Many like to yell "systemic collapse" in a crowded system. Most studies of complex systems or networks of interactions realize there are 3 zones in which a system exists.
Economies are systems. In this example I label the 3 zones, stable, interesting and random. Some use the term chaotic, for the 3rd zone, but this or "complexity" is really a gradient between zones 2 and 3. Lets not get started on a debate about Kolmogorov constants.
The energy or system inputs in the model above range from 0 to infinity. A basic example of this model is given by the carbon in #2. The states are Diamond (Stable), interesting (organic) and finally gaseous (CO2) which could be paired on temperature gradient.
Moving out of stable interesting in the middle to the right or to the left is usually considered a "collapse" by those within the system. We still don't know where the US is going to end up, Hyperinflation volatility Latin America style on the right or Japan Land of the Living dead economic tomb on the left. The dollar shoe has yet to drop (pun intended).
Ideally we get back to the middle zone rapidly, but US debt and unfunded government liabilities causes this to be a challenging assumption.
In an economy or any complex system it is impossible to understand the relationships and dynamics in the interesting zone. This impossible understanding of "inter" relationships however doesn't stop the lazy practitioner from modeling co-relationships and thus assuming detailed knowledge of the system dynamics. Co-relationship analysis is useless and flawed in systems, such as economies or portfolios of loans or portfolios for that matter. We are now witnessing these flaws of mis-understanding in the early stages of the US meltdown. Common sense never got in the way of a large vocabulary or a good "working" model.
Don't even bring up options models, which are only really lucrative if you understand their flaws. The myopia of the model builders can be quite a gold mine for the good gambler.
Common sense rules of thumb were overrun by correlations. For example, imagine if the continuum above were leverage for a bank or economy. Leverage runs from 0(no leverage) to infinite leverage. Historically banks would use a rule of thumb such as 12:1 for our cultures fractional banking system. Think of this as the tier 1 capital requirement. A regulator can't understand the dynamics of the system so tells everyone to operate within a safe zone. There are no further assumptions made such as sub zones etc. The entire system has a single regulating variable, namely leverage.
This left the system "economy" safely in the interesting zone where it could evolve and grow maybe not the fastest, but safely. No one really understood all the components or co-relations and if they did they were just deluded economists or some other garden variety charlatans.
The dangers in the system occurred when US banking regulators 5 years ago allowed entities to leverage up 30:1 etc. based on correlation or covariance analysis. Basically a lot of very bright people who had no idea about system dynamics and lacked the humility to utter the words "I don't know" (Jim Crameritis), or god forbid, "I can't model it". When a person tells you they can model a CDO, they may as well tell you they can model a meadow. You should thus leverage your CDO or bank accordingly. Basel 2 is a disaster waiting to happen, for it assumes knowledge of interactions within the system.
This leverage pushed the system to the point where it would inevitibly suffer collapse. The pertinent question today is how much of the system is collapsing, housing, banking etc. The system dynamics are too complicated for me to understand, but months ago on this blog, I made suggestions of more currently "stable" regions for investment consideration.
A person who understands the edge of statistics is Nassim Taleb. Check out his great essay on "the limits of statistics" His pen and mind are far sharper than mine in these areas.