By Sell on News, a global macro analyst. Cross posted from MacroBusiness
A question that has been asked, but not nearly often enough, is why did the complex risk defrayal methods fail so completely during the global financial crisis? The GFC proved that risk measures based on INTERNAL measures, i.e. measures within the system, will fail. At the time of the GFC many participants thought they had defrayed their risk only to find out that they had not.
What is needed is a measure of risk that is EXTERNAL to the system. This is a logical necessity. The financial system works of the assumption that risk can be shifted from individual exposures. But risk cannot be eliminated, it can only be moved, something that was obvious to many outside observers but not to financial practitioners. What happens is that the risk is moved on to the system, which exposes all participants in ways they cannot anticipate. That defeats risk management. This is particularly evident with the proliferation of high frequency trading, where we are seeing stop loss orders fail because of system melt down.
An anthropologist friend, Dr Larry Cromwell, calls this vulnerability. When added to risk measures, it can give a much better chance of being able to defray risk effectively. Certainly something is needed that is outside the financial system. It is analogous to the famous incompleteness theorems of Kurt Godel, the Austrian philosopher which is stated thus:
If the system is consistent, it cannot be complete.
The consistency of the axioms cannot be proven within the system.
This should be mandatory reading for any financial analysis of risk. As Godel showed, even a basic arithmetic equation requires some external element to be validated for it to be proved. Anything else is self delusion and it was just that self delusion that plagued practitioners in the lead up to the GFC. They believed they had covered all the options that were possible, only to get a nasty surprise.
Godel has profound implications for many areas of science, including physics, mind and machines, computerisation . But it is rarely, as far as I am aware, used in financial analysis. It should be, if only to serve as a reminder that even if the mathematical systems are pure on their own terms, they cannot be complete. And when we add in the fact that financial markets are full of people who understand the risk analytics and can exploit them, we can see how far we are from having risk defrayal systems that will actually work.Which means getting fee income prove difficult. Can’t have that.
First step in looking at vulnerability is to scale the potential risks, which means looking at a measure that comes from outside the system itself. Cromwell sayswhat is needed is an assessment of scale and scope:
”Scale” has to do with the degree of destruction; “Scope” has to do with spread of effect. Low-Scale Wide-Scope vulnerabilities and High-Scale Limited-Scope vulnerabilities can be ranked in the same ranking table: Low-Scale Wide-Scope would be e.g. pandemic virus, where huge numbers of workers were afflicted with few or no deaths but very wide economic disruption; High-Scale Limited-Scope would be e.g. World Trade Centre. The GFC has been Low-to-Medium Scale Very Wide-Scope–and as such, like the Boiling Frog, it did not notice its own approaching demise (“it” being the denizens of the financial system, collective in their observations and awareness).
So-called “systemic risks” are all Wide Scope; some of them ‘open the door’ to some very High Scale consequences here and there, until the door is closed (through recognition of the systemic nature of the vulnerability.
The point about measures such as scope and scale is that the point of perspective comes from outside the system. If Godel were a financial trader, he would no doubt agree that is essential.
43 comments
Comments are closed.
when wall street captures the regulators and both political parties there is no “outside the system” powerful enough to give proper perspective. rules are hastily rewritten or ignored and risk is spread to those not powerful enough to resist it. word is corzine is looking to start a hedge fund.
Introducing interest into a monetary system coincidentally introduces the concept of risk premiums. That’s why people are theoretically willing to loan money to, say, Greece at a 500% yeild when they know that they will absolutely never be paid back instead of not at all.
This there is no other end state possible for a financial system with an interest component on it but collapse. That is the ultimate reason why all of these risk-shifting/hiding strategies are failing, one after another.
Uh, not yeild. Bummer there’s no edit feature. Well I suppose as insane as things are today such a yield isn’t impossible.
The problem with interest bearing debt is not the idea itself, but who it is that takes the damage when there is a default.
The current issues are:
no outside control. This in the sense that there is nothing that constrains the lender from lending beyond a certain point. Reserves do not do so, as they have a time lag of up to 1 month.
All risk is with the borrower. Disregarding systemic collapses like the present, where the market is flooded with defaulted on products, the lender can either collect interest or the proverbial pound of flesh to be resold somewhere else.
End result is a predatory lenders market, where they will lend to anyone and everyone that walks in the door.
