The Reverse Black Swan, Part II

Posted by: Michael Mandel on April 17

In the Reverse Black Swan Part I, I drew three conclusions from Taleb’s work.

1) Unexpected technological breakthroughs are possible. That’s good

2) The timing and nature of the breakthroughs cannot be controlled. That’s bad

3) Unexpected large bad events are possible as well. That’s bad. In fact, we can get bad events which have as big an impact, in the negative direction, as the technological innovations.

These principles help frame a very important policy question: How can we design our economy and financial system to decrease the odds of the negative Black Swans wiping us out, while doing the most to maximize the positive Black Swans of technological change? Or to put it another way, what kind of regulation do we need?

Reading and thinking about Taleb leads me a different answer than I would have given 6 months ago. Yes, we need more regulation—but it should be ‘regulation by simplification’ rather than ‘regulation by supervision’.


Why? As a technological optimist, I believe in innovation as the major force for growth over the long run. So one important part of policy is setting up the conditions under which technological innovation can occur. That means supporting R&D and education, encouraging venture capital and entrepreneurship, and making sure that government regulation does not get in the way of 'good' innovation.

But here's the kicker: Innovation is fundamentally unpredictable, both in timing and scale. That means no matter how much we spend on R&D and education, we cannot guarantee big economy-moving innovations--the reverse Black Swans--will arrive on any timetable. The best example: For all the money lavished on biotech and healthcare science over the past decade, turning scientific advances into economically viable and clinically useful products has been very difficult. Eventually it will come--but eventually could be tomorrow or it could be ten years from now.

That means innovation policy needs a second focus: While we are working hard on finding "The Next Big Thing," we also need to avoid or minimize the odds of negative Black Swans wiping us out in the mean time. The right way: 'Regulation by simplification', which means forcefully simplifying enough of the financial system and economy so that we actually understand what's going on, and can see problems occurring before they get too big.

This is the way that Taleb put it in a recent Financial Times piece:

Counter-balance complexity with simplicity. Complexity from globalisation and highly networked economic life needs to be countered by simplicity in financial products. The complex economy is already a form of leverage: the leverage of efficiency. Such systems survive thanks to slack and redundancy; adding debt produces wild and dangerous gyrations and leaves no room for error.

What does this mean in practice? In his book, The Black Swan, Taleb lays out the idea of a "barbell" strategy:

taking maximum exposure to the positive Black Swans while remaining paranoid about the negative ones.

For an investor, Taleb explains the barbell strategy this way:

you need to put a portion, say 85 to 90 percent, in extremely safe instruments, like Treasury bills...the remaining 10 to 15 percent you can put into extremely speculative bets.

The implication, as Taleb says, is that "you are 'clipping your incomputable risk, the one that is harmful to you." At the same time, you are maintaining your exposure to the upside, and putting yourself in situations "where favorable consequences are much larger than unfavorable ones." Some examples of businesses with big upside include:

some segments of publishing, scientific research, and venture capital. In these businesses, you lose small to make big.

The principle of the barbell strategy extends from the investor level to the economy level. That is, we should put ourselves in a position to enjoy the benefits of reverse Black Swans, while reducing the odds of out-of-control crises sneaking up on us.

The first step: Simplify big portions of the financial system so that we actually understand what is going on. Why is that important? It's becoming clear in retrospect that no one--literally no one--understood all the interactions between the new financial instruments, the massive flows of money sloshing around the globe, the huge trade deficits and the world housing boom, not to mention fifteen other factors. When the regulators let Lehman go bust, they had no idea it was going to lead to a run on European banks.

In other words, the global financial markets made up a complex system, defined as a system that produces results outside the bounds of what anybody could reasonably foresee. That's not a good thing. To the degree possible, we need to turn the financial markets into a simpler system, which is easier to understand.

That may mean partitioning the financial system and going back to something like a Glass Steagall world. There would be one part which consists of simple banks that only take deposits and make simple loans, and which are heavily insured by the government. The other part would do the more complicated and harder to understand financial deals, with explicitly no guarantee from the government and measures put in to restrict leverage.

