Kill Candle Trading: A Rule for the Candle and a Test of the Short
A checkable rule for the kill candle (body against ATR, volume against its average), the three it marked on the S&P 500 ETF in 2025, and the short tested on 89 to 200 cases per market.

On this page
A kill candle, also called a death candle, is traders' slang for a single large red candle, on heavy volume, that wipes out days or weeks of gains in one session. TradingSim's guide (May 2026) describes it as "a large bearish bar that engulfs and invalidates the prior rally" and says it is most commonly seen on intraday charts of low-float momentum stocks at the end of a squeeze. The name has no standard definition: some traders mean a stock that drops after a pump, others any outsized down day in an index. Kill candle trading means selling short after such a candle, on the idea that the selling has only started. This guide gives the candle a precise rule, shows it on a year of the S&P 500 ETF, and tests the short on every case since 1993 on three US stock ETFs and since 2017 on Bitcoin and Ether.
Kill candles on the S&P 500 ETF, 2025

- Price, daily candles (hollow up, filled down)
- Kill candle: down bar, body at least 1.5 × ATR(14) of the previous bar, volume at least 2 × the previous 20-bar average
- Average volume of the previous 20 bars (solid line, second color)
- Volume of each bar, in millions
A rule for the kill candle
To test the idea, the candle needs a definition anyone can check on a chart. The one used on this page:
- A down candle: the close is below the open.
- A large body: open minus close is at least 1.5 times the average true range (14) of the previous bars, so the size is judged against the market's recent volatility, not in dollars or percent. The long line candlestick pattern uses a similar idea, but compares the body with the average body of the previous candles.
- Heavy volume: the session's volume is at least twice the average of the previous 20 sessions.
The kill candle rule used on this page

Both thresholds use only the bars before the candle, so the signal is known at its close. A looser version, body at least 1 ATR and volume at least 1.5 times the average, gives three to eight times more cases and is tested too.
Kill candles on the S&P 500 ETF in 2025
The chart above covers the 250 sessions of the S&P 500 ETF from January 2 to December 31, 2025, with dividend-adjusted prices. The rule marks three candles. What followed each one, measured from its close, with the return of a short computed as (entry minus exit) divided by entry:
| Candle (day, close) | Short, 5 sessions | Short, 20 sessions |
|---|---|---|
| Apr 4: −5.85%, 497.04 | −5.67% | −11.52% |
| Oct 10: −2.70%, 646.05 | −1.74% | −2.75% |
| Nov 20: −1.52%, 645.57 | −4.73% | −4.61% |
All three shorts were losing after 5 and after 20 sessions: after the April candle the fund closed at 525.23 five sessions later and 554.32 twenty sessions later. Before that rally, it made a lower close of 488.38 on April 8, so a short taken at the April 4 close was up 1.7% on a closing basis two sessions later. After the October candle the fund never closed below 646.05 again in the following 20 sessions; its lowest close was 653.59 on October 16. Three candles prove nothing either way, which is why the next section tests every case on longer histories.
Kill candle shorting tested since 1993
For each kill candle, we took the close and measured the short's return after 5 and 20 sessions. The loose rule is used here to get enough cases; histories run from 1993 for the S&P 500 ETF, 1999 for the Nasdaq 100 ETF, 2000 for the Russell 2000 ETF and 2017 for Bitcoin and Ether, all to 2026. As a comparison, the chart below the table also gives the share of all sessions, kill candle or not, after which the close 20 sessions later was lower: a signal that helps a short should beat it. No costs are counted, and borrowing a stock or an ETF costs money.
| Market (cases) | Mean short, 5 sessions | Mean short, 20 sessions |
|---|---|---|
| S&P 500 ETF (200) | −0.67% | −1.18% |
| Nasdaq 100 ETF (170) | −0.69% | −1.32% |
| Russell 2000 ETF (135) | −0.38% | −1.49% |
| Bitcoin (89) | −0.71% | +0.30% |
| Ether (104) | +1.01% | −1.12% |
Price lower 20 sessions later: after a kill candle and after any day

