Reading the evidence
Regression to the mean is the tendency for an extreme noisy observation to be followed by one closer to the typical level, even without a causal intervention.
Make it click
The highest score in one noisy round often combines skill with an unusually favourable fluctuation. A repeat keeps the skill but is unlikely to repeat all the luck.
Now put that into markets
Extreme company, fund or indicator results attract attention. Later moderation can be mistaken for reversal caused by whatever happened between measurements.
The effect appears when repeated measures are imperfectly correlated and selection uses an extreme first value. Its size depends on reliability.
Use control groups, repeated observations and pre-selection baselines. Avoid explaining every move back towards normal as a new mechanism.
True structural change can also follow extremes. Regression is a statistical expectation, not a rule that every observation must reverse.
Takeaway
