The Golden Cross and the Death Cross
Two named events describing the 50-day moving average crossing the 200-day. They are widely reported, heavily lagging, and describe a trend that has already turned.
MadStockAlerts Research · Updated August 28, 2026
What to take away
- A golden cross is the 50-day crossing above the 200-day; a death cross is the reverse.
- Both are lagging by construction and confirm a change that already happened.
- They generate frequent false readings when price is ranging.
- Their prominence owes much to how often they are reported.
- The slope of the longer average carries more information than the crossing itself.
MAD Academy Training Video · 0:43
The Cross Prints After the Move
Why the 50/200 cross is a confirmation event, what the lag actually measures, and the one use that survives it.
This lesson is part of a Stock Alerts + Tools plan.
The definition
When the 50-day simple moving average rises through the 200-day, the average of the last fifty closes has moved above the average of the last two hundred. That is the entire content of the event: the medium-term picture has become stronger than the long-term one.
Stated that plainly, it is obviously a statement about the past. Two averages of historical closes crossed; nothing about the future has been asserted, and nothing about it can be.
The lag
Both inputs are averages of past data, so the crossing cannot happen until price has already moved enough, for long enough, to drag one average through the other. By the time a golden cross prints, a substantial advance has usually already occurred, frequently twenty percent or more.
Scroll the chart sideways to see all of it.
- Faster average
- Slower average
This is not a criticism so much as a description. These are confirmation events, not early ones, and reading them as anything else misunderstands what an average of two hundred days can do.
There is a defensible use for a confirmation event: it can be a condition for a strategy rather than a trigger for a trade. A rule that only takes long setups while the 50 is above the 200 is using the cross as a regime filter, which is what a slow signal is actually good for.
Whipsaws
When price moves sideways for an extended period the two averages converge and can cross repeatedly. Each crossing is reported as a named event and none of them marks a trend change, because there is no trend.
A partial filter is the slope of the 200-day itself. A cross that happens while the long average is flat is happening inside a range; a cross that happens while it is already turning up is describing something with more behind it.
Why they are watched anyway
They are unambiguous and easy to communicate, which is why financial media report them so consistently. There is no judgement involved: two lines either crossed or they did not, and everyone computing them gets the same answer.
That coverage itself draws attention to the security, and attention is flow. Part of whatever effect these events have is the effect of being reported, which is a circular mechanism but not an imaginary one.
What the record actually shows
The 50-day crossing the 200-day is among the most widely reported technical events, and it is worth being precise about what studies of it have found rather than repeating either the bullish or the dismissive version.
- The signal is heavily lagging by construction: both averages are computed from past prices, so the cross occurs well after the move that caused it.
- In sustained trends it stays on the correct side for long periods, which is where its reputation comes from.
- In range-bound conditions it produces repeated false signals, and ranges are common.
- Tested mechanically on indices, the results are broadly similar to holding throughout, with lower volatility and lower returns.
- The comparison depends enormously on the period chosen, which is the usual difficulty with any small number of signals.
The fourth point is the honest summary for index-level use: the crossover behaves less like a forecast and more like a volatility filter, keeping capital out of some large declines at the cost of missing early recoveries.
It is also among the most reflexively watched signals in the market, which raises a reflexivity question. Some of any observed effect may be participants acting on the cross rather than the cross describing anything, and the two are difficult to separate.
Why the two averages, and what changing them does
There is nothing special about 50 and 200. They are round numbers approximating a business quarter and a business year, adopted by convention and reinforced by every platform defaulting to them.
| Pair | Effect |
|---|---|
| 20 and 50 | Many more crosses, much earlier, and far more false signals |
| 50 and 200 | The convention. Slow, and stays on one side through a trend |
| 100 and 300 | Slower still. Almost no signals, and each is very late |
| Exponential rather than simple | Earlier crosses, and more of them |
Every row is the same trade-off in a different position: earlier signals mean more of them are wrong, and later ones mean fewer signals that capture less of each move. No pair escapes it, and searching for the pair that historically performed best is the definition of curve fitting.
The conventional pair has one genuine advantage that has nothing to do with its statistical properties: everybody watches it. Whatever coordination effect exists attaches to 50 and 200, and it does not transfer to a pair chosen because it tested better.