Sylvain Raynes: The State of Financial Engineering

[DHaynes]

Stern's Joel Hasbrouck's Power Point presentation [pdf]

The preliminary version of Donald MacKenzie's argument:
An Equation and its Worlds: Bricolage, Exemplars, Disunity and Performativity in Financial Economics [pdf]

I don't think it's the same power point presentation

 

[doug reich]

I appreciate the follow-up; I suppose the motivational tone the author intended was missing from the first piece. However, (not surprisingly) I take exception to this:

The most interesting puzzle is why someone who apparently teaches FE himself would bad-mouth his own occupation. It seems counter-productive and nonsensical at best. Perhaps you should reflect on this for a while before rushing in with sour grapes. By the way, please leave Baruch College out of this. Institutions don't teach classes, professors teach classes. I wrote this on my own time, and nobody else but me needs to feel responsible. Taking responsibility is, however, what we are really talking about.

Complaining only gets you so far. As an instructor in an FE program, I would expect him to offer solutions -- something he is uniquely qualified to give -- and not just criticism, or demanding "responsibility". I believe bigbadwolf, Michael Lewis, (to name a few) and others who are not involved in FE programs take very similar positions to this piece, but I expect something extra from an insider.

And Baruch is not irrelevant, I'm not asking the college or Bernard himself to reply. I'm asking the author, as a privileged insider, to offer his greater insight. It seems he --or the program in which he teaches -- has nothing special to offer above those he deplores. Why does he lend his name to an institution he thinks is a poor reflection of his profession? Or is he holding back on us?

 

[bigbadwolf]

Complaining only gets you so far. As an instructor in an FE program, I would expect him to offer solutions -- something he is uniquely qualified to give -- and not just criticism, or demanding "responsibility".... but I expect something extra from an insider.

He teaches it and gets paid: a source of income. Surely he doesn't have to believe in it as well, does he? This reminds me of a discussion here a few months back, when the subject was the fact that Paul Wilmott (and some quants on this forum) put their money in a savings bank account (rather than some risk- and return-optimised portfolio or some derivative products). Yet surely the same charge of -- dare I use the word? -- hypocrisy could be leveled against them as well? That they don't believe in their own work? I think Raynes is making an explicit distinction between what he does in his private time and his institutional persona. Is this not acceptable? Does he have to act like every other cowardly professor who dare not speak truth to power?

 

[doug reich]

I suppose his teaching distinguishes "private" from "professional", but I think the average observer (myself included) expects some level of earnestness in all actions. This does not preclude the possibility that he keeps his corner tidy but considers the rest of the ship a shambles. His profit motive as a teacher doesn't seem to have stopped him from writing this piece, so why stop there and hold out on his perspective on Baruch? Must we actually infer from his silence his true intentions rather than reading them forthright?

 

[bigbadwolf]

... but I think the average observer (myself included) expects some level of earnestness in all actions.His profit motive as a teacher doesn't seem to have stopped him from writing this piece, so why stop there and hold out on his perspective on Baruch?

Is his opinion on Baruch -- whatever it happens to be -- relevant to his general discussion of quant programs? Unless possibly he wants to illustrate some general point with a (representative) example.

For my part I'm glad he's opened the field to a more critical discussion of what does get taught in quant programs. MFE programs teach a grab bag of special tricks that have "worked" (I use the term rashly) because of a particular constellation of political and economic circumstances. If these circumstances change -- which appears to be the case now -- can we not discuss how appropriate that toolkit is, on what philosophical basis it has rested, and whether any math toolkit can replace it?

My own ever-so-humble opinion on the matter is clear: mathematical modeling of fundamentally irrational social phenomena such as finance and economics is not going to work. This is not electrodynamics or fluid mechanics. And just as the economics literature is replete with worthless mathematical formulations that don't explain anything (let alone make accurate predictions), the same holds, I maintain, for finance as well. This is why I don't buy the argument that if the tools don't work, a better set is needed. My point is that no toolkit will work: economics and finance are completely artificial phenomena, governed by political decisions, arbitrary and fluctuating rules, and irrational individual and group behavior.