Maybe under normal circumstances (i.e., before de-regulation) criminal prosecution by the justice system might have been the outside influence able to stop the risk-taking. Now, one wonders what the outside influence will be to stop the on-going financialization of the economy.
death is always with us… that shadowy caricature of our useless lives… do we want to leave a legacy of zero..? perhaps imagined as beneficial to our offspring…? economically…? or rather… a tale of difficult and offensive assertion? i really do wonder what the hell ‘life is worth’… i love people; but i hate you in digital effigy… the more abstract it gets, the less i can see a way out.. the more ‘realistic’ we get- the more I am ‘SURE’ we are goners… plenty to live in the….
Risk? How about the risk of putting all our eggs in one basket – into a single government/private sector money supply that we all must use?
How about the risk of government insuring the depositors in private lending institutions?
How about the risk of a money system that requires exponential borrowing to provide the interest for existing loans?
How about the inherent instability of a money supply that is lent into existence? And which is destroyed when the loans are repaid?
How about the risk of mocking God with a money system based on usury for stolen purchasing power, especially from the poor and helpless?
How about the risk of trusting in a money system that is a proven killer (eg. WWII, a major cause of which was the Great Depression)?
You’re a one-note wonder, aren’t you?
G-Major.
In the realm of unprovable axioms, this one tops the list.
Well, isn’t that how the glass is shattered? :)
The financial system is huge and complex but it totally lacks diversity. Not a good recipe for survival. The complexity comes in the form of arcane rules and vested interests. Deregulation didn’t help. We didn’t create diversity. I don’t understand how there can posssibly be a viewpoint from outside this system so that risk can be precluded. Risk will always be shifted off illegally until there are laws, imposed by the social system itself, to intercept it. To think that an analysis of scale and scope isn’t already being done on a minute scale, is a little fluffy.
but it totally lacks diversity. Susan the other
Eggs-(all in one basket)-actly!
There is no excuse for a nationwide (much less worldwide) synchronized boom-bust cycle. Yet we have one.
note to webmaster … some irritating popevers on this site … when i have to close a window that opened just because my mouse was in the neighborhood, i lose interest in the content.
If referencing the share menu, it goes away after the mouse has not been on the button for a bit.
So called risk management was nothing but religion dressed up with mathematics. Its purpose was to justify the gambling in derivative contracts that enabled looting in the short term by traders and executives. You don’t need fancy analytics to puncture this bubble of delusion. Contemporary finance is about two things and two things only: usury and looting. Those who dominate the field are simply polished self aggrandizing bullshit artists and connivers adept at bribing legislators, manipulating politicians and dodging prosecutors. They have created a house of cards that inevitably will collapse on the heads of everyone else and the only thing which can be done is to kick the can and hope to buy some time, never knowing how much we will actually get.
What Jake said….
To which I would add: This leaves the rest of us in the unavoidable position of the Prisoner’s Dilemma. Do we cooperate (follow the rules and get taken advantage of) or defect and join the stampede to get our own piece of the action before the whole sorry mess collapses?
Jake-“bushbama sin-drome”…
But the converse is true as well, is it not? (It’s been longer than I care to think since I read up on this stuff.)
That is, if a system is complete, it cannot be consistent.
So where does that leave algorithmic trading?
And is a level of fraud a systemic consequence of the lack of consistency?
Goedel proved that self-consistent systems can’t be complete. But I do not think that he said whether or not complete systems can exist. I suspect no finite set of rules can be complete whether self-consistent or not. Seems to me that if a system can be proven to be complete, it would have to be self-consistent, but then it couldn’t be complete!
With regard to Financial Risk Management: The problem really is one of Uncertainty not Risk. Risk is quantifiable as a probability because all possible outcomes are known. Risk is just the number of desired outcomes divided by the total number of possible outcomes (assuming equal probability). Uncertainty can not be quantified because the outcomes are truly unknown.
Gambling on Black Jack, dice or roulette, for example, is amenable to risk management because there is a finite number of outcomes with computable probabilities. Gambling on horse races, football games or poker is not amenable to risk management because no probabilities can be assigned all the possible outcomes.
Financial markets fall into the latter category. Uncertainty dominates and risks cannot be computed. There is simply no such thing as financial risk management.
