• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Nothing Has Changed, Paul Wilmott Says

Joined
12/29/06
Messages
140
Points
28
It would be foolish to paraphrase Paul Wilmott, a researcher, consultant and lecturer in quantitative finance who has strong opinions about Wall Street's risk management practices, a sharp wit and a way with words. So here's a lightly edited version of our recent interview with him about how to fix risk problems on the Street.

WS&T: How are you?

Wilmott: Tired. I've got a crazy Dutch film crew following me around, everything's a bit manic and I'm exhausted.

WS&T: Concerns about risk management have been growing on Wall Street post-crisis and pre-regulation. How might Wall Street firms change the way they do risk management, are there new metrics, models or methods that will be or should be considered? For instance, some say Value at Risk is no longer relevant as a measurement of risk.

Wilmott: There are plenty of good models out there, it's just that nobody's incented to use them. If you used the wrong model to design airplanes, you'd kill passengers and end up in prison. But do the people using financial risk models care what happens? Of course they don't.

WS&T: So there should be a direct link between the risk modelers and...

Wilmott: Prison, I agree. There's people who think we should go after these people. There should certainly be more lawsuits going around. Someone I know moved to a new bank recently. He was doing a risk management procedure and the traders came up to him to say, 'You're telling me the risk is too big, can you just fudge the figures and make it smaller?' Nothing has changed. I don't know why anybody thinks it has.

That's why no one with half a brain has ever liked VaR, because it can be used to hide risk. I've been talking for years about how people hide risk rather than hedge, because that allows them to trade bigger and bigger and if they're lucky, they get a big bonus. If they're not lucky, nothing happens to them. The more sophisticated your tool, the greater potential for pretending there's no risk.

WS&T: Do you think Value at Risk should be thrown away?

Wilmott: No. VaR is fine for for day-to-day activities, but there are lots of other things people should be doing. For example, worst-case scenarios. I don't believe in trying to refine the probability of crashes happening. People are surprised that crashes are happening every few years that according to theory are only supposed to happen once every 10,000 years. Who cares whether it's once every few years, once every six years or once every 7.3 years? What matters is taking steps that will protect you when there's a crash, because there will be one.

Did you expect anything was going to change because of the subprime crisis? People say that's the end of everything, quantitative analytics is dead, CDOs are dead. I say it will be business as usual and faster than you think. Nothing's changing. People deserve everything they're going to get in the next crisis because they're not complaining enough. But there are really good models out there.

WS&T: Such as?

Wilmott: The trick is to not have models that are too sophisticated. Don't get too caught up in the details of the mathematics. One of the main culprits in all this has been the Masters in Financial Engineering courses. They're sold by universities to 22 year olds who have no experience in life and banking. Professors who never worked a day in their life in the real world are teaching all of these poor, unsuspecting fools who are paying $80,000 a year for the degree, because they know they can get a job in a bank and they'll be making millions. They're happy to pay $80,000. I'm not happy because I know the education they're getting is substandard, to say the least. It may be fantastically mathematical, but it's got nothing to do with finance in the real world. So there are swarms of these people out there, many tens of thousands of people have come out of these degree programs and are put in charge of derivatives, valuation and risk management and they've never seen the real world. This is where I plug the Certificate in Quantitative Finance. CQF is the only financial engineering course in the world that hasn't had to rewrite any lecture notes because of the recent crisis. We've been warning about these things since we started in 2003.

WS&T: Have the graduates of your course performed better than the average quant?

Wilmott: Sadly, we haven't reached the critical mass where my CQF alumni could save the planet. Had we started the CQF five years earlier, then we would have saved the world, but we haven't. Maybe we'll stop the next crisis.

WS&T: Do you try to introduce common-sense thinking into your courses?