Taleb is not alone here. Paul Krugman recently wrote a NYT piece called "Making Banking Boring," which was heading in the same direction.

Reducing leverage and debt is an essential part of the simplification process. The reason? Excess use of leverage can turn almost any system 'complex', in the sense of making it hard to predict. If I take on a lot of debt, I've become vulnerable to relatively small negative events. If my trading partners have taken on lots of debt as well, we've set up a situation where there's the potential for chain reactions.

In the broadest sense, taking on big debt forces you to make an assumption about future growth--but as we saw before, growth depends on technological innovation, which is itself unpredictable. In other words, too much debt amplifies the fundamental unpredictability of innovation. Indeed, Taleb would like to go the other way and delink most of the economy from the financial system, arguing that "economic life should be definancialised."

Now, that goes against prevailing wisdom. The Free Exchange blog at the Economist magazine objected to Taleb's FT article, saying:

While limits on leverage are necessary and tax policy which encourages firms to issue debt rather than equity is misguided, the villain here is not complexity. Perhaps Mr Taleb and lawyers hired as regulators do not understand complex derivatives, but many people do. Should we outlaw innovation for the benefit of people who lack quantitative skills? These products do indeed provide a means to hedge risk. The crisis may have been much worse without some of the financial products that did pay off.

Further, the crisis was not caused by complex products. It was the result of a terrible assumption—that housing prices would continue to rise....Wishing away complexity does not change that or prevent future incompetence.

But Taleb is making a different point. He would argue that as complexity rises, you have to make more and more assumptions, with less and less understanding of the underlying system. As a result, the chance of making a 'terrible assumption' rises. This time it was the future path of housing prices. Next time it will be something different. As Taleb writes in his book. "Remember that infinite vigilance is just not possible."

Whew...sorry this took so long. The next piece in the reverse Black Swan series will look at innovation at the corporate level.



TrackBack URL for this entry: http://blogs.businessweek.com/mt/mt-tb.cgi/

Reader Comments

CompEng

April 20, 2009 12:09 PM

Interesting. I'd agree, with a couple caveats:


1. We need to somehow supply a real incentive for risk containment. The financial instruments employed didn't act to contain risk, but to ground it against systemic risk. The thinking appears to be along the lines of "We won't go down unless everyone else does too, in which case we still won't be in any worse shape than anyone else... and the government will bail us out". This is part of the risk of a hyper-competitive system. Your suggestion is a good one, but may end up being a heavy hammer. I think there's more brainstorming to be done here.

2. This is a very hard sell. The money in many fields, including sales and investing, is made by knowing something the competition doesn't. Obfuscation through complexity is an important part of that competitive advantage. There's a *lot* of money to be lost here. Given the amount of political clout financial guys still have this is going to be awfully hard. For all the media buzz on "populist rage", we still lack the political will to do anything useful instead of merely symbolic.

ekzept

April 20, 2009 02:04 PM

The present (former?) market economy has a problem properly pricing long term effects. It does not know how to anticipate them, and nothing in the standard discount rate approach encourages anyone to consider huge good or bad effects at a temporal range of a decade or more. This phenomenon might subsume some of the Swans effects. While perhaps this is an alien idea for finance, but the same people who object to Professor Taleb as being quant-adverse might be able to adapt the rate estimating techniques which biologists, geophysicists, and astrophysicists have devised for the expected long term rate of bolide impacts on Earth in order to estimate a background rate of good and bad huge events. Surely, the dearth of data on bolides suggests the scarcity of data on financial eruptions like the explosion of PCs, or the Web, or Google isn't out of league.

LAO

April 20, 2009 04:36 PM

Michael, In a round-about way, I think I agree with your conclusions.

Yes, simplify, to say the least. Being prone to over-simplification, I cannot understand why the most critical infrastructure of civilization -- the monetary system -- should be subjected to extraordinary leverage and gambling. I would feel the same if my nation invested sovereign wealth funds in the stock market, especially some other nation's stock market. I would fear food shortages if my nation pushed the farmers into towns and agreed to import all food without any control over exchange rates, and that is what happened to some. There are some gambles that should not be taken -- it is hard enough to manage the risk that nature and limited knowledge deliver without deliberately inviting more.