But quant finance makes pretty camouflage: it presents the illusion that finance is rational, that risks can be ascertained, that capricious reality can be captured by equations. The derivatives bubble would never have reached the size it did without these delusions successfully being sold to both regulators and a credulous public. Nor would other areas of finance such as hedge funds and portfolio management. Yet it was all smoke and mirrors.

Quant finance exists largely because universities can make a pretty penny peddling meretricious programs on shoestring budgets, "adjunct faculty" can make some extra money moonlighting, and because credulous students think it's going to lead to lucrative employment. And of course because Wall Street has been peddling snake oil, and it helped to hire some graduates of these "programs." The whole thing has been done in a spirit of cynicism. And not because quant finance is a developed academic discipline inherently worthy of respect. Now that the quant employment terrain is changing, I expect many of these programs to quietly fold (with some lag factor). Maybe fifty years from now some historian can write an account of how people went into these one-year and 18-month programs and then commanded high salaries and stellar bonuses -- while the party lasted.

 

[Stefan Zota]

Many opinions ready to dismiss all current models. If a model is based on some wrong assumptions, you would think, you would correct the assumptions, change the model, monitor it better, adjust the risk. No, the solution that you are proposing is to burn all models.
Ok. All are burned. We don't have any pricing theory as it is now.

What is your alternative? I am a regular trader and I would like to hedge some investment in a simple call. What do I do? I cannot trade using fundamentals because options are some functions of an underlying asset.

Also, on a different note, I have seen you are an advocate of a solid Math foundation. Me too. How do you plan to use this knowledge in finance if not building other mathematical models. If trading is all deal-making and bargain, then why do you need to know all that Math?

Bottom line is that you have a contradictory position. This is what Doug has tried to explain. I respect a position or one side or another. I respect an MBA student that says financial engineering is useless and does not care about it. He/she made a clear choice.

 

[Andy Nguyen]

I don't think it's the same power point presentation

The titles of the slides you download here and in the original blog are now the same and so I believe this is the same ppt as cited originally.

 

[samiam]

I disagree with the notion that finance cannot be mathematically modeled because it is governed by irrational group behavior, political policies, social phenomena, at least in the equities space. I'll just speak from my experience as I'm not qualified to talk about derivatives or mortgages and why so many deals were made. I'll just place that situation in a similar one to the south sea bubble and because at the time everyone was doing it and it was profitable.

There are certain technical patterns that occur almost everyday in the realm of plain equities combined with a certain unique set of news where the outcome always plays out in a certain way with odds way better than even.

For example, a gap up at the open after two or more consecutive days of solid gains always leads to a close of the gap on that day of trading without the existence of any more fundamental news and still even more than even with more fundamental news that is positive

I believe there is no reason why occurences like this which are governed by irrational group behavior can't be extended to longer time frames or be mathematically modeled. There is a method to the madness and there is still some order in chaos. It's a matter of finding these patterns that statistically happen enough that you'll gain more than lose. I look forward to applying what I learn in the programs with my experience in discretionary trading and the tendencies of the market I've learned.

 

[bigbadwolf]

No, the solution that you are proposing is to burn all models.
Ok. All are burned. We don't have any pricing theory as it is now.

No, use them now if you can and they help you. But the force of changing circumstances may force you to jettison them or replace them. If government regulation and restriction comes back with a vengeance -- and I think it will -- then many models will be consigned to the trashcan. For example, it's conceivable that CDS be banned altogether -- what happens to the body of theory associated with it? The point is that much of what is being taught to MFE students is contingent, of ephemeral value and importance.

What is your alternative? I am a regular trader and I would like to hedge some investment in a simple call. What do I do? I cannot trade using fundamentals because options are some functions of an underlying asset.

Use what you can right now, but expect substantial changes in the regulatory environment.

