An interesting question on Slashdot: What is the future for Financial Mathematics?

[Supercilious]

Re: An interesting question on iTulip: What is the future for arithmetic?

My outrage is aimed at the accountant-charlatan putting society at risk using arithmetic. We lost trillions of taxpayer dollars and the world financial system is in danger because of arithmetic. And our kids are brainwashed with this stuff starting from kindergarten. It does not add up. :rolleyes:

I think you are right. Misused mathematics was an excuse and cover for fraudulent profit extraction.

[allenjs]

Re: An interesting question on iTulip: What is the future for arithmetic?

My outrage is aimed at the accountant-charlatan putting society at risk using arithmetic. We lost trillions of taxpayer dollars and the world financial system is in danger because of arithmetic. And our kids are brainwashed with this stuff starting from kindergarten. It does not add up. :rolleyes:

I'm very pessimistic about the chances that we'll be able to rid ourselves of these charlatans, for two reasons.

First -- this is how the last few generations have been indoctrinated. We've brainwashed the kids to believe that numbers and mathematical models are the only thing that can be trusted, so we shouldn't be surprised that they stick to what they were taught. Who will teach future generations not to place too much faith in models? I don't see anyone stepping up to the plate, so it's foolish to assume things will get better. In fact, materialistic reductionism has become the new world religion, and is rapidly encroaching on territory once reserved for religions.

Second -- people seem incapable of learning from reality. Nobody paid attention to Mandelbrot when he said for years that the models were not worthy of our faith. He's like the guy who said that the king's statue was made of clay, and nobody listened. When Enron blew up, you would have expected people to start questioning their models, but it seems to have just led people to cling to the models even more. When LTCM blew up, you really would have expected people to learn.

Now that the entire freakin' world economy has blown up, we sit around and tell ourselves, "NOW people will see how powerless their models are!". We suppose that these people, so religious about "empirical" data, will at last respond to the empirical fact that the whole world blew up. In my opinion, that is just wishful thinking. I see very little evidence that the general public has come to grips with this reality, and lots of evidence that they are all crying out for new models to believe in.

History suggests that people will increasingly degrade themselves in fealty to worthless models that become increasingly absurd. There is no apparent countervailing force that I can see anywhere on the horizon to reverse this trend.

[Spartacus]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

In one of his speeches Taleb demands proof, and simultaneously criticizes Dawkins for demanding proof.

Taleb's ancestors were Lebanese Christians, but he is not a Christian. He expresses more sympathy toward stoicism than toward Christianity. He has addressed this point directly.

Regarding Dawkins, I cannot understand how people think the guy is God. His contributions to science stopped 20 years ago, and he's produced nothing but strident evangelical screeds since then. It's amazing that the guy can make an entire career off of finding new ways to say: "Stop the presses! Everyone has been terribly wrong for centuries! The truth is, MAN WAS FORMED FROM DUST!"

The whole hero-worship of Dawkins, and the cult of militant materialistic reductionism, shows a sort of brainwashed mentality similar to the religious nuts.

Don't hold back ... tell us how you really feel.

What's FAR more amazing is the careers of creationists, religious lobbyists, broadcasters and apologists - all careers based completely on chimera. The software industry could learn a lot from these guys - vapourware? HAH!

But yeah, you gotta watch out for that Dawkins guy - he's a reg'lar Jim Jones in the making.


Taleb is greek orthodox. I had remembered reading it, and he implied it several times in interviews and speeches (including the one where he showed a slide of a Greek Orthodox or Catholic looking ceremony) but I had forgotten where I read it, and the vagueness of my memory turned him maybe into a Catholic.

Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets

vbulletin is not letting me post the URL directly, so go here and click on the hits

http://www.google.ca/search?hl=en&nu...G=Search&meta=

"full of crowds ensnared with the chants (I am Greek-Orthodox)"

[unlucky]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

Mathematicians create models and then analyze the models to see how they behave. The craft is useful because it allows reasonably complex models to be analyzed in a rigorous way without making mistakes.

But the biggest mistake that folks make with math is to forget that it is just a model. If you build the right model, you can make it say whatever you want. This is apparently what all the credit-risk analysts did, because that was the way for them to make money.

Non-mathematicians are easily impressed by a pageful of equations with lots of greek letters and funny symbols. Whenever you see such a page, you need to think: "this is a model - what is it modelling and is the model correct?". These questions cannot be answered by the math itself.

