Nassim Nicholas Taleb on the Financial Crisis
Mar 23 2009

Nassim Taleb talks about the financial crisis, how we misunderstand rare events, the fragility of the banking system, the moral hazard of government bailouts, the unprecedented nature of really, really bad events, the contribution of human psychology to misinterpreting probability and the dangers of hubris. The conversation closes with a discussion of religion and probability.

Nassim Nicholas Taleb on Black Swans, Fragility, and Mistakes
Nassim Taleb, author of The Black Swan and Fooled by Randomness, talks with EconTalk host Russ Roberts about his latest thoughts on robustness, fragility, debt, insurance, uncertainty, exercise, moral hazard, knowledge, and the challenges of fame and fortune.
Charles Calomiris and Stephen Haber on Fragile by Design
Charles Calomiris of Columbia University and Stephen Haber of Stanford University, co-authors of Fragile by Design: The Political Origins of Banking Crises and Scarce Credit, talk with EconTalk host Russ Roberts about their book. The conversation focuses on how politics...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Mar 23 2009 at 7:01am

Not again…

Nothing against him as a bright dude, but his idea is a one trick pony and not original. What we are experiencing is not a black swan. geez.

John S.
Mar 23 2009 at 9:39am

Like Alex I had grown tired of Taleb, but I found this interview to be a breath of fresh air. Mr. Taleb seems a whole lot more humble this time around. I suspect that his vigilance for black swans has been completely useless during the current crisis, and that he too has taken some big losses.

Alex C.
Mar 23 2009 at 10:49am

Fellows, deriding fundamental insight such as his as a “one trick pony” does not do it justice. By that logic, Einstein and Maxwell were one trick ponies as well because you could sum up their entire life’s work with a few simple equations. Capturing complex behavior with a few equations or heuristics is the height of achievement.

Had we all followed Taleb’s investing advice (80% T-bills, 20% stocks,high risk, etc.) our losses would have been much more palatable.

You can’t be “vigilant” for black swans because by definition, you can’t see them coming. I believe what Taleb is saying is that you need to set yourself up to succeed big if and when they do come and lose only a little if they don’t come.

Alex D.
Mar 23 2009 at 11:07am

I’m sorry, but you may recall that the easiest thing in the world to do is theorize…in this case regarding the success or failure of a prominent figure in today’s financial media. I agree with Alex C., his idea may not be original, but it is definitely relevant and grossly under appreciated.

Mar 23 2009 at 11:26am

There is much about too big to fail in this podcast. It makes me ask: are the USA Federal Reserve and FDIC to big to fail? Should we have multiple currencies in the USA and multiple Federal Reserves? Should the FDIC have acted like an intelligent insurer and started to threaten to drop insurance of depositors in banks that were taking too much risk? If so do we need 5 or 6 competing FDICs?

Mar 23 2009 at 11:28am

Excellent podcast w/ Taleb Russ. Your discussions of late on knowledge, uncertainty and bias have been terrific. Keep pushing!

Re: John S on suspect of Taleb taking losses: There were several articles published in November 2008 about Taleb’s funds making remarkable profits (WSJ, Financial Times, so possibly just tabloid). If true and if Taleb has since been holding 80% cash equiv. and 20% buying out-of-the-money options he has probably done quite well since then also.

Taleb has been consistent. If we find his advice useless during the current crisis, remember: 1. that we had the opportunity to follow it before the crisis started; and, 2. we have no way of knowing what the rest of the crisis has in store for us.

Ted C.
Mar 23 2009 at 11:35am

He put his money where his mouth was and look at the results…

Mar 23 2009 at 11:35am

Universities having a negative impact on knowledge is interesting to me. I see in schooling testing squeezing out education. E.G. almost everyone can understand the principles of physics without the math but everyone wants the class to be rigorous and so we put enough math in it to make it rigorous and so withhold the education value to many.

BTW great pod cast thanks.

Cap Khoury
Mar 23 2009 at 1:09pm

I’m a mathematician with a developing interest in economics.

I’m very interested in Taleb’s remark that the long-run statistics tend to look “rosy” because of “survivor bias”. Can anyone recommend some literature with some detailed analysis of this phenomenon? (I don’t want a brief overview “in layman’s terms”.)

Dan K.
Mar 23 2009 at 1:44pm


“Separate funds in the Universa’s Black Swan Protection Protocol were up between 65 percent to 115 percent in October [2008] alone.”

The question remains: Could he keep up those returns or was he ‘calibrated’ to short-term shocks like Oct. ’87?

My guess would be that the returns were closer to zero (i.e. cash only position) in current months after the options were exercised, but that is still better than a negative return.

Mar 23 2009 at 1:50pm

I have really enjoyed your shows dedicated to epistemology and philosophy. We simply do not get enough exposure to these topics in media so your show is a real breath of fresh air. Keep up the great work!

John Thurow
Mar 23 2009 at 2:10pm

I enjoyed the podcast.

The insight about the internet causing global fads is interesting in lieu of the current “financial crises”. I am surprised that in the analysis of mortgage default risk there was no thought in how the loans were being used i.e. it sounds like the analysis prior to 2006 was based upon passed default rates, etc. and if I understand Mr. Taleb’s argument, the amount of losses due to the low probability events were not being correctly accounted for. I am not sure what percentage of the housing market in the past was based on house flipping but in the years 1998 to 2006, a lot, if not the majority of the loans and marginal loans were going to house flippers, who were borrowing money for practically nothing at no risk – they could walk away from the loans in at least 8 states that have laws protecting them without penalty or having the banks pursue them. To do this the house flippers borrowed from the least interest loans. These house flippers were all tied together via the internet: blogs, newsletters, chatrooms, etc. so when the first signs showed the housing market prices flattening, the house flippers all walked away from their no risk loans in masse causing the start of the current “crises”.
Perhaps one of the reasons the risk analysis failed was that the analysis failed to take into account the entire user community of mortgage loans and how that mortgage loans were being used. Such analysis would of shown that the loans were not as diversified as initially thought due to a large quantity going to people who were more interested in flipping and making money from the flipping then living in the home. The amount of defaults at the same time in this case would be a lot higher then in a standard model that each home was individually owned and was the primary residence leading to a greater chance of a precipitous collapse.
Thus, was the risk analysis wrong solely because of not characterizing the risk and the dollar values associated with what was considered a low probability event or was the risk analysis wrong because it failed to understand the problem correctly and take into account the entire system and the players therein including the house flippers, the number of, how much was being loaned to them and how they would behave as a group?

Robert York
Mar 23 2009 at 2:14pm

Oh please. Free enterprise works. Taleb is a socialist. He’s just trying to regulate markets. Markets don’t need regulation. They self-regulate. This is self-evident.