Also, on a different note, I have seen you are an advocate of a solid Math foundation. Me too. How do you plan to use this knowledge in finance if not building other mathematical models. If trading is all deal-making and bargain, then why do you need to know all that Math?

To me mathematics is a queen and I her devoted slave. This is why I push math foundations all the time. And there are areas such as physics where it's indispensable. But I'm sceptical about areas like finance. The models are ad hoc, limited and dependent on changing regulatory environments. Not quite the same as General Relativity. But the real object of my fury is the propaganda around these limited models, the hucksterism involved in pimping them by a greedy finance industry, and the way mercenary universities provide a shoddy education in a spirit of cynicism. In this, Raynes has hit the nail on the head.

We're all grown-ups here. People have to make a living. If someone can do it as a quant, more power to him (or her). But the necessity of making a living doesn't mean the quant has to believe in what he or she does -- any more than Raynes may believe in the lasting worth of what he's teaching.

 

[Andy Nguyen]

It should be noted that Prof. Raynes has single-handedly placed a large number of his students in structured finance roles the past few years; once as many as half of the graduating class. They almost always end up in jobs involving structured finance deal-making so this is where his perspective on this opinion piece should be seen.

 

[Stefan Zota]

No, use them now if you can and they help you. But the force of changing circumstances may force you to jettison them or replace them. If government regulation and restriction comes back with a vengeance -- and I think it will -- then many models will be consigned to the trashcan. For example, it's conceivable that CDS be banned altogether -- what happens to the body of theory associated with it? The point is that much of what is being taught to MFE students is contingent, of ephemeral value and importance.



Use what you can right now, but expect substantial changes in the regulatory environment.



To me mathematics is a queen and I her devoted slave. This is why I push math foundations all the time. And there are areas such as physics where it's indispensable. But I'm sceptical about areas like finance. The models are ad hoc, limited and dependent on changing regulatory environments. Not quite the same as General Relativity. But the real object of my fury is the propaganda around these limited models, the hucksterism involved in pimping them by a greedy finance industry, and the way mercenary universities provide a shoddy education in a spirit of cynicism. In this, Raynes has hit the nail on the head.

We're all grown-ups here. People have to make a living. If someone can do it as a quant, more power to him (or her). But the necessity of making a living doesn't mean the quant has to believe in what he or she does -- any more than Raynes may believe in the lasting worth of what he's teaching.

I do share some of your opinions and I do take everything with a grain of salt. However the main difference is a clear position.

You say in one post that Math cannot be applied in finance, then you say you think current models should still be used up to a point, then you say that all areas of Math should be explored to find appropriate uses to finance, then you say that models are not necessarily bad just the propaganda around them, then you say that models apply only because of current regulatory conditions.

You may want to review your conflictual posts.


PS: Andy's note is something that should make you think ...

 

[joe_bradley]

Just my $0.02 as someone who took both of Prof. Raynes class - Prof. Raynes frequently indicated his dissatisfaction with "business as usual" in finance in his classes, and you could figure as much by reading the R & R Consulting blog. I doubt any of his former students would be shocked at this column.

As far as whether he does anything different, in fact the methods in the courses he teaches don't seem to be typical. Speaking as a broad overgeneralization, his method for valuation and rating seems to be based on the characteristics of the loan portfolio, rather than working backwards from a target rating structure. If you can't take his class and are curious, all the details are in the Raynes & Rutledge book.

Finally, Baruch is unusual in that this is not a cash cow for the university. The MFE curicculum costs no more than any other MS course offering. If it's a racket, its a rather poorly-paying one.

 

[charlesdwright]

I'm inclined to think a curriculum in a hot new field is at least partly industry-driven. I'm inclined as well to suspect that the industry said, "give us people who can do the fancy math."

Separately, nobody can take a good idea and run it into the ground like Wall Street when in a regime of easy money.

In the regime of easy money, when the demand for product was perverse, the industry wanted bodies, not minds.

Easy money taints everything.

*note: the term "easy money" is a pointer to wider concept of private profit at the expense of socialized risk. That is, for example, the "Greenspan Put," and the "Bernanke Put."

