Imagine the crypto crowd is euphoric. Prices have been climbing, social media is buzzing, and the mood sits firmly at “greed.” While most people are piling in, a quieter group is doing the opposite: placing bets that the price is about to fall. These are short sellers, and a new analysis asks whether their behavior in Bitcoin markets looks anything like the way short sellers operate in ordinary stock markets.
The answer, according to research published in the Journal of Behavioral and Experimental Finance, is that they are informed traders, but they play by a different set of rules. They tend to build up their positions when Bitcoin sentiment reaches extreme optimism, and their trades offer evidence of a short-lived ability to predict falling prices.
A new window into crypto trading
Short selling is a way to profit when an asset’s price drops. A trader borrows shares, sells them, and hopes to buy them back later at a lower price, pocketing the difference. In traditional stock markets, short sellers have long been viewed as sophisticated players who bet against overpriced assets and help push prices back toward reasonable levels.
They also tend to follow a specific pattern. Research has shown that stock short sellers often act as “contrarians,” meaning they ramp up their bets right after prices have run up, wagering that the gains have gone too far.
Whether any of this carries over to cryptocurrency has been hard to study, in part because so much crypto trading happens on unregulated exchanges with limited data. That changed in January 2024, when the U.S. Securities and Exchange Commission approved the first spot Bitcoin exchange-traded funds. An ETF is a fund that trades like a stock, and a “spot” Bitcoin ETF holds actual Bitcoin, letting investors get exposure through a regulated brokerage account. Because these funds trade on regulated exchanges, their short-selling activity is recorded and reported.
Olena Onishchenko of the University of Otago in New Zealand used this newly available data to examine three questions. Do Bitcoin ETF short sellers behave like contrarians, betting against recent price gains? Do they time their trades to shifts in investor mood? And does their activity predict where prices head next?
How the study was built
The analysis draws on transaction-level short-selling data from the Financial Industry Regulatory Authority, an organization that oversees U.S. brokerage firms and publishes daily short-sale volumes. The data covered eleven spot Bitcoin ETFs from the launch date of January 11, 2024, through October 31, 2025, spanning 450 trading days.
Onishchenko combined this with market information from Bloomberg, including prices, trading volumes, and fund flows. She also pulled in a widely used gauge of crypto mood called the Bitcoin Fear and Greed Index, which runs from 0 to 100. Low readings signal fear among traders, while high readings signal greed. The index blends together measures of volatility, market momentum, trading volume, and public interest.
One detail worth noting is scale. The dollar value of short positions in these ETFs peaked at about $0.20 billion, roughly 2 percent of their total market value. U.S. stocks, by comparison, tend to carry short interest closer to 4 percent, so shorting activity in Bitcoin ETFs remains relatively modest.
These funds are also still owned mostly by retail investors rather than large institutions. As of the third quarter of 2025, institutional ownership across the eleven ETFs ranged from 4 percent to 31 percent. That retail tilt matters, because the author argues that everyday crypto investors form their price expectations differently from stock investors, which shapes how short sellers respond.
Not the usual contrarians
The first finding stands apart from what happens in stock markets. Onishchenko found no evidence that Bitcoin ETF short sellers act as contrarians. When she measured whether shorting increased after five-day price run-ups, the relationship was statistically insignificant across her models.
The author interprets this through the lens of how crypto prices tend to move. Drawing on earlier work by Kogan and colleagues, she notes that retail crypto investors often read recent price increases as a sign that Bitcoin adoption is spreading, which can feed self-reinforcing momentum. Betting against that kind of rising tide is risky, because prices can keep climbing far longer than a short seller can comfortably wait. That risk, she suggests, likely explains why the contrarian playbook does not show up here.
Riding the mood instead
What short sellers did respond to was sentiment. The study found that shorting activity tends to rise on days following an increase in the Bitcoin Fear and Greed Index. Specifically, a 10 percent jump in Bitcoin sentiment over the prior five days was associated with a 0.08 percent increase in next-day short selling.
This link held up even after accounting for other factors that might influence shorting, such as price volatility, trading costs, fund flows, turnover, and market size. The pattern suggests that short sellers treat extreme optimism itself as a signal of overvaluation, stepping in when the crowd looks most exuberant.
A short-lived edge
The third question was whether all this shorting actually predicts anything. It did, but only briefly. Higher short-selling activity was linked to negative returns over the following two to six trading days. Push the window out to about a month, and the effect vanished entirely, with the relationship turning slightly positive and statistically meaningless.
To illustrate the size of this effect, Onishchenko simulated a strategy that took short positions whenever the sentiment index hit “Extreme Greed,” defined as a reading above 75. That strategy produced a risk-adjusted return of 0.24 percent over five trading days, which works out to roughly 12.10 percent on an annualized basis. Beyond that five-day window, the advantage disappeared, which the author reads as a sign that optimism-driven mispricing gets corrected quickly.
Important caveats
The author is direct about the limits of this work. The reported returns are “gross,” meaning they do not account for the cost of borrowing shares to short in the first place. Citing other research, she cautions that ignoring these lending fees can greatly overstate how profitable short strategies look, and that the apparent gains might shrink or disappear once those costs are included. For that reason, she frames the simulated strategy as a way to measure the information content of short selling, not as a recipe anyone should trade on in real time.
The sample is also short. Spot Bitcoin ETFs have existed only since early 2024, so the findings cover a single stretch of market history. Onishchenko presents her results as early evidence rather than a settled account, and points to future research examining how short-selling behavior might shift across different market cycles as more data accumulates.
Taken together, the study offers a portrait of short sellers in crypto markets who look informed but operate on a different logic than their stock-market counterparts. Where fundamentals like earnings and dividends are absent, the author argues, the edge comes not from analyzing a company’s books but from anticipating and timing the swings of investor mood.