Mark K
Mar 23 2009 at 2:16pm

Since Russ did not seem to press Taleb on the point about government regulators pushing the ‘value at risk’ model, I would like to know if anyone here as any evidence to support this assertion? Your explanation should not include the past precedence of government bailouts, which is an entirely different thing than endorsing (or promoting) the VAR mathematical model. He says government wanted VAR, but does not offer any support for this, except for a vague reference to rating agencies, which to my knowledge are private profit making companies, not government agencies. Give citations to support any assertion you make.

Brad Hutchings
Mar 23 2009 at 4:11pm

The remarkable thing about Taleb is that he drives his critics bat guano crazy. His message isn’t about trading securities, nor is it’s appeal about trading. It’s about hedging against unseen risk and recognizing the kinds of things of can’t be known. The options markets are his laboratory and a context for his stories.

Here is my take-away from Taleb: Hedge against the mob, because the mob creates its own black swans. One way I’ve done that in the past two years with a software product I develop is to resist the tech fads. Take “Web 2.0”. Our product has some elements that Web 2.0 cheerleaders might call “Web 2.0”. But it’s not. And to be seen as “running in the cloud” takes away about 4/5 of the value proposition of our product, specifically the ability to use very inexpensive desktop computing power (CPU, RAM, disk, printing) to locally produce and publish children’s books.

Last month, I read about how “Web 2.0” as a term was beyond cliche and near death, according to Google search stats. Google says the term is growing in eastern Russia, of all places, while on the decline pretty much everywhere else. And it’s probably still on the rise in education IT circles, where acronym tossing has always substituted for knowledge ;-). While we keep doing what we do, grow our user base, grow our collection of children’s books, our competitors who embraced “Web 2.0” soon have to find the new fad to wrap their services in. Not only do we save the rebranding effort, we’ve taken a steady course. That differentiation is quite valuable to education customers who want long term value from their investments. And when everyone agrees that Web 2.0 was the mullet haircut of the Internet age, I will happily chide competitors up for going down that road.

Mar 23 2009 at 4:39pm

Mark K,
You might find what you’re looking for here, at Taleb’s notes page, especially near the body of the page.

Mar 23 2009 at 4:40pm

*near the BOTTOM of the page, that is.

Russ Roberts
Mar 23 2009 at 5:27pm

Mark K,

Google “regulators VAR” and you’ll find stuff on how regulators began using VAR because it was used by the industry. Hope to take a closer look at VR in a future podcast.

Russ Roberts
Mar 23 2009 at 5:29pm

Cap Khoury,

Google “survivor bias mutual fund active management” and you’ll find studies and news reporting on the topic. It is a standard claim of index funds, for example, that the performance of managed funds is biased by ignoring funds that have gone out of business.

Lee Kelly
Mar 23 2009 at 5:45pm

There are many people who know how to manipulate the numbers, and who understand the formal language of mathematics better than me, but I know enough about logic to know that few people understand what probabilities actually mean, their informative content, and limitations. The desire to probabilify knowledge is born from the failure of naive empiricism, logical positivism, and inductive epistemology generally. It is the same research program except for the substitution of certainty for partial certainty. But such does not even begin to solve the problems therein.

People cannot measure their own ignorance.

Mar 23 2009 at 6:12pm

So just assume anything with risk can go to zero, and be 100% equity funded for that, the rest in T-bill like securities? That’s not an equilibrium, because there’s no fine line on what is ‘risky’.

The assertion that leverage in Europe increased three fold is simply not true, unless he is talking about notional values of derivatives, but that is silly because the vast majority of these contracts offset each other.

Mar 23 2009 at 6:13pm

I enjoyed this talk with Taleb and am fond of many of his ideas. I do have a few problems. I think in his effort to counteract some biases he has possibly created his own. For example, I’m willing to admit that Universities and other institutions of “higher” education are probably not the “centers” of learning that they are touted to be. However, if most people were told that their brain-surgeon, or the the lead engineer on the plane they are in, had not received such an education I don’t think they would be so fast to scoff at their role.

This is definitely not the place to get into religio-philisophical discussions but I think the ideas of looking towards tradition, the wisdom of the crowds, and avoiding hubris can be used to make many opposing arguments. I personally don’t see how not being raised in a religious community, or joining one, is a form of hubris. Although the whole “God knows” thing is a nice story it makes for a rather silly argument about humility and I think many religious people would be surprised to find out their religion isn’t based on any specific beliefs just practices.

He spends some time deferring to the wisdom of the “Ancients.” Now don’t get me wrong I think you can learn a lot reading writings from throughout history but to suggest that they were some bastion of lost knowledge and wisdom is pure romanticism. Taleb is quick to proclaim the dissolution of the nation-state. It is a system, however imperfect, that has developed by trial and error throughout the world as the dominant paradigm. I, personally, am not in a hurry to live someplace in the world that is not a nation state.

Mark K
Mar 23 2009 at 6:50pm

Thanks Russ and Miguelito,
I had already investigated this after my post and found that government was changing to this system in the 2006-2007 time frame, long after overzealous housing consumers, dumb banks, unregulated hedge funds, Wall-street bank insurers (e.g. AIG), and Fannie and Freddie had set the crash events in motion. But listening to the podcast, one gets the impression that regulators pushed VAR at industry and created the problem, which as near as I can tell is completely false. Industry, as Russ points out in his response to me, took the lead on this, and government, lagging many years behind, bought into the BS and decided to regulate using the model. Of course, it took a free market supporting government regime (Bush administration) to put this garbage in place. So to blame government regulation for that part of the meltdown seems ludicrous.

Good regulation does not involve adopting industry schemes, but rather about creating a set of rules where transparency is encouraged and fraudulent reporting is severely punished, so stocks are correctly priced. Also, they need to protect the investor from theft. Speaking of regulation, what ever happened to anti-trust regulation, where too big to fail becomes a non-issue? Instead of that, we got the repeal of Glass-Steagell, which is pro-trust and the dumbest thing to which Clinton ever agreed.

Mar 23 2009 at 9:49pm

Mark K

The rating agencies were granted monopoly power by the government. See here:

Mar 23 2009 at 9:59pm

It wasn’t heroin we gave to the heroin addict. We gave them the cash for rehab and sent them on their way. Why then are we surprised to see they stopped at the first street corner to score?

Mar 24 2009 at 12:12am

Great podcast with much wisdom. Let me push back on a few points though. It’s great that Taleb sees the transformative power of the internet starting in the 90s, but I wonder about his analysis of stars taking the gains. While that is no doubt true, the question is how much that trend contributed to income inequality. Aren’t Tiger Woods, Luciano Pavarotti, and J.K. Rowling a drop in the pond compared to the richest 1-5% that skewed the income inequality distribution recently? Rather, I suspect that the inequality came from professional and technical positions being highly renumerated in the last decade or two, to try and approach the productivity of these highly skilled people. For two people digging ditches or working on a factory floor, the productivity of any worker is fundamentally limited compared to any other worker, at most 2-3X. Whereas with information work this multiple explodes to 10-50X- this has been shown to be true in computer science- I believe much of the income inequality came from the attempt to compensate in equal proportion to this productivity difference. It is likely that employers then got into the star mentality and overpaid, just as you often see sports team owners overpay for star athletes, and might have been better off trying to raise the productivity of everyone by spreading best practices from the most productive to everyone else. Either way, the opera singers and golf superstars that Taleb talks about will soon see a great reverse in their unique positions. The internet initially augmented their positions greatly, now it will tear them down. One Katie Couric on CBS will be replaced by hundreds of news anchors reading the news online, helping flatten out the income distribution in the process.

