Peter Cebon on innovation and the financial crisis

I was sufficiently taken with this piece in the Fin that I asked it’s author Peter Cebon of the Melbourne Uni Business School if I could republish it here.

Were told that the root cause of the current financial crisis is a few regional financiers selling dodgy mortgages to poor people. How can that be? The sub-prime mortgage market is a small portion of the U.S. mortgage market, which is a small portion of the U.S. credit market, which is a small portion of the global credit market. How can defaults in a sub-sub-sub-market destroy banks in two continents and send countries bankrupt?

In his landmark analysis of the Three Mile Island nuclear accident, sociologist Charles Perrow argued that that meltdown was not caused by any particular component or operator failure. Rather, it was caused by a number of small component failures that interacted in unpredicted ways to escalate to a catastrophe. Those failures included a valve that didnt reseat properly, a meter which gave deceptive information, another meter that didnt work properly, and some perceptual errors by the operators.

If a system has high interactive complexity, it is hard to predict the impact of a particular action on the other elements within the system. If a system is tightly coupled, it means that an action at one point propagates rapidly throughout. Perrow argued that systems with high interactive complexity and tight coupling are more prone to systems accidents.

The financial sector meltdown is also a systems accident. However, in this case, no failures are interacting. With the possible exception of a U.S. Securities Exchange Commission decision, in 2004, to eliminate the capital reserve requirements of investment banks, it is hard to say that anyone erred, anywhere up the food chain from the mortgagees to the banks. Rather, people just innovated so they could do their job better, and those innovations interacted. One group innovated by creating the sub-prime mortgage category; another by finding a way to lend people the deposits on their homes; a third by working out how to slice and dice the mortgages and combine them with bonds to turn them into securities; a fourth by working out how to price those securities using mathematical formulae. The list goes on.

Innovators deploy resources in novel ways to produce desired outcomes. Innovations come in two varieties: radical and regular. Radical innovations bring new resources to bear on an old problem. For instance, LEDs produce light using fundamentally different technologies and materials than Tungsten filaments. Similarly, repackaging mortgages and bonds into tradeable securities was a radical change from banks holding individual mortgages on customers properties. In addition to fundamentally transforming a core technical process, radical innovations change the way the artefact interacts with the rest of the system. Just as LEDs produce much less heat than Tungsten bulbs, but require a much more stable voltage, securitized debt instruments created fundamentally different relationships between the mortgage and the rest of the financial system. One change was that the mortgage issuer had much less interest in the debt they had bought, because they were rid of it as soon as they sold it to the investment bank that repackaged it. Another is that the instruments are so complicated that it is no longer possible to trace where the risk associated with a particular mortgage is located, which means the securities cannot be turned back into mortgages. A third is that instead of being held locally, the mortgage is now smeared across the globe. It is highly unlikely that the mathematicians who invented these instruments, or the regulators who regulate them, even considered these impacts, let alone thought about their implications. That is, a secondary consequence of radical innovations is that they change the linkages within the larger system, and make it harder to understand and analyse. Radical innovations increase interactive complexity and increase the possibility of mis-matches between reality and what regulators assume they are regulating.

Regular innovations generally involve using the same resources more efficiently. As a result, regular innovations tighten coupling. Every generation of computer chip is basically the same as the one before, but the chips are faster because the engineers find ways of making the lines in the Silicon thinner, and packing the transistors more tightly. Similarly, a sub-prime mortgage is basically a regular mortgage, pitched at a different market segment, as is a loan provided so people dont need to actually have the deposit. By deploying resources more efficiently, innovators remove slack from the total system. That is, each regular innovation makes the system more likely to propagate, rather than absorb, things that go wrong.

Consequently, no smoking gun explains the meltdown. Everyone, with the possible exception of the regulators, just did their job. The meltdown occurred because of the way the mis-matches between the innovations and the regulatory environment interacted. The consequence is almost incomprehensible, given the root sources a collapse in the market for inter-bank lending.

These days, everyone seems to think innovation is a great thing. While innovation can be incredibly valuable, all innovations, whether in genetically modified organisms or financial instruments, can have dire and unpredictable consequences. Consequently, an innovation economy needs a strong and intelligent regulatory regime that manages the systemic risks to ensure that the system as a whole has adequate buffers and safeguards.

This entry was posted in Business, Economics and public policy. Bookmark the permalink.
Subscribe
Notify of
guest
7 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Chris Lloyd
Chris Lloyd(@chris-lloyd)
13 years ago

The general contention of this article that innovations make systems more complex and lead to unintended consequences, is well made. But I cannot agree with Peters statement: “It is hard to say that anyone erred.”