Risk may have failed as a control because the barrier to entry to cloaked and highly leveraged investments was based not on the financial status of the investor, but on the amount of pooled money that often represented thousands of individuals who could not afford that kind of risk. Every state pension fund that is almost broke, for instance, was run by someone over whom retirees and taxpayers had no real control -- so now future retirees have to give up what was promised, or the taxpayers have to replace what was lost, though neither group should have had their individual finances exposed to that level of risk.

I arrived at Taleb's 85% safe / 15% high risk split before I ever heard of Taleb. I did it by playing "what if" with every type of investment, using the Monte Carlo simulator at Nobel winner Bill Sharpe's "Financial Engines" to review different ways of accomplishing what I wanted over the long term with 95+% certainty. I could go a more traditional conservative way, or I could go this way and have a much higher upside potential and the same bottom-line risk. I'm no doubt overly smug about it, except for one sucker move into an ultra-short bond fund (lowest risk rating from Sharpe) that turned out to be secretly loaded with toxic mortgage assets. Caveat emptor. 2008 calendar year losses were still only 5%, so my husband and I are still ahead of plan in the long run and have capital losses to spare. Of course, if everyone invested that way, I doubt that anyone would make any money. I also keep converting chunks of IRA and rolled over 401K to Roth, just to reduce the unknown of future taxes -- even if the hedge is a bad bet, it's worth it to have the uncertainty reduced.

The important thing surely is to let individuals take the risks if they are up to it, but don't arbitrarily impose high risk (and losses) behind their backs. I would not expect anyone to innovate in such an uncertain world.

Yes, let the economy become a boring and transparent place, and let the self-selected high risk takers see what they can do, without granting them the right to take down the rest of us.

Ajay

April 20, 2009 08:25 PM

To begin with, your three opening conclusions are either obvious or overstated. Of course, there are positive black swans and their timing cannot be forecast, why would you even think they can be "controlled"? As for your third overstated conclusion, negative black swans have been nowhere as bad as the positive ones have been good, for at least the last 60 years. We do need simplification of financial instruments, but not through regulation. Investors have been burnt already, they will no doubt demand simplification where it's necessary. As for Taleb's recommended investment strategy, it's as dumb in its own way as the conventional wisdom. If you really wanted such a strategy, there's an easy and much better way to accomplish it: put all your savings into index funds. That way you have to do no work to figure out what's a better investment, will do better over the long run than 99% of current strategies, and are open to the upside from some random stock in the index taking off, which is presumably why actively-managed funds never beat index funds over the long run. I don't agree that the problem was that the market was a complex system, the problem was that people put money in risky investments without appropriate due diligence. I agree that partitioning financial institutions along the lines you put forward is the best route, I just don't see any role for regulation in making it so. No doubt investors have learned their lesson and will partition future institutions this way anyway. The big problem has always been depository institutions that take in money and lend it out for longer-term loans, but then claim that depositors can be paid back at any time, such as savings deposits or money market funds. Such institutions are fundamentally unstable and prone to runs; they should be replaced with time deposits, or at least explicit disclaimers that no govt assistance will ever be provided for such funds. Unfortunately, the govt just did the opposite last year by guaranteeing such deposits and money funds. I agree that leverage is a problem but the only solution is for investors and creditors to be aware of how much leverage they're exposed to. I don't think anybody taking on big debt was thinking that innovation would bail them out someday, it was just stupidity on their part. Arguing that they were implicitly betting on innovation and then blaming the unpredictability of innovation would just be stupid, though I'm not sure you're saying that.

