[FRIAM] The Blunders that lead to the catastrophe - Humpty Dumpty's modeling school
The main article ( : ( : pete Peter Baston IDEAS www.ideapete.com The blunders that led to the banking crisis 25 September 2008 From New Scientist Print Edition. Subscribe and get 4 free issues. Rob Jameson WHAT'S the quickest way to kill a bank? As recent events in the financial world have shown, the answer is to deny them access to ready cash. Over the past year, a string of banking institutions have found themselves in such a "liquidity crisis": unable to convince the market they can honour their promises to pay back money they owe. The result has been a series of high-profile failures, from Northern Rock in the UK last year to Lehman Brothers last week. The crisis did not come without warning. Ten years ago this month, a giant hedge fund called Long-Term Capital Management collapsed when it too suffered a liquidity crisis. Yet banks and regulators seem not to have heeded the lessons from this wake-up call by improving the mathematical models that they use to manage their risk. That raises two key questions. How did the risk modellers get it so wrong? And what can they do to prevent similar crises in future? Banks are vulnerable to liquidity crises because they borrow money that may have to be repaid in the short term, and use it to back up more lucrative longer-term investments. If depositors withdraw their money and other lenders refuse to lend the bank the funds they need to replace it, the bank ends up in trouble because it can't easily turn its long-term assets into cash to make up the shortfall. Banks pay enormous sums to lure researchers away from other areas of science and set them to work building complex statistical models that supposedly tell the bankers about the risks they are running. So why didn't they see what was coming? The answer lies partly in the nature of liquidity crises. "By definition they are rare, extreme events, so all the models you rely on in normal times don't work any more," says Michel Crouhy head of research and development at the French investment bank Natixis, and author of a standard text on financial risk management. What's more, each liquidity crisis is inevitably different from its predecessors, not least because major crises provoke changes in the shape of markets, regulations and the behaviour of players. “Liquidity crises are rare, extreme events, so all the models you rely on in normal times don't work any more” On top of this, banks wrongly assumed that two areas of vulnerability could be treated in isolation, each with its own risk model. When the two areas began to affect each other and drive up banks' liquidity risk there was no unifying framework to predict what would happen, explains William Perraudin, director of the Risk Management Laboratory at Imperial College London. False assumption The first set of models covers the bank's day-to-day trading. These models typically assume that market prices will continue to behave much as they have in the past, and that they are reasonably predictable. Unfortunately, while this assumption may hold for straightforward financial instruments such as shares and bonds, it doesn't apply to the complicated financial instruments which bundle up different kinds of assets such as high-risk mortgages. What's more, information about the market prices of these products usually goes back only a few years, if it is available at all. "Statistical models based on short time series of data are a terrible way to understand [these kinds of] risks," says Perraudin. The models also assumed that the bank would be able to sell "problematic" assets, such as high-risk sub-prime mortgages, and this too turned out not to be true. "It's the combination of poor price risk modelling and being unable to sell out of the position that has produced the nightmare scenario," Perraudin says. The second set of risk models is intended to estimate the risk from borrowers failing to repay money they owe the bank. Because it's harder to sell off loans than bonds or stock, these models assume that the banks may have to bear the risks for longer. Such models were often regarded as the cutting edge of risk modelling, using sophisticated mathematics to predict how different debtors might be affected by economic conditions. However, Perraudin says these models mostly overlook how bad news can affect banks' ability to raise funds. "The real risk," he says, "turns out to be a cycle of drops." It plays out like this: word gets around that banks have got something on their hands that has dramatically lost value; this makes other institutions reluctant to lend them money to help them out, which in turn makes the value of their assets shrink further. The overall effect is to suck liquidity out of the market. Perraudin is working on a model for a hedge fund that takes account of this feedback, but he says it's a fiendishly difficult problem, partly because the models have to include a factor that captures the relationship between a bank's
[FRIAM] Blinded By Science - When models FAIL taking all the humans with them
The best piece in this entire article ( Posted in full because its subscribers only ) is " How could so many smart people have got it so wrong? One reason is that their faith in their models' predictive powers led them to ignore what was happening in the real world." What does that say about our entire approach to modeling especially when our clients believe in the model, thinking it reflects the real world as most of these poor ( pun ) bankers and financial geniuses did Ill post the master text from the main article " The Blunders that lead to the catastrophe " - How the brightest and the best mathematical modelers fail at their task of keeping disaster at bay in the next post To be fair the main quant defense is we didn't have enough related data but when the result is a financial loss to the financial world equal to all the sums made in the systems lifetime that excuse frankly sucks ( :: ( : pete Peter Baston IDEAS www.ideapete.com Editorial: When the numbers don't add up Blinded by science - Financial regulators have allowed themselves to be bamboozled ( Ably assisted by quants and techies on huge bonuses - my insert ( : ( : pete ) 24 September 2008 From New Scientist Print Edition. Subscribe and get 4 free issues. ONE of the most alarming things about the crisis in the global financial system is that the warning signs have been out there for some time, yet no one heeded