What Is a Bear Market?
Bear market is the conventional label for a sustained decline, usually defined as a fall of twenty percent or more from a recent peak. The twenty percent figure is a convention rather than a measurement, and the arithmetic of recovering from one is asymmetric.
A bear market is a label applied to a decline after the decline has reached a size somebody once decided was significant. Understanding how arbitrary that threshold is makes the term far more useful.
How it works
The conventional threshold is a 20% fall from a recent high. Below that it is called a correction; past it, a bear market.
There is nothing behind the number. Twenty percent was not derived from any study of market behaviour — it is a round figure that became standard through repetition in financial journalism.
Which means the label tells you about magnitude and nothing about cause, duration or what comes next. A 19% fall and a 21% fall are treated as different categories and are the same event.
The recovery arithmetic
Losses and gains are not symmetric, because the base changes. Falling 20% leaves 80; getting back to 100 from 80 requires a gain of 25%, not 20%.
And it steepens. Down 50% needs 100%. Down 70% needs 233%. The deeper the hole, the more disproportionate the climb out, which is the same arithmetic risk of ruin works through.
This is why avoiding large losses matters more than capturing large gains. Not as a temperament preference — as arithmetic. A portfolio that never falls 50% never needs to double to stand still.
A worked example
Take this site’s shared series. Ninety-five percent of bars sat below a prior peak. The deepest drawdown ran 3.76%, the longest wait for a new high was 73 bars, and the whole stretch finished +3.61%.
That first figure is the one worth sitting with. For 95% of the time, the series was below a level it had already reached. Being underwater was not the exception — it was the condition.
So the emotional experience of holding and the reported result point opposite ways. The stretch made money and almost all of it felt like losing.
And the 73-bar wait is the part that breaks plans. A decline recovered quickly is survivable on resolve alone; one that takes 73 bars outlasts most people’s conviction, which is why positions get abandoned at the bottom rather than at the top.
You cannot time either end
The start is invisible. Every bear market began as an ordinary decline that did not stop, and every ordinary decline that did stop looked identical at the time.
The bottom is only visible afterwards. Direction runs on this series average 2.01 bars, so several consecutive up bars — which is what a bottom feels like — is a routine event that happens constantly during declines.
Which is what makes a bear market rally so convincing, and why it gets its own term.
The original data
On this site’s shared series: 95% of bars below a prior peak, deepest drawdown 3.76%, longest recovery 73 bars, stretch finished +3.61%. Direction runs average 2.01 bars with a longest of 11. A round trip costs 0.0098, about 2% of the median bar range of 0.493.
Those numbers describe a positive period that felt overwhelmingly negative. Any plan built on the assumption that declines are brief interruptions is planning for 5% of the observed time.
Where the twenty percent came from
Nobody is quite sure, which is itself informative. The threshold appears in financial journalism through the twentieth century without any founding study behind it, and it survives because it is round, memorable and produces a satisfying number of events per decade.
A different threshold would produce a different history. At 15% there are far more bear markets; at 30% there are very few. The events did not change — the categorisation did, and every claim about how often bear markets occur is really a claim about where somebody drew a line.
This matters when a statistic is quoted at you. “The average bear market lasts X months” is computed over a set of episodes selected by the 20% rule, so the average is a property of the rule as much as of markets.
And it explains why the label feels arbitrary in practice. A portfolio down 19% and one down 21% are in the same situation, and only one of them is in a bear market according to the convention.
When it fails
The characteristic failure is selling once the label is applied. The 20% threshold is crossed, the media begins using the term, and the decision to reduce is made at a point defined by an arbitrary round number rather than by anything about the position. The selling happens after the fall that triggered the label, into the conditions where the spread is widest, and the buyer is somebody who did not organise their decisions around a journalistic convention.
A second failure is waiting for an official end. There is no announcement, and by the time the recovery is obvious it is priced.
A third is assuming bear markets are brief. Some are; the 73-bar recovery on this series is the ordinary case rather than the extreme one.
A fourth is judging a strategy by its behaviour in one. A single decline is one observation, and the sequence matters more than the average.
And a fifth is confusing a bear market with a permanent loss. A decline in a diversified holding is recoverable; a decline in one company may not be, which is the distinction business risk covers.
Related
Bull market covers the mirror label and the same identification problem. Risk of ruin covers the recovery arithmetic in full. And volatility covers the bar sizes a decline is built from.
The thing nobody tells you about a bear market is that you cannot identify the start of one while it is happening, and you cannot identify the end either. Both are named afterwards, which makes the label useful for describing history and almost useless for deciding anything today.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.