Harvard Business School professor Niall Ferguson includes the following key concept (among others) in his historian’s view on globalization and the current financial crisis:
Optimized global networks may be vulnerable to crises.
This reminded me of a general optimization problem: we always try to find a special state (optimum) for a special situation under special circumstances. Quite special, isn’t it 
And so the “optimization circuit” starts throwing … optimize, optimize, optimize and asymptotically approach the optimum. In the meanwhile the circumstances change. Due to the fact that optimization is always just a response at one time is has to be too slow. It’s that point where your current model doesn’t have enough space for further optimization to properly answer to the new situation and then … crash … jump down about 100 optimization steps but win a lot of (degrees of) freedom to follow different pathes. And maybe that’s what we often praise as the “chance of a crisis”.
