Bitcoin Nears a Golden Cross: What the Last 9 Actually Delivered
A Bitcoin golden cross is within days of printing. As of our study computed September 7, 2026 from daily closes, Bitcoin (BTC) has a 50-day simple moving average of $69,699.48 USD against a 200-day of $69,815.70 USD. The gap is $116.22, or 0.2%. If price holds, the faster average crosses above the slower one and the signal chartists treat as the start of a new uptrend fires for the tenth time since 2016.
We did not want to write about that on vibes, so we computed the full record instead: every BTC golden cross in 3,653 sessions of daily closes since September 7, 2016, and what each one actually delivered. Nine completed regimes. The signal has usually paid. It also lost money the last time it fired.
The setup this morning
Bitcoin trades at $79,388.94 USD, down 1.2% over the past day, or $109,681.30 CAD, down 1.34%. Ethereum (ETH) is at $2,487.83 USD, down 1.06%, or $3,437.48 CAD, down 1.19%. Data as of 6:05 a.m. ET, September 7, 2026. Crypto runs through the holiday, but the TSX and US equity markets are closed for Labour Day, so there is no spot ETF trading today and no fresh flow data until Tuesday. Our Sunday crypto recap covers the weekend drift into this setup and the macro week that follows.
The regime this cross would end has been a long one. A death cross printed November 16, 2025 at $94,177.08 USD and has run 295 days. Bitcoin fell as low as $58,558.86 USD inside it, never made a new all-time high, and still sits 15.7% below the price where that cross printed. The last close in our series was $79,373.98 USD.
What the last 9 golden crosses actually delivered
Every figure below is computed from BTC daily closes (source history: Yahoo Finance) using simple moving averages. Conventions matter here: studies built on exponential averages produce different dates and different numbers, so this table should not be compared against one.
| Cross date | Price at cross (USD) | Peak gain | Days to peak | Regime ended | New ATH |
|---|---|---|---|---|---|
| 2017-03-25 | $972.78 | +1,904.3% | 266 | +608.3% | yes |
| 2019-04-24 | $5,464.87 | +138.2% | 63 | +69.2% | no |
| 2020-02-19 | $9,633.39 | +3.0% | 4 | -30.6% | no |
| 2020-05-21 | $9,081.76 | +599.2% | 327 | +292.2% | yes |
| 2021-09-15 | $48,176.35 | +40.2% | 54 | -10.5% | yes |
| 2023-02-07 | $23,264.29 | +35.3% | 156 | +11.0% | no |
| 2023-10-30 | $34,502.36 | +111.8% | 135 | +76.6% | yes |
| 2024-10-28 | $69,907.76 | +51.8% | 85 | +13.3% | yes |
| 2025-05-22 | $111,673.28 | +11.7% | 137 | -15.7% | yes |
“Regime ended” is the return from the cross price to the day the regime flipped back.
The median peak gain after a golden cross was +51.8%. The best was +1,904.3% in 2017 and the worst was +3.0% in February 2020. The median run to that peak took 135 days, with a range of 4 to 327. Six of the nine regimes produced a new all-time high. Three ended below the cross price.
Forward returns measured from the cross date, n=9 at every horizon:
- 30 days: median +9.7%, best +46.2%, worst -35.7%, positive 67% of the time
- 60 days: median +22.0%, best +151.2%, worst -25.4%, positive 89%
- 90 days: median +21.8%, best +182.2%, worst -3.2%, positive 89%
- 180 days: median +35.4%, best +273.3%, worst -17.7%, positive 78%
The shape of that is the interesting part. The first month is the least reliable stretch, with a third of instances negative and a worst case of -35.7%. Hit rates improve at 60 and 90 days, then slip back at 180. Nine observations is a small sample and the spread across them is enormous, from a +1,904.3% peak in 2017 to +3.0% in February 2020, so the medians carry more information than the extremes.
The case against reading too much into it
Three of the nine regimes ended in the red, and the details matter more than the count.
February 2020 is the cautionary one. The cross printed on February 19 at $9,633.39 USD, peaked four days later at +3.0%, and then the COVID crash arrived. That regime ended -30.6%. A golden cross offered no protection whatsoever against a macro shock, and this time a macro event is already on the calendar.
September 2021 fired at $48,176.35 USD into deteriorating macro conditions, managed a +40.2% peak, and still ended -10.5%. And then there is the most recent one: May 22, 2025, at $111,673.28 USD. It peaked at just +11.7% and the regime ended -15.7%. The last time this signal fired, buyers lost money.
The mirror signal is a useful check on how much any of this is worth. After BTC’s nine death crosses, the 180-day forward return was positive 56% of the time with a median of +33.8%. The bearish signal was followed by gains in five of the nine cases. Moving-average crosses lag by construction. They confirm trends that already happened; they do not predict them.
One more thing that should keep the enthusiasm in check: the cross has not printed. At $79,388.94 USD and falling 1.2% on the day, a continued slide pushes the crossover out or cancels it outright.
The macro overlay
The event risk is dense, and it lands inside the first two weeks of any new regime, well before the 30-day horizon in the table above would close.
Friday’s US jobs report came in hot. Nonfarm payrolls rose 162,000 in August against a consensus of roughly 56,000, July was revised from -23,000 to +21,000, and unemployment held at 4.1% (BLS Employment Situation). That print revived rate hike risk and knocked Bitcoin back under $80,000 USD on Friday.
August CPI lands Friday, September 11 at 8:30 a.m. ET, the last inflation reading before the Federal Reserve decides on September 16 at 2:00 p.m. ET, with updated projections. So this cross, if it prints, prints into a Fed meeting that carries live hike risk rather than into a settled policy backdrop. February 2020 is the reminder of what an unpriced macro event does to a chart signal four days old.
On the demand side, US spot Bitcoin ETFs took in $986.9 million USD last week, capping their best three-week run of 2026 at $3.8 billion USD in net inflows, per flow figures reported September 6 by CoinCentral. That is not our own data, and it pauses today, since the holiday closes the exchanges those funds trade on.
What would confirm the signal, and what would cancel it
Confirmation is mechanical: the 50-day closing above the 200-day, which needs price to hold well enough to close a $116.22 gap. The more informative test comes afterward, in the form of a hold above the cross price through the first 30 days, the window where a third of past instances went negative.
Cancellation is equally mechanical. Continued selling of the kind seen Friday and this morning keeps the 50-day underneath and no cross prints at all. If Friday’s CPI or the September 16 decision hardens hike expectations, that is the route by which this setup expires without ever becoming a signal.
Bitcoin was under $78,000 USD earlier this month after a 25% August, and we asked then what comes next. This study is one answer with its limits attached: nine observations, a median that looks encouraging, and a most recent instance that lost 15.7%. For Canadian readers who would rather take the theme through an equity account than the spot market, our Canadian crypto stocks page covers the listed names.
Disclaimer: The content on bestcanadianstocks.ca is for informational and entertainment purposes only and does not constitute financial advice. Past performance is not indicative of future results. Always consult a qualified financial advisor before making investment decisions. Market data as of 6:05 am ET, September 7, 2026; crossover study computed 2026-09-07.