It was interesting to hear Taleb make Kling’s point that regulators helped cause this mess. It would have been interesting to hear Taleb expand on his solution of nationalization followed by most of the market becoming unregulated and why he felt some portion of the market would need to stay nationalized. Regarding his skepticism of theories and directed research, isn’t it possible that the scientific frontier expanded greatly in the last century and that both directed research and tinkering would have had much lower success rates as a result? Although, the point he might be making is that researchers always have to be on the lookout for secondary effects that they weren’t looking for, rather than throwing that secondary data away, because that’s much more likely to be beneficial than the original hypothesis. I think we ultimately have to theorize, Taleb’s point is simply that we need to emphasize empirical evidence much more where we can and not be so confident in existing theories. As for the benefits of religion, it would be nice if it were merely used as a moral code but one can point to religious wars as an example of religious theorizing and not just following the practices. We have to theorize, we just have to be much more thoughtful, empirical, and humble when we do.

Mar 24 2009 at 1:30am

Mark K,

the 3 rating agencies were given monopoly power by the government:

Mar 24 2009 at 3:42am


Let’s examine the relative economic arguments available in this particular podcast.

I’ve read about Taleb, but never heard him (reading him is impossible, sorry, I’ve tried). If you play the horses, there is always one guy near your seat who swears by some method. He claims that intuitively, he can pick a winner. Maybe he can, but he refuses to give anyone a hint as to how he ostenstibly does it.

I’m siding with some other guy who at least explains his method, even if he had a bad day at the track. Even if that guy loses, I’ll understand why.

Mar 24 2009 at 6:26am

Towards the end there the podcast really drifted off into crazy-land. They both started to blab about strange hypothesis without any evidence, like two drunk sophomores at 3am.

Also, I prefer the old “shit happens” to the trendy “black swan”.

Mar 24 2009 at 8:11am

Another solid podcast that was quite thought-provoking, especially the last 5 or so minutes on economics and religion. I think you’ve got a future podcast topic there…the tough part will be selecting the right guest with whom to do it.

The one question that I would have asked Mr. Taleb was in response to his criticism of the models that didn’t take the magnitude of expected losses into account. Suppose instead of a 10%, 20%, or a 42% loss when taking on risk, they had factored in a 100% loss….not a loss just of that one project/process, but of the whole company? I understand much of his black swan argument, but it seems to me that if his criticism is that the magnitude is always underestimated, then increase the magnitude significantly in the models. A pefect solution? No. But wouldn’t that be a better model?

Mar 24 2009 at 8:52am

Congrats. Another excellent interview with Nassim Taleb. For those who want more specifics on the current financial crisis, here’s a link to a great discussion between Nassim and Daniel Kahneman:

In the latter, Nassim really gets rolling about half way through the discussion. But it’s all worthwhile.



Mark K
Mar 24 2009 at 8:55am

What does the government giving monopoly power to credit rating agencies, which could have several meanings and is debatable, have to do with my central point, which is that it appears the government did not push Value at Risk at the Financial Industry as Taleb claims, since Industry began adopting it about 10 years earlier than government regulators?

In my opinion Credit Rating Agencies, which is an entirely different subject to the point I was making, should not be bought by the companies creating the Securities. As it is, they were are large part of the problem owing to collaboration with private Financial firms.

Russ Roberts
Mar 24 2009 at 10:08am

Mark K,

For me, VAR’s use by the regulators is an example of how regulation is not some magic bullet. Not sure how Taleb sees it other than his insistence that it wasn’t a reliable measure. It wasn’t the cause of the crisis. It’s simply another example of human frailty or malfeasance depending on one’s perspective.

I don’t think too-big-to-fail is an antitrust issue. It was really too-entangled-to fail that justified the interventions we’re seeing now. I don’t think anti-trust is part of that.

BTW, you call the Bush Administration a “free market supporting government regime.” Given the behavior of that administration I think the right phrase is “pretending to be a free market supporting administration.” Bush increased the size of government dramatically, put tariffs on steel, and created the bailout strategy we’re still following right now. When Bush defended that strategy (the few times he spoke publicly about it) he said something about sacrificing his principles to save the free market. I saw his interventions as being consistent with his principles–but they weren’t free market ones.

Mar 24 2009 at 10:24am

I’d like to briefly comment on three semi-related concepts discussed in this net cast: too big to fail, history of medicine, and religion.

First, at about 31 minutes into the discussion, the summaries of the arguments for and against nationalization and bail out brought out a great point. Taleb mentioned that “government does not understand that you do not give heroin to heroin addicts.” Analogies, by definition, are weak in most cases. There are too many dissimilarities . This one, in particular, falls apart quickly because public policy has favored treating heroin addicts with heroin since the 60s ( I believe). It’s a subsidized form of heroin called methadone. Public policy does, in effect, treat heroin addicts with heroin, in principle in the form of methadone maintenance. Clearly, the financial markets and the medicalization of lifestyle choices greatly differ, but it is the underlying rationale that says “government knows best.”

Does it? I am begining graduate studies in financial economics, so I do not have much knowledge of financial markets. Based on my experience though, providing methadone to a heroin addict can be considered an iatrogenic malady. (Thanks for the definition, Hanson). In other words, the cure can be worse than the disease. My descriptions of how this occurs go beyond the scope of my comment.

Suffice it to say that I am simply relating back to Taleb’s thoughts on his current research, which was discussed at about 35 minutes into the net cast. He is looking at the literature on how our knowledge of treating illness can come as an accident. The benefits of methadone maintenance, an alternative therapy for heroin addiction, were tertiary or latent functions of methadone therapy . It was originally designed to replace the shortage of morphine in WWII. It’s effectiveness for treating heroin withdrawals was later observed.

As someone who was addicted to heroin and later treated with methadone, I doubt that I would be drug-free (including methadone) today without compelling arguments set forth by religious doctrines that emphasized the hopelessness of living the rest of my days taking methadone, as that is the purpose of methadone maintenance: to keep you as a government junkie.

An afterthought to our knowledge of how to cure a disease. Take a look at the history of Viagra. Not designed for ED; that was an accident.

Keep up the good work EconTalk; informed dialogue is great. All the best to Taleb in his continued research.

