Golden Cross and Death Cross Explained
The golden cross and the death cross explained: the 50-day and 200-day averages, what each cross means and why it comes late, with our tests on S&P 500 stocks.
The Test Bench › Moving averages
How SMAs, EMAs and crossover rules work, and what our historical tests found on S&P 500 stocks.
A moving average smooths the price and follows it with a delay. Traders use it to judge the trend, and to build rules that buy as the price strengthens and sell as it weakens. Trading rules compare two lines: the price against an average, or a fast average against a slow one. In a common crossover rule, the strategy buys when the faster line crosses above the slower one and sells when it crosses back below. The length of the averages decides how early, and how often, that happens; in a market moving sideways, the lines can cross back and forth within days.
This page covers:
A simple moving average (SMA) adds the last n closes and divides by n. An exponential moving average (EMA) starts from yesterday’s EMA and moves it towards today’s close by a fixed share, 2 ÷ (n + 1). Six invented closes show both, with n = 5:
| Value | How | |
|---|---|---|
| Closes | 20, 21, 22, 21, 23, 25 | Six days, oldest first |
| 5-day SMA on day 5 | 21.40 | (20 + 21 + 22 + 21 + 23) ÷ 5 |
| 5-day SMA on day 6 | 22.40 | (21 + 22 + 21 + 23 + 25) ÷ 5 |
| 5-day EMA on day 6 | 22.60 | 21.40 + 0.33 × (25 − 21.40), starting from the day-5 SMA |
On the jump to 25, the EMA moved 1.20 and the SMA 1.00: the new close gets a third of the weight in this update (k = 2 ÷ 6 ≈ 0.33), against a fifth in the SMA. This example starts the EMA from the day-5 SMA; platforms start it in different ways (GU Analyser starts it from the first close of the history); the starting value’s influence decays with each new close, and how long the early values differ depends on the platform and the period.
Three common crossover rules on daily charts:
| Signal | Fast line | Slow line | Common use |
|---|---|---|---|
| Golden cross / death cross | 50-day SMA | 200-day SMA | Long-term trend, months to years |
| The 200-day average | The close | 200-day SMA | Long-term trend filter; Faber’s timing rule |
| 9/21 EMA crossover | 9-day EMA | 21-day EMA | Short-term swings, weeks |
Each signal can be read two ways. As an event, it asks what usually followed the day it fired. As a rule, it holds the stock while the fast line is above the slow one and sells when it drops below, which is how a backtest uses it.
Moving-average rules are trend following: they buy after a rise has started and expect it to continue. Mean reversion does the opposite and buys after a fall. Breakouts to new highs, as in our Donchian study, are another trend-following family.
| Trend following (moving averages) | Mean reversion | |
|---|---|---|
| Typical entry | After a rise has started | After a fall |
| Typical hold | Often weeks to years | Often days |
| Common return pattern | Can rely on a few large gains | Can rely on frequent smaller gains |
| Main risk | Repeated losses in choppy markets | A fall that keeps going instead of rebounding |
| Main attraction | Staying in long trends and stepping aside for part of long declines | Catching rebounds after sharp falls |
The golden cross and the death cross explained: the 50-day and 200-day averages, what each cross means and why it comes late, with our tests on S&P 500 stocks.
The 200-day moving average explained: crossings, whipsaws and Faber’s 10-month timing rule, with our tests on S&P 500 stocks and SPY.
The exponential moving average and the 9/21 EMA crossover explained, EMA against SMA, with our tests on S&P 500 stocks.
How to read these results. An event study compares the days after a signal with ordinary days of the same stocks; a backtest runs one complete rule with costs and reports a median total return per stock over the whole period; the two answer different questions. “Clear” is our own rule: the 95% interval excluded zero under every test setting we ran, which is not proof that a gap will last. All tests use today’s S&P 500 members projected back, which can flatter strategies that hold through deep falls, such as buying and holding (see survivorship bias in backtesting), and price returns without dividends.
Golden and death crosses on 483 S&P 500 stocks, 2006 to September 2026: no clear difference from ordinary days over the next 100 trading days in any of the four signal tests.
Price crossings of the 200-day average on 483 S&P 500 stocks, 2006 to September 2026: no clear difference from ordinary days over the next 100 trading days in either of the two signal tests.
The 9/21 EMA crossover on 483 S&P 500 stocks, 2006 to September 2026: no clear difference from ordinary days over the next 20 trading days in any of the four signal tests.
The golden cross, the 200-day average and the 9/21 EMA as trading rules on 483 S&P 500 stocks, 2006 to September 2026, with next-open fills and costs, against buying and holding.
For the opposite approach, see our mean reversion backtest.
The average of the last n closing prices, recalculated every day, so it moves with the price. It smooths out day-to-day noise and shows the direction of the trend. The 20-, 50- and 200-day averages are among the most commonly used on daily charts.
A simple moving average (SMA) gives each of the last n closes the same weight. An exponential moving average (EMA) gives the newest closes more weight, so it turns sooner after the price changes direction.
A rule that buys when a faster moving average crosses above a slower one and sells when it crosses back below. The same idea is used with the price itself as the fast line, as in a close above the 200-day average.
There is no single best length. Longer averages give fewer, later signals and suit long-term trends; shorter ones react faster and give more signals that reverse soon after. The 50/200 pair, the 200-day line and the 9/21 EMA pair are common examples.
In our event studies, none of the three signals was followed by a clear difference from ordinary days. That does not prove every moving-average rule fails: these signals, horizons and stocks showed no clear difference under our rule. In our backtest of the three rules, buying and holding had the higher median total return per stock (+361%, against +137% for the best rule), and the rules’ deepest falls were somewhat smaller; on SPY the 200-day rule cut the deepest fall to −22% against −56%.
Buying what is rising and selling what is falling, on the expectation that a move continues for a while. Moving-average rules, breakouts to new highs and momentum ranking are the common forms.
A signal that reverses soon after it fires: the price crosses above its average, then back below within days. Whipsaws are a major problem for fast moving-average rules: each one can be a losing round trip, and each pays trading costs.
An average of past prices can only change after the prices have changed, so a crossover always comes after the move has started. The longer the average, the longer the delay.
Yes. In GU Analyser you add the averages, set the crossover as the buy and sell rule on a visual canvas and run a backtest or an event study on daily bars.
Want to test your own moving average rule?
Choose the averages and the crossover, then measure what followed the signal across your stocks, or backtest the rule with a cost on each trade.
Create a free account and build your own strategies → or read how to backtest a strategy without code →The methods: how a backtest is built · how an event study works