 

[Andy Nguyen]

On a somewhat related article, NYT has an article today about Moodys. It has a picture and quote of Prof. Raynes in it. Feel free to discuss it.
http://www.nytimes.com/2008/12/07/business/07rating.html?hp=&pagewanted=all

 

[bigbadwolf]

You say in one post that Math cannot be applied in finance, then you say you think current models should still be used up to a point, then you say that all areas of Math should be explored to find appropriate uses to finance, then you say that models are not necessarily bad just the propaganda around them, then you say that models apply only because of current regulatory conditions.

Use the models because your livelihood depends on it; maybe they even shed some qualitative insight on your contingent financial reality. Understand that those who expect you to use them may decide they are no longer applicable if conditions change. Did I say all areas of math should be explored to find appropriate uses for finance? I may have said the conceptual toolkit might change -- but not that I have any faith in it. My consistent line is that math modeling in finance and economics is doomed to failure; that there are fundamental philosophical problems with the attempt; and that the attempt to mathematise them hides ulterior motives.

 

[Andy Nguyen]

I contacted Prof. Joel Hasbrouck, the author of the slides and he put it back on his website at http://pages.stern.nyu.edu/~jhasbrou/Miscellaneous/JH_Quantitative finance.pdf

 

[dstefan]

Thanks, Andy!

 

[PatM]

From Raynes in the NYT:
“The mistaken notion that Moody’s was a company like any other, that was very fundamental,” said Sylvain Raynes, a former Moody’s analyst who is co-founder of R&R Consulting, a firm that helps investors gauge debt risks. “It is not just a profit maximization entity like Exxon or Microsoft. Moody’s has a duty to the American public. People trusted it.”

This is odd. Don't people implicitly trust all the companies that they do business with? Are we not supposed to trust Exxon or Microsoft? They have no "duty" to their public? If Exxon started watering down the oil, or Microsoft started stealing people's credit card information, would we just say, "Oh, well, sure, but they're not like Moody's. We *TRUST* Moody's, but XOM and MSFT are just profit-maximizers." That's sort of a false distinction?

People trust all the companies they do business with, to provide a decent product, worth the money paid, and not lie about it.

As for math on Wall Street, there's no going back. It's here to stay, and it will only get bigger. We might have a slow patch over the next several years, but at some point, someone will come up with something else, and it will speed up again. Raynes is arguing for more math (the Nyquist sampling theorem??), not less.

 

[bigbadwolf]

People trust all the companies they do business with, to provide a decent product, worth the money paid, and not lie about it.

Do they really? After twelve years in the USA, I don't trust my health insurance company to cough up when payment is due, I don't trust pharma to introduce properly tested products, I don't trust automobile reliability, I don't trust my ISP not to jack up rates if they can get away with it, I don't trust the high-fructose corn syrup that's used ubiquitously in the USA (and just about nowhere else), and I don't trust Microsoft not to steal their next big idea.

As for math on Wall Street, there's no going back. It's here to stay, and it will only get bigger.

Why? Or is it an article of faith analogous to the belief that private enterprise will not swindle you to the hilt?

Raynes is arguing for more math (the Nyquist sampling theorem??), not less.

He's arguing for a more complete and versatile toolkit for quants, rather than the meretricious, disjointed and superficial rubbish served up in so many quant programs. Here, as in corporate America, Caveat Emptor.

 

[doug reich]

Raynes is correct in pointing out that the rating companies have a privileged role in the marketplace. I do not get to look at a company's financials with the same depth they do; I do not get to know every bond in the securitized portfolio they are rating. The market exists in the shape it does because there are these pseudo-regulators out there. If they can't be trusted, then the government is going to have to take that role, because the public has some recourse if they don't hold up their end of the bargain (or so the thinking goes).

By contrast, if Microsoft started unabashedly stealing people's credit card numbers, they would be out of business shortly after. They are not special or unique in the marketplace except that they control the dominant market share. They aren't the only ones who know how to program computers, for example.