them. Exactly 10 years ago a hedge fund called Long-Term Capital Management failed to convince investors that it could repay its debts, thereby bringing the world to the brink of a similar "liquidity crisis" to the one we now see. Disaster was averted then only because regulators managed to put together a multibillion-dollar bailout package. LTCM's collapse was particularly notable because its founders had set great store by their use of statistical models designed to keep tabs on the risks inherent in their investments. Its fall should have been a wake-up call to banks and their regulatory supervisors that the models were not working as well as hoped - in particular that they were ignoring the risks of extreme events and the connections that send such events reverberating around the financial system. Instead, they carried on using them. Now that disaster has struck again, some financial risk modellers - the "quants" who have wielded so much influence over modern banking - are saying they know where the gaps in their knowledge are and are promising to fill them (see "How the risk models failed the world's banks"). Should we trust them? Their track record does not inspire confidence. Statistical models have proved almost useless at predicting the killer risks for individual banks, and worse than useless when it comes to risks to the financial system as a whole. The models encouraged bankers to think they were playing a high-stakes card game, when what they were actually doing was more akin to lining up a row of dominoes. How could so many smart people have got it so wrong? One reason is that their faith in their models' predictive powers led them to ignore what was happening in the real world. Finance offers enormous scope for dissembling: almost any failure can be explained away by a judicious choice of language and data. When investors don't behave like the self-interested Homo economicus that economists suppose them to be, they are described as being "irrationally exuberant" or blinded by panic. An alternative view - that investors are reacting logically in the face of uncertainty - is rarely considered. Similarly, extreme events are described as happening only "once in a century" - even though there is insufficient data on which to base such an assessment. “Bankers' faith in their models' predictive powers led them to ignore what was happening in the real world” The quants' models might successfully predict the movement of markets most of the time, but the bankers who rely on them have failed to realise that the occasions on which the markets deviate from normality are much more important than those when they comply. The events of the past year have driven this home in spectacular fashion: by some estimates, the banking industry has lost more money in the current crisis than it has made in its entire history. Can modellers do better? There are alternatives to the standard approach: models based on people's real-world behaviour (New Scientist, 30 August, p 16) and on "virtual agents" (New Scientist, 19 July, p 32) have shown promise, though these are still fringe fields in economics. Most quants, while acknowledging the shortcomings of their models, tend to argue that approximations are necessary, given the difficulty of modelling extreme events, which are in any case rare. That may be true, but it is dangerous to assume that the approximations are sound. Sometimes even small modelling deficiencies can have huge consequences. Nassim Taleb, an expert on chance and co-director of the Decision Rese
[FRIAM] Fwd: sfx Events: Occam's Razor, Saturday, October 18, 4:00 pm
Stop by this Saturday--this looks like a stunning show. Begin forwarded message: From: Don Begley <[EMAIL PROTECTED]> Date: October 15, 2008 10:30:17 AM MDT To: [EMAIL PROTECTED] Subject: sfx Events: Occam's Razor, Saturday, October 18, 4:00 pm Reply-To: [EMAIL PROTECTED] Simple but No Simpler: Occam's Razor Opens This Week Saturday, October 18 at 4:00 pm "All other things being equal, the simplest solution is the best." William Oakham, 14th century "Things should be made as simple as possible, but not simpler." Albert Einstein, 20th century Santa Fe Complex is located in the Railyard Art District within walking distance of the hotels, restaurants and shops at the plaza downtown. We're housed in two facilities, the project space at 624 Agua Fria and the work space at 632 Agua Fria. The conference area contains meeting rooms and facilities for short- term use associated with on-going sfComplex projects. The project space houses the great room, where we hold events and offer Internet access, working facilities, a coffee lounge and work carrels for laptop users. While there is parking at 624 Agua Fria, the Romero Street parking lot is more conveniently located for the 632 facility. Romero St. is an old-style Santa Fe ox-cart road just east of the 624 driveway. Follow it until it opens up to two lanes and turn hard right into the parking lot for 632. Here's a map to our location. For more information, call Don Begley at 505/216.7562. Forward email This email was sent to [EMAIL PROTECTED] by [EMAIL PROTECTED] Update Profile/Email Address | Instant removal with SafeUnsubscribe | Privacy Policy. Email Marketing by Santa Fe Complex | 624 Agua Fria | Santa Fe | NM | 87501 FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] The Evolutionary Basis of Depression
I found it to be an interesting read, but the ending left me kind of sad. ;-} On Wed, Oct 15, 2008 at 1:18 PM, Jochen Fromm <[EMAIL PROTECTED]> wrote: > There is a new book from Paul Keedwell named "How Sadness Survived: The > Evolutionary Basis of Depression". The thesis is that depression has evolved > to avoid the pursuit of unachievable goals. What do you think of it? See > http://blog.cas-group.net/2008/10/depression-as-adaptation/ > > -J. > > > FRIAM Applied Complexity Group listserv > Meets Fridays 9a-11:30 at cafe at St. John's College > lectures, archives, unsubscribe, maps at http://www.friam.org > FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
[FRIAM] The Evolutionary Basis of Depression
There is a new book from Paul Keedwell named "How Sadness Survived: The Evolutionary Basis of Depression". The thesis is that depression has evolved to avoid the pursuit of unachievable goals. What do you think of it? See http://blog.cas-group.net/2008/10/depression-as-adaptation/ -J. FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Selection, Reproductive rate, and Karrying Kapacity.