The central, but seldom unmentioned, villains in this saga are the rating agencies. They failed to measure the risk of the securities, and deliberately adjusted to their ratings formulas to under-estimate risk. Why? Because they were chasing business. The consequence was, as Peter says, that the mortgage issuer had much less interest in the debt they had bought, because they were rid of it as soon as they sold it to the investment bank that repackaged it.” They could just pass the risk onto others under false pretences with the imprimatur of Moodys. It was basically a scam.

Has there been any public discussion of government taking over the role of rating? They might be better at this than regulating. The government, at least, has no incentives to under-estimate risk since they end up picking up the pieces. . I cannot imagine any one has much faith left in private ratings agencies any more. US Ratings Inc might actually make a decent profit.

p.cebon
p.cebon
13 years ago

I considered Chris’s argument when I wrote the piece, and am planning on writing a longer version in which I explore it in more detail. However, there are two key points to be made.

First, unless I’m misinformed, my understanding was that the rating agencies did commit acts of fraud, as Chris’s post seems to imply. Rather, I understand they relied on highly mathematical formulations of risk that were quite disconnected from reality, and those who thought about what the numbers were saying were quite happy if the risk numbers seemed a bit low. Rather than being fraud, there was a serious lack of integrity.

Second, is it reasonable for us to expect that they would act with integrity and try to produce unbiased risk estimates? I would argue that it would not be. Quite simply, true professionals act with integrity because they are taught to. They are members of professions with codes of conduct and standards of practice. These people, in contrast, have an entire “professional” education that tells them that the world runs on agency theory. They simply acted as they have been taught that agents should.

Ingolf Eide
Ingolf Eide(@ingolf)
13 years ago

Peter, a most interesting article. Thanks for allowing Nicholas to republish it here.

Like Chris, I entirely agree with the proposition that “systems with high interactive complexity and tight coupling are more prone to systems accidents”, although I’d add that unless the result is a high level of systemic gearing, the consequences are likely to be quite manageable. I also reluctantly accept your point that it may be unrealistic to have expected more of the players to act with integrity, although you do seem to me much too kind in some of your judgments. Do[ing] their job better and deploying resources more efficiently seem to me unduly generous interpretations of what was actually happening. The incentive structure that progressively evolved may have made many of the subsequent actions understandable on a purely pragmatic, short term basis but I don’t think this changes the fact that errors of both conception and execution were legion at almost every stage of the process.

What I see as very valuable in your piece is the focus on securitisation. It was indeed a radical innovation, a game changer. Taken to its extreme, as it was from 2003-20006 by an unholy combination of mortgage companies, investment banks, ratings agencies and gullible institutions, it ensured the debacle would be off the charts. It still seems remarkable that so few grasped the longer term consequences of separating origination from ongoing responsibility. It’s not easy to understand how otherwise intelligent people could imagine this would end in anything but disaster.

The obvious culprits are those in charge of regulating the financial markets. Their abdication of responsibility will be intensely studied for decades to come. Greenspan will probably bear a disproportionate (albeit in my view justified) share of the burden since he placed all of his reputation at the service of policies that ought to have been viewed as snake oil. He conferred a sort of legitimacy where none should have been possible. In any case, there’s plenty of blame to go around.

Still, enjoyable though it may be, indulging in righteous rage probably doesn’t help all that much. Shifting the focus to a closer analysis of the failures in system design and implementation, as you’re doing, is a more constructive approach. This is of course a very large topic (indeed, short of the meaning of life they don’t come much larger), but it seems to me taking in a much longer perspective than the last decade is vital if we want to come away with any durable lessons. The sheer extravagance of recent excesses was as I see it merely the final blow off stage of a process that’s been underway for generations.

Rather than go on at length here, if this is of any interest to you a brief overview of why I see things this way can be found in this recent post, starting about half way through with the phrase: A market based economy is a complex, self organising system. In order to adapt and prosper, such systems must have constant real world feedback . . . . . Also, some 18 months ago Nicholas and I (and James) had an extended discussion about secular influences in monetary and economic matters which covered some of this territory in a bit more depth.

In any case, Peter, whether you decide to have a look at those or not, many thanks for the article.

NPOV
NPOV
13 years ago

Yes, I have to take issue with the idea that it’s “hard to say that anyone erred”.

