The comparison to the NASDAQ in 2000 may be inappropriate.
Firstly, those 350 “internet stocks” may have comprised a significatn fraction of the market capitalization.
And other not completely “internet stocks” were at very rich valuations: e.g. Cisco, AOL, even Microsoft.
Many of the internet stocks went down 99%.
By the way, the description of the “self organized criticality” of the sandpile system doesn’t exactly mean (as the article states) “It was indeed completely chaotic in its unpredictability.” That’s not really meaningful scientifically.
But the description of the sandpiles is reasonably apropos as to financial markets.
The best “take home message” in nerdspeak is that second moments (variance) can be unbounded, and assurances that anything is “contained” or “will blow up” aren’t to be trusted.
For policy makers it means that taking smaller hits sooner is better. If one were a nonlinear-dynamics informed physicist running the Fed, then a reasonable policy to pursue might be to intentionally have capricious, but small, policy changes at random intervals. (This is “stochastic resonance” or noise-induced stability). Maybe dynamically justified but politically impossible.
The Fed is going of course to “transparency” and “predictability” which is probably the worst thing in this setting since outside participants can predict its ‘control function’ and will themselves attempt to maximize profits and probably thereby go into a state where the catastrophic outcome is more likely.
