25 — Market Cycles

Mean Reversion Psychology#

Debated1 min read
◆ The PatternExtremes revert to the mean — in statistics AND in psychology.

Mean reversion psychology — the principle that extremes in both price and sentiment revert toward their long-term average.

Statistical basis: Regression to the mean (Galton, 1886). Extreme observations are followed by less extreme ones — not by "correction" but by probability.

Market application:
• Extreme euphoria → sentiment reverts toward neutral (prices fall)
• Extreme panic → sentiment reverts toward neutral (prices rise)
• The VIX and credit spreads tend to mean-revert; valuation ratios such as P/E do too, but slowly, and where their “mean” lies is debated

Patience is the key: wait for the extreme, then let mean reversion work in your favor.

Worked example

10,000 managers whose results are a little skill and a lot of luck. Last year’s top 10% averaged 3.93; the same managers average 0.83 this year. They kept 21% of their lead, close to the theoretical one in five — the share of the variation that is skill. Nobody lost their touch; most of the lead was luck, and luck does not repeat.

Where it misleads
  • Regression to the mean is about noisy measurements, not a force pulling prices back. A price has no fixed mean to return to.
  • Evidence for mean reversion in stock prices is weak over short horizons and debated over long ones (Poterba & Summers 1988; Fama & French 1988).
  • Waiting for “the extreme” assumes you know where the mean is. If it has moved, what looks extreme is the new normal.
In code
rng = np.random.default_rng(3)
skill = rng.normal(0, 1, 10_000)                 # 10,000 managers: a little skill,
year1 = skill + rng.normal(0, 2, 10_000)         # a lot of luck
year2 = skill + rng.normal(0, 2, 10_000)
top = year1 >= np.quantile(year1, 0.9)           # last year's top 10%
kept = year2[top].mean() / year1[top].mean()     # theory: var(skill) / var(total) = 1 / 5
Python, in your browser — edit the code and run it again

Assumes import numpy as np and import pandas as pd. Each snippet carries its own example numbers; the comments say which are made up.

Sources
  • D. Kahneman & A. Tversky, “On the Psychology of Prediction”, Psychological Review 80(4), 1973
  • J. M. Poterba & L. H. Summers, “Mean Reversion in Stock Prices: Evidence and Implications”, Journal of Financial Economics 22(1), 1988
  • E. F. Fama & K. R. French, “Permanent and Temporary Components of Stock Prices”, Journal of Political Economy 96(2), 1988
  • M. M. Carhart, “On Persistence in Mutual Fund Performance”, Journal of Finance 52(1), 1997

Last reviewed October 2026 · How this site checks its content · Report a mistake

Pattern bridge: The belief that extremes return to average — a psychological trust in the central limit theorem. In ML, normalization layers enforce mean reversion on activations.
Linked from 1
← Previous
Smart Money vs. Dumb Money
Open in the full reader, with the topic sidebar