Here’s another interesting article from the NY times covering what was obviously a major contributing factor behind the crisis: overeagerness of certain politicans to encourage home-ownership in ways that were often fairly obviously economically unsustainable.

http://www.nytimes.com/2008/10/19/business/19cisneros.html?em (free registration required)

(I have to say, the fact that it was so difficult for policemen, firemen, teachers etc. to afford homes in the first place – and this in a time when house prices were not particularly high – points to trying to treat the symptom than treating the cause.)

I also caught the end of a story on, I think, 60 minutes last night where there was Wall St ‘observer’ Jim Grant who supposedly foresaw the crisis, and was quite confident that at least ‘someone’ would go to jail over the whole affair. It’s hard not to be extremely skeptical that it actually will be those that genuinely deserve to be.

p.cebon
p.cebon
13 years ago

Dear Nicholas, Chris, Ingolf, and NPOV. Thanks very much for the extremely useful comments. Rather than reply to the the last two at the time, I’ve incorporated my responses into a longer version of the piece, which I’m preparing for Sloan Management Review. If you’ld like to see it, or even better still give me comments, please send an email to my Melbourne Business School email address, and I’ll send you a copy. (You can find the address at http://www.mbs.edu/go/faculty-and-research/faculty-profiles?researchOnly=1).

Tel_
Tel_
13 years ago

In his landmark analysis of the Three Mile Island nuclear accident, sociologist Charles Perrow argued that that meltdown was not caused by any particular component or operator failure. Rather, it was caused by a number of small component failures that interacted in unpredicted ways to escalate to a catastrophe. Those failures included a valve that didnt reseat properly, a meter which gave deceptive information, another meter that didnt work properly, and some perceptual errors by the operators.

What I’ve noticed is that both workers and plant managers are reluctant to rush at replacing a broken meter, because it’s only a meter and there are plenty of others and the plant does not actually require that meter to keep working. Probably eventually it will get replaced, but not as a high priority item, and only when there are staff around with time on their hands. I’ve seen exactly the same thing with computer backups, vehicle maintenance, etc.

Everyone, with the possible exception of the regulators, just did their job.

Everyone always “just did their job”, see also: Milgram experiment.

The thing that bothers me, is that the financial industry “just did their job” at extraodinarily elevated remuneration (both bankers and regulators), as compared with most of the more directly productive vocations. We were told that financial management was special wizardry and only artfully qualified people could be trusted in these most important positions.

If we are now to be told that all their risk management and stabilising regulations didn’t actually give any tangible result and we have a boom-and-bust economy full of market bubbles, no better than the Tulip Craze of 1637 (or many more recent examples), then might as well be a bit frugal and NOT buy the overpriced and ineffective service that the modern financial markets offer.

After the pile of equations settles and the noise of high speed paper shuffling dies down, we learn once again that you don’t get something for nothing. And we paid WHAT for this lesson? I can sell you tea leaves much cheaper, they won’t tell the future either, but you will get a nice refreshing cup of tea.

Still, enjoyable though it may be, indulging in righteous rage probably doesnt help all that much. Shifting the focus to a closer analysis of the failures in system design and implementation, as youre doing, is a more constructive approach. This is of course a very large topic (indeed, short of the meaning of life they dont come much larger), but it seems to me taking in a much longer perspective than the last decade is vital if we want to come away with any durable lessons. The sheer extravagance of recent excesses was as I see it merely the final blow off stage of a process thats been underway for generations.

Well, I won’t promise to be able to run a global financial system that is anything other than a chaotic mess… but then again, no one else can honestly promise anything better and what I can offer is a chaotic mess with VERY LOW OVERHEADS. My method is easy to learn and offers enjoyable righteous rage in the process. I learned the words from Donald Trump, so I know they must be the right words. Line up all the bankers earning over $1M per year and say, “You’re fired!” then line up all the regulators on similar income and say the same. Then wipe off every bit of banking regulation written between the 17th century and today, because it was all useless. Peter Cebon has kindly provided proof of why it was all useless (although to be fair, Paul Omerod gave a similar explanation years ago).

Arcelia Siders
11 years ago

Mortgages, in general, are secured debt. They are secured by the home that you bought with the home loan. It can be not standard to have any further rights in a mortgage loan than foreclosure (it can happen, though). About the only way they can come after you is if they can prove that you were trying to cheat them when you got that mortgage (fraud). If you let the home go to foreclosure, your credit will be absolutely trashed for several years. You might want to see if you can negotiate a short sale on that property (with that lender). A short sale won’t trash your credit like the foreclosure would.