Post-1987 margin rules replaced log-normal pricing assumptions with fat-tailed stable distributions

Evidence

In part, the memory of 1987 has been institutionalized as a shift to margin requirements based explicitly on levels of risk, requirements that are underpinned not by the log-normal distribution of Black-Scholes-Merton pricing but by stable Paretian distributions with infinite variance, in which the probability of extreme events is much higher. — Donald MacKenzie & Yuval Millo, Constructing a Market, Performing Theory, 170

Categorization

• XREF: The lognormal-versus-Paretian contrast maps onto Mandelbrot’s critique of financial orthodoxy — connects to the wider fat-tails literature the reader would already hold. — connects to: Stock price changes are modeled with thin-tailed lognormal distributions, not fat-tailed ones — Donald MacKenzie, An Engine, Not a Camera: How Financial…, 138; Fat-tailed price distributions were documented decades before the 1987 crash — Donald MacKenzie, An Engine, Not a Camera: How Financial…, 1163; Infinite-variance estimators resist both calm and crisis, stabilizing margin requirements — Donald MacKenzie, An Engine, Not a Camera: How Financial…, 940 (+2 more)

Related

• Stock price changes are modeled with thin-tailed lognormal distributions, not fat-tailed ones (similarity 0.83)
• The 1987 crash was a multi-sigma event under the log-normal assumption. (similarity 0.83)
• The bell-curve shape of Gaussian distributions is what banks’ risk models assume (similarity 0.83)
• Fat-tailed price distributions were documented decades before the 1987 crash (similarity 0.82)
• Practical statistical convenience, not empirical fit, drove finance’s retention of the normal distribution (similarity 0.81)
• Infinite-variance estimators resist both calm and crisis, stabilizing margin requirements (similarity 0.80)
• The Levy distribution’s failure to survive time aggregation cooled academic interest in fat tails (similarity 0.80)
• Mainstream statistics training in the 1970s ignored Levy distributions (similarity 0.78)

Broader Connections

• Stock price changes are modeled with thin-tailed lognormal distributions, not fat-tailed ones #contradicts (0.90) — Donald MacKenzie, An Engine, Not a Camera, 138
• Infinite-variance estimators resist both calm and crisis, stabilizing margin requirements #expands (0.85) — Donald MacKenzie, An Engine, Not a Camera, 940
• The Levy distribution’s failure to survive time aggregation cooled academic interest in fat tails #complicates (0.80) — Donald MacKenzie, An Engine, Not a Camera, 509
• Fat-tailed price distributions were documented decades before the 1987 crash #qualifies (0.78) — Donald MacKenzie, An Engine, Not a Camera, 1163
• The bell-curve shape of Gaussian distributions is what banks’ risk models assume #responds-to (0.75)