Short-Squeeze Forensics: The Method
The repeatable checklist for telling a real short squeeze from a re-rating wearing one's clothes: anchor the tape, separate catalyst from positioning, read the borrow market first, decompose headline short interest, respect what short volume cannot show, price what the options market already charged, and write your kill triggers before you pay. Built from a real trade we graded in public — including the half we got wrong.
01Anchor the tape before any story
our own analysisBefore any story about WHY a stock moved, pin down WHAT moved: how big was the day, against how many times normal volume, and where did it close? Get those from your own data, not from a headline — even the percentage move itself can be quoted wrong when someone's feed carries a stale prior close.
The anchor set is: the move measured close-to-close from bars you trust, the volume multiple against the recent average, the close's position in the day's range, and the after-hours or premarket follow-through with its own volume context. Two-sided heavy premarket trade is information; a thin quote drift is not. This is also where you log what your own systems saw and when — the honest lag measurement later tells you whether your coverage, or your judgment, was the gap. Daily Short volume prints belong in the anchor too, with their limits kept in mind.
02Catalyst and positioning are different questions
our own analysis"Why did it move?" is really two questions. One: what NEWS arrived — an earnings beat, a product surprise? Two: who was POSITIONED wrong when it arrived? A violent rally can be buyers choosing to re-value the company, or sellers forced to retreat, and the difference decides whether the move holds.
Treat the catalyst and the positioning as separate findings with separate evidence. A Re-rating shows up as new economics in the filing — a new revenue line, changed unit economics, a credible raised guide — and survives on normalizing volume. Forced positioning shows up in the borrow market and the short-interest decomposition, and decays once the forcing exhausts. The two can stack: covering can amplify a genuine re-rating. Label the move only after both columns are filled, and beware circulating "catalysts" that were public weeks earlier — priced news is not news.
03The borrow market is where squeezes physically happen
our own analysisIf shorts were really being forced out, borrowing the stock would get expensive within hours — the people lending shares reprice fast when their inventory gets scarce. So the first live check is always the borrow desk: what does it cost, is the rebate still positive, and are shares available?
A real squeeze prices the borrow. The Borrow fee leaping from cheap single digits toward the tens of percent, the Rebate (stock loan) flipping negative, and availability evaporating — that triad is the physical signature of shorts being cornered. Its absence through a violent up day is close to disqualifying: whatever the narrative says, the market with actual inventory at stake is lending calmly. Conversely, a fee that suddenly multiplies AFTER you concluded "no squeeze" is the signal to re-open the question — write that revival threshold down as a standing trigger.
04Decompose the headline short interest
public domainThe scary headline number — "a third of the company is sold short!" — is a gross total. Read one filing deeper before believing it: company financing deals can park tens of millions of hedged shares inