Revoke Nobel Prize in Economics?

Ronald Bailey | October 14, 2008

Does the cause for the current world financial chaos lie in the following equation?

That's what Nasim Taleb, author of The Black Swan: The Impact of the Highly Improbable, asserts about the above formula devised by economics Nobelists Robert Merton and Myron Scholes. As Taleb explained to NPR:

Their formula for evaluating stock options laid the ground work for risk-management in modern financial markets.

In its press release for the 1997 prize, the Nobel Committee wrote:

Robert C. Merton and Myron S. Scholes have, in collaboration with the late Fischer Black, developed a pioneering formula for the valuation of stock options. Their methodology has paved the way for economic valuations in many areas. It has also generated new types of financial instruments and facilitated more efficient risk management in society....

A new method to determine the value of derivatives stands out among the foremost contributions to economic sciences over the last 25 years....

Black, Merton and Scholes thus laid the foundation for the rapid growth of markets for derivatives in the last ten years. Their method has more general applicability, however, and has created new areas of research - inside as well as outside of financial economics. A similar method may be used to value insurance contracts and guarantees, or the flexibility of physical investment projects.

Taleb claims that since recent events have shown that Merton-Scholes formula doesn't work, the Nobel Committee should ask for the medals back.

Comments

JP|10.14.08 @ 11:29AM|#

Every theory will eventually be shown "not to work." That's the whole basis of modern science.

|10.14.08 @ 11:37AM|#

Yes, it should be revoked. Oh, you weren't talking about Krugman's.

|10.14.08 @ 11:38AM|#

They worked for the vast majority of option under the vast majority of situations. That's like saying that if Newton had a Nobel prize, we should revoke it because Newtonian physics doesn't work at speeds (relatively) close to c.

robc|10.14.08 @ 11:38AM|#

Did the formula "not work" or did people plug in the wrong values into the formula?

Chuck|10.14.08 @ 11:39AM|#

Every model comes with simplifying assumptions. If the model doesn't fit, the first thing to do is re-examine the assumptions to see if they are met in this case or not. My guess is not. (Black)-Merton-Scholes is for option pricing, which is a fairly simple instrument compared to the MBS we've been hearing about. If the assumptions about default rates are incorrect, whatever model you try to use is going to fail. Garbage in, garbage out, and all that...

|10.14.08 @ 11:47AM|#

I've always thought people paid way too much attention to the Nobel prizes anyway. The only difference between them and hundreds of thousands of similar prizes is name recognition. There's nothing particularly special or better about the selection process. Particularly with the "political" prizes- peace and economics.

T|10.14.08 @ 11:47AM|#

They're gonna have to get busy yanking medals back if subsequent events prove them wrong. I can see a whole bunch of peace prizes that need to go back to Stockholm.

I also don't think you blame the guys who discovered/wrote the formula for the subsequent misapplication of the formula by a bunch of yahoos.

|10.14.08 @ 11:47AM|#

Efficient risk? Well, I may be an engineer and not an economist, but I know what efficient implies.

Think if it like a factor of safety. If I use less materials in a design it is more efficient but it is more likely to fail. Newer analytical models attempt to justify more efficent designs. Since the things I work on (hydro dams, power plants) are crucial to public safety the models tend to err away from efficiency and more toward conservative design.

I don't think you can strip a nobel prize, but certainly this model led to investment design failure. You can't necessarily blame the model or it's author though. It is now the job of the forensic economist (they all are forensic economists aren't they) to evaluate and update the model and find the range of investments for which it does apply. The models will improve, but efficiency implies increased risk.

Past performance is no guarantee of future results.

|10.14.08 @ 11:53AM|#

I thought Black Merton Scholes just made a model for pricing options given that price movements are normally distributed.....it isn't their fault that prices aren't normally distributed.

|10.14.08 @ 11:53AM|#

The fact that so many people I run into have never heard of Taleb is depressing. More people need to be introduced to TGIF...

