Is Greed a Sell Signal? What the Data Actually Says

Everyone knows the line. "Be fearful when others are greedy." So when the Crypto Fear and Greed Index, a 0 to 100 gauge of crowd mood, lights up red hot at 75 or above, the instinct is to sell. We wanted to know if the instinct is correct.

Key facts

• From February 2018 to June 2026, the first day the Crypto Fear and Greed Index climbed to 75 or higher was followed by Bitcoin closing higher 7 days later 72.2 percent of the time, versus a 52.6 percent baseline on a normal day (Crypto Fear and Greed Index, alternative.me, and Bitcoin daily closes, Binance).
• After a greed reading of 75 or higher, Bitcoin's average 7 day gain was +4.71 percent and the median was +3.82 percent.
• Tightening the greed cutoff to 80 raised the 7 day win rate to 88.9 percent with an average gain near 9 percent, drawn from 9 signals at 80 versus 18 signals at 75.
• Over 90 days after a greed 75+ signal, Bitcoin closed higher 61.1 percent of the time with an average gain of +23.38 percent and a median of +18.54 percent.
• Charles Kindleberger's five stage bubble model, built on Hyman Minsky, runs displacement, credit expansion, euphoria, distress, then panic, placing euphoria in the middle rather than at the end (Vault: Manias, Panics, and Crashes, Kindleberger).

So we ran the mirror image of our fear study. We pulled the full Fear and Greed history and Bitcoin daily closes from February 2018 to June 2026. We marked every first day the index climbed into greed, at 75 and higher, and deduped them so a fresh signal only counted once the crowd had cooled back below 60. Then we measured what Bitcoin did over the next 7, 30, and 90 days, and compared it against the average of every single day in the sample.

Here is the one nugget that breaks the reflex. On a normal day in this window, Bitcoin closed higher 7 days later 52.6 percent of the time. After the crowd flipped to greed, it closed higher 72.2 percent of the time, with an average 7 day gain of nearly 5 percent. The greed reading did not mark the top. It marked the middle of a move. The question is why, and that is where it gets interesting.

Here are the real numbers. Same dataset, no cherry picking.

Tighten the cutoff to 80 and it gets stronger near term, not weaker. The stricter greed signal won 88.9 percent of the time over 7 days with an average gain near 9 percent. That is 18 signals at 75, 9 at 80. Across every horizon, walking in when the crowd was greedy beat walking in on a random day.

This is the opposite of what the slogan promises, and the masters in our research stack explain it cleanly. Howard Marks, the distressed debt investor who built Oaktree, describes sentiment as a pendulum that swings between greed and fear and is almost never at rest. His most useful observation for us is the asymmetry. Bottoms form fast, because fear is violent and exhausts itself quickly. Tops form slowly. Euphoria lingers, builds, and rolls over in stages rather than snapping. A single greed reading fires early in that slow process, while the trend is still feeding on itself.

Charles Kindleberger, the economic historian, gives the anatomy. His five stage bubble model, built on Hyman Minsky, runs displacement, then credit expansion, then euphoria, then distress, then panic. Notice that euphoria sits in the middle, not at the end. It is the phase where the mainstream finally piles in, and prices often keep climbing for a long stretch after greed first appears. The top does not arrive until distress, when sophisticated money quietly sells into the crowd. That handoff, often called distribution, is slow and hard to time precisely. George Soros frames the same thing as a boom bust sequence where a self reinforcing loop accelerates into a parabolic move before a twilight phase of ignored warnings. Greed is the accelerant, not the brake.

So the honest read is this. A greed reading is the smoke, not the fire. It tells you the loop is running and momentum is real. It does not tell you the loop is about to break. The break shows up later, in distress phase tells like insider selling, fading volume, and price stalling while sentiment stays hot.

Limitations, stated plainly. This is one asset across one era, a roughly eight year window dominated by a structural bull market in Bitcoin, so the deck is tilted upward and 18 signals is a small sample. Forward returns overlap, which inflates apparent significance. And greed marking the middle of past trends is not a promise it marks the middle of the next one. The euphoria phase is real, but where you sit inside it is the part nobody times cleanly.

Takeaway. Greed alone is not a sell signal. It is a momentum signal. Selling the first hot print would have left large gains on the table in this history. The discipline the masters teach is not to flee greed, but to flee distress, the slow, quiet handoff that comes after the crowd is already euphoric. Watch for the second domino, not the first.

Sources: Crypto Fear and Greed Index (alternative.me), Bitcoin daily closes (Binance). Vault: Mastering the Market Cycle (Marks), Manias, Panics, and Crashes (Kindleberger), The Alchemy of Finance (Soros). Past performance is not indicative of future results. Editorial research. No financial advice.

Frequently asked questions

Should I sell Bitcoin when the Crypto Fear and Greed Index hits greed? Not based on this study. From February 2018 to June 2026, the first greed reading of 75 or higher was followed by Bitcoin closing higher 7 days later 72.2 percent of the time, so greed alone is a momentum signal, not a sell signal.

What return did Bitcoin show after the Fear and Greed Index hit 75? After a greed reading of 75 or higher, Bitcoin's average 7 day gain was +4.71 percent (median +3.82 percent), the average 30 day gain was +14.55 percent (median +8.35 percent), and the average 90 day gain was +23.38 percent (median +18.54 percent).

Does a stricter greed reading of 80 give a stronger signal than 75? Yes, near term it gets stronger. The 80 cutoff won 88.9 percent of the time over 7 days with an average gain near 9 percent, based on 9 signals at 80 compared with 18 signals at 75.

Why does greed mark the middle of a move rather than the top? Howard Marks notes bottoms form fast while tops form slowly, so a single greed reading fires early in a slow process. Kindleberger's five stage bubble model places euphoria in the middle, and the top arrives only at the later distress phase when sophisticated money sells into the crowd.

What are the limitations of this Bitcoin greed study? It covers one asset across one roughly eight year era dominated by a structural Bitcoin bull market, with only 18 signals, a small sample. Forward returns overlap, which inflates apparent significance, and past behavior is not a promise about the next trend.