Wilmott: Everything we do is real world. Everything is practical, based on data, common sense. When did common sense disappear from the planet? We try to knock some common sense back into these people. We try to instill confidence that they can do their own modeling and think for themselves. Another one of my pet peeves is how sheep-like people who work in banks are. They copy each other so much. We teach skeptical thinking, to question assumptions, to figure out if the model is wrong.

WS&T: What about the data itself? One of the problems with the subprime mortgage mess was that for many years prior, housing prices hadn't dropped and U.S. homeowners had a history of paying their mortgages and being unwilling to foreclose. So if someone was analyzing the past four years of historical data, they wouldn't have any indicator of trouble.

Wilmott: Why are they using the last four years worth of data?

WS&T: How far back to they need to go? No historical information really tells you what's going to happen in the future.

Wilmott: I've owned some properties, I bought my first house when I was 25 and in my experience, there's a one in three chance of losing money from property. You do have to go back a few years, but not a long time, to see falling house prices. When you go through bubble after bubble after bubble in every single walk of life, what a lack of imagination a person has to have to think house prices never fall. I have to hit myself in the head with a hammer to get into that frame of mind.

WS&T: But I've heard models like that are used in financial firms.

Wilmott: 2010 I hope will be the year in which people finally start taking moral hazard seriously. People have an incentive to say house prices will keep on rising if it means they can do the trade. But you're also right that there are people who believe in all of these models and the idea that house prices don't go down. I'm incredibly blessed or cursed by always questioning everything. You have to do constant rethinking.

Nothing Has Changed, Paul Wilmott Says by Wall Street & Technology
 
Wilmott: "The trick is to not have models that are too sophisticated. Don't get too caught up in the details of the mathematics. One of the main culprits in all this has been the Masters in Financial Engineering courses. They're sold by universities to 22 year olds who have no experience in life and banking."

He says of these programs, "So there are swarms of these people out there, many tens of thousands of people have come out of these degree programs and are put in charge of derivatives, valuation and risk management and they've never seen the real world." Of course, he's also plugging his Certificate in Quantitative Finance course.

"Paul Wilmott is the smartest of the quants," says his friend and fellow quant Nassim Nicholas Taleb. "He may be the only smart quant."

Also, according to this Portfolio article, Wilmott magazine costs 395 pounds a year for six prints with around 2000 subscribers. His CQF program graduates 500 students a year, easily the biggest quant diploma mill in the world, rivals even Poly.
 
Has he really looked into all MFE programs to make a statement like that with certainty? I can't imagine he's really done the due diligence on that statement.

Comes across as a salesman, a bit.
 
Wilmott is a very smart guy, but that interview reads more like an ad for CQF than something else.
 
Have you read any Wilmott's interview lately that he didn't plug the CQF?
500 graduates a year for 18K a pop, CQF is a cash cow so why not, right? In 5 years, he will get a Nobel for saving the world from another crisis.
 
Wilmott: The trick is to not have models that are too sophisticated. Don't get too caught up in the details of the mathematics. One of the main culprits in all this has been the Masters in Financial Engineering courses. They're sold by universities to 22 year olds who have no experience in life and banking. Professors who never worked a day in their life in the real world are teaching all of these poor, unsuspecting fools who are paying $80,000 a year for the degree, because they know they can get a job in a bank and they'll be making millions. They're happy to pay $80,000. I'm not happy because I know the education they're getting is substandard, to say the least. It may be fantastically mathematical, but it's got nothing to do with finance in the real world.

He's hit the nail on the head there, though. It's not just Wilmott who knows teaching financial engineering is a cash cow; it's just about every American university. Intellectually threadbare programs have been set up overnight on shoestring budgets, with ill-planned curricula (I use the term rashly) and with "faculty" literally pulled off the street. I don't see much difference between some of these programs and what online diploma mills like University of Phoenix are doing. The programs are frauds -- and I use the term soberly.
 
