A golden cross is a chart pattern. It forms when a faster, short-term moving average crosses above a slower, long-term one. The most common setting is the 50-day moving average crossing above the 200-day. Traders read it as a sign that a downtrend may have turned into an uptrend. It is a lagging signal: it confirms a move that has already happened and does not predict the next one. StockCharts' ChartSchool says that, taken alone, the pattern "can't forecast bullishness with a reliable degree of accuracy". Its opposite, the death cross, forms when the 50-day falls below the 200-day.
How a golden cross forms
A simple moving average (SMA) is the average closing price over a set number of periods. It is recalculated with each new close. The 50-day SMA follows recent prices closely. The 200-day SMA covers about a trading year of data in stock markets and moves slowly. When the faster line rises through the slower one, recent prices have been running above the long-run average for long enough to pull the short average past it.
The pattern usually builds in three phases:
- Downtrend. The 50-day SMA sits below the 200-day SMA.
- Crossover. The 50-day SMA rises through the 200-day SMA. This is the golden cross.
- Follow-through. The 50-day SMA stays above the 200-day SMA. Some traders then watch both averages as possible support on pullbacks.
The 50/200 pair is a convention, not a rule. ChartSchool notes that traders vary the lengths. Shorter pairs react sooner and give more false crossings. The pattern can be read on weekly or hourly charts as well, and the same trade-off applies.
How traders use it
ChartSchool lists three common uses:
- Confirmation that a reversal is under way, for someone already watching the trend.
- An entry signal for long positions.
- A stop-loss reference. The 50-day or 200-day average serves as the line where the bullish reading would be wrong.
ChartSchool also warns that conditions around the cross matter. If buying activity and volume fade after a golden cross, the momentum behind it may be weakening.
Type 1 and Type 2 crossings
Technical analyst Tom McClellan sorts crossings by where price sits when the averages meet. He wrote about this in February 2023, after the S&P 500 completed a golden cross on 2 February 2023:
- Type 1. Price is already stretched far from the level where the averages crossed. McClellan says such a moment "most often marks a reversal point", which can be temporary or larger. His example is the March 2022 death cross: prices turned at the crossing and rallied 11% before falling again.
- Type 2. Price pulls back toward the crossover level after the cross. What usually follows, in his reading, is a resumption of the trend that came before the pullback.
This is one analyst's framework, offered as a way of reading crossings rather than as a tested trading rule.
A worked example: the S&P 500 in 2020
- March 2020. The S&P 500 bottomed on 23 March during the pandemic sell-off. A death cross followed on 30 March, about a week after the low, according to Dow Jones Market Data as reported by ETF Trends. Anyone who sold on that signal sold near the bottom.
- July 2020. On 8 July 2020, SentimenTrader noted that the 50-day average was about to cross above the 200-day for the first time in months. Before the turn, the 50-day had been more than 9% below the 200-day. SentimenTrader found 8 earlier signals since 1950 with a gap that wide. In the year after those signals, the index's maximum loss was 5.6% and its average maximum gain was more than 22%. The same analysis called the golden cross in general "only a moderately useful technical signal".
Both crosses came weeks after the actual turning point. The death cross was the more misleading of the two.
What is this 50/200-day cross telling you?
Answer a few questions about the crossing on your chart to see how the article's sources would read it.
Has price been moving sideways within a range, with the 50-day and 200-day SMAs crossing back and forth?
- Yes, it has been ranging → A sideways-market crossing
- No, there was a clear trend before the cross → next question
Since the cross, have buying activity and volume faded?
- Yes, they have faded → Momentum may be weakening
- No, or too early to tell → next question
Where is price now compared with the level where the two averages crossed?
- Already stretched far from that level → McClellan's Type 1 crossing
- Pulled back toward that level after the cross → McClellan's Type 2 crossing
- Neither yet → A lagging confirmation, not a forecast
When price moves within a range, the 50-day and 200-day averages can cross back and forth. Each crossing counts as a signal, but none carries much information.
ChartSchool warns that if buying activity and volume fade after a golden cross, the momentum behind it may be weakening. The conditions around the cross matter as much as the cross itself.
Price is already stretched far from the crossover level. Tom McClellan says such a moment 'most often marks a reversal point', which can be temporary or larger. In March 2022, prices turned at a death cross and rallied 11% before falling again. This is one analyst's framework; no published backtest of the two types was found.
Price has pulled back toward the crossover level. In McClellan's reading, what usually follows is a resumption of the trend that came before the pullback. This is one analyst's framework, not tested in published work.
The cross confirms a move that has already started. ChartSchool says that, taken alone, the pattern 'can't forecast bullishness with a reliable degree of accuracy'. In 2020, both S&P 500 crosses came weeks after the actual turning point.
This sorts a crossing by the conditions the article describes; it does not predict price, is not trading advice, and rests on research into US stock indices, not crypto, where 24/7 trading and higher volatility produce more false crossings.
What the research says
There is little rigorous published work on the golden cross itself. The academic work tests broader moving-average rules:
- **Brock, Lakonishok and LeBaron (1992), Journal of Finance.** The authors tested simple moving-average and trading-range-break rules on the Dow Jones Industrial Average from 1897 to 1986 and reported "strong support for the technical strategies". Returns after buy signals were higher and less volatile than returns after sell signals. Returns after sell signals were negative.
- **Sullivan, Timmermann and White (1999), Journal of Finance.** The authors took the 26 rules of the 1992 study, widened the set to 7,846 rules and adjusted for data-snooping, the risk that a rule looks good only because many rules were tried. The best rule held up over the original 100-year sample of daily Dow data, but its edge did not carry over to the following ten years, 1987 to 1996.
Neither study is specific to crypto. Both cover US equity indices.
Where the golden cross fails
- Lag. Two averages of past prices can only cross after the move has started. Signals tend to come late.
- Sideways markets. When price moves within a range, the 50-day and 200-day averages can cross back and forth. Each crossing counts as a signal, but none carries much information.
- Crowded readings. A cross on a widely watched index draws headlines. That attention does not show the signal works, and the 2020 death cross shows how badly a well-publicised cross can time a market.
- Crypto specifics. Crypto trades 24/7, so a "50-day" average covers 50 calendar days, not the roughly 10 trading weeks it covers in stocks. Higher volatility also means more crossings. Neither study above tested crypto assets.
Golden cross vs death cross
| Golden cross | Death cross | |
|---|---|---|
| What crosses | 50-day SMA rises above 200-day SMA | 50-day SMA falls below 200-day SMA |
| Usual reading | Possible shift to an uptrend | Possible shift to a downtrend |
| Shared weakness | Lags the turn; unreliable in sideways markets | Lags the turn; unreliable in sideways markets |
| 2020 S&P 500 case | July 2020, months after the March low | 30 March 2020, a week after the low |
Frequently asked questions
Is a golden cross a buy signal?
Many traders treat it as one. The analysts quoted here treat it as a confirmation that needs context: volume, what price does after the cross, and the wider market trend. It is not a forecast on its own.
Does the golden cross work in crypto?
The arithmetic is the same on any price series. The research covers US stock indices, not crypto. Crypto's round-the-clock trading and higher volatility tend to produce more crossings, and so more false ones.
What timeframe is standard?
The 50-day and 200-day SMAs on a daily chart. Other pairs and timeframes exist. Faster settings react earlier and give more false signals.
