The Fourth Quadrant: a Map of the Limits of Statistics (by Nassim Nicholas Taleb)(edge.org)
edge.org
The Fourth Quadrant: a Map of the Limits of Statistics (by Nassim Nicholas Taleb)
http://www.edge.org/3rd_culture/taleb08/taleb08_index.html
7 comments
I really like this essay. It almost makes too much sense, like it is too simple an idea to have gone ignored for this long, but the premise is hard to argue with: you can't accurately predict events that happen very rarely (on the time scale that we have accurate measurements for) that have extreme outcomes.
Also, reading this reminds me how doomed younoodle.com is.
Also, reading this reminds me how doomed younoodle.com is.
Ah, but just because you can't accurately predict events doesn't mean you can't convince people you can accurately predict events.
Perfect timing given the current financial crisis which was triggered in part by an over reliance on (inaccurate) statistical models.
I'm pretty sure the timing is planned rather than coincidental. He mentions the "busts" of both Bear and Lehman in the notes on figure 2.
Or maybe I'm misunderstanding you and you mean the timing of his release of 'black swan'.
If so, I still have to say that from what I can tell, Taleb has been espousing these same ideas for several years, so it's still not just coincidence that his ideas are getting a wider audience at the same time that the meltdown he warned about is taking place.
Or maybe I'm misunderstanding you and you mean the timing of his release of 'black swan'.
If so, I still have to say that from what I can tell, Taleb has been espousing these same ideas for several years, so it's still not just coincidence that his ideas are getting a wider audience at the same time that the meltdown he warned about is taking place.
Talk about being at the right place at the right time. Taleb may have even timed the release of this new essay.
The essay makes this assertion a couple of times: "The banking system .. just lost > 1 Trillion dollars (so far) on a single error, more than was ever earned in the history of banking."
Really, more than was ever earned in the history of banking? To me that would seem to imply that at this point every banking institution would be bankrupt.
Shouldn't a guy who chides people for sloppy thinking be rather more careful, or am I just missing something and being stoopid?
Really, more than was ever earned in the history of banking? To me that would seem to imply that at this point every banking institution would be bankrupt.
Shouldn't a guy who chides people for sloppy thinking be rather more careful, or am I just missing something and being stoopid?
It's easy to confuse the value banks create with the money they have on hand. For example, take a look at bank of america. They earn about 1.80 for every share outstanding, and there are about 4.5 billion shares. so BofA creates about 8.1 billion of value every year (given todays rates). So with a loss of a trillion dollars, it would take BofA 123 years to make that back.
You have to remember, banks take a tiny fraction of a ton of money. They make the promise, give me a buck today, and i'll give you back $1.02 in a year, with no risk. You created all the value in the initial dollar, they just make the 2 cents. So, when a bank doesn't even give you your dollar back, they wipe out a ton of value.
You have to remember, banks take a tiny fraction of a ton of money. They make the promise, give me a buck today, and i'll give you back $1.02 in a year, with no risk. You created all the value in the initial dollar, they just make the 2 cents. So, when a bank doesn't even give you your dollar back, they wipe out a ton of value.
[...] More than was ever earned in the history of banking? To me that would seem to imply that at this point every banking institution would be bankrupt.
Actually, all you would need is some subset of banking institutions to go _spectacularly_ bankrupt, and end up owing, in aggregate, more than the others have ever made. Of course, the institutions would have to be big to make that happen. Say, the size of Bear Stearns, Lehman brothers, or Merrill Lynch...
I'm not a domain expert (I'm sure we have a number of them around), but it sounds at least within the realm of plausibility.
Actually, all you would need is some subset of banking institutions to go _spectacularly_ bankrupt, and end up owing, in aggregate, more than the others have ever made. Of course, the institutions would have to be big to make that happen. Say, the size of Bear Stearns, Lehman brothers, or Merrill Lynch...
I'm not a domain expert (I'm sure we have a number of them around), but it sounds at least within the realm of plausibility.
Because when they went under they took all of their depositors money with them, and left their creditors in the lurch. This point is really why his thesis is so important.