TGIF : The Great Inellectual Fraud (aka the bell curve assumption).

|10.14.08 @ 11:54AM|#

Black Scholes has been well known not to work since the 87 crash. After that point in time, there has been a volatility smile - options that cover extreme events are more expensive than they should be based on the BS formula. It's even one sided - the downside ones are always more expensive than the upside ones.

the distribution is not normal, there are tons of studies that show it isn't. the formula is still useful for lots of things like hedging.

|10.14.08 @ 11:59AM|#

Gabe, what Taleb argues is that it is intellectually fraudulent to propose a theory based on assumptions that you know are false and pass it off as a result. Which I agree with only if you perpetuate the myth that your assumptions are correct - Merton and Scholes did this at LTCM. My physics techer in HS let us assume people were point masses to simplify problems - not a fraud, since he clearly stated assumption was ridiculous.

egosumabbas|10.14.08 @ 12:09PM|#

LOL. Amazing that you can place a value on derivatives using an equation without actually buying or selling them. Pure hubris. Repeat after me: SOMETHING ONLY HAS VALUE WHEN TWO PARTIES AGREE ON A PRICE. Duh!

Also, when you think you've found an equation that you think tracks the value of some item, imagine what happens when idiots start making gambles on this equation... your original assumptions will undoubtedly get screwed up and the equation won't apply anymore!

No wonder I only listen to Austrian economists anymore. When you use deduction you're *always* correct.

a name before submitting the f|10.14.08 @ 12:22PM|#

No, I don't think it should be revoked. It is utterly fitting that they give a prize for "economics" to a theory that utterly failed. "Economics," as popularly understood, is a failed concept.

|10.14.08 @ 12:29PM|#

SOMETHING ONLY HAS VALUE WHEN TWO PARTIES AGREE ON A PRICE. Duh!

It's like when someone comes up to you and says, "Hey, I got this $500 watch for $100!"
Answer: "No, fuckdick, you got a $100 watch."

Kolohe|10.14.08 @ 12:32PM|#

It must be noted that Taleb founded a hedge fund which was basically a bet that the Black-Merton-Scholes model does not work.

It went bankrupt.

(For balance, it's a little more complicated than that)

Tym|10.14.08 @ 12:36PM|#

The formula uses volatility as a proxy for risk. Volatility value from the past may not be accurate in the future. It is just an approximation that under normal circumstances holds.

egosumabbas|10.14.08 @ 12:39PM|#

"It must be noted that Taleb founded a hedge fund which was basically a bet that the Black-Merton-Scholes model does not work.

It went bankrupt."

If enough people believe (and bet) that the model is true, and you bet against it, and people keep plowing money into the bet who have more money and religious fervor than you do, well yeah, even if you're "correct" you'll still go bankrupt.

It's amazing how high some things will go (e.g. internet bubble) before they come crashing back down onto the ground.

|10.14.08 @ 12:41PM|#

egosumabbas - many derivatives can be priced based on the component pieces. Moreover their theoretical prices can (sometimes) be enforced by arbitrage. That in no way prevents people from doing stupid things like trading them at the wrong price. Agreement on price does NOT imply correctness.

egosumabbas|10.14.08 @ 12:48PM|#

@domoarrigato

Uh no, price is what people agree it to be. This is by definition.

Jerry|10.14.08 @ 12:50PM|#

Tym, like a chicken who thinks that every day he gets fed, he'll assume more and more that his feeder is benevolent. Until the day the butcher shows up :) That basically how risk models work these days.

|10.14.08 @ 12:51PM|#

then your statement is a tautology, and is pretty uninsightful.

If someone is willing to sell me some security for 90 and someone else is willing to buy it for 95 at the exact same moment in time - what is your definition based view of it's price?

egosumabbas|10.14.08 @ 12:58PM|#

The two parties hash out a deal and agree to a new price. I could make a *bet* that it's really one or the other, but then I'm assuming a risk. There's no way to know which one is correct until a deal is made.

mike|10.14.08 @ 12:59PM|#

What rubbish. This is even worse than saying guns kill people - its more like saying that bullets kill people. Yes of course BS, risk neutral and every other tenet of quantitative finance can be said to be 'dangerous'. If put in the wrong hands and misused.

egosumabbas|10.14.08 @ 1:00PM|#

Also, I could make a deal with both parties at both prices without letting each party know, and make money as the middle man. That's called arbitrage.

|10.14.08 @ 1:03PM|#

LOL. Amazing that you can place a value on derivatives using an equation without actually buying or selling them. Pure hubris. Repeat after me: SOMETHING ONLY HAS VALUE WHEN TWO PARTIES AGREE ON A PRICE. Duh!