This is not what Godel’s theorems actually say. For example, the system of Euclidean geometry is both complete and consistent. ‘complete’ in this context means that every statement made within the scope of Euclidean geometry is either provably false or provably true. Many such systems exist; another example would be first order predicate calculus.
What Yves is probably thinking of are the theorems which say that sufficiently powerful systems cannot be both complete and consistent. One such sufficiently powerful system would be the formal arithmetic based upon Peano’s axioms.
So wasn’t Godel just saying that if the argument is circular it contains no logic? I think MacroBz wants to shine us on.
No. He wasn’t saying that. Suppose you have a formal logic system with axioms stated in some formal language, and with rules of inference that let you combine axioms to make provable statements, and let you combine axioms and provable statements to make other provable statements. Goedel proved that if one of these systems is complex enough (and a system that describes ordinary arithmetic is complex emough,) then one of two things will happen:
1) there will be a statement constructed according to the language’s grammar that can’t be reached by any combination of the axioms and proof rules — either to prove that it’s true or to prove that it’s false. (i.e. The system is incomplete.)
or 2) there will be a statement that can be reached in at least two ways, one that proves the statement, and one that proves the statement’s negation. (i.e. the system is inconsistent.)
This gets important if you try to, say, do a computer analysis of the effects of High-Frequence-Trading. The set of programs that you create to infer the HFT behaviour will fall into one of the classes above, Class 1 risks vanishing down a rabbit-hole and never coming back to you with an answer; class 2 risks giving you a wrong answer — no help if it comes back later and also gives you the opposite answer; you used the computer programs because you couldn’t answer the question without them.
We were talking over coffee this morning about _The Hitchhiker’s Guide to the Galaxy_, and Deep Thought, and the program to compute the question to the great answer of Life, the Universe, and, Everything. Douglas Adams nailed a lot of important ideas there.
…”front-running”…is where algorithmic trading is left..
This is an intellectually interesting piece and it makes a point that will be necessary to keep in mind when the time comes to design and implement a financial system that works for all.
However, I think it is quite unwarranted to assume that risk management failed due to incompetence rather than quite knowing theft or at a minimum venal negligence.
Sometimes, it seems to me that there are two separate conversations that go on. One assumes that the system was sincere but needs to learn some lessons. The other tends to ignore the detailed functioning of the system and concentrate on the actual kleptocratic nature of the system as a whole. Either of these alone, even when it has useful insights, leaves me feeling frustrated.
The possibility of fraud is one of the reasons financial risks cannot be computed. Uncertainty overwhelms risk.
Hear, hear, Jessica! Never attribute to chance and incompetence what can be better explained by calculated malice, greed, and powerlust.
It’s obvious that what we need gong forward to prevent all uncertainty and risk, is a global central bank with an HFT version of Bernanke, who very calmly dispensed his comforting blue pills throughout the subprime crisis. This would be an AI Matrix, an electronic version of the plunge protection team that buys stocks at 3:45 PM in volatile markets. We would call it Skynet, and it would render all future crises obsolete — not only financial, but ecological, energy, and military as well. The elite are finalizing the details of such a system now in order to ensure the new world odor.
Excellent.
Godel’s theorem is about specific formal mathematical systems. Financial and economic systems are in no way formal mathematical systems. So trying to apply Godel’s theorem here is at best a very loose metaphor and at worst meaningless.
See the book Godel, Escher, Bach by Douglas R. Hofstadter.
A better take-away from GEB would be the Prisoner’s Dilemma. More applicable to financial markets and easier to get a handle on, since we’ve all faced our own PD at one time or another. Or you could just watch Breaking Bad.
Goedel talked about number theories, which are the most self-consistent systems possible. What is true of number theories must be true for other self-consistent systems of rules. The bottom line is that there are no complete systems of rules. You can’t set up a set of rules (or regulations) and then sit back and let the system run without intervention. In Goedel’s terms there will be true statements that the system cannot generate, i.e. the rules won’t cover every possibility.
Applied to financial systems, Goedel’s theorems only tell us that they can not be complete. This means that they require outside intervention by regulators to keep them functioning the way we want them to. Neither finance nor the economy as a whole can self-regulate.
I said it was an analogy … The obvious difference between markets and a mathematical system is the former is full of self conscious humans. It was intended to stimulate discussion, which here is fascinating.
