The article mentions CMO’s, which were one of the first packaged mortgage instrument outside the Freddie/Fannie secondary market. They weren’t all that liquid, and whatever protection that might have been used on them in the 80’s was even less liquid.
When LTCM went down, it was (again) due to the deadly combination of leverage and illiquid instruments, particularly with exposure to Russian debt markets.
Buffett’s choice of the timeframe is right (when they started going public in the 80’s), but identified only one of the reasons I think are behind this. The article makes an oblique reference to one of the other issues that is responsible here, and I think that is an over-reliance on computer modeling of risk on Wall Street. Before computers (or computing cycles) were cheap, portfolio analysis was based much more in business instinct and common sense, and much less on computer models. Computer models, as we all know, are GIGO - garbage in, garbage out. Or as is more the case now, “selective data in, selective projections out.” The [over] reliance on computer modeling has seduced Wall Street pro’s into thinking that they have risk completely quantified and accounted for, when the truth is far different.
Another problem in the late 80’s was that the computers used at brokerages were overwhelming the exchanges. At some points in ‘87, the exchanges were running a half-hour behind in filling orders. Here’s some additional information in a paper by the Fed:
http://www.federalreserve.gov/Pubs/FEDS/2007/200713/200713pap.pdf
Indeed! There were some grand financial threads here at FR from 2006 on concerning just such quants and their "standard deviations".
The only quant we hear about these days is "quantitative easing". Ah, ha, ha. YIKES!!!
yitbos
Cut a fat hog in the fanny for a long time off that one.