re: The Obama Bank Shakedown- "It has nothing to do with justice or restitution"

quote:

---

Econometric models will show one some 'risks' but not the 'risk' of using the model.

---

If you're suggesting that institutions follow models blindly without knowing the limitations of them... you're incorrect.

I'd also contend that much of the "mis-evaluated" risk you perceive had more to do with government backstops and an over abundance of people willing to purchase MBS and CDOs with NO evaluation of risk. That's not a modeling failure.

quote:

---

Also, a mathematical model (econometric or any model) by definition is constructed with a finite set of variables.

---

Yes. Which is why I do not understand your advocacy of an unlimited DOF approach. No such thing exists.

quote:

---

The rub is that in the real-world, there are infinite sources of risk, so any model is therefore a crude reduction of economic reality.

---

That's a bit silly. There are innumerable things that can make a plane crash. But that doesn't mean that a wind tunnel can't accurately mimic an airplane's performance.

And while there is an infinite number of risks, they are not all created equally. For exapmle, there is a risk a that a gigantic meteor will crash into earth. But in an economic model... that risk can generally be ignored, as there are far more likely discrete and stochastic events.

Again, I'll pose the question... in lieu of using models with finite variables... what do you suggest be used to evaluate risk?