[biren34]

I think this is just an obvious example of a much larger phenomenon. The trend over the past 40 years has been to put efficiency above all--but that means there is no excess capacity left to ensure robustness. Even "normal" variation in patterns can cause outsized disruptions, especially of the extreme edge of the range only occurs on a timeframe much longer than the optimization cycles used to achieve efficiency.

Efficiency also tends favor exactly one configuration over all others, so any changes that impact the viability of that one path leave the system with few, if any, alternatives--and those tend to have been atrophied.

In each case, the specifics can be quite different, but the general principal is that, in order to avoid such periods of supply chaos and constraint, someone has to be willing to pay a higher price so that more than just the lowest-cost suppliers can survive for when they're needed.

In most situations, the periodic supply crunch may actually be preferable to the consequences of subsidizing the existence of less efficient players. But the costs of efficiency are just as real as the costs of robustness, they just tend to be paid out via a very short period of large excess of profits, as opposed to a long period of slightly better margins.

I'm not sure if the areas under the curve end up being equal, but "efficiency" can certainly be overdone to the point of being inefficient over the full cycle.