Zenith

Mean reversion

Why stocks often fall after a big one-day spike

The pattern behind Zenith's whole premise — what the historical record shows, and the large asterisk that goes with it.

Last updated: September 6, 2026

The pattern

Stocks that post very large one-day gains have historically been more likely to close lower the following session than to close higher. The tendency is real, it is measurable, and it is the premise Zenith is built on.

It is also much weaker than people assume, and describing it as a rule rather than a tendency is how students lose simulated portfolios.

Why it happens

A few mechanisms overlap, and they don't all apply to every stock.

The move outruns the news. A piece of genuine information is worth some amount to a company's value. When a stock moves far more than that amount in one session, a portion of the move is momentum rather than valuation, and momentum has nothing holding it up.

Short-term buyers take profits. Much of the volume in a one-day spike comes from traders with a horizon of hours, not years. They sell into the following session.

Small floats exaggerate everything. A company with very few tradeable shares can be moved a long way by a modest amount of buying — and moved back just as easily when that buying stops. This is why the effect is strongest in the smallest companies.

Some spikes have nothing underneath at all. Promotional campaigns and coordinated buying produce charts that look identical to real news on the day, and behave nothing like it afterwards.

What the numbers look like

Zenith groups historical top-gainers by company size and by how unusual the day's volume was, then measures what each group did the next session. Those groupings are what produce the base rate quoted in a thesis and on each ticker page.

The honest summary of that data: the fade is more likely than not, the edge is measured in single-digit percentage points of probability rather than certainty, and it is meaningfully stronger for very small companies on abnormal volume than for large established ones.

A rate near 60% means roughly two in five of these setups go the other way. In a competition scored on a handful of trades, two in five is not a rare event — it is something you should expect to happen to you.

Why a base rate is not a prediction

A base rate describes a population. It says what a large group of similar past situations did on average. It says nothing about which specific stock in front of you belongs to the majority and which to the minority.

This is the distinction that matters most and is missed most often. "Stocks like this closed lower 60% of the time" is a fact about history. "This stock will close lower" is a claim about the future that no base rate supports.

Companies also genuinely change. An earnings beat, a takeover offer or a drug approval can revalue a business permanently, and a stock that spiked for one of those reasons is not the same animal as one that spiked on nothing. Distinguishing the two is what the thesis attempts, and it does not always get it right.

Related