Russ & Nick, Regarding multilevel selection, aren't there multi-level systems involved? Certainly a change in cell behavior affects the organism, and the local pack, and larger population, and the local ecology too. But you also have reverse effects in that the larger scale orders greatly alter what each lower order differences will make a difference. Then there's the interesting aspect that some kinds of complex systems overlap in lots of ways, like complexly varied ecosystems with many intersecting levels, and so a simple hierarchy is not what is operating either. What can, if you follow it through, straighten all that out is considering systems as individual exploratory networks. Then you can still have independent ones that overlap and they still work fine, and all of them can have a role in mediating selection for all the others. Phil Henshaw .·´ ¯ `·. ~~~ 680 Ft. Washington Ave NY NY 10040 tel: 212-795-4844 e-mail: [EMAIL PROTECTED]explorations: www.synapse9.com "it's not finding what people say interesting, but finding the interest in what they say" FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
Re: [FRIAM] Selection, Reproductive rate, and Karrying Kapacity.
Russ, That's a good example about the difference between breeding for the best bird vs. the best bird environment, but they don't immediately seem to address whether variation is developmental or random. It's tricky to find the hard evidence, but I don't know of anyone saying they could show statistically that random variation would be constructive either. My hint is that the organizational processes we can observe the workings of generally do exhibit developmental variation, like we use in any programming or other design process. Once you think of the first part in the design, the process that seems to work better for people is adding a second related part, *if the first seemed to work*, and that way extending variations from prior variations experimentally, rather than randomly.It takes some effort to imagine how genetic variation could be 'tree like' instead of helter skelter. but there a number of ways. What you need is for competitive advantage to multiply related variations. In any case individual organism growth and development is clearly a branching process, and speciation seems to clearly be an extension of a prior branching process. Maybe speciation occurs by a branching process too.In speciation the form of the organism appears to extend its developmental trees as whole, all at once, something that a tree like variation process could do and a random variation process very likely not. So that's what I think would be sensible to look for. Besides, tree-like development could do one thing that random variation can't, produce developmental step changes that begin and end. That's what is apparently displayed by my little plankton. I'd really love to have the $'s to do a photo animation of how the smooth to then bulgy shapes on it's shell changed through the dips and turns of it's dramatic changes in size from one to another stable form. Phil From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Russ Abbott Sent: Monday, October 13, 2008 5:15 PM To: [EMAIL PROTECTED]; The Friday Morning Applied Complexity Coffee Group Subject: Re: [FRIAM] Selection, Reproductive rate, and Karrying Kapacity. One of my favorite books of the year is David Sloan Wilson's Evolution for Everyone. Wilson has been arguing for multi-level selection for quite a while -- and as far as I'm concerned he makes very good points. The fundamental insight is that everything is both a group and an individual. And hence virtually anything can evolve at the individual level -- even if it's a group. Wilson likes talking about religions (or religious groups united by religious practices) as an example of a group that competes evolutionarily. He argues that religious that promote hard work, support of fellow members of one's religious community, etc. tend to succeed. He also tells the story of the experient in which groups of hens were allowed to evolve. It was done in two ways. 1. Start with (say) a dozen cages, each with a certain number of hens. At the end of a given time, the best egg-layer in each cage were bred to create a second generaation of cages. Continue for a certain number of generations. 2. Start the same way, but after each generation, breed the best cage, regardless of how its individual members performed. Continue for a certain number of generations. The result: breeding cages was much more successful than breeding individuals. In this case it turns out that breeding individuals produced macho hens who pecked each other to death. Breeding cages produced cooperative hens who lived happily with each other and produced lots of eggs. The larger lesson is that groups often embody structures that support the group's success. To enable those structures the group needs members who play various roles. Simply selecting the most productive members of a group and rewarding them breaks down the group structure. -- Russ On Mon, Oct 13, 2008 at 11:18 AM, Nicholas Thompson <[EMAIL PROTECTED]> wrote: All, Here are some comments on various comments. I succumb, reluctantly, to the community norm about caps. [grumble, grumble] Glen Said > The idea of expansion and contraction is interesting: rapid expansion of populations (when selection is relaxed) vs. rapid contraction of populations (when selection is intensified). The human population went indeed through a phase of rapid expansion in the last decades while natural selection was released through cultural and technological progress. Seed Magazine has an article about human evolution and relaxed selection, too http://www.seedmagazine.com/news/2008/10/how_we_evolve_1.php <=== Nick Replies ===> I think this is a confusion between carrying capacity and selection. When, for some reason, carrying capacity is increased, the whole population can expand, but this does not stop selection. It may change the nature of selection from tracking how well individu