What kind of mathematical model could successfully predict the behaviour of the economy? I'm thinking it would be at least as complex as the models that are used (successfully) for short-term weather prediction. Consider the components currently used by meteorologists:

1. Solid understanding of the underlying "microscopic" phenomena and proven mathematical models that capture these phenomena.
2. A vast network of data gathering stations providing constant updates on conditions in every part of the world.
3. Vast computer resources to gather all the data and crunch it according to the model to produce predictions.

Not only are meteorologists more successful than economists (quanitative or otherwise), they also have a good understanding of their current limitations.

If you started paying meteorlogists for predictions of sunny weather - would they need either good math models or all of the above infrastructure to produce the required results?

[allenjs]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

http://www.google.ca/search?hl=en&nu...G=Search&meta=

"full of crowds ensnared with the chants (I am Greek-Orthodox)"

Wow, thanks for sharing that. I'm really surprised. I could have sworn that he wrote about this specifically in "The Black Swan" or "Edge", something to the effect of "Ever since 'Fooled By Randomness' people keep accusing me of being a Christian, and asking me about my religion, but I am just a philosopher".

But yeah, you gotta watch out for that Dawkins guy - he's a reg'lar Jim Jones in the making.

I know, right?

If these financial priests had just worshiped clay idols of Ganesh, or based their choices off of sheep entrails, they would have done a lot less damage.

My problem with Scrappy Dickie Dawkins, and (for example) the moral crusaders against superstition who venerate Sagan by calling themselves "Saganists", is this: They spent so much energy picking easy and irrelevant targets and congratulating themselves on being smarter than "those superstitious nuts", and as a result they made themselves incapable of realizing the massively absurd superstitions upon which they were basing their own lives. The moralistic crusade of "materialist reductionism versus ignorant superstition" is indeed a purple kool-aid that puts its drinkers to sleep and renders them incapable of realizing the dangerous limitations of reductionism.

These crusaders have lost all sense of scale or proportion. They will fight until their last breath against anyone who teaches kids to burn incense for luck, but when teachers tell our children absurd and profoundly damaging lies like "Put your money in the bank, and the magic economic God it will double every 18 years", these defenders of truth say nothing. Yes, it's superstitious to burn incense for good luck. But most people who burn incense for luck already know this somewhere in their hearts. They don't make confident and sustained policy decisions that blow up the whole world.

[GRG55]

Re: An interesting question on iTulip: What is the future for arithmetic?

My outrage is aimed at the accountant-charlatan putting society at risk using arithmetic. We lost trillions of taxpayer dollars and the world financial system is in danger because of arithmetic. And our kids are brainwashed with this stuff starting from kindergarten. It does not add up. :rolleyes:

I'll try to remind myself about the arithmetic "brainwashing" in the education system every time I run into a store clerk with a high school education that can't seem to count out the proper change...even with the help of a computerized till...:rolleyes:

[This happens a lot more frequently than I would care to admit. Years ago I had a temp assistant that couldn't figure out how to calculate the percentage change between two numbers. Maybe less "brainwashing" and more 3Rs is in order? ]

[Chris Coles]

Re: An interesting question on iTulip: What is the future for arithmetic?

Maybe less "brainwashing" and more 3Rs is in order? ]

When I was a kid, some six decades ago, all schools rang out with the chants of "two twos are four, three twos are six" and so on, but today, teachers are too inclined to believe that their status is reduced if they resort to the simplistic chanting of older times. But to this day, I can calculate instantly using those old rymes... food for thought??

[*T*]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

I strongly disagree that math is a subset of science.

Totally true. Mathematics is the queen and servant to the sciences. It is not a science. It is self-contained.

[*T*]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

how do you distribute the the benefitsof this research then?

Put research in the public domain.

[*T*]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

Mathematicians create models and then analyze the models to see how they behave. The craft is useful because it allows reasonably complex models to be analyzed in a rigorous way without making mistakes.

But the biggest mistake that folks make with math is to forget that it is just a model. If you build the right model, you can make it say whatever you want. This is apparently what all the credit-risk analysts did, because that was the way for them to make money.

Non-mathematicians are easily impressed by a pageful of equations with lots of greek letters and funny symbols. Whenever you see such a page, you need to think: "this is a model - what is it modelling and is the model correct?". These questions cannot be answered by the math itself.

Yes. Maths is just a language. Saying maths is a science is like saying English Language is a science. It is not how you say it that is important, it is what is being said.

[jiimbergin]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

Mathematicians create models and then analyze the models to see how they behave. The craft is useful because it allows reasonably complex models to be analyzed in a rigorous way without making mistakes.

But the biggest mistake that folks make with math is to forget that it is just a model. If you build the right model, you can make it say whatever you want. This is apparently what all the credit-risk analysts did, because that was the way for them to make money.

Non-mathematicians are easily impressed by a pageful of equations with lots of greek letters and funny symbols. Whenever you see such a page, you need to think: "this is a model - what is it modelling and is the model correct?". These questions cannot be answered by the math itself.

What kind of mathematical model could successfully predict the behaviour of the economy? I'm thinking it would be at least as complex as the models that are used (successfully) for short-term weather prediction. Consider the components currently used by meteorologists:

1. Solid understanding of the underlying "microscopic" phenomena and proven mathematical models that capture these phenomena.
2. A vast network of data gathering stations providing constant updates on conditions in every part of the world.
3. Vast computer resources to gather all the data and crunch it according to the model to produce predictions.

Not only are meteorologists more successful than economists (quanitative or otherwise), they also have a good understanding of their current limitations.

If you started paying meteorlogists for predictions of sunny weather - would they need either good math models or all of the above infrastructure to produce the required results?

I agree, especially about understanding the limitations. Without models we would have no insurance industry. Before I retired I was an insurance actuary. Back in 1995 2 companies had come out with deferred annuities that were "linked" to the SP500 index. My company wanted to jump in. I ran models that I created and figured out how they came up with the idea. However, there were still too many outliers for my comfort. The head of investments got me together with Lehman. Lehman took the risk away from me. Of course that was assuming Lehman would be around if needed

[Sharky]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

An interesting question on Slashdot: What is the future for Financial Mathematics?

The problem with today's Financial Mathematics is that it was not developed in a true mathematical way. It makes assumptions about the nature of financial investments that simply aren't true (things like the lognormal distribution of prices that underlies the Black-Scholes option pricing model). It's like assuming that black is white, and then deriving a whole structure from that.

Financial Mathematics will continue to exist. Modeling will continue to evolve. In fact, I predict that the models will become much more sophisticated, and will include many more real-world effects -- yet they will still be totally wrong, because the underlying philosophy won't change.

This approach of grabbing onto some apparently random "fact" and building a structure from it, is a key characteristic of rationalism -- which is one of the most rampant philosophical viewpoints in the financial world. The solution is to ground those facts in reality, to not let them dangle unsupported. Models should be based on measurable realities, not unprovable assumptions.

[petertribo]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

...even Einstein felt bad after he made the bomb.

I like this quote from Einstein, it goes to the heart of the matter:

If we knew what it was we were doing, it would not be called research, would it? Albert Einstein

and on Mathematics:

As far as the laws of mathematics refer to reality, they are not certain, and as far as they are certain, they do not refer to reality. Albert Einstein

Many eminent philosophers claim that Science has become the new God with the usual taboos. Along with Bill Joy, ex-chief scientist at Sun Microsystems, and Einstein, I have serious reservations about the direction and dangers of Science. As for Mathematics, subset or not, it's corruption has cost this nation trillions of dollars and set up a phony Economy based on Snake Oil Economics. A forerunner of some more serious Disasters which might follow.

[marvenger]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

that would be very interesting, could be very destablising to the established order, they may not like.

[Supercilious]

Re: An interesting question on Slashdot: What is the future for Financial Mathematics?

The problem with today's Financial Mathematics is that it was not developed in a true mathematical way. It makes assumptions about the nature of financial investments that simply aren't true (things like the lognormal distribution of prices that underlies the Black-Scholes option pricing model). It's like assuming that black is white, and then deriving a whole structure from that.

Financial Mathematics will continue to exist. Modeling will continue to evolve. In fact, I predict that the models will become much more sophisticated, and will include many more real-world effects -- yet they will still be totally wrong, because the underlying philosophy won't change.

This approach of grabbing onto some apparently random "fact" and building a structure from it, is a key characteristic of rationalism -- which is one of the most rampant philosophical viewpoints in the financial world. The solution is to ground those facts in reality, to not let them dangle unsupported. Models should be based on measurable realities, not unprovable assumptions.

Good points but I'm not sure it was just worshipping of models. I posted in the Video section a great interview with Janet Takavoli. When she was asked about those funny financial models she said:

"There weren't any outliers... there were just a lot of outright liars."