In an analysis of the global financial system, Duncan Watts says that we should limit the complexity of these sorts of systems because "once everything is connected, problems can spread as easily as solutions".
Traditionally, banks and other financial institutions have succeeded by managing risk, not avoiding it. But as the world has become increasingly connected, their task has become exponentially more difficult. To see why, it's helpful to think about power grids again: engineers can reliably assess the risk that any single power line or generator will fail under some given set of conditions; but once a cascade starts, it's difficult to know what those conditions will be - because they can change suddenly and dramatically depending on what else happens in the system. Correspondingly, in financial systems, risk managers are able to assess their own institutions' exposure, but only on the assumption that the rest of the world obeys certain conditions. In a crisis it is precisely these conditions that change in unpredictable ways.
No one, for example, anticipated that an investment bank as old and prestigious as Lehman Brothers could collapse as suddenly as it did, so nobody had that contingency built into their risk models. And once it did fail, then just as the failure of a single power line increases the stress on other parts of the system, leading to further "knock on" failures, so too did Lehman's unlikely collapse render other previously unlikely failures suddenly much more likely.
This is essentially the same point that Nassim Taleb makes in The Black Swan re: Extremistan and Mediocristan.