Fooled by Randomness

[Martinghoul]

Quote from johnny88:

I know he doesn't believe in CAPM etc. This question was asked in class and have to do a paper on it. I think I know the answer here now. He basically says we don't focus on the unpredictable risk in a firm, and we only focus on things like CAPM and M&M model for finding risk. How you would quantify unpredictable risk is beyond me. What would you say is an alternative to CAPM? Since we live in a imperfect world, capm with its assumptions is useless. We haven't really learned any other models extensively as M&M and CAPM. Thanks.

I don't know any widely used alternatives to CAPM, personally. I am sure you can google and possibly find some literature, but I can't give you names off the top of my head.

Point is that CAPM is a model. It's based on some assumptions. Everyone apart from Taleb seems to realize that a model is exactly that, no more no less. His ravings against these models (CAPM, VAR, BS etc) are completely idiotic. His ravings against people who forgot that these models have flaws are sensible.

Personally, I think one of the main problems of all arbitrage models lies in the risk-neutrality assumptions. If I were to dig anywhere, that's where I would start. However, people much smarter than myself can't figure out how to mathematically deal with models, where risk neutrality of behavior is violated.

 

[Swan Noir]

One of Taleb's problems is a common one. He took a point of view that, if properly presented, could have been anice 1,200 or so word article that might have made him a few G's and secured his reputation in the minds of his mother and wife.

The aunt will always have her doubts.

The payoff for stretching it to a book was huge. In both financial and non-financial terms. I must say that rarely has someone perverted anything so much as he stretched it. No sexual pun intended.

At least half of what he had to tell us could have been done by the following sentence.

"Good (or even great) models will produce absolute crap if you take your vested interest (think investment banks and rating firms) and use assumptions that are not just suspect but that you know are the flip side of reality."

So much of the agony this society will eat for years to come has been produced by assumptions that had not a shred of honestly or reality to them.

It is by definition fraud. But fraud on a widespread basis perpetrated by those close to the seats of power and part of that strata of society are never investigated. A "commission" when formed will have the unspoken mandate to put this thing to bed. After all they will say -- "there is enough blame to spread around". The code for no one in our social set need worry.

We were not hit by an outside the box event -- think meteor destroying Beaumont, Tx and causing a spike that resulted in a crash-- we were destroyed by one of the most predictable results of a society passing the line of reason and stepping into the abyss.

The assumptions were lies. No model can correct that.

 

[Thunderdog]

Quote from Martinghoul:

...Personally, I think Taleb's completely useless where constructive analysis of any sort is required. He hates CAPM, but doesn't offer an alternative analytical framework.

Nevertheless, there is value in identifying bullshit where it exists.

 

[Martinghoul]

Quote from Thunderdog:

Nevertheless, there is value in identifying bullshit where it exists.

Yes, but, as I keep saying, there are people in academia that do it a lot better than NNT. His only claim to fame, as far as I am concerned, is that he screams the loudest.

 

[trendlover]

Quote from johnny88:

Only reply if you know what Capital Asset Pricing Model (CAPM) is and have read the book Fooled by Randomness. Thanks.


Question: How would you use CAPM and the Black Swan Principle to explain the severe decline in equity values in late 2008 and early 2009? One obvious is hidden risk that wasn't included in calculation of projected equity values. What are some others?

Maybe you can challenge "hidden risk" to say in your paper that rating agencys for morgage back securities gave good ratings to bad securities. Rating is measure of risk, right? So you can say is a black swan principle what happened from 2008 to early 2009, or was that something that can be seen, but people want profit today, then pass that risk to someone else, and who care about the future of the company they work for? Greed.
You can say the rating "model" is flawed, and everything that measure risk after that will be flawed too for equitys of 2008 to 2009.
I did not read Taleb book, but I understand he thinks some risk can never be seen, so it can not be measured.

 

[Thunderdog]

Quote from Martinghoul:

Yes, but, as I keep saying, there are people in academia that do it a lot better than NNT. His only claim to fame, as far as I am concerned, is that he screams the loudest.

Judging by recent history, perhaps no one was screaming loud enough? What constructive analysis were the others offering?

 

[Thunderdog]

Quote from Swan Noir:

One of Taleb's problems is a common one. He took a point of view that, if properly presented, could have been anice 1,200 or so word article that might have made him a few G's and secured his reputation in the minds of his mother and wife.

The aunt will always have her doubts.

The payoff for stretching it to a book was huge. In both financial and non-financial terms. I must say that rarely has someone perverted anything so much as he stretched it. No sexual pun intended.

At least half of what he had to tell us could have been done by the following sentence.

"Good (or even great) models will produce absolute crap if you take your vested interest (think investment banks and rating firms) and use assumptions that are not just suspect but that you know are the flip side of reality."

So much of the agony this society will eat for years to come has been produced by assumptions that had not a shred of honestly or reality to them.

It is by definition fraud. But fraud on a widespread basis perpetrated by those close to the seats of power and part of that strata of society are never investigated. A "commission" when formed will have the unspoken mandate to put this thing to bed. After all they will say -- "there is enough blame to spread around". The code for no one in our social set need worry.

We were not hit by an outside the box event -- think meteor destroying Beaumont, Tx and causing a spike that resulted in a crash-- we were destroyed by one of the most predictable results of a society passing the line of reason and stepping into the abyss.

The assumptions were lies. No model can correct that.

That's pretty good.

 

[Random.Capital]

Quote from johnny88:

Risk free rate...

...doesn't actually exist and...

+ Beta(Market Risk Premium).

...it has been demonstrated over and over that past beta and future beta are second cousins at best.

So what does this tell you in accordance with the mayhem we saw in 2008? [/B]

That when you build a model on assumptions that are often orthogonal to market reality, you have to expect the occasional strong divergence between what happens and what "should" happen.

 

[Random.Capital]

Quote from Martinghoul:

Everyone apart from Taleb seems to realize that a model is exactly that, no more no less.

I wish this were true. The typical salesman - sorry, "advisor" - servicing retail customers trots that model out on a regular basis. Usually it's not expressed explicitly as "CAPM", but rather as a set of vague principles that just happen to align with the assumptions that go into CAPM.

 

[sosueme]

Quote from Swan Noir:

-- we were destroyed by one of the most predictable results of a society passing the line of reason and stepping into the abyss.

The assumptions were lies. No model can correct that.

Ahhhhh a thinking man, the only basis for the creation of a great Trader.

sosueme