I agree that economic basics should be definancialised, meaning no money in subprime CDOs or Bear Sterns, but there are always idiots who want to take that risk. You hear all the stories about middle-aged morons or retirees who exposed themselves heavily to the recent equities boom in the hope that they'd have a richer retirement as a result. Stupidity springs eternal, your hope of caging it with regulation only leads to it leaking out elsewhere. The Free Exchange blog's argument that complex derivatives are only opaque to those without quantitative skills is nonsense. No more esteemed an investor than Warren Buffet got out of the market 7 years ago because of the opacity and risk of widespread derivatives usage:

http://www.berkshirehathaway.com/letters/2002pdf.pdf (p. 13-15)

But that blog is correct that the main problem was the idiotic basic assumption that housing prices would never fall, which was compounded by the secondary problem of financial instruments that were essentially insurance with zero collateral, but for which the insured could then demand collateral in certain conditions. It is not that investors have to make a "terrible assumption," it is that lazy idiots would rather make a terrible assumption than do the work to make a safer choice. Infinite vigilance is hugely overstating what's necessary, Buffett saw that people were taking unnecessary risks based on easily understood analysis- he never submits to the same collateral requirements that many like AIG did, for example- and got the hell out. No reason others couldn't have done the same, especially considering that he openly and freely published his analysis. The fundamental problem with Mike's analysis is that he takes the bureaucratic approach that the economy, which has been a highly complex system for centuries, can be managed, through "innovation policy" or whatever. The truth is that the economy has always been a highly complex, self-regulating system, into which the govt bureaucrats have intervened in the past to on the whole make things far worse, not better.

LAO

April 21, 2009 12:58 AM

Ajay,

The S&P 500, or any number of stock indexes, would have returned zero over the last 10 years.

Ajay

April 21, 2009 01:52 AM

LAO, yep, doing better than actively-managed funds, most of which lost money over those 10 years plus took out hefty fees: your point is? In a decade of flat index funds, where the overall gainers like Google where equaled by losers like Lehman Brothers, you need to put in more investment work to only invest in the gainers. Since Taleb's advice is geared towards the average retail investor who doesn't want to do any investment work, better to be exposed to positive black swans through an index than by picking some random "exciting" market, like tech, and blindly throwing money down a black hole, like VC funds. Bond funds would have been another good choice. I'm not saying these are great investment strategies, only that they're better than many of the alternatives and are the only options for those who don't want to do any investing work themselves.

CompEng

April 21, 2009 10:10 AM

Ajay,

Didn't Warren Buffet make the exact same point you're making several times? :)

I think part of the problem with government regulation is it creates a dependency. Once you start regulating everything, it's a long hard road out. I think we would be better off with a substantially smaller government, but it's not trivial to get there, given where we are, without doing more harm than good. If we have to bail out the banks, we also have to regulate them. It would be better if we didn't have to do either, but I haven't seen a great analysis of the practical consequences if we didn't.

LAO

April 21, 2009 04:11 PM

Ajay,

I'm not sure that Taleb was trying to create a no-work strategy for the average investor. Even so, assuming that there are any "safe" places to park 85%, then it pretty much guarantees that losses cannot exceed 15% -- in actuality it will be less than that, because the high risk portion is not going to drop to zero, and the safe portion will usually generate a few percent.

It depends upon the investor's objective and time frame and risk tolerance. I had the attitude at that time that I wouldn't want bond funds just to follow conventional wisdom when they were being slaughtered by Fed actions. I am probably older than you and thus less tolerant of big losses that could take years to recoup. The riskier portion included specialized index funds focused in more volatile markets.

I certainly don't view this as a forever strategy, but a very effective one for my needs and for the times. While broad index funds were earning decent returns, the higher risk areas were generating stunning returns, so the fact that my execution of my own strategy was less than perfect is of little consequence. It is a humble strategy that covered for my growing nervousness about the economy. The time had come for me to take greater ownership of our finances and develop some higher certainty about where they were headed. The certainty can be more important than the potential when you are trying to make financial decisions, and I think that was Taleb's point.

Kartik

April 21, 2009 06:51 PM

More important than black swans, is the Impact of Computing :

http://www.singularity2050.com/2009/04/the-impact-of-computing-78-more-per-year-v20.html

I hope Michael will be the first economist to accept the 'good' variety of deflation as a major economic force (rather than just the bad variety of deflation).