- After a kill candle
- After any session
At 20 sessions, the last candles of 2026 have no result yet: 199 cases on the S&P 500 ETF and 168 on the Nasdaq 100 ETF.
On the three stock ETFs the result is consistent: the short lost money on average after 5 and after 20 sessions, and the price was lower 20 sessions later less often after a kill candle than on an ordinary day. In about two cases out of three the price was higher 20 sessions later, as often as after any session or more often (69% against 60% on the Russell 2000 ETF): the candle did not mark the start of a decline. On Bitcoin the price was lower after 20 sessions 56% of the time, against 46% on any day, but the mean short gained only 0.30%, and on Ether the 5-day and 20-day results point in opposite directions. Crypto shows no stable edge for the short either.
The strict rule of the chart (1.5 ATR, twice the volume) gives few cases: 48 on the S&P 500 ETF, 31 on the Nasdaq 100 ETF, 17 on the Russell 2000 ETF, 30 on Bitcoin and 37 on Ether. On the three stock ETFs the 20-session short lost on average there too (−1.53%, −1.41% and −1.77%). With samples this size, we do not draw a separate conclusion from them.
Mean result of a short held 20 sessions after a kill candle

- Loose rule
- Strict rule
Kill candles on single stocks
On single stocks, the textbook kill candle hits a small company: a low-float stock that ran up, then collapsed in one candle. Those are the cases where the idea comes from, and where it is hardest to trade. Shares may be hard to borrow and expensive to borrow, and a large fall in one session can trigger the short sale restriction (SSR): under the SEC's Rule 201, a decline of at least 10% in one day restricts the price at which short sales can be executed or displayed for the rest of that day and the following day. We have not tested single small stocks on this page; the ETF test shows only that, on liquid markets, the candle alone was not a reason to sell short.
How to read a kill candle
- Measure it against volatility. A 2% red day is large in a calm market and ordinary in a volatile one; the ATR test handles that.
- Check the volume. Heavy volume shows that many holders sold; it does not say whether they are done. Our trading volume guide covers how to read it.
- Look at what comes next. Traders give more weight to a further close below the candle's low than to the candle itself, and a close back above its midpoint is often read as a sign that selling has dried up; we have not tested either reading. Classical candlestick analysis even has a bullish reversal that starts with two large black candles, the concealing baby swallow, though it is extremely rare.
- Check the context. A kill candle after a long rise, a blow-off top, is not the same as one in the middle of a breakout that failed.
Other candles and setups are covered in candlestick patterns and in trading concepts. A long down candle that opens at its high after a rally is read differently, as a bearish belt hold.
Frequently asked questions
What is a kill candle?
A single large down candle on heavy volume that erases days or weeks of gains. There is no standard definition; on this page it means a down candle whose body is at least 1.5 times the 14-bar average true range, on at least twice the 20-session average volume.
Is a death candle the same as a kill candle?
Yes, both names are used for the same idea: one large bearish candle, usually on heavy volume, that wipes out a rally. TradingSim's guide gives death candle as another name for the kill candle.
Is shorting a kill candle profitable?
Not on the stock ETFs we tested. On the S&P 500 ETF from 1993 to 2026, the kill candles under the loose rule (199 with a result at 20 sessions) were followed by an average short loss of 1.18% after 20 sessions, and the price was lower 20 sessions later only 33% of the time, against 35% on any day.
How do you trade a kill candle?
If you trade it, do not short the close of the candle by reflex: wait for a further close below its low, place the stop above its high and size the trade for a sharp rebound. On the markets we tested, rebounds were more common than follow-through.
What is the difference between a kill candle and a bearish engulfing candle?
A bearish engulfing pattern is defined by two candles, the second body covering the first. A kill candle is a single candle defined by its size and volume, whatever came the day before.
