Complexity that truly is unpredictable

[koganbot]

A Mountain, Overlooked: How Risk Models Failed Wall St. and Washington

Piece by James G. Rickards in the Washington Post. I know nothing about financial risk models, but assumed that they failed in the current crisis because people were plugging in risk values for subprime mortgages that were based on fantasy without really knowing what the actual risk was. But this article is arguing that even if analysists had known the precise risk of the individual mortage-based securities, the models themselves were flawed and that there is no way to create a model that will accurately predict the risk to the system, except you can be sure that the overall risk is far greater than the sum of the individual risks.

The article is too short, but here are a couple of potent quotes:

Lurking behind the models, however, was a colossal conceptual error: the belief that risk is randomly distributed and that each event has no bearing on the next event in a sequence.

...

Beyond chaos lies complexity that truly is unpredictable and cannot be modeled with even the most powerful computers. Capital markets are an example of such complex dynamic systems.

Comments

[koganbot]

Duncan J. Watts would surely be relevant here (I wrote about him in a piece entitled "The Social Butterfly Effect"); a section in his book, Six Degrees, dealt with bubbles and panics, though I don't recall that section directly addressing risk. I wonder if what might be a factor here is that small events can have outsized effects, and there's no way to predict which event will have the effect, and therefore what the direction of the effect will be.

[koganbot]

And Hyman Minsky is getting mentioned on the blogs, his theory that stability evolves into crisis:

"A fundamental characteristic of our economy," Minsky wrote in 1974, "is that the financial system swings between robustness and fragility, and these swings are an integral part of the process that generates business cycles."

[petronia.livejournal.com]

Or Terry Pratchett: "The million-in-one chance happens nine times out of ten."

[dubdobdee.livejournal.com]

newberry explains why you need to think about apocalypse in order to understand what money is

(interestingly, he is less apocalyptic than usual here -- and than the markets currently)

(also he spells it "apocolyptic") (but i think this essay is clearer than he usually is -- perhaps the arrival of "his" emergency has brought him zen calm)

usual caveats apply

[koganbot]

newberry explains why you need to think about apocalypse in order to understand what money is

Well, the way he connects money to value via his apocalypse is:

What do you get if everything else goes bad? What is it that is the ultimate good a society produces? The answers must be political and social, because the market can only price what people value, it cannot tell them what they should value.

Problem here is that, while I agree with the answer ("must be political and social"), I reject the question. Which is to say that I see absolutely no reason why there must be "an ultimate good that society produces," any more than there must be an ultimate set of values that everybody agrees on or ultimate congruence on what people designate as good and bad. In fact, my ideas about Superwords etc. are that partial disagreement is built into the functioning of words like "good" and "bad" and "value." And I'd say that partial disagreement about what to value is built into the exchange of goods and services. And of course social and political life is the means by which we negotiate/enforce our agreements and disagreements and the balance between them. And the "answer" as to what to value is social and political because the answer is always in partial flux. And what I've just said borders on the platitudinous, since all that "social and political" means really is "the lives we lead and the relations we have," so all I'm really saying is that the social relations evolve from social relations. I don't think that when Newberry says "social and political" he means "but excluding the markets," though maybe that's exactly what he means. That would seem pretty dumb, since markets are one way of "discussing" what is of value.

Also, maybe I'm misunderstanding his apocalypse idea, but I don't see that "everything else goes bad" is necessarily instructive, though I think it's a useful thought experiment. That is, I don't think that Somalia in breakdown is a better indicator of what East Africans value than Uganda trying to pull itself together is; nor do I think that civilization in breakdown is somehow a more basic or primal state than civilization when it's relatively stable.

A self-conscious, articulate discussion of - and argument over - what to value is probably a Good Thing. But some such discussion/argument (whether or not it's self-conscious and articulate) is always underway.

Reading Plato and Augustine for my philosophy course, I decided (which of course is a notion I went into it with) that by trying to make the ideas/forms (in Plato's case) and God (in Augustine's case) the source and cause of everything, they made them vacuous, but that this was hardly all that was going on in their philosophy and their use of those notions. (Cf. in Why Music Sucks #7 where I say that "He went looking for El Dorado and it wasn't there" is not a good account of what Sir Walter Raleigh actually did.) So I find "the ultimate good a society produces" just as vacuous, but I can still think that Newberry might put the bad idea to interesting use. Or not.

Don't see where putting some other commodity in place of oil and electricity necessarily changes the ballgame. But we'll see. The buzzwords of the day are "information" and "environmental technology." In any event, the piece ends up as the usual fakeout, since the obvious next step is to say what the new assets are, and Newberry doesn't, not at all.

[koganbot]

His account of the current mess still has way too many ellipses. I still don't understand what he's saying, though my concentration today probably isn't what it should be either.

And maybe this is a misreading, but he seems to be implying that Bernanke has a hidden agenda that he - Newberry - can discern though the rest of us can't. In any event, whether Newberry is implying this or not, I doubt that it's true.

Have Krugman, DeLong, Galbraith fils, or Roubini ever commented on Newberry? I might understand what he's saying better if I understood what other people take him to be saying.

[dubdobdee.livejournal.com]

john robb also very animated -- usually he talks about warfare rather than economics

(not especially the stuff about unstable systems)

[dubdobdee.livejournal.com]

(i mean NOTE especially)

[koganbot]

He's writing almost in note form. His idea seems to be that somewhat unstable systems are actually more stable than totally stable systems (that can suddenly go down), but that the somewhat unstable systems also need continual counterbalancing or else they suddenly go down as well. (Or is that an entire misreading. My first clause was actually a guess.)

But maybe an airplane isn't a good metaphor for a society.

Here are some assumptions that underly all current activities related to the global financial crises:

--Optimistic long term outcomes. Things are going to get better. Pollyanna mindsets.

--The system, as it is currently configured, is the best possible or the only possible system. Faith in ideology.

--Nation-states, singularly or collectively, are bigger and stronger than the financial/market system. The US is the guardian of this system. 20th Century legacy thinking.

Really? Are these the only three assumptions?

[koganbot]

But do you have any opinion on the actual piece I actually linked?

[dubdobdee.livejournal.com]

(yes but too mazed and busy to write it up yet)

[koganbot]

E.g., as I read down further I see that Robb actually does talk about information cascades and butterfly effects (which, when the effects are big and unexpected, he calls "black swans").