Back to blog

Golden Cross vs Death Cross: How EMA Crossovers Work in Crypto

Every 50/200 EMA cross on Bitcoin since 2019 came late. Learn how golden and death crosses work, which filters help and how to set crossover alerts.

Bitcoin chart showing golden cross and death cross EMA crossover patterns with AI signal overlay

On November 17, 2025, Bitcoin’s 50-day EMA closed below its 200-day EMA for the first time in more than two years. BTC closed that day at $92,215 on Binance. By then it was already 27% below the all-time high of $126,200 it had set six weeks earlier, on October 6.

The opposite cross came on September 12, 2026, when the 50-day EMA closed back above the 200-day with BTC at $77,279. That was 34% above the low of $57,800 it had touched on July 1.

Both crosses described the trend that was underway, and both arrived long after it began. That is how moving average crossovers work: they confirm a trend that is already in place. They do not predict one.

This guide explains how golden and death crosses form, lists every 50/200 EMA cross on Bitcoin since 2019 and what followed it, compares the EMA and SMA versions on real data, and shows how to build crossover alerts on TraderSpy that add the confirmation a cross lacks on its own.

What Is a Golden Cross?

A golden cross occurs when a short-term moving average crosses above a long-term moving average. In the most commonly used configuration:

  • Short-term MA: 50-day EMA (or SMA)
  • Long-term MA: 200-day EMA (or SMA)

When the 50-day crosses above the 200-day, recent prices have stayed far enough above the long-term average, for long enough, to pull the faster line through the slower one. Traders read it as a sign that the longer trend has turned up.

Three Phases of a Golden Cross

  1. Decline and bottoming: Price is below both moving averages. The short-term MA is below the long-term MA. The downtrend is intact.
  2. Convergence: The short-term MA begins rising and approaches the long-term MA from below. The gap narrows. Momentum is shifting but has not confirmed yet.
  3. Crossover: The 50-day crosses above the 200-day, usually well after the price has turned.

What Is a Death Cross?

A death cross is the inverse: the 50-day EMA crosses below the 200-day EMA. This signals that short-term momentum has turned bearish relative to the long-term trend.

Three Phases of a Death Cross

  1. Rally and topping: Price is above both MAs. The short-term MA is above the long-term MA. The uptrend is intact.
  2. Convergence: The short-term MA begins falling and approaches the long-term MA from above. Momentum is fading.
  3. Crossover: The 50-day crosses below the 200-day, usually after the price has already fallen a long way.

Every 50/200 EMA Cross on Bitcoin Since 2019

Most crossover guides quote a few well-chosen examples. Here is the full list instead: every time Bitcoin’s 50-day EMA crossed its 200-day EMA since January 2019, computed from Binance’s BTCUSDT daily closes (UTC).

DateCrossBTC closeWhere the cross cameBTC 90 days later
Apr 28, 2019Golden$5,30868% above the low+79%
Nov 23, 2019Death$7,31248% below the high+32%
Jan 31, 2020Golden$9,35345% above the low−8%
Mar 15, 2020Death$5,36149% below the high+77%
May 13, 2020Golden$9,309110% above the low+22%
Jun 23, 2021Death$33,67848% below the high+21%
Aug 14, 2021Golden$47,06961% above the low+36%
Jan 10, 2022Death$41,82239% below the high+1%
Feb 18, 2023Golden$24,63259% above the low+9%
Sep 24, 2023Death$26,24817% below the high+66%
Oct 3, 2023Golden$27,4266% above the low+61%
Nov 17, 2025Death$92,21527% below the high−25%
Sep 12, 2026Golden$77,27934% above the lowNot yet 90 days

“Where the cross came” compares the close on the day of the cross with the lowest low (golden cross) or the highest high (death cross) since the previous cross.

What the table shows:

  • Every cross came late. Each death cross arrived 17% to 49% below the preceding high. Each golden cross arrived 34% to 110% above the preceding low, except October 2023, which came nine days after a death cross.
  • Death crosses were often followed by higher prices. After four of the six, BTC was at least 20% higher 90 days later. The two exceptions were the ones that mattered: January 2022, flat after 90 days and 48% lower after 180, and November 2025, 25% lower after 90 days.
  • The golden crosses rode a rising market. After five of the six that have 90 days of history, BTC was higher 90 days later. But BTC was also higher 90 days later on 58% of all days since 2019, so part of that is the market’s own rise, not the cross.
  • Whipsaws were rare on the daily EMA. Only one cross was reversed within a month: the September 2023 death cross, undone nine days later.

Six or seven events per side in almost eight years is history, not odds. Use the table to calibrate your expectations of a cross, not to predict the next one.

The Lag Problem: Why the Cross Comes After the Move

A 200-day average is built from 200 days of closes. For the 50-day line to cross it, a trend has to run long enough to drag both averages with it, so the cross can only confirm what price has already done.

The two most recent crosses show the delay in days and dollars:

  • Death cross, November 2025: BTC peaked at $126,200 on October 6, 2025. The cross came 42 days later, at $92,215.
  • Golden cross, September 2026: BTC fell to $60,000 on February 6, 2026, and to $57,800 on July 1. The cross came 73 days after that low, at $77,279.

