Who invented the follow-through day?

William O'Neil, founder of Investor's Business Daily, described it in How to Make Money in Stocks based on his study of every market bottom since the 1880s. IBD tracks it daily as part of its market outlook and has refined the thresholds over the years.

Does the follow-through day have to be on day four?

Day four or later. Day four is the earliest; the count goes on as long as the rally attempt's low holds. A follow-through on day ten is valid. Follow-throughs that come later in the count, once the low has held for a while, have on average been more reliable than the earliest ones.

Can a follow-through day happen in a bear market?

Yes, and many do. Bear market rallies produce follow-through days that fail within weeks. The signal marks the point where a new uptrend becomes possible, never where one is assured. That is why the response is to start small and let the market prove it.

What is a distribution day and how does it relate?

A distribution day is the mirror image: a down day on higher volume than the prior day, showing institutions selling. A cluster of them, five or six in a few weeks, is the warning that an uptrend is under pressure. The follow-through day opens an uptrend; distribution days close it.

How big does the up day have to be?

O'Neil used a gain of about one and a quarter percent on a major index, and later work loosened that toward one percent when the market was quiet. The volume condition matters more than the exact size of the gain.

Why does the volume have to rise?

Because the point is to see large investors come back. A gain on light volume can come from an absence of sellers. A gain on heavier volume than the day before means money was put to work, which is the confirmation the signal is after.

When does a rally attempt end before any follow-through?

When the index closes below the low of the first rally day. The count resets and a new attempt starts from the next low.

Is a follow-through day enough to buy on its own?

No. It says the market may have turned. The names you buy still have to pass their own test, and the usual approach is to add exposure in steps rather than all at once.

What happens if the index undercuts its low after a follow-through day?

The signal is dead and the clock starts again. A failed follow-through is common, and it costs little when exposure was added in steps. The traders who get hurt are the ones who treated the signal as certainty and put the whole position on the first green day.

Does the follow-through day work outside the US indexes?

The logic travels, since it only asks whether large money returned on rising volume. The practical problem is data. Volume on some foreign exchanges is less comparable from day to day, so the count is noisier than it is on the American indexes where the rule was developed.

What is a follow-through day, and how do you spot one?

By Daniel, The Philosopher Investor · updated September 18, 2026

A follow-through day is a strong up day, on volume higher than the day before, that arrives on the fourth day or later of a rally attempt off a market low. William O'Neil used it to confirm that a correction had ended. It does not guarantee a new uptrend, but in his research every major bull market began with one.

What is a rally attempt?

It starts when a major index, after a decline, closes up for the day. That is day one. The attempt stays alive as long as the index does not close below the low of day one. Day two and three can be up or down; what matters is that the low holds. If it breaks, the count resets and you wait for the next up day off the new low.

The point of the count is patience. Bear markets are full of one-day rallies that fail. Requiring the low to hold for several days filters out the reflex bounces and leaves the ones where selling has actually dried up.

What makes a day a follow-through?

On day four or later of a live attempt, the index closes up strongly. O'Neil's original threshold was one percent; Investor's Business Daily raised it over the years and currently uses a gain of about 1.25 percent or more, with some flexibility for very volatile markets. The volume on that day must be higher than the previous day's. Not necessarily above average, just higher than yesterday.

The volume condition is the whole idea. A big up day on light volume is short covering. A big up day on rising volume means institutions are buying, and institutions are who start uptrends. Either the S&P 500 or the Nasdaq can deliver the day; it need not be both.

How often does it fail?

Often. Roughly a quarter to a third of follow-through days are followed by a return to the lows within weeks, and in choppy markets the rate is higher. O'Neil's claim was never that a follow-through day guarantees a bull market. It was that no bull market has started without one, which is a different statement. It is a necessary condition, never a sufficient one.

The failures tend to share a shape: the follow-through comes on day four exactly, on modest volume, and the following days see distribution, heavy-volume down days, right away. A follow-through on day seven or later, on clearly higher volume, followed by quiet up days, has a much better record.

What do you do when one appears?

Start buying, in small size, the leading stocks that are breaking out of sound bases. Not the index, and not everything. The follow-through day says the market is likely to support new positions; the stock selection still does the work. If the first few buys work, add. If they fail quickly, the follow-through was probably false and you stop.

This is the part most people skip. The signal is only useful as a permission to start, with the expectation that some starts will be wrong. Traders who buy heavily on the day itself, into the strongest close of the month, are the ones who get hurt when it fails.

How does it compare to other bottom signals?

Breadth thrusts, where the share of stocks advancing over ten days crosses a high threshold, have a similar logic and a similar record: rare, and reliable when they fire. The follow-through day is simpler to track because it needs only the index price and volume. It is also earlier, sometimes by a week or two, which is why it fails more often.

Used together they are stronger. A follow-through day that is confirmed within a couple of weeks by a breadth thrust has been the shape of most durable bottoms. A follow-through day that breadth never confirms is the shape of most failures.

Common questions

Who invented the follow-through day?: William O'Neil, founder of Investor's Business Daily, described it in How to Make Money in Stocks based on his study of every market bottom since the 1880s. IBD tracks it daily as part of its market outlook and has refined the thresholds over the years.
Does the follow-through day have to be on day four?: Day four or later. Day four is the earliest; the count goes on as long as the rally attempt's low holds. A follow-through on day ten is valid. Follow-throughs that come later in the count, once the low has held for a while, have on average been more reliable than the earliest ones.
Can a follow-through day happen in a bear market?: Yes, and many do. Bear market rallies produce follow-through days that fail within weeks. The signal marks the point where a new uptrend becomes possible, never where one is assured. That is why the response is to start small and let the market prove it.
What is a distribution day and how does it relate?: A distribution day is the mirror image: a down day on higher volume than the prior day, showing institutions selling. A cluster of them, five or six in a few weeks, is the warning that an uptrend is under pressure. The follow-through day opens an uptrend; distribution days close it.
How big does the up day have to be?: O'Neil used a gain of about one and a quarter percent on a major index, and later work loosened that toward one percent when the market was quiet. The volume condition matters more than the exact size of the gain.
Why does the volume have to rise?: Because the point is to see large investors come back. A gain on light volume can come from an absence of sellers. A gain on heavier volume than the day before means money was put to work, which is the confirmation the signal is after.
When does a rally attempt end before any follow-through?: When the index closes below the low of the first rally day. The count resets and a new attempt starts from the next low.
Is a follow-through day enough to buy on its own?: No. It says the market may have turned. The names you buy still have to pass their own test, and the usual approach is to add exposure in steps rather than all at once.
What happens if the index undercuts its low after a follow-through day?: The signal is dead and the clock starts again. A failed follow-through is common, and it costs little when exposure was added in steps. The traders who get hurt are the ones who treated the signal as certainty and put the whole position on the first green day.
Does the follow-through day work outside the US indexes?: The logic travels, since it only asks whether large money returned on rising volume. The practical problem is data. Volume on some foreign exchanges is less comparable from day to day, so the count is noisier than it is on the American indexes where the rule was developed.

Primary sources

Nasdaq market activity and volume · www.nasdaq.com
NYSE market data · www.nyse.com
FRED, historical index series · fred.stlouisfed.org

See it on a real name

Distribution days explained → Breadth thermometer, the confirmation → Market extremes, the setup for a bottom →

Try the app free

How the picks are made →