Mar 24 2009 at 12:40pm

I agree on the premise that it can be dangerous to rely primarily on quantitive methods. History is not a very reliable guide for the future and the information content of (time series) data is in most cases anecdotic.
Regulations and risk management methods using such flawed measures created perverted incentives which lead to this crisis. Most risk measures are designed to deal with exogenous shocks, not for risks coming from inside the system.

However, I cannot understand the conclusions Mr. Taleb draws from this. Frankly speaking they seem reactionary and not very construtive.

‘Pragmatism, idea that Descartes dangerous because he thought reason could triumph over everything. Trial and error may be wise. Data that universities had negative contribution to knowledge compared with practitioners.’

I really don’t see how practical trail and error is inherently superior to scientific trial and error (research).
Practical and scientific knowledge will both always lag behind the reality in this ever-changing system.

Although Mr. Taleb’s analysis is appealing, his image of humanity seems rather odd to me.

Mar 24 2009 at 12:51pm


I thought it was interesting your comment about the Bush administration “pretending to be a free market supporting administration.” It makes me wonder about a problem that I see with the dogma of free market philosophies and not so much the theories. It seems to me that people often do things in the name of the free market as a way to make money or serve some other extraneous goal. Enron pushed for supposedly freer markets yet was basically running a racket with the governments help.

A deeper problem I have with this is the fact that one could easily point to the way these markets were not free (although they could possibly be considered less regulated than before), in other words it was not a failure of free-market theory just free-market practice. But can’t those in favor of regulation make the same argument? Sure you can point to regulatory practices that were harmful, to which a support could say that it was an example of bad regulatory practice, not theory.

Just some thoughts.

Mark K
Mar 24 2009 at 2:13pm

Interesting points. I accept that ‘anti-trust’ is a bad term for the AIG situation. The primary point, as you point out, is entanglements, but it also has to do with size of assets and domination of the derivatives insurance market, which may or may not fall under anti-trust law as a monopoly. However, it could also simply be that AIG was among the few dumb enough to insure sub-prime mortgage-backed Securities.

I was taught in Econ 101 that a ‘large number of relatively small firms’ and low barriers to entry were good preconditions for efficient markets that produced the widest societal benefit in terms of goods and services. It seems as though over the last few decades American economy has crept toward a small number of relatively large firms in many markets (e.g. defense companies and financial institutions), which beside tending to produce higher priced product at lower output, also seems to produce government bailouts.

Mar 24 2009 at 3:26pm

Thanks again for another great podcast. I’ve read both of Taleb’s books. His ideas are both timely and thought provoking. Like Ajay, however, I find his example of the opera singer somewhat wanting. The opera singer in Taleb’s example probably cannot command the market share she held 10 or 15 years ago because the internet and other technologies have opened up the market to thousands of others with comprable talent. In some domains — entertainment, most prominently — technology has leveled the playing field and increased competition. Any given opera singer may be the biggest star at the moment, but her share of the proceeds of stardom (wealth, fame, and power) is less. She has a larger set of competitors, who now have the tools to reach the same audience. Doesn’t this phenomenon, after all, help explain the economic difficulties of the broadcast and recording industry? I wonder if we will ever see another phenomenon like the Beatles again. As alluded to by Ajay, America does not go home and tune into Walter Cronkite any more. Instead, if we want to watch the news, we chose between Fox, MSNBC, CNN, Katie Couric, Amy Goodman, etc. In this kind of environment, it is increasingly difficult for one person or entity to take all — or even a largely disproportionate share — of the marbles. The interesting question is why, given increased competition and loss of market share, some professions continue to earn such disproportionate salaries.

Mar 24 2009 at 3:43pm

So, this has me a bit confused. Taleb’s point seems akin to Knight’s point a while back – that there are these ‘unquantifiable risks’ that throw all of our models into confusion. Alright, fine, you can say that, but how is it at all helpful? We have to quantify some kind of risk to be able to make a decision about quantitative issues (e.g. how much to buy of this asset, whether to take out a loan to start a new entrepreneurial venture. Otherwise, we are trapped with anxiety and stagnation.

I’m not saying that Taleb is wrong in asserting that we can’t quantify these risks well. But we need to quantify them somehow, don’t we (even if it is only unconscious, as revealed through our actions)

Mar 24 2009 at 5:02pm

Another interesting point I think to note is Taleb advocating the nationalization (not exactly sure if he means the receivership model) of trouble banks, something heard from many people during this crisis on various sides of the economic spectrum.

Also interesting is a point made on planet money about the fact that the government would most likely not announce before hand if they were going to indeed do this.

Ben Itri
Mar 24 2009 at 5:05pm

I am not an economist but just an “engineer” and have enjoyed these podcasts having attempted to educate myself on the current financial crises. The mathematical models were simply used as a rationalization for taking on excessive risk. The decision making process lacked any “common-sense”. More concerned with stock price performance and not being left behind by the other firms. How does it make sense to sell insurance (CDS) multiple times on the same underlying security? Would CEOs and “risk managers” actually use their own money to make these bets? If so they were probably in the wrong postions from the start. Clearly there was a complete failure to regulate these off balance sheet transactions that eluded the regulatory agencies. This was greed run amuck without a completed understanding of the ultimate consequences.

Hans Lempka
Mar 24 2009 at 5:42pm

Further to NT’s comments about so called lumping, could it be that excessive financial regulation encouraged financial institutions to merge due to the high cost of compliance? Another unintended consequence perhaps.

Mar 24 2009 at 8:32pm

The history of medicine comments are thin, to say the least. Advances in the last seventy years easily outstrip all prior history by any measure. Some of this is accidental and “tinkering,” but hardly hindered by expanded biological knowledge. Simply preposterous and should have been challenged.

Hubris is also mistaking your understanding of one topic for an understanding of other topics.

The fascinating paradox here is the example of a man on display who always has a firm and quick answer for any question on any topic while explaining his theory of uncertainty and of the limits of knowledge.

Is he serious, or doing sophisticated comedy? Prof. Roberts, you make a good straight man.

Russ Roberts
Mar 24 2009 at 10:24pm


The internet simultaneously allows greater blockbusters like Harry Potter but also allows for a much wider variety of folks to flourish–see the first EconTalk podcast with Chris Anderson on The Long Tail.

So share can go down, but the absolute rewards going to the biggest stars are way up. That’s one reason why athletes and entertainers make so much money today relative to the past–the market is much more global and people still want to see Lebron James rather than an incredible player who is almost as good.

Mar 24 2009 at 10:35pm

Would have loved to hear him compare his position of how we tend to abuse “knowledge” vis-a-vis Wittgenstein’s similar position with that of our utility of words/language. From my pedestrian position, there seems to be a parallel. Both categorically and with regard to religion.

Mar 25 2009 at 12:33am

Thanks for another great podcast Russ. I have a comment/question on the “too big to fail” discussion.

