Find Them: What a Squeeze Is and How to Screen

A short squeeze is not a prediction that a stock is going up. It is a description of who has to buy. Short sellers owe shares back. When the price rises against them, some are forced to buy, and their buying pushes the price up against the rest. The trade is to find stocks where that forced buying could be large relative to the shares that actually trade, then wait for a reason it could start.

1. The mechanics in one paragraph

A short seller borrows shares, sells them, and later buys them back to return. The loss on a short is unlimited and grows as price rises. Pressure to cover comes from four places: stop orders, margin calls, the borrow fee, and the lender recalling shares. If enough of that pressure arrives together, in a stock with a small float, buying demand exceeds the shares offered and price jumps. That jump creates more pressure. For the full borrow-and-locate mechanics, see Short Selling and Short Squeeze Mechanics.

Two things people call a "squeeze" are different:

• Options hedging (often called a gamma squeeze) comes from market makers adjusting their stock hedges as option prices move. It can add to a squeeze when dealers are net short the options, but it has no official definition, and the SEC staff found no evidence of one in GameStop in January 2021.
• The Bollinger Band squeeze is a volatility pattern on a chart. It has nothing to do with short sellers. See Bollinger Bands.

2. Four gates and the numbers that rank them

A full list of metrics and abbreviations is in the Squeeze Metrics Cheat Sheet.

These thresholds are working heuristics, not laws. Nobody has a validated cutoff that predicts a squeeze. Use them to sort, not to predict.

Worked example. A stock has 40 million shares in its float and 14 million sold short. Short interest is 14 ÷ 40 = 35% of float. Average daily volume is 2.8 million shares, so days to cover is 14 ÷ 2.8 = 5.0. Float is under 50 million. That is one gate cleared (short interest). Float and days to cover will help it rank, but rankers are not gates. It needs two of the other three (borrow, utilization, a dated catalyst) before it makes the list. S2 picks the same stock up with its borrow and catalyst. Making the list does not earn a trade.

3. A repeatable screen

Run it weekly, not hourly. The goal is a list of 10 to 20 names, ranked, that you know well enough to act on.

1. Set a price and liquidity floor, for example no stocks under $2 or under 500,000 shares of average daily volume. The cheapest, thinnest stocks are where the worst fills and the most dilution live.
2. Check each name against the four gates: short interest of 20% of float or more, cost to borrow of 50% or more, utilization of 90% or more, and a dated catalyst inside the next two weeks.
3. Keep the names that clear at least three of the four.
4. Rank what is left by days to cover (above 5 is better), float (under 50 million is better) and the chart.
5. Remove anything with an active offering, shelf registration being used, or a pending bankruptcy or delisting notice. These are supply events. See S3.

The stock profiles and Stock Pulse pages are useful for news context. Short interest and borrow data come from your broker, an exchange, or a data vendor. They differ by source. Always compare at least two.

4. What the screen cannot tell you

• Short interest is reported twice a month, as of a mid-month and a month-end settlement date, and published about seven business days after each settlement date. It is always stale.
• Some shorts are hedges. A fund short against a convertible bond has no reason to cover when price rises. Raw short interest overstates the real pressure.
• A crowded short with no catalyst can stay crowded for years.

Key Takeaways