenchX
Posts: 5911
Joined: March 29th, 2010, 6:54 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 2:39 pm

Did they talk about a higher regularization of the financial operations ? Or even the creation of an institutionnal supervisor?After the subprime crisis, there were a plenty of talks about financial control but it becomes more and more discrete by now.The point that really surprised me in your brief sum up Paul is the "physics envy" one. My question is maybe very naive but why those people are in competition with physicists ? I mean they earn probably far much more than any scientist and have more political power. I term of social gratitude traders are considered as money hungers and physicists as lazy nerds so we are even . In my case, I'm a physicist and I really believe that some very good collaborations could be made. Of course quants used models developped by their own or by physicists but maybe physicists could also use some models develop by quants .
Last edited by frenchX on April 15th, 2010, 10:00 pm, edited 1 time in total.
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 3:31 pm

I think 'physics envy' refers to the sublime math and repeatable experiments which physics enjoys. That is economists envy the predictability of many physical systems.The problem is that economists don't study elementary particles but something rather more complex and non-stationary. Physics enjoys the fact that a proton in France behaves like a proton in China and a proton in the year 2010 behaves identically to a proton from 2000 or 1929 or 3000 BC for that matter. The same cannot be said for people, corporations, or government.Of course, the deeper problem is that protons don't profit from clever marketing, advancing technology, regulatory arbitrage, or committing fraud. That is, the protons don't try to outsmart the physicists. In contrast, economist face 6.6 billion people, many of whom are trying to outsmart the economists and make some extra profit. Many of these out-smarters join together to have rather more collective brain-power, computers, and assets than the economists.
Last edited by Traden4Alpha on April 15th, 2010, 10:00 pm, edited 1 time in total.
 
User avatar
hamster
Posts: 216
Joined: October 12th, 2008, 3:51 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 6:40 pm

i heard that INET was just an opportunity for economists chumming up with Sorros in order to get some funding.ok i will chum up now. here is my "great theory of disaster": (i) observation: market participants believe that writing thresholds in business contract (e.g. stop limits, options, oc tests, etc.) increase the certainty of the securitized payouts. (ii) consequence: thus these market participants feels save to hold a fraction of non-defaultedable claims to a certain extent that can be deposited as security in order to borrow money to finance something else. (iii) reasoning: as market participants takes threshold levels as reference point to guess a second moment, they become blind to consider scenarios lower than thresholds, and thus are vulnerable to such events. (iv) interpretation: from economic point of these market participants violate the condition of being an economic agent who is described as trying-to-estimate-(bounded)rationally-the-economic-profit-of-his/her-(investment)strategy.... bla bla bla... blabla ..blabla blaa...Dear Mr Sorros, please transfer all the INET funding to my bank account XXXXX... ok, i will bore you with another one, that i call "the short circuit within the fundamental theorem":(i) economic agent X_1 is assessing the value of assets (set B) by using RNPs some another assets (set A)(ii) as the asset values of set A might be valued by agent X_2 using RNPs beforehand. thus agent X_1 is assessing with the RNPs of RNPsresult 1: as agent X rely on "correct" absolute valuations of other kinds of agents Y, a tiny mispricing in set A caused by Y amplify to mispricing in set B while using RNPs.(iii) if you relax the assumption that there just two agents X_1 and X_2, then many X_i will assess sets A_i and B_i.(iv) as a result set A and B used in (i) and (ii) are no disjoint sets anymore (i assumed X_1 is assigned to set B, and X_2 to set A). (v) Thus agents X_i will assess wildly assess their sets A_i with RNPs by using sets B_j (i not j) that where assessed by agents X_j using RNPs, who in turn used ... etc.result 2: if many agents using the fundamental theorem, or relative pricing, circular references emerge automatically (an electrician would say: short circuit). result 3: it might happen that agents Y (or anyone else using absolute pricing) vanish, and no agent X_i recognize, that the assumption/requirement of using fundamental theorem vanished with these agents Y.
Last edited by hamster on April 15th, 2010, 10:00 pm, edited 1 time in total.
 
User avatar
trackstar
Posts: 27450
Joined: August 28th, 2008, 1:53 pm

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 6:52 pm

When you are approaching a potential donor for assistance, one thing that you need to get right is his or her name.SOROS
Last edited by trackstar on April 16th, 2010, 10:00 pm, edited 1 time in total.
 
User avatar
ChicagoGuy
Posts: 455
Joined: April 13th, 2007, 1:45 am

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 10:10 pm

Perhaps the problem is with the calibration of the models?I feel like there is a giant leap from the theoretical part to the implementation of a model. Thats when most of the hand waving begins, and probably where finance and math finance most differ. Perhaps it's really impossible to implement most math finance models correctly. Remember how the "simple" B-S model is calibrated to market data: a complete model is calibrated to a market that is really incomplete. Which risk neutral measure should one pick? I think there are many unanswered questions on how a model should be calibrated, but maybe I am not looking in the right places for the answers.
 
User avatar
spacemonkey
Posts: 443
Joined: August 14th, 2002, 3:17 am

FT - Mathematicians must get out of their ivory towers

April 16th, 2010, 11:34 pm

A bunch of banks, who thought they could consistently beat the market, blew up wiping out all the anomalous profits they made. And this is evidence that the EMH is flawed?
ABOUT WILMOTT

PW by JB

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...


Twitter LinkedIn Instagram

JOBS BOARD

JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...


GZIP: On