The “too big to fail” idea has been criticized in past podcasts as well (Meltzer, etc.). Given that both conservative and liberal politicians so quickly accepted these bailouts and the “too big to fail” rationale behind them, I wonder if that points to some type of institutional or behavioral explanation. It seems like there is an institutional tendency to prioritize avoiding large, visible, and immediate losses (collapsing banks) over the invisible future gains of letting these institutions fail (more disciplined and cautious behavior by firms in the future). Compound this with a general fear of uncertainty, i.e., we know what the world looks like with a large, solvent AIG, and since we don’t know what the world will look like without AIG, we should avoid that outcome if at all possible. Maybe the pure economic arguments against “too big to fail” are not conducive to the existing political incentives, and therefore politicians just will not respond to them, regardless of how intellectually persuasive they are. As a consequence, free market proponents may need to either figure out a way to change the political incentives to make it easier for politicians to reject “too big to fail”, or they need to assume that “too big to fail” is a permanent feature of our politics and consider the implications of that for free market policies. Do you know whether public choice or behavioral economics has spoken to this issue?

Mar 25 2009 at 12:52am

Mark K,

I was only replying to what you said here:

“…except for a vague reference to rating agencies, which to my knowledge are private profit making companies, not government agencies.”

Sorry for not making that clear.

Sri Hari
Mar 25 2009 at 1:43am

The discussion in the financial press about the path to economic nirvana- is it the one paved by the Austrian school of economics or is it the one shown by Keynes?- is increasingly irrelevant !!

The fact of the matter is, the world economy is made up of Austrian & Keynesian models – or to put it simply we work and trade in a multi-disciplinary (or model) economies. When one model tries to dominate the system the balance of the world economy as a whole, is lost. And it appears the panacea perhaps is the less dominant economic model!

This happened in 70s, when Keynes was in vogue and later in mid 80s the pendulum swung back to the Austrians represented by Milton Friedman and Alan Greenspan. And now we are again talking about Keynes. Perhaps the solution may be, if we can clearly demarcate the economic activity that will function well within each model and not let either of them to dominate the economy as a whole.

Make working of each model transparent and have a system to ascertain accountability and calculability of risk in real time.

Chris H.
Mar 25 2009 at 1:50pm

I enjoyed the Podcast, and Mr. Taleb’s decidedly non-mainstream views.

I don’t have a lot of familiarity with economics (I’m a beginner, you could say), but it seems to me that Mr. Taleb’s remarks about religion toward the end of the program clashed with his economic view.

Traditionally, Western religions have had “prophets” who’ve purported to know the true nature of God and creation, and religious literature has claimed to recount actual historical encounters with the divine.

So, if religion isn’t about “belief,” as Mr. Taleb suggests, than isn’t it at least about treating the divine as a known (and witnessed) entity? Isn’t it Mr. Taleb’s general thesis that there is just so much that we can’t claim to know (or haven’t yet discovered, despite millenia of having tried)?

I get the point about being humble about what we think we understand, but I failed to see how that was an arguement for religious practice.

Mar 25 2009 at 7:43pm

Mark K,

“…my central point, which is that it appears the government did not push Value at Risk at the Financial Industry as Taleb claims…”

From the Joe Nocera article in NY Times linked on econtalk website,
“In the late 1990s, as the use of derivatives was exploding, the Securities and Exchange Commission ruled that firms had to include a quantitative disclosure of market risks in their financial statements for the convenience of investors, and VaR became the main tool for doing so. Around the same time, an important international rule-making body, the Basel Committee on Banking Supervision, went even further to validate VaR by saying that firms and banks could rely on their own internal VaR calculations to set their capital requirements.”

VAR was accepted by the government to monitor risks. What gives you the confidence government has government has a better ability to monitor risks than the some of the smartest, highest paid people in the world. What makes government smarter? They aren’t.

Keelan Downton
Mar 25 2009 at 11:16pm

As a theologian who has been a long time listener to supplement my research in entrepreneurial approaches to churches, it was fascinating to hear this discussion veer into religious territory. I’d be interested in hearing Larry Iannaccone as a guest sometime.

[See The Economics of Religion for an EconTalk podcast interview with Larry Iannaccone.–Econlib Ed.]

Mar 26 2009 at 11:22am

I found the podcast one of the best ones in recent memory. Taleb is actually very friendly with people and interviewers who like to discuss ideas with him. And I did not find a great deal of repetition with previous remarks by Taleb. The discussion about religion and the worship of scientism resonated with me. Taleb is turning into a systematic anti-positivist, and many of the criticisms of mathematical modeling and theorizing are spot-on (and very close to Hayekian epistemology). Reading at the comments above, I think that the commonly presented alternatives (Keynes or Hayek?) are misplaced. The alternatives are between modes of knowledge and approaches to synthetic statements. Taleb is, I believe, one of the few public intellectuals who has identified them correctly.

Mar 26 2009 at 3:13pm

I wonder if Taleb has read Mises? If he had he would appreciate that what we are going through is a predictable event following a classic Austrian anticipated credit boom which would not be possible without fiat and fractional reserve banking. Taleb might acknowledge Hayek but it might dethrone his black swan tag if he opened people’s eyes to Mises et al. To use his language, the financial markets would be a lot more ‘mediocristan’ than ‘extremistan’ in a hard money world where every dollar loaned represented a dollar actually saved. Imagine that.

Lee Kelly
Mar 26 2009 at 4:48pm


Don’t be absurd. Credit comes from a magical bottomless well. Saving just deprives someone else of income. Why are you so selfish, Jonathan?

Mar 26 2009 at 9:40pm

Re the ‘one trick pony’ remark, I think there is a lot of merit in Taleb repeating his ideas, after all no one listened the first time.

Mar 27 2009 at 12:59am

Adam, thanks for that video link. In there, Taleb seems to call for the banks to be nationalized while leaving the hedge funds completely unregulated, with it made clear that they’ll never be bailed out. He never really talks about what bank nationalization would look like, so I think it’s just a reactionary response to the fools who ran them. As for those confused by Taleb, I think he’s a strict empiricist who dislikes all the theorizing that is done without careful analysis of the data. This can be hard to tell sometimes, as he’ll often focus on how current practitioners are wrong and how some fat tails are unknowable, but if you actually read his stuff he’s pretty data-driven. I think his point about the ancients is the essential conservative viewpoint that practices that have evolved over the centuries are based on what worked and we throw those evolved practices away at our peril. His point about religions seems to be that they’re merely sets of these best practices, with God thrown in to scare the disinclined into following. fischer, as to models, I think he’s simply saying that people were using a known probability distribution to model rare events that are very difficult to figure out the real probability for. He simply says that we should stay away from such unquantifiable events, not that we should throw away all modeling. Ben Itri, the reason the investment banks blew up is because they used their own money on those bets. In order to sell CDSs, they kept the riskiest portions themselves, presumably cuz nobody else wanted them. Many of them owned huge amounts of stock in their own firms, most of which they lost when they blew up.