If you wait for the cross to act, you act late, every time. This is why experienced traders do not trade crossovers in isolation. They use a cross to confirm a trend they have already seen in faster indicators: momentum (RSI, MACD), trend strength (ADX) and volume on shorter timeframes.

EMA vs SMA: Which Should You Use?

Both Exponential Moving Averages (EMA) and Simple Moving Averages (SMA) are used for crossover strategies. They differ in how they weigh the past:

FeatureEMASMA
WeightingMore weight to recent closesEqual weight to every close in the window
How the line movesTurns sooner after a price changeTurns later, and also moves when an old close drops out of the window
50/200 crosses on BTC daily since 20191317
Crossed first, in the 13 turns both caught6 times7 times

The textbook says the EMA signals earlier. On Bitcoin’s daily chart the record is split: in the 13 turns both versions caught since 2019, the EMA crossed first six times and the SMA seven times, by anywhere from 1 to 32 days. In November 2025 the SMA death cross came one day before the EMA’s.

The SMA also made four crosses the EMA never did, all between August 2024 and May 2025: a death cross, a golden cross, another death cross and another golden cross. Through all of it, the 50-day EMA stayed above the 200-day without interruption, from October 2023 to November 2025.

So neither version is better in general. What matters is consistency:

  • Many news reports and charting tools use the SMA version, so the “death cross” in a headline may not match an EMA chart.
  • Use the same version everywhere: on your chart, in your alerts and in your backtests, so the cross you are alerted to is the one you see.

TraderSpy’s alert strategies offer both, an EMA cross and an SMA cross, with the fast and slow periods you choose.

Crossover Settings for Different Trading Styles

The faster the chart and the shorter the averages, the more crosses you get and the more of them reverse. Measured on Binance’s BTCUSDT perpetual:

StyleChartEMA pairCrosses on BTCUSDTFilters worth adding
PositionDaily50/2002 in the last two yearsFaster early-warning alerts on the 4-hour chart
Swing4-hour21/5571 in the last two yearsMACD direction, volume on the breakout candle
Active swing4-hour9/21206 in the last two yearsADX above 25, volume above 2x its average
Day trading1-hour9/21389 in the last yearRSI momentum, ADX above 20, volume

On the 4-hour chart, 79 of those 206 EMA 9/21 crosses were reversed within ten candles, less than two days. On the 1-hour chart it was 156 of 389. At those speeds, the filters are the strategy.

Five Common Crossover Mistakes

Mistake 1: Trading the Crossover Alone

The crossover is a confirmation signal, not a prediction. Acting on it alone means acting after the move is well underway. Faster tools (RSI, MACD, ADX and volume on shorter timeframes) show the shift earlier; the cross confirms it.

Mistake 2: Expecting Filters to Fix the Daily Cross

The usual advice is to require RSI above 50 for a golden cross, ADX above 25 or a volume surge. On Bitcoin’s daily chart, those filters changed little:

  • RSI: at every golden cross since 2019, RSI(14) was already above 50 (55 to 68), and at every death cross it was below 50 (26 to 45). A cross comes after a move, and RSI has registered the move by then.
  • ADX: it was below 25 at only two crosses, the September 2023 death cross and the golden cross that reversed it nine days later. An ADX filter would have skipped both: the false start and the turn that followed it.
  • Volume: every golden cross since 2019 closed on volume at or below 1.04 times its 20-day average. A volume-surge rule would have skipped all of them.

Filters earn their keep on the faster crosses, where noise is the problem. The 4-hour numbers further down show how much they remove.

Mistake 3: Using the Wrong Timeframe

Crosses on the 5-minute chart are mostly noise. Even the 50/200 pair crossed 58 times on BTCUSDT’s 1-hour chart in the last year and 24 times on the 4-hour chart in the last two years, against 13 times on the daily chart since 2019. Match the chart to how long you intend to hold a view: daily crosses for the big picture, the 4-hour chart for swing trades, the 1-hour chart only with strict filters.

Mistake 4: Reading Volume on the Wrong Candle

Volume confirms a move when it expands on the candles that break a range or start a trend. The day of a daily cross is usually not one of them: as Mistake 2 shows, BTC’s golden crosses came on ordinary or quiet days. Check volume on the breakout candles of the shorter timeframes instead, where it separates a real break from a thin one (our breakout study measured the difference).

Mistake 5: Watching a Single Timeframe

The strongest crossover setups line up across timeframes: a bullish EMA cross on the 4-hour chart while the daily trend is already up, with momentum building on the 1-hour chart. Each timeframe answers a different question. The daily chart sets the direction; the faster charts set the timing.

Building Crossover Alerts on TraderSpy

TraderSpy does not ship a ready-made golden cross alert that you switch on. Its AI signals come from TraderSpy’s own presets on the 1-hour and 4-hour charts; each is checked before it is published and comes with an entry, three targets and a stop. You don’t pick or enable those presets, and none of them is a daily golden or death cross.