That's true in a sense, but it's not what this formula is about.

The formula figures out the expected value of an option or derivative at the time when the option/derivative is exercised, and then comes up with a price based on that expected value and the "time -value of money". It doesn't assert that people will always buy/sell at that price.

For a simple (and unrealistic) example, suppose I am selling a call option on a barrel of oil, (exercise date 1 year from now). The strike price is $100.00 and we know for a fact that the going price for oil will be $110.50 in a year. And let's also say the going rate of interest is 5%.

It would not make sense for someone who wants to maximize his/her profit to buy my option for more that $10, because that $10 now will get you $10.50 (110.50-100.00) in a year (a 5% return). Paying more would mean you get less than the 5% return.

Similarly it would not make sense for me to sell for less than $10, since I'd be effectively borrowing money for more than the 5% interest rate.

Option pricing methods would put the price at $10. But that isn't a statement that I won't be able to get some crazy person to buy the call option from me for $1000, or that nobody will buy or sell at different prices (due to error or whatever reason).

Of course, the real world is more complicated; with a range of possible future going prices, changes in interest rates, different rates available to different individuals, uncertainty in what will happen to cause these changes, etc. But doing the math to estimate expected values can still be a useful exercise in many cases, and its not the same as making up numbers.

egosumabbas|10.14.08 @ 1:06PM|#

Also, I think it's funny that you say that tautology (e.g. the truth) is uninsightful. I guess lies are far more interesting. Just keep throwing your money away on your subscriptions to The Economist and The Wall Street Journal.

|10.14.08 @ 1:07PM|#

Taleb is getting to be insufferable.

The great achievement of the BSM model is not the assumption of normality, but rather that it paved the way for arbitrage-free valuation models for options -- some of which may have false assumptions, and some of which may not.

The analogy to Newtonian mechanics seems apt. It may be strictly false, but it works fine sometimes, and if not for that breakthrough, we'd never have got to relativistic mechanics.

|10.14.08 @ 1:12PM|#

yes its an arbitrage, genius. The question - which remains unanswered - is which one is the price?

Saying a tautology is true is another dumb tautology. Get a dictionary, and figure out where you screwed up.

|10.14.08 @ 1:13PM|#

If someone is willing to sell me some security for 90 and someone else is willing to buy it for 95 at the exact same moment in time - what is your definition based view of it's price?

Also, I could make a deal with both parties at both prices without letting each party know, and make money as the middle man. That's called arbitrage.

Yes, you could; and yes it is called arbitrage.

One things that these sorts of formulas could be used for is to identify opportunities for risk-free arbitrage (or nearly risk-free, near-arbitrage), in more complicated circumstances, based on the "going prices" of securities/derivatives/etc.

egosumabbas|10.14.08 @ 1:13PM|#

@BG

How can you say that the equation is true when you have no idea what future prices will be? All you're doing is making an educated guess (as opposed to a wild-assed one).

I would wager that this equation is part of the reason why we're in the mess we're in today. People would make unrealistic estimations of rate of appreciation way out into the future, set prices on those projections, and continue ad infinitum.

For example, assuming that "houses go up forever" or that "oil goes up forever since we're running out of it". How much you wanna bet that those assumptions royally screwed up the "values" of derivatives? If you commit a logical fallacy in your assumption, your fancy economic model will also result in a fallacious conclusion.

|10.14.08 @ 1:17PM|#

"People would make unrealistic estimations of rate of appreciation way out into the future, set prices on those projections, and continue ad infinitum."

If you are willing to bet that estmations of future prices impact blackscholes pricing in anyway, I will take your bet up to the limit of your net worth.

egosumabbas|10.14.08 @ 1:19PM|#

"The analogy to Newtonian mechanics seems apt. It may be strictly false, but it works fine sometimes, and if not for that breakthrough, we'd never have got to relativistic mechanics."

Or to 700 billion dollar bailouts.

But I swear that my economic model said that the values I set for those derivatives was true! Not fair!

"yes its an arbitrage, genius. The question - which remains unanswered - is which one is the price?"

I guess it depends if there's another buyer or not. Say I bought two contracts at 90, and only sold one at 95, and another buyer came along. The price would be 95, wouldn't it?

Also, it helps if you create a point of reference. I'm obviously assuming myself and a future buyer as a point of reference.

egosumabbas|10.14.08 @ 1:22PM|#

"If you are willing to bet that estmations of future prices impact blackscholes pricing in anyway, I will take your bet up to the limit of your net worth."

I was actually planning on short selling home builders and mortgage lenders a couple years ago, but since my net worth and credit is low (I'm in my 20's), and I only had a ROTH account, I wasn't able to make my bet. I would have made a shit ton of money though.

|10.14.08 @ 1:24PM|#

"I guess it depends if there's another buyer or not. Say I bought two contracts at 90, and only sold one at 95, and another buyer came along. The price would be 95, wouldn't it?"

We are getting closer. obviously whoever is willing to trade more wins. This is called the law of one price. Now what happens if there is a seller at 95 and no one is willing to buy? what is the price?

|10.14.08 @ 1:25PM|#

How can you say that the equation is true when you have no idea what future prices will be? All you're doing is making an educated guess (as opposed to a wild-assed one).

I would wager that this equation is part of the reason why we're in the mess we're in today. People would make unrealistic estimations of rate of appreciation way out into the future, set prices on those projections, and continue ad infinitum.

For example, assuming that "houses go up forever" or that "oil goes up forever since we're running out of it". How much you wanna bet that those assumptions royally screwed up the "values" of derivatives? If you commit a logical fallacy in your assumption, your fancy economic model will also result in a fallacious conclusion.

Derivative-pricing formulas don't tell you what assumptions to make about the range of possible, or likely, future prices of the relevant asset. They merely are a way to do the math once you have your assumptions.

Of course, if your assumptions are wildly incorrect, you are likely to also get a wildly incorrect estimate of the profit/loss potential from an investment strategy using pricing formulas.

egosumabbas|10.14.08 @ 1:30PM|#

"Derivative-pricing formulas don't tell you what assumptions to make about the range of possible, or likely, future prices of the relevant asset. They merely are a way to do the math once you have your assumptions.

Of course, if your assumptions are wildly incorrect, you are likely to also get a wildly incorrect estimate of the profit/loss potential from an investment strategy using pricing formulas."

Okay, now we're getting somewhere. How do I set up a falsifiable experiment to determine that this equation holds water?

"Now what happens if there is a seller at 95 and no one is willing to buy? what is the price?"

If nobody is willing to negotiate (possible but unlikely), 90 would be the price. If either is willing to negotiate, the value would be somewhere near 92.5, but that isn't a guarantee.

|10.14.08 @ 1:39PM|#

Calumnies.
It is NOT true that Taleb's fund had problems. This is a myth. Do you have SOURCES?

|10.14.08 @ 1:42PM|#

egosumabbas -

"If nobody is willing to negotiate (possible but unlikely), 90 would be the price."

no one is willing buy at 90 anymore. there is no buyer, only sellers. There is no price.

To belabor the already tedious, the point is that there can be multiple prices (briefly, thanks to arbitrage) and no prices (when markets are disfunctional, like in mortgage CDO shlock). Regardless, neither price predicts or is based on any certainly of what future prices will be - moreover, the prices may be based on sellers legal requirement to buy/sell rather than some economic assement of their worth. Clearly "value" as you stated it is not at all as simple as "what people are willing to transact" even without nifty derivative pricing problems.

egosumabbas|10.14.08 @ 1:46PM|#

I should be more specific: if nobody is willing to buy at 95, and there is another selling for 90, the only information that you can gather is that the new price would be less than 95.

egosumabbas|10.14.08 @ 1:51PM|#

"no one is willing buy at 90 anymore. there is no buyer, only sellers"

I'm sorry, you didn't make that clear. All you said is that are no buyers for 95. No need to get snarky.

