Golden Cross & Death Cross
Overlay two moving averages and, most of the time, one line sits above the other. The folklore reads a short line (say the 20-bar EMA) above a long line (say the 50-bar EMA) as 'the recent mood is better than the big picture,' and below as the reverse. The golden cross and death cross capture the moments these two lines swap places.
In numbers: the EMA20 was at 99.5 million won, below the EMA50 at 100 million. After a few up days, the EMA20 climbs to 100.6 million and crosses above the EMA50 (at 100.3 million) — that's the golden cross. The short line sinking back below the long one is the death cross.
The signal is famous because it's intuitive. 'The short-term flow beat the long-term flow' looks like the starting point of a turn upward. The daily 50/200 cross is famous enough that 'death cross forms' makes news headlines.
But it's structurally late. For two averages to swap places, price must already have moved a long way. By the time a golden cross prints, price is often well off the low; by the time a death cross prints, well off the high. Which is why, awkwardly often, a death cross lands right near a short-term bottom.
The other weakness is sideways markets. When price chops without direction, the two lines keep swapping places inside a narrow band, alternating golden and death crosses. Follow every cross in that environment and you bleed a little on each turn.
What the data actually shows
Barobara counts what happened after every past EMA20/EMA50 cross, timeframe by timeframe — golden cross (1h), death cross (1h), and the golden cross (1d). Despite the famous name, the outcome distribution sits around a coin flip. This is a statistic of what happened, not a direction prediction — and even where a combination looks high-win-rate, you have to weigh the target width and the size of wins and losses together. That comparison is in the setup catalog.
Common misconceptions
'Golden cross = buy signal, death cross = sell signal?' It's spread around like a textbook formula, but the measured distribution doesn't lean hard either way. And since crosses print after price has already moved — a lagging signal — selling on a death cross has often meant selling near a short-term bottom.
'The daily golden cross is almost never wrong?' That's closer to memorable successes being retold. Failed crosses don't make the news. Count every instance without exception and the story changes.
FAQ
Q. Should I buy when a golden cross prints?
A golden cross is the past fact that 'the short average moved above the long average,' not a guarantee of upside. The past distribution doesn't stray far from a coin flip, and the print itself arrives after price has already risen.
Q. Is this different from a MACD golden cross?
They're cousins — both built on averages — but the math differs. This one is two EMAs crossing directly; MACD looks for crosses in an indicator derived from the gap between two averages. Each should be verified as its own separate signal.