David, I think you may have misunderstood my point a bit. What I was saying is that I don’t think the richest .001%, the superstars, contributed that much to rising income inequality recently. Rather, it was the richest 1-5% that caused it, for the reasons given. You’re right that Couric now has to compete with cable news stars, but there’s still only 5-10 of them: all you’ve done is make the richest group bigger, from .001% to .003%. When the internet can be monetized through micropayments, that changes as hundreds of online news readers will spring up. As for the richest 1-5% that actually caused the shift in income inequality, that will change too but not enough to go back to the old income distribution. Education is currently a bad joke, the internet will destroy the current education system and mitigate how sharp the current income inequality trend is, but those high multiples are likely to stay fairly high as they represent real productivity differences between information workers.

Mar 27 2009 at 12:43pm

Taleb’s books led me to Benoit Mandelbrot’s “The (Mis)Behavior of Markets,” which is terrific. Those interested in fat tails should check it out (especially the footnotes).

Gringo, your comment on playing the horses is interesting to me. You’re siding with the guy that explains his method… even if he is wrong. CNBC does this all day long. Of course we risk mixing up correlation with cause, and worse, mixing up garbage with cause. One of Taleb’s arguments is that most of this explaining is noise, creating meaningless information and methods that do more harm than good. It gives us false confidence and leads us into making even bigger mistakes.

I was surprised to see Kahnamen disagree with Taleb on this point in the Edge video asserting that man is better off with a “bad map” than he is with no map at all.

On “too big to fail”, here is Taleb speaking briefly on moral hazzard at Davos 2009 – a few days after the Edge video w/ Kahneman.

Mar 27 2009 at 1:35pm


Cap Khoury wrote:

“I’m very interested in Taleb’s remark that the long-run statistics tend to look “rosy” because of “survivor bias”. Can anyone recommend some literature with some detailed analysis of this phenomenon? (I don’t want a brief overview “in layman’s terms”.)”

You replied:

“Google “survivor bias mutual fund active management” and you’ll find studies and news reporting on the topic. It is a standard claim of index funds, for example, that the performance of managed funds is biased by ignoring funds that have gone out of business.”

I’m not sure this is what he was asking. Taleb says (at least my interpretation) that the next negative market deviation is likely to be larger than the last due to “survivorship bias” and the fact that we see the “rosy part of the distribution.” He declines to provide evidence because he says it’s beyond the scope of the podcast (probably true).

I’m also interested in what exactly he means the evidence for it. Could explain what you think he meant or ask him for clarification/evidence? I sent him an e-mail and received an auto reply saying he’s not responding to e-mail right now. I hope you’d have better luck :).

Also, excellent job on the podcast as usual. You and Taleb have very comfortable conversations, and he certainly opens up to you much more than almost any other interviewer. I’d grown somewhat tired of his “sound bite” responses; this podcast was almost all new material. You should have him on again.


Mar 27 2009 at 5:46pm

Quick thought:

For those who are interested in religion and economics, it is interesting what almost all religious texts (especially in the Abrahamic traditions) say about money lenders and the rich.

Mar 28 2009 at 10:02am

Taleb’s argument about knowledge is self-defeating. He is not a practitioner, he is a theorist. So by his own reasoning, his ideas are not useful knowledge.

Mar 29 2009 at 12:59pm

Mike R:

Successful traders and successful horseplayers have much in common. However, let’s not confuse method with analysis. Analysis can be full of errors and misinformation, such as your CNBC example. And there is an advantage in being able to recognize such faulty analysis, traders can use this information (or misinformation) to their advantage. At the track, for example, a handicapper can use “track buzz” about a certain horse to their advantage, especially when the buzz can be considered as inaccurate.

Method is entirely different. In trading, or even in any transaction that involves insight into economics, method is path taken to arrive at a position. In race betting, method would be how a handicapper arrived at a point of considering a wager. If the handicapper tells me why he or she likes a certain horse in a certain race, that’s analysis, and while I might take mental notes I’m not interested in basing my own wager on someone else’s analysis. But if the handicapper gets up to make a wager, I’d love to know why – this is the method they use in taking a position.

Taleb seems overly careful not to give up his method.

John D
Mar 29 2009 at 9:46pm

This was a great podcast and im always interested to hear/read what Taleb has to say. I would have to agree that i really appreciated the increased humility compared to his books and last interview.

p.s. AHBritton there is a great book that among other things looks at a the development of capitalism in relation to the abrahamic religions by Muller called The Mind and the Market: Capitalism in Western Thought

Mar 30 2009 at 12:32pm

Lots of great comments this week. I did not have time to read them all. A couple of people have mentioned this, in particular, Mark K. I agree with Mark, and have to add that the SEC didn’t exactly ram VaR down the throats of the financial industry. NOT AT ALL. Go to the Basel Banking Committee’s website and search VaR and search Michael Alix. Alix was Bear Stearns chief risk officer. You’ll find a letter from BSC to the commission suggesting that they use VaR as a way to regulate risk, rather than some other method, which would have forced Bear to revise their VaR models. Alix, and Cayne, and various others, all were selling the kool-aid of VaR. The regulators sought a way to regulate the amount of risk that the investment banks could take on, and the financial industry campaigned for VaR. Blame the regulators, yes, but blame them for not doing their job. I should also point out the complete and utter failure of the CRE program. Again, this was systemic risk caused by regulators, but because the regulators were nether strong enough, nor willing enough to make it worthwhile. And Russ, I do blame the Bush admin on that. I see your point on a lot of issues, but on this one thing, the appointment of Chris Cox to the SEC, and Hank Paulsen to the treasury (And I don’t think Paulsen was necessarily a bad choice, at least he didn’t appoint Christine T. Whitman) were completely in line with conservative, free market principles.

By the way, I think my accounting professor has been posting here! That’s awesome!

Mar 30 2009 at 4:52pm

Looking forward to listening. I try and listen to Taleb whenever he is interviewed. Yes, he’s arrogant. Yes, he’s repetitive but he is also RIGHT!!

He has a right to a little arrogance. I’ll take his arrogance over the smugness of most of the people he’s criticizing. I think his area of study is crucial to our societies “getting things right”. We are lost in old paradigms, floundering around with false theories of knowledge, convinced we live in a top down world unable to truly understand how life, and all it encompasses, just bubbles up. Illuminating our biases wont eliminate them but will make us live with some humility. I hope he stays repetitive and never shuts up about this stuff, its important.

In regards to something Alex said earlier, Taleb does NOT say this current event is a Black Swan, only that our flawed modeling has placed the likelihood of such event into Black Swan territory.
This is something I’ve noticed many people wrongly claim about Taleb.

And gringo, he has given up his (investment) method, 80% super safe, 20% in very high risk out of the money options.

Jim F
Mar 31 2009 at 7:52am

It was surprising to me during the podcast, when within about a 5 minute span both Russ and Nicholas 1) can’t understand why banks invested in the most senior tranches of the CDO’s, with their very low returns of pennies, and 2) Can’t understand why they levered up 30 times. One was a product of the other.