Obviously, if there are no buyers, then all the sellers prices are too high. Until there are buyers the price will come down gradually to zero (like that $1 house in detroit).

However, I would disagree that the price IS zero. You don't know what buyers will become interested until the prices start dropping first

|10.14.08 @ 2:00PM|#

i don't think the price is 0 either - i think there is no price. With a gun to my head - GAAP, SEC, Congress - I have to estimate it. Moreover, pulling a number out of my ass is likely to be viewed unfavorably. So I use some crazy model - it might not be any better than the ass-number, but at least I don't get sued. Note: Black scholes, which is a crazy model fom 30 years ago is way more defensible than the more recent crazy models that were used to make CDO's, credit derivatives, etc. BS has its limitations, but it is far from totally useless which was my only point.

|10.14.08 @ 2:07PM|#

The value of an investment something to you is the sum of the cash flows you receive after applying the time value of money. If others have a higher time value of money than you, then no one will buy it from you, but it still has a value.

brec|10.14.08 @ 2:42PM|#

Ron, you copied NPR's misspelling of Nassim Taleb's first name.

egosumabbas|10.14.08 @ 3:08PM|#

So, Mr. Roboto, the government FORCES economists to come up with crazy models for prices, simply because an economist's word is better than mine? Can you see what's wrong with this picture?

|10.14.08 @ 4:21PM|#

Okay, now we're getting somewhere. How do I set up a falsifiable experiment to determine that this equation holds water?

Well, I guess you could look at the performance of hedge funds (well the option/derivative portion of the funds anyway) that used that method versus the performance of funds that didn't.

If the ones that used the equation did better, then the ones that didn't, that could be evidence that there is something to it. If not, that could be evidence that it's bogus. Although there is some ambiguity about what the correct control group would be (all hedge funds that don't use this formula? all that don't use any formula? some measure of overall market performance?).

It's not as scientifically precise as I'd like, but I guess its something.

|10.14.08 @ 4:28PM|#

egosumabbas - you are attributing an argument to me that I didn't make. I pointed out that crazy models (not BS) are sometimes used to value illiquid securieties because of regulations. Of necessity, the models predate the securities - not the other way round.

Paul Davis|10.14.08 @ 6:25PM|#

Taleb's first book was great. His second book made me want to drive to New York and ask him personally for my money back. What a disappointment.

|10.14.08 @ 6:35PM|#

The Black-Scholes formula has two very serious mathematical flaws. The first flaw is that it treats historical investment returns as if they constitute an independently determined data set for statistical analysis purposes. Historical investment return data is clearly conditional data. The second flaw is that "continuous dynamic hedging" is not arbitrage. This point probably should have been the key lesson learned from the Long-Term Capital Management disaster, but evidently was not. Real arbitrage can occur through the application of the put-call parity theorm, but not hedged equity trading based on a single option contract.

These errors are fully and completely described, and to the extent possible, corrected in two articles found on the website www.cpmvals.com. The articles are posted, and available for a full and complete mathematical review.
By RRJoss

Simon Cranshaw|10.15.08 @ 3:28AM|#

Merton-Scholes formula doesn't work

I've been an options trader for ten years and it seems like this is nonsense. It may well be that the Nobel was not deserved. That's a question for historical analysis and maybe Taleb has a point.

But the formula is fine and it is still in general, even universal use. True, you can't use one volatility. Duh! But once you plug in a volatility surface and some way to model that surface you can do valuation and hedging the best way that is available to my knowledge. You can fit market data. You can model all the kinds of extreme events that we see. You can reflect the real data distributions. It also gives you simple ways to look at relative options spread values through the implied surface. I don't see any real weaknesses to the formula when used this way and I don't see it going anywhere.

|10.15.08 @ 10:41AM|#

How well would this work applied to, say, the Zimbabwean economy?

Now, there's an outlier.