Ajay

April 21, 2009 07:11 PM

CompEng, Which point are you referring to? I've made several points and the one point that Buffett made first, I linked to. It is not hard to get out of regulation, it's only that entrenched interests like politicians or those receiving subsidies will fight it tooth and nail. However, they're easily defeated if any voters paid attention at all. The banks were not actually bailed out in most cases, it was their creditors that were bailed out. The argument was that certain banks were so big and had so many creditors that their failure would lock up those creditors' assets in bankruptcy for too long. The govt created this bailout culture by previous unnecessary bailouts from past decades, for the same "too big to fail" reasons. That's why Citigroup was created a decade ago, to explicitly become too big to fail, as John Gutfreund said back in the '90s, "because of the size of the new Citicorp they'll be bailed out again when the cycle goes the other way, too big to fail will be the thesis as it was last time." http://video.google.com/videoplay?docid=4047250746566601531

The best policy would be to setup appropriate policies to take care of bankruptcy faster and make it known that nobody will ever be bailed out next time, "too big to fail" or not. Of course, that would mean politicians not using these emergencies to grab more power and make things worse precisely when they're already bad, so fat chance of that happening.

LAO, if that's not a no-work strategy for the retail investor and is aimed at the professional investor instead, his suggestion goes from merely idiotic to ridiculous. Of course your downside is protected with Taleb's strategy, but the loss of the high-risk part is almost guaranteed. Given that bond funds have done better than almost every equity strategy over the last 10 years, with the possible exception of certain international funds if you really got in that early, you'd probably have been better off not following the herd. Are you seriously talking about big losses from bond funds? It is all the equity funds that ended up taking big losses. I'm not sure what to make of your veiled statements about your own strategy. Agreed that avoiding risk is the goal of the retail investor, hence indices and bond funds as preferred investments.

CompEng

April 21, 2009 07:42 PM

Ajay,

I'm referring to the comment on the value of index funds. I read on CNN that Buffet had made that point and then followed through by making a bet with an exclusive hedge fund that, after fees, a common index fund would do as well or better.

Interesting point on the too big to fail argument: I'll definitely watch the video when I have a bit more time.

Ajay and Lao,
The best defense I can make for the non-professional investor keeping a smaller percentage of earnings in risky investments of their choosing is that if you're a non-professional investor, you may actually be able to spare the time and effort on getting a smaller number of bets right. For example, I've watched my company stock long enough that I'm confident I could day-trade it and earn the equivalent of a weekend part-time job, even without inside information. But I would make mistakes occasionally, and it's good to keep the magnitude of your gambles within what you can afford. My grandfather did the same with Standard Oil and managed to pad his retirement account significantly. But that's very different from taking 15% of your cash and putting it into risky adventures you know little about. You might as well go play slots every weekend as do that.

Ajay

April 21, 2009 08:15 PM

Kartik, you're dead right about the potential impact of computing. I myself am amazed when I got a Powerbook four and a half years ago for $1600 that could barely handle 720p video (and only after the latest Quicktime came out that squeezed as much out of the G4 cpu as it could) and just bought my next computer six months ago for $300, a Shuttle PC with a $50 Celeron dual-core CPU, that could easily handle 1080p video on a single core. That's approximately an order of magnitude price-to-performance ratio increase in 4 years. However, you have to honestly admit that this computing explosion has not had the full effect it should have had. The reason is monetization, we cannot easily and seamlessly trade a small bit of our work for a small bit of computing work, as we would be able to do with micropayments. So while I agree with you about the great potential for computing, it's unfortunately pent up at the moment by the bottleneck of a lack of micropayments, and will only explode once we get micropayments deployed. I despair that a smart guy like you, who can see the computing revolution coming, cannot seem to see that mp is the bottleneck right now.