Quite simply, they had AAA floaters paying just over LIBOR that they could fund slightly under LIBOR. A handful of basis points isn’t a great return but when you lever it up 30 times you have a ‘virtually risk free’ investment that’s pretty attractive.

Now these weren’t necessarily stupid guys, they were just wrong. A AAA floater is a very safe security. It’s insensitive to the overall level of interest rates (since the coupon is directly linked to the discount rate). Though it is subject to basis risk, that is the change in their required return relative to the risk free rate, a AAA security should have very slight basis risk.

The thinking by many was that while the leverage made the positions more risky, they were less risky than an unlevered position with a similar return in the same asset class. To exploit that situation is sound financial economics.

I suppose the point is (and I think Taleb could also be paraphrased) mean/variance optimization is not enough; skew and kurtosis must be considered.

Russ Roberts
Mar 31 2009 at 2:26pm

Jim F,

I don’t fully understand your point. If you’re interested in explaining it a bit more via email, contact me at mail at

Dave S
Apr 14 2009 at 7:26am

This idea of building for robustness against extremely infrequent big-impact events seems to me to be a compelling argument for spending a lot of money, probably through government spending, on seemingly kooky things like asteroid defense (and on space exploration in general for that matter). It probably won’t pay off, but on the off chance that it does it could save all of humanity.

It’s an enormous, waaaaay out-of-the-money call option.

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Books:


Web Pages:

Podcasts and Blogs:



