Golden Cross and Death Cross — The Moving Average Crossover Strategy

Few technical signals make financial-media headlines. The golden cross and death cross — every single time. "Death cross on Bitcoin!" reads like a warning of the apocalypse, and "golden cross" like a rocket launch. Underneath it all, though, sits one of the simplest and slowest strategies in all of technical analysis: two moving averages crossing.

That simplicity is both its strength and its weakness. This article shows the numbers from multi-year backtests — including the ones that spoil the media myth — and answers a question more important than "does it work": who it works for.

What the Golden Cross and Death Cross Involve

The strategy uses two simple moving averages on the daily timeframe:

• SMA50 — the average closing price of the last 50 sessions (the medium-term trend),
• SMA200 — the average of 200 sessions (the long-term trend).

Golden cross: the SMA50 crosses above the SMA200 from below. Interpretation: medium-term momentum has caught up with the long-term trend — the market is shifting into an uptrend regime. The classic buy or hold-long signal.

Death cross: the SMA50 crosses below the SMA200 from above. Interpretation: the market is entering a downtrend regime. A signal to close longs (and, in more aggressive variants, to go short).

A key feature: both averages are calculated from the past, so the signal always arrives late — after a large part of the move has already happened. This isn't a tool for catching bottoms and tops. It's a tool for classifying the market's regime: "we're in a bull market" / "we're in a bear market" — with full awareness that you'll be one of the last to find out, but in exchange for a relatively low rate of false alarms in markets that trend.

What the Numbers Say

The most widely cited golden-cross backtest (QuantifiedStrategies, S&P 500 since 1960 — buying on the golden cross, exiting on the death cross) paints a picture far from both hype and media panic:

How to read that honestly? The strategy didn't beat plain buy-and-hold on returns — its value lies in what happens to risk: a drawdown almost half as deep, because the death cross pulled the strategy out of the market before the deepest part of the major bear markets. You pay for that with part of the gains from bull markets you enter late.

A second number the media stays quiet about: a study of moving-average crossover systems on the S&P 500 (1960–2025) found that basic crossover systems generate 57–76% false signals — especially in sideways markets, where the averages cross back and forth without a trend ever forming. Interestingly, shorter average periods had a higher single-signal win rate (up to ~43%), while very long ones could drop to ~24% — but the rare signals from the longer averages, when they did hit, caught moves many times larger than the losses.

Mandatory caveats: we're citing QuantifiedStrategies data from summaries — verify the exact numbers at the source before making any decision. A sample of 33 signals is statistically very small. And, as always: historical results don't guarantee future ones — 66 years of the US stock market is one specific dataset, not a universal law. On BTC, whose history offers barely a dozen or so crosses, every statistic is even weaker.

How to Apply It Step by Step

Variant 1 — the classic positional system:

1. D1 timeframe, two averages: SMA50 and SMA200, signals only from closing prices (don't react to intraday crossovers).
2. Entry: the close of the day the SMA50 crosses above the SMA200, following the opposite setup.
3. Exit: the close of the day of the death cross (SMA50 below SMA200).
4. Size the position as always, from risk — except that in a system without a classic stop, the death cross itself serves as the emergency exit, so position size has to assume the exit could land tens of percent below the entry (position sizing — how to calculate it).

Variant 2 — a regime filter (for active traders): you don't trade the crosses themselves; you use the average alignment as a filter: only take longs when the SMA50 is above the SMA200, and prefer shorts (or cash) in the opposite alignment. Entry signals come from other tools — e.g., RSI or MACD. This is the most practical use of the crosses: the slow indicator does what it does best (classify the regime), and faster tools handle the timing.

Variant 3 — support for a DCA plan: an investor accumulating via DCA can use the golden cross/death cross as a simple signal for modulating purchase amounts (e.g., larger buys in a downtrend regime require strong nerves, but historically bought cheaper). This is an add-on, not a necessity.

Numerical example: a $20,000 account, a golden cross on BTC at $85,000. You plan to exit no later than the death cross and accept a −20%-from-entry exit scenario; at 4% account risk ($800 — positional, not intraday) the position = 800 / 17,000 = ~0.047 BTC (~$4,000 notional). Conservative? Yes — because the exit in this strategy can be far away.

Variants and Modifications

The 50/200 pair is a convention, not a law of nature — it's worth knowing the most common variations:

• EMA instead of SMA. Exponential averages weight recent prices more heavily, so crosses appear earlier. The cost: more false crossovers in consolidations. In markets as volatile as crypto, some traders consider this trade-off worthwhile — others don't; only testing settles it.
• Shorter pairs (20/100, 20/50). More signals and less lag, but a rising whipsaw rate — you step straight into the zone the 1960–2025 study found most prone to false crossovers.
• The weekly timeframe. A cross on W1 is an extremely rare, extremely slow signal — used more as macro background than as a trading system.
• Confirmations. Popular extra filters: rising volume on the crossover, the slope of the SMA200 (only signal when the two-hundred is climbing), a trend-strength reading from ADX. Every filter shrinks an already small number of signals — on a small sample it's easy to "optimize" a strategy down to nothing this way, so add them sparingly.

On BTC, the history of the crosses is instructive qualitatively: golden crosses confirmed every major bull run — but usually long after the bottom; death crosses warned ahead of the 2018 and 2022 bear markets, but sometimes printed right before sharp rebounds too. The conclusion matches the equity data: this is a regime-confirming tool, not a regime-leading one.

When It Doesn't Work / Who It's For and Who It Isn't

It doesn't work in a sideways market. This is its biggest weakness: in consolidation, the averages cross repeatedly, generating a string of losses (the aforementioned 57–76% false-signal rate). A filter can be ADX or simply watching the SMA200's slope — a flat average means no regime, means no trade.

It doesn't work as a timing tool. Entering after a golden cross usually lands after a 15–30% rally from the low; a death cross on volatile assets (BTC!) can print close to a local bottom more than once. Anyone expecting to buy the bottom will be frustrated every time.

A small sample means big uncertainty. 33 signals in 66 years on an index, and a dozen or so on BTC, isn't a big enough base to "know" a strategy works. It's a hypothesis with a nice track record, not a certainty. For comparison: strategies generating hundreds of trades can be tested statistically in a meaningful way; with 33 observations, it only takes two or three of the 20th century's biggest bull runs lining up favorably for the whole statistic to look better than reality actually is.

The media makes the crosses out to be something they're not. A "death cross on BTC" headline generates clicks precisely because it sounds like a forecast — when it's actually a delayed description of something that already happened. If reading a headline like that makes you feel an urge to act immediately, that's the best moment to go back to the table with the signal count: one every two years.

Who this IS for: a long-term investor who wants a simple, mechanical rule that limits exposure to the deepest bear markets and accepts a lower return in exchange for shallower drawdowns; a trader, as a regime filter layered under faster strategies; someone who doesn't have time for daily analysis (a decision roughly once every two years).

Who this is NOT for: day traders and scalpers (a signal once every two years isn't trading), anyone looking to "buy the bottom," anyone trading short-history altcoins (not enough data to judge), and anyone making decisions in five minutes because of a "death cross!" headline. The crosses are slow by nature — used as a quick emotional signal, they work exactly backwards from how they should.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero