Bear Market

A bear market is a period of sustained decline in the prices of an index or a market, conventionally described as beginning once prices have fallen roughly 20 percent from a recent high and continuing until they have recovered. The threshold is not something the page has invented or that only commentators use: the Securities and Exchange Commission's own investor glossary describes a bear market as a time when stock prices are declining and sentiment is pessimistic, occurring "generally" when a broad market index falls by 20 percent or more over at least a two-month period. That is investor education rather than a rule. The phrase appears nowhere in the Code of Federal Regulations, nothing turns on it legally, and no body publishes dates for the episodes it names.

The convention deserves a careful correction rather than a debunking. It is genuinely useful, in the same way that "a hard frost" is useful, because it separates the ordinary noise of markets from the episodes people remember. It is simply not a measurement anyone is obliged to make in a particular way, and four choices sit inside every use of it: which index is being measured, which prior high counts as the peak, whether the prices are closing prices or intraday highs and lows, and how long the decline must persist to count. The SEC's description answers the last of those with two months and leaves the rest open, and plenty of accounts that quote the 20 percent figure drop the duration condition entirely. Two commentators using different answers can date the same episode differently, or disagree about whether it qualified at all.

The most important property of the label is that it is retrospective. A decline is only a bear market once it has already reached the threshold, and it has only ended once prices have already climbed back, so both dates are known too late to be acted on. That is the same structural problem the business cycle has, where NBER's dating committee announces peaks and troughs many months after the turning points they name, and for the same reason: certainty and timeliness trade against each other.

It is worth separating the label from the economy, because the two get conflated constantly. A bear market is a statement about the prices of traded securities. A recession is a statement about economic activity, dated in the United States by the Business Cycle Dating Committee of the National Bureau of Economic Research using monthly indicators such as employment and personal income. The two can coincide, and they can also arrive apart, because share prices reflect expectations about the future while economic indicators measure what has already been produced and earned. Reading a falling market as confirmation of a recession, or a rising one as proof that a recession is over, is reading one measurement as though it were the other.

The vocabulary of declines has more than one word in it, and they differ only by size. A fall of roughly 10 percent from a peak is conventionally called a correction, and a fall of roughly 20 percent a bear market. The two sit on different footings, which is worth knowing. The larger figure is at least written down by a regulator, in the SEC's investor glossary; the smaller one is not, and the SEC's glossary carries no entry for a correction at all. Both are applied after the fact, neither is binding on anyone, and both carry the same ambiguity about index and price basis. The practical point is that the smaller category is far more common than the larger one, so most declines that get named do not go on to become the thing people fear. The precise historical proportions are quoted widely and inconsistently, and the figures in circulation are calculated on different indices over different periods, so this page states no ratio.

What the label cannot do is the thing readers most want from it. Knowing that a market is in a bear market says nothing about how much further it will fall or when it will turn, because the threshold is a property of what has already happened. Anyone offering a forecast built on the label is dressing up a guess. What is knowable, and what actually determines the damage a household takes, is whether the household will be forced to sell into the decline. That depends on the emergency reserve, on job security, and on how much of the portfolio is earmarked for spending in the next few years, none of which are facts about the market.

For someone already drawing on a portfolio, the exposure is more specific than the label suggests. What harms a retiree is not the decline itself but having to sell shares into it to fund spending, which permanently removes those shares from the recovery. That is sequence of returns risk, and it is why the same market decline is a headline for one household and a plan-altering event for another.

The calculation is simple, which is part of why the convention spread. Take an index level at a prior peak, take the level now, and express the fall as a percentage of the peak. Once that figure reaches about 20 percent, the period is conventionally described as a bear market, and it is conventionally described as having ended once the index has recovered.

A hypothetical illustration of why the choices matter. Suppose an index peaks at 5,000 on a closing basis and later closes at 4,050. The decline is 950 points, which is 19 percent of the peak, so on closing prices no bear market has occurred. If on that same day the index traded as low as 3,950 at some point during the session, the intraday decline is 1,050 points, or 21 percent of the peak, and on an intraday basis it has. Nothing about the market changed between those two descriptions. Only the measurement convention did, which is why two sources can report the same week differently. All figures are illustrative.

A second choice is which index. A broad US index, a large-company index, a small-company index and an international index will not cross the threshold on the same day or necessarily at all in the same episode, so "the market is in a bear market" is always shorthand for a specific measurement someone made. And the third choice is the peak. Measuring from the highest point ever reached gives a different answer from measuring from a recent local high, and neither is wrong, because there is no rule to be wrong about.

What the term is useful for

• It separates the routine movement of prices from the episodes large enough to change how people behave, which is a real distinction worth naming.
• The arithmetic is transparent and anyone can check it against published index levels.
• It gives a household a concrete scenario to test a plan against before the scenario arrives.
• Because it is symmetric with the way people talk about rising markets, it keeps the discussion of falls in proportion rather than treating each one as unprecedented.

Where it misleads

• The threshold is a convention, so the same episode can be described differently depending on the index, the peak and the price basis chosen.
• It is only ever applied in hindsight, so it cannot serve as a signal to act on.
• It says nothing about what comes next. The label describes what has already happened and carries no information about depth or duration from here.
• It is routinely read as a statement about the economy, which it is not.
• Treating the label as the risk misdirects attention. What determines the damage is whether a household has to sell into the fall, which is a fact about the household.