Podcast Episode Highlights
0:36Intro. [Recorded Mar. 15, 2009] Taleb: One of the few people who can say "I told you so." Critic of mastering risk and uncertainty with mathematical tools. Where right and where wrong? Not so much a critique of mathematics as of methods to estimate rare events. We don't know much about rare events because by definition they are rare. If an event happens once in 10,000 years and the person is less than 10,000 years old, you know he is getting the probability from outside his experience. Smaller the probability the more you have to rely on some a priori specification or on a model, a representation of the world. The smaller the probability the more fragile and in some areas, the larger the impact. A ten year flood has a higher probability than a 100 year flood, but the 100 year flood will be massively more consequential. You care about the probability times the size of the impact, the expectation of these events. Small-probability events can have in some domains, fat-tailed domains, a big impact and we don't know how to estimate them. Critique of the models was basically that they relied a lot on some theories to produce small probabilities, and second that these models were very fragile and lent to the build-up of humongous positions. Banking system extremely fragile, and also anyone who had exposure to rare events. Black Swan: companies looking at their risk reports were sitting on a lot of dynamite and ended up exploding, maybe a little later than Taleb thought. Joe Nocera article on value at risk: these investment banks and hedge funds had made their estimates of how exposed they were to risk and had made errors on two counts: probabilities not right; more importantly, the size and magnitude of the consequences. Were heads of Bear Sterns and Lehman Brothers not aware of how much they were gambling or did they not care how much they were gambling? Combination. Three things: 1. fooling themselves, psychological dimension. 2. Had an interest in building huge risks and tail because if you blow up every 10 years, you will make 9 bonuses and the 10th year someone will pay the cost, not you. Vicious: taxpayers are paying retrospectively for the bonuses of the first 9 years. Banks are insolvent, have lost more than their capital base, but managers have kept their bonuses. Some of them have been wiped out because they went a little further than normal blow-up cycle. What about the ones who didn't do it? Lower returns year after year; now should be doing extremely well, but now unable to buy up some of the firms that have made the mistakes because the government is propping them up. Too big to fail: 1981, if we did not save the banks, or in 1998 with Long Term Capital Management (LTCM), we wouldn't have let something so fragile grow to be so large. Capitalism is about failing early, not having some government prop you up so when you fail you fail big. Moral hazard problem is very real? Started early, in 1982. Need punishment in capitalism, not bailouts. If you have incentive without punishment, someone has a free option. Risk methods used: maybe bankers were foolish enough to believe in them; on the part of scientists, some breaches of ethics.
8:36Value at Risk (VAR); Expected Shortfall. Value at risk told you that with 99% probability you won't lose more than a million dollars. Problem it had was that should you lose more, the expected loss could be $50 million. Not just a bad quarter--you get wiped out. Care about how much you can lose when you lose. Intractable. Didn't the people using those models know that? Started using VAR in 1993, problem in 1994. Warning people. Another argument against using these models that rely on the past. If you are looking at large deviations in the past and you are in Oct. 1987, right before the 1987 crash, the worst deviation in history would have had close to -10% on any given day. Calibrate your VAR at close to 10%. Had an event close to 23%. Now calibrate to 23% on any given day. Next largest deviation is going to be different and worse. Bias: random variables tend to show the rosier scenarios, linked to survivorship bias. Greenspan, to apologize for letting the banks take so much risk, said "We never saw these events before." Large deviations do not have predecessors. The first great war did not have a predecessor. Cold War did not have a predecessor. Very easy to be robust against it: can have two portfolios, one gentleman exposed to rare events and his brother exposed to same variables without so much toxic payout. Conditional on insurance company existing or can construct portfolio in the same way. Linear combination of 80% no risk and high risk for 20% is more robust to rare events than 100% allocated to medium risk portfolio. Robustness is extremely easy to build, but have to abandon these metrics. Standing on the shore as the Titanic heads off to sea. That's where the party is, where the champagne is flowing; difficult to have lower returns than you year in and year out but there is one year when I will trounce you. Cash: in this crisis, cash performed the best.
14:38Not having a predecessor: People are saying we will have another depression. If economy takes another bad tumble, but if it does, it is going to be different. Rare things happen sometimes that are not good; but we want to learn more than that--learn principles--but sometimes that's not true. Form of autism with regard to history. Don't look at the past in relation to its own past: When you look at 1987, you don't say because worst event went from -11% to -23%, don't transfer that to today, don't say the next worst event could be 50%. Today's past, today's future, don't transfer. May increase with use of mathematics. Cab driver would get the point. Hubris. Some ratings agencies did build into their forecasts the possibility that default rates on housing would reach levels of the Great Depression. Seems to be very safe. But we could have default rates even worse than that. Exposure to these tail events was preventable. Take banks: lost on subprime initially then on other areas not particularly profitable. Increased exposure. Why did people buy the senior tranches? Only making pennies. Before LTCM, using VAR; anyone using it afterwards should have been convinced that you can have these events. Those who blew up were those who made pennies. Most risk with least return. Returnless risk, not riskless return. Interesting episode in history; in 1991-1992, kept building risk, particularly with Greenspan's Novocain. Disappointing: Ben Bernanke's calling the era of low volatility the Great Moderation. Didn't realize that what we call the tails are getting fatter; risk coming from jumps increasing. He was referring to the longer period, 1982-2000. Central Banker, system getting more complex, rising after 1995, after Internet. Why complexity rising after 1995? Internet, take a book like Harry Potter. Environment that doesn't have complexity is an environment that doesn't produce large-tail events. The more tail events, the more complexity. Before the Internet couldn't spread cultural ideas very fast. After Internet, Harry Potter effect, whole planet reading the same book. Planetary-wide fads. Leads to large right-hand tails for small group of people; could be negative tails. Concentration of income for some people in sports or arts: fewer people making larger and larger incomes. Local opera singer who performs at weddings. If no way to store your voice, opera singer in Italy has audience because people who love opera have to go to Italy to hear it. Now, gramophone, later the Internet, technology allows you to hear anybody, so why listen to that poor guy. Whole planet will have a handful of opera singers making huge amount of money. Dentist cannot store, leverage, or scale his work; same for massage professional. Black Swan. The more informational economic life is, the more it can deliver large deviations. Good overall. Get great opera singing. Have to expect to see spikes. Can get very rich if you are on the right side, but blow up if you are on the wrong side. Especially if leveraged.
25:17Leverage: complexity cannot tolerate leverage, since no room for error. Since 1980, tripling in leverage, in U.S. and in Europe, ratio of leverage to GDP. High exposure to errors. Commodity prices, similar argument: can have rise in wheat prices in response to very small imbalance in demand, followed by rapid collapse. Not long ago talking about inflation, now talking about deflation. Speculative issue in commodity prices: how much comes from similar mistakes to Greenspan. Globalization has side effects; produces fat tails. In competition, consider two banks, one robust and one not. With globalization, everybody pushed to do outsourcing; concentration of U.S. and German firms outsourcing to, say, Bangalore. Probability of failure low because only one source of randomness. But suppose there is a problem in Bangalore--political, storms--that disrupts communication there? What happens to all these firms? French bank, lost money on a rogue trader. If you had 10 small banks each with 5 billion, wouldn't have a problem because easy to liquidate. With one bank, should a rogue trader happen, his position will be 10 times the size. Assume no linearities in execution cost. Liquidating $50 billion is more costly than 10 times liquidating $5 billion. Lumping makes you more vulnerable to large events. One large error is a lot worse than 10 smaller errors.
30:53Where do we go from here? Lessons we won't learn because government won't let people pay those prices. Nationalization versus bailing out: you don't give heroin to heroin addicts. Stop problem early rather than late. Whatever is fragile, it should break as early as you can. Second: people ask for more regulation. Risk was implemented because the regulators wanted it. If choice between bailout and ownership, ownership probably better. Problem is: government can't keep their promise to not bail out the other part. Civil servants, politicians, different incentives from the public. Like regulation. Nobody in Washington talked about that this crisis was caused by regulators, rating agencies. Triple A assets, big premium put on having your asset rated AAA. Pension funds couldn't hold some of their assets in certain categories. Fannie Mae: VAR. Regulators like VAR.
34:45Will we continue to live in this fantasy that we can create a riskless world? No government is big enough to save large banks. System will break whatever is fragile no matter how big it is. Seeing this with AIG. Failure of Swiss bank would be a concern to the U.S. taxpayer. Gordon Brown in Washington talking about stabilizing all the banks. Nation-state, globalization: nation-state ceases to exist. Governments have no control. Still have control over people's pocketbooks, though, through physicality of where you live. What does future hold for Nassim Taleb? Back to trading, involved financially in this crisis, back to being a philosopher. Doesn't need to be heeded. Small minority will understand the points. Moving on to other projects. Linked to probability. Don't make decisions based on true/false, heads/tails. Set confidence interval. We make decisions based on payoffs and other judgments. Don't really use probabilities in decision-making. The ancients didn't use probabilities: much fuzzier. probability, Cicero coined the word: probability was something you approve of regardless of the odds. Would you drink water from a glass left on the table left the table by a stranger? No. But do you have evidence that that water would hurt you? No, but you won't do it because it is not something you approve of. Decision making under uncertainty without using probabilities. What we do versus what we know. Closer to Hayek. History of medicine, anti-academic, anti-knowledge. Medicine: whenever we use knowledge as a driver instead of tinkering, we get in trouble. Examples: Our understanding of biological processes led to a decrease in cures. When just tinkering we did better than with directed research. Directed research gives us a strong bias and blinds us to things we don't know are there. In medicine, most medicines are used to cure something completely different from what the intention was. Side-effects dominate. Try to collect positive black swans. Hubris problem, overestimating our knowledge. Difficult to think rationally about uncertainty and risk. Trust the science part, ask the doctor, but the doctor has no idea about the probabilities. Each person, disease is different. Minimize the harm coming from theories. Empirical doctors were successful until eliminated after the rise of Arabic medicine. Western medicine was rationalistic after the Arabic tradition. Improvements after that came from the barbers, not from the doctors. Thinking has not helped us a lot. Evidence in option trading. People think that quants make option formulas, therefore the market uses them. Bogus. Supply and demand. Did a lot better before the Black and Scholes formula.
46:01Pragmatism, idea that Descartes dangerous because he thought reason could triumph over everything. Trial and error may be wise. Data that universities had negative contribution to knowledge compared with practitioners. Hope not, but could be. Lecturing birds how to fly and later on claiming credit. Interview with Peter Singer: probably okay for people to choose smarter children--or is it? IQ: local and applies to exam-taking, important skill between 16 and 23.
48:08Religion and probability. Most people think that religion is about belief, but it is about practice. Greek Orthodox but Arabic-speaking. The way Arabs say is not "I don't know" is "God knows." Allows you to say you don't know, transfers from yourself to another entity. Allows you to be humble. History of medicine: accounts of giving a fortune to the Temple of Apollo: You saved me when my doctors failed me. Doctors gave negative contributions, particularly by bleeding; or more recently, delivered a baby after going to the morgue. Brought in religion. Error we have in believing religion is about belief, but it's about commitment, the system, living with something. We're not yet good with ideas. Can see from this crisis. Great Moderation turned out to be not so great. What does probability have to do with religion? Idea of true/false; degree of belief. May do something against the odds because the consequences or large or even without analyzing it. Probability is not opaque; and even if it were, we wouldn't use it because of the consequences. Pascal's wager: Since God might exist, I might as well be a religious person. Payoffs from being religious are much higher than negative payoffs if God doesn't exist. Had to also assume that God doesn't exist and also not know about gaming the system. Maybe actions are more important than beliefs--in some religious systems the actions count. Greek Orthodox, Easter; Judaism, Maimonides, not every Jewish sage lists belief in God as one of the commandments because there is a debate about whether you can mandate belief. In Arabic, the name for religion, din, is the same as the word for law in Hebrew. To be a law-abiding citizen, keep the rules. Hard to keep the rules if you do not keep the faith. Motivating people without faith. Losing patience about people who are skeptics about religion, and at the same time are not skeptical about economics or VAR. Solving our own problems. Hayek, lot of instinct, contributions part of philosophical thinking.

More EconTalk Episodes