CompEng, I was unaware that Buffett was also such a proponent of index funds, I guess non-idiot minds think alike. ;) The part of that video with Gutfreund is only the first 15 mins, well worth your time. I agree with you about playing slots, we're making the same point there. As for investors doing part-time work, I actually think there's a lot of scope there and it's likely to be the future of investing, though not through the insider or technical investing routes you mention. Certain retail investors could spend some time every week investigating certain companies or sectors and making investment picks based on their work, and selling their info too. I might choose to focus on tech companies, somebody else might choose retail companies. This type of distributed information-gathering and investment analysis has a lot of potential, for those retail investors who are willing to put in the time and effort. For the majority who don't, broader indices and bonds are the better route.

LAO

April 22, 2009 12:34 PM

Ajay,

Sorry I was not clear about bond funds. I don't view bond funds as capable of delivering the same magnitude of losses as stocks. It was a time frame when bond funds were pretty much guaranteed to go down, after a long good ride, so I chose not to go heavily in that direction.

By the way, there is definitely a well populated middle ground between the professional investor and the retail investor who isn't willing to focus on investing -- I always presumed that the entire Investment section of Business Week was aimed at them.

LAO

April 22, 2009 02:25 PM

Kartik,

Thanks, your computing impact link allowed me for the first time to envision a future health care solution that could actually do something dramatic for the economy at large -- an intelligent skin patch, for instance, that would tell me about worrisome trends, let me learn from nutrition experiments, and automatically send for help if a bee sting caused a severe reaction. I'm not saying that the thought has any merit, just that perhaps by harnessing computing power there is greater economic potential for health care research than I've been able to imagine, not just limited to specific health problems and the companies that succeed in addressing them.

I think the progress is not quite as rosy as depicted, though, since operating systems and applications consume an increasingly abusive amount of power and storage, and a lot of our personal computing power is absorbed by purposes that our not truly our own, such as advertising and defending against viciousness.

I also remain somewhat baffled at some of the things that don't materialize -- chargers and power supplies that truly prevent wasted electricity, ultra-cheap lap tops for this nation's own students, and any number of mundane ways of putting money in the pockets of business and consumers. There seem to be limits of another nature at work that are not obvious to me, plus you have to admit that even Moore's law will reach a limit. Your optimism is contagious, though.

Kartik

April 22, 2009 04:49 PM

Lao,

"ultra-cheap lap tops for this nation's own students,"

I'd call $300 Netbooks pretty darn cheap. What was available a decade ago?

"plus you have to admit that even Moore's law will reach a limit. "

Moore's Law will expire in 2018, to be precise. But other technologies pick up from there (quantum, molecular, graphene), without missing a beat.

Computation was exponential even before 1965's Moore's Law. Moore's Law is actually the 5th generation, not the 1st. If you really want to read up more on this, Google for Ray Kurzweil's Law of Accelerating Returns, and Moore's Law : The Fifth Paradigm.

Sean

April 24, 2009 05:39 PM

You might find this post I wrote last summer on the dangers of scale-free networks and desirability of loosely coupled networks in financial systems of interest:

http://www.parkparadigm.com/2008/08/23/averting-financial-ecological-disasters/

empedos

May 5, 2009 04:28 PM

For those who are familiar with the casino game of 21 (blackjack) Taleb's advices make perfect sense. There are several occasions during blackjack shoe games that the player has an advantage over the house. At these moments, the educated player splits his large bet into several smaller bets and places them on several table spots in order to minimize his risk for the same expected win rate. Nevertheless, the so called "covariance" (which is far from zero due to the fact that all these spots lie on the very same table) may lead to a disaster if proper money management techniques are not applied.
The complexity of today's financial products prevented the majority of the investors and professionals from realizing that they were placing bets "on the same table" i.e. that the same underlying assumption was hiding behind all those seemingly uncorrelated bets. This led them to catastrophic money management schemes.

Joey

May 30, 2009 01:30 PM

there is a post on "positive black swans" that is one point at www.venturepopulist.com

Thank you for your interest. This blog is no longer active.

 

About

Michael Mandel, BW's award-winning chief economist, provides his unique perspective on the hot economic issues of the day. From globalization to the future of work to the ups and downs of the financial markets, Mandel-named 2006 economic journalist of the year by the World Leadership Forum-offers cutting edge analysis and commentary.

BW Mall - Sponsored Links

Buy a link now!