And on the same breath, he loses all the credibility he has
CQF is the only financial engineering course in the world that hasn't had to rewrite any lecture notes because of the recent crisis. We've been warning about these things since we started in 2003.
As a mathematician, he should know better that it's impossible for him to examine "all financial engineering course in the world" to make such a claim.
He may trick the naive crowd into buying his world-saving CQF diploma but for people in the field already, they can see through the same bandwagon that he and Taleb are riding to make a buck out of their "I told you so" tale.

As a quant, you need a pair of critical eyes to see through all the sales pitches, all the BS because after all, everyone is trying to sell you something. They aren't out on a mission to save the world.
 
He's hit the nail on the head there, though. It's not just Wilmott who knows teaching financial engineering is a cash cow; it's just about every American university. Intellectually threadbare programs have been set up overnight on shoestring budgets, with ill-planned curricula (I use the term rashly) and with "faculty" literally pulled off the street. I don't see much difference between some of these programs and what online diploma mills like University of Phoenix are doing. The programs are frauds -- and I use the term soberly.

No surprise coming from you.
However a bit abnormal to come from the leading CQF guy, a quantitative finance degree by all means. He has in the curriculum same "deeply mathematical" models that other MFE programs use. In fact his papers&books are used in his program as well as others.
So it sounds like criticizing your own career&profession. Interesting ...
 
Though he may be a smart man, and obviously a bit of an entrepreneur, in every almost every interview I've read he's spent half of it pimping his program, and all of that makes it very very hard to read something like "All quant programs (except the one I'm the director of) are worthless" with a straight face.
 
I'm always wondering about the relative supply/demand for quant careers. Are all these courses turning out "too many" quants to fill the roles?
 
I'm always wondering about the relative supply/demand for quant careers. Are all these courses turning out "too many" quants to fill the roles?

'Course there are too many. That's obvious. Unlike, say, a decade ago there is now a plethora of schools offering MFEs (or close variants) and if anything, the number of jobs has not kept pace (probably sharply declined during the last two or three years). But the schools have a disincentive to reveal the true situation: they're too busy hustling their threadbare offerings. Scant consolation though it may be, the same situation prevails in law and MBA programs.
 
You're treating the supply and demand as though all MFEs were interchangeable. Quality is a big separator.
 
You're treating the supply and demand as though all MFEs were interchangeable. Quality is a big separator.

Quality is the single most important separator. But the first-rank programs can be counted on two hands (okay, three hands), while the crud programs are ten times that number. Same as in other areas like MBA and law. The point is it's a waste of time and money to go to a second-tier program.
 
So with job quantity in decline and a saturation of new "talent" on the hunt, it would seem that now is a suboptimal time to hunt for a quant job?

Unless one is exceptionally good of course...!
 
I believe the circulation of Wilmott mag is a bit higher, but it is reassuringly expensive :)

I agree with those that say the quality of MFE grads varies a lot, I have personally spoken with a greater % of the world's MFEs than anyone else, so I know this to be true.

But a critical problem is that there is no consistent definition of what should be in a MFE. I now teach occasional lectures on ethics for quants on Paul's CQF, and when researching this I found almost nothing I could use as a reference.

Most 'authorities' are stuck in about the 7th century AD.

Here's a modern ethical dilemma for you.

The time of year comes to renew your expensive, but excellent health insurance.
The cost is enough to make the burden heavy.

You decide to pay for a thorough check on your health.

If it says 'oh dear....' you renew your insurance. No legal or deceit issues here because you are already under cover.

If it says "you are actually from planet Krypton and will never die", you cancel.

Is that acceptable to you, as a course of action ?

This is wholly legal, involves no deception, but has a clear victim, and you benefit from it.

I'm not interested in your answer which I suspect will be the same as mine (we all hate insurers), but what reasoning can you apply that supports your position ?

Also, if you are religious, can you find any text in your holy books that cover this ?

Fact is that any ethical judgement requires that you integrate probability into the decision.
The Christian Bible has a passage "suffer not a witch to live".