Even the people who pretend/are supposed to understand the system very often do not have a clue. Example: fractional reserve banking which “temporarily” keeps the banks from collapsing due to a run. What happens when “temporarily” lasts longer than the reserves – what’s the plan. Is there a technocrat out there who can translate it into semi-comprehensible English (or Greek or whatever)?
Once again we wind up with “Bart at the Bank”:
http://www.youtube.com/watch?v=dLtyV1WkFFo&feature=my_favorites&list=FLXtDAJXiwNNR7yV-kFnAviQ
So do we sell Bill’s house – and Fred’s house. And what to we do when we know the money has been stolen and they won’t give it back.
By the way, one of the restatements of Godel’s Theorem is that from within the system, you cannot determine whether the system is going to crash.
What we have is worse than fractional reserve lending. The banks are able to lend so long as they have “creditworthy” borrowers and meet tiny capital requirements. And during the boom (surprise, surprise!) more people become “creditworthy.” But then the bust inevitably comes and suddenly more and more people become “non-creditworthy” as the economy shrinks.
Actually I knew that fractional reserve banking was quaint and obsolete. It just made an easy example.
I am curious what real constraints there are on lending – if any, and I am curious as to how the money in insured demand deposit accounts is used in today’s system.
and I am curious as to how the money in insured demand deposit accounts is used in today’s system. Gerard Pierce
Because of government deposit insurance and the lack of a practical alternative such as a Postal Savings Service, the banking cartel has a government enforced monopoly on the storage and transfer of fiat. That means that the banking system as a whole rarely has to redeem its liabilities with reserves. And even when it has to, the Fed stands by as a lender of last resort to provide new reserves as needed.
But what if there was no government deposit insurance, no legal tender lender of last resort and a risk-free place for the general population to store reserves (eg. a Postal Savings Service)? In that case, a bank would quickly find itself in trouble if it tried to extend credit; its reserves would quickly drain to the risk-free reserve storage service.
It is ridicolous to compere stockmarket to gambling
do you know anything better to colerct riskviling capital to strt up new companies???
Buying stocks is very much a form of gambling. And it is gambling on uncertainty, not on risk. There is no way to quantify the probability of possible outcomes i.e. past performance is no predictor of future performance.
Nanex, Tyler Durden, et al. would snicker if they heard you say that. Stock market capital doesn’t found new companies. Private capital founds new companies. Stock market capital cashes out old companies. If things get really brutal, that IPO could turn out to be the last thing Facebook ever did.
Required reading: _The Folklore of Capitalism_, Thurman Arnold.
Risk? What risk? Steve Keen says the system requires accelerating debt to keep from collapsing. But nothing can accelerate forever. Systemically then, the risk of recurrent collapse is 100%. Why do we tolerate such a system?
The philosopher that comes to my mind reading through the comments
(always excellent) on this article about how to reform the system/can it be made risk free (gross oversimplification, I know) is P.T. Barnum: “There’s a sucker born every minute.” That’s the only thing anyone needs to know about the global financial complex.
Goedel. As if….
Ah, glad to see someone else saying that risk cannot be eliminated, only moved (I’d say “transformed”, which is a bit wider term). I’d add to that – if you believe you eliminated risk, you just don’t know yet what you transformed it into.
Unfortunately, the problem is that we (humans) are horribly risk averse (as a society, not necessarily individually), and this got worse and worse the more complex government structures we created. After all, the goal of more complex gov’t structures is to (seemingly) eliminate some risks – be it risks of not being able to afford healtcare (or anything at all being unemployed), or risk being run over by a drunk driver. The (relatively) recent anti-terrorist laws and measures are the perfect example (the optimal strategy to beat terrorists is to ignore them. Even 9/11 total casualties count (3000 +/-) is less than 1/10th of trafic caused deaths in US during the year – so if you want to save lives, there’s more efficient way of doing so. Much less visible and much more contentious of course, but politicians do what’s right, not what’s visible, right?)
We crave no-risk world, even if unachievable, and will go to great lengths to pretend so.
In a system as fraught with risks and as central to our current mode of living as banking, a claim of “eliminating risks” should be a capital offence.
The Romans have some pretty effective method for dealing with risk: if you are a building builder, you should be living in the houses you have built. If they collapse, you are dead.
Of course this method cannot be applied to any situation, but in essence it involves dissuading risk-shifting with an istantaneous effect.