I just want to say something about the banking crisis though, to get it off my chest - it will happen again, sooner or later. Even though new financial regulation will likely be introduced to prevent a crisis like this from happening again, over time institutions will develop ever more complex instruments and trading rooms to find a way around them - as they have done so historically - and make turkeys out of all of us, once again. The fundamental problem is the lack of political will to maintain oversight and keep regulation up-to-date.. politicians (and the rest of us) only really take notice when something goes badly wrong. How do we solve this problem?
I just want to say something about the banking crisis though, to get it off my chest - it will happen again, sooner or later. Even though new financial regulation will likely be introduced to prevent a crisis like this from happening again, over time institutions will develop ever more complex instruments and trading rooms to find a way around them - as they have done so historically - and make turkeys out of all of us, once again. The fundamental problem is the lack of political will to maintain oversight and keep regulation up-to-date.. politicians (and the rest of us) only really take notice when something goes badly wrong. How do we solve this problem?
this part is relevant to startups: (see figure 7)
"A biotech company (usually) faces positive uncertainty, a bank faces almost exclusively negative shocks. I call that in my new project "concave" or "convex" to model error."
concave to model error: Bank, Short volatility, security
convex to model error: Long options, biotech, technology, entrepreneurship, etc
"A biotech company (usually) faces positive uncertainty, a bank faces almost exclusively negative shocks. I call that in my new project "concave" or "convex" to model error."
concave to model error: Bank, Short volatility, security
convex to model error: Long options, biotech, technology, entrepreneurship, etc
I find his ideas fascinating. I also wish he had a much better editor, with good grammar.
So I guess he is saying startups have a small but real chance of massive success. True, maybe obvious, but still music to my ears!
So I guess he is saying startups have a small but real chance of massive success. True, maybe obvious, but still music to my ears!
Taleb is the man.
I think he would totally agree with you. Try to read Fooled by Randomness to see what I mean. I gave up after the first few chapters because I couldn't stand his unstructured - almost incomprehensible at times - arrogant, self-referential prose. Perhaps he has a lot to say, I just wish he could write too.
In fact, I believe totally otherwise. He is not only a great thinker but an excellent author also. While reading Fooled by Randomness, I literally wished the book could go on and on.
If he is pompous, he is so because he wants to drive home points which people think are unconventional and idiosyncratic. But in fact the points are real. I just wish people could understand him more, without any biases whatsoever.
If he is pompous, he is so because he wants to drive home points which people think are unconventional and idiosyncratic. But in fact the points are real. I just wish people could understand him more, without any biases whatsoever.
I slogged through The Black Swan with the same attitude. He had some good insights, but he was so pompous I had difficulty following his reasoning. I'd continually get derailed when I encountered something that I thought was pure self-promotion.
There is no greater crime in this country than creating something of value and being proud of it.
Other people often called "arrogant" - Paul Graham, Steve Jobs, etc.
Other people often called "arrogant" - Paul Graham, Steve Jobs, etc.
Be proud all you want. Just be able to explain your idea clearly. Taleb intersperses self-promotion in his explanations, which muddies them.
His writing style in general is meandering and lacks focus. The book I read had a handful of insights amid pointless narration. I read books like The Black Swan to learn new ideas, not to learn how great the author is.
His writing style in general is meandering and lacks focus. The book I read had a handful of insights amid pointless narration. I read books like The Black Swan to learn new ideas, not to learn how great the author is.
Often it seems that people don't have a problem with the "arrogant" people themselves, but those who choose to surround them.
Totally agree - pompous. That's the word I was looking for. The kind of author who creates a fictional character based 100% on himself, makes said character the intellectual and professional superior to his peers, the only one who survives the ups and downs of the market; and to top it all off names the character Nero! Seriously!
I found that "When Genius Failed" gives a great story-telling journalistic perspective on some of the same incidents Taleb talks about in his book while "The Drunkard's Walk" provides a good mathematical intuition (sans math sadly) behind seemingly rare events.
I found that "When Genius Failed" gives a great story-telling journalistic perspective on some of the same incidents Taleb talks about in his book while "The Drunkard's Walk" provides a good mathematical intuition (sans math sadly) behind seemingly rare events.
<3
You have to be willing to deal with terrible grammar, but aside from that this essay makes the necessary points about as succinctly as you could, I think. It still takes a while to get through, but very worthwhile.