But in a criminal justice system, you never have 100% confidence, but can you imagine a holy book saying "after due process, when you have achieved a confidence level of 90 % you should kill the *****" ?
 
Here's a modern ethical dilemma for you.

...

I'm not interested in your answer which I suspect will be the same as mine (we all hate insurers), but what reasoning can you apply that supports your position?

In reality, the injunction people follow is "Looking out for #1." If some ethical position runs counter to looking after one's interests, involves a real sacrifice on one's part, then people look for rationalisations to circumvent the ethical position (Just this once; everyone else is doing it; when I become rich and powerful I won't need to do this, it all depends on interpretation, it's all relative, etc.). This quibbling aside, what ethical system is being assumed, being employed? There's no one canonical system. One might use Kant's categorical imperative:

"Act only according to that maxim by which you can at the same time will that it should become a universal law."

An ethical person, as I see it, would be one willing to make tangible sacrifices for what he or she believes to be right, in consonance with the moral structure of the cosmos. This is the acid test. How many such people do you know? I may have stumbled across one or two during my lifetime.

For the example you cite, the insurance company should be assuming I would undergo a medical exam to decide whether to continue -- there's no subterfuge involved and they must assume I'm a rational agent. So this thinking must be built into the premiums they charge. I am therefore within my rights to make my rational decision based on the results of the test and yes -- I am comfortable with this practice being a universal law.
 
The problem with universal law is that not only are humans not smart enough to choose one, undergrad maths can show no such decision can be made.

Acceptance of your own limitations should be a critical aspect of ethics.

You are asserting that if the insurer is acting rationally and competently, then there is no conflict, but if one is an expert in a domain, is it fair to utilise that ?

In my work I often deal with people who are smart, but have zero experience of finance. Is it appropriate for me to use that against them ?
I can't know the expertise of everyone I deal with, especially since some of them involve people I know nothing about.

Also I misled you on who the victim was (I am of course a headhunter you should have expected that).
By following my procedure you push up the premiums for people who just trust to luck. Most of them will be people with no education in probability, and of course many will be gravely ill.

You happy to do that ?
 
You are asserting that if the insurer is acting rationally and competently, then there is no conflict, but if one is an expert in a domain, is it fair to utilise that?

What is the use of expertise if it's not utilised? In financial engineering, people are incessantly looking for an edge over other parties. To the extent that they keep their acquired expertise proprietary, secret, under wraps. If you saw a way to make a trade that netted you $10m, would you not seize it? Or would you tell the counterparty the secret you have so he might capitalise on it? In many areas expertise is used in zero-sum games where your expertise hurts someone else. It's like telling a chess grandmaster that the thousands of hours he's spent studying the game and honing his skills is an unfair advantage and he should gallantly offer his opponents draws.

A deeper philosophical question is whether human social existence can be just about cooperation and not competition. I doubt it. And if you accept the necessity of some competition, then the advantage offered by earnt expertise (or native talent) becomes inevitable.

In my work I often deal with people who are smart, but have zero experience of finance. Is it appropriate for me to use that against them?

Sure. Their problem, not yours. Are you your brother's keeper? If you choose to do otherwise, that's your prerogative.

I can't know the expertise of everyone I deal with, especially since some of them involve people I know nothing about.

Don't lose any sleep over it. We have to survive in this imperfect world. Nit saying you should be determined to be a baddie (or any such puerile nonsense), but pragmatism suggests you not waste time on this.

By following my procedure you push up the premiums for people who just trust to luck. Most of them will be people with no education in probability, and of course many will be gravely ill.

The fit and smart tend to survive; the others tend to die. You're helping evolution along.

You happy to do that ?

I may not always act the way I'm prescribing above but that's my personal choice -- not because I adhere to some abstract code of morality. Or it could be because I read too much Nietzsche in my impressionable years. I also think Spinoza said somewhere that he who feels remorse is twice weak.
 
Back
Top