What you build instead is an alert strategy: one timeframe, the coins you choose (or all coins, every crypto perpetual TraderSpy tracks) and up to three groups of rules. TraderSpy watches it around the clock and notifies you when the rules hold, with a push on the TraderSpy app and an in-app notification. Strategies send alerts, never orders.

Rules are judged on closed candles only. A daily rule is checked when the daily candle closes at 00:00 UTC, the same close the table above was built from, so the alert comes once the candle that completes the rules has closed. TraderSpy reads Binance futures candles; on the BTCUSDT perpetual, every 50/200 EMA cross since 2021 fell on the same day as in the table.

Layer 1: Early Warning on the 4-Hour Chart

The shift that eventually produces a daily cross shows up first on faster charts:

  • EMA 9/21 cross on the 4-hour chart, with ADX above 25 and volume above 2x its 20-candle average.
  • MACD turning positive on the 4-hour chart (the MACD line crossing above its signal line), with RSI above 50 and ADX above 25.

The filters are what make a fast cross usable. On BTCUSDT’s 4-hour chart over the last two years, the EMA 9/21 crossed upward 103 times. Requiring ADX above 25 left 22; adding the volume rule left 3. The MACD rule went from 166 upturns to 34 with both filters.

Each strategy watches one timeframe, but a 4-hour rule can still carry the daily trend. Add price above EMA(1200): on the 4-hour chart, a 1,200-candle EMA follows the 200-day EMA within a fraction of a percent (0.06% on average on BTCUSDT over the last two years).

Layer 2: Confirmation on the Daily Chart

  • RSI(14) crossing above 50 on the daily chart, or below 50 for the bearish side: daily momentum turning with the faster signals.
  • Price crossing above the 200-day EMA: on BTC since 2019, the last close back above the 200-day line came 3 to 30 days before each golden cross. It also fires on reclaims that fail, which is why it confirms rather than triggers.

Layer 3: The Cross Itself

  • Golden cross: daily chart, EMA cross 50/200 crossing above.
  • Death cross: daily chart, EMA cross 50/200 crossing below, with the strategy’s direction set to short.

You can add RSI above 50 (below 50 for the death cross), but as Mistake 2 shows, on BTC it has always been true at the cross. The cross alert is the slowest in the set: it tells you the trend has been confirmed, not that it is starting.

Layer 4: Invalidation

  • After a golden cross: price crossing below the 200-day EMA on the daily chart weakens the bullish case.
  • After a death cross: price crossing back above the 200-day EMA weakens the bearish case.

A strategy has no memory of earlier alerts, so this one fires on every daily close through the 200-day EMA; read it together with your cross alerts.

Each of these alerts is its own strategy. The Free plan includes one; paid plans include more (pricing).

Or Describe It to the Agent

You can also type the rule into the AI agent in a sentence, for example: “Alert me when the 50-day EMA crosses above the 200-day EMA on the daily chart for BTCUSDT and ETHUSDT.” It backtests the idea on up to 1,000 candles per coin, tells you how often it would have alerted and how price moved afterwards against the coin’s own baseline, and offers the strategy as a card you create with one tap.

Expect thin numbers on the daily chart. A 50/200 cross is rare, so the stored history usually holds one or two per coin. The backtest says so when a rule fired fewer than ten times, and running it on all coins gives it more events to read. The 4-hour rules above fire often enough for a meaningful test.

Why Crossovers Get So Much Attention in Crypto

  • 24/7 markets: No weekend gaps or overnight sessions. Every close feeds the averages, so the lines reflect continuous trading.
  • Leverage: When a widely watched level breaks, leveraged positions get liquidated, which can extend the move in either direction.
  • Everyone watches the same lines: The 50/200 cross is one of the most quoted signals in crypto news, which is why it can move sentiment on the day it prints, even though it describes the past.
  • Correlated assets: Altcoins often move with BTC, so it pays to watch the same cross across many pairs. An alert strategy set to all coins does that with one rule.

Getting Started

  1. Know the current regime: Check where BTCUSDT trades against its 200-day EMA and which way the 50-day line is heading. That tells you which cross is even possible next.
  2. Build the cross alert on TraderSpy: daily chart, EMA cross 50/200, for the coins you follow.
  3. Add a 4-hour early-warning strategy: EMA 9/21 or MACD with the ADX and volume filters, so you see the shift before the daily lines cross.
  4. Test before you trust it: describe the rule to the AI agent and read how often it would have fired and what followed, before you create it.
  5. Check smart money positioning: see how the top traders TraderSpy tracks are positioned, on the Smart Money feed.
  6. Never act on the cross alone: require at least one confirmation (RSI direction, ADX trend strength or volume on a shorter timeframe), and decide your risk before any alert fires.

A golden cross or a death cross sums up what the last few months did in one line. Used alone, it is late by design. Used as the slow layer under a set of faster alerts, it keeps the big trend in view while the faster signals do the timing. Past crosses describe what happened, not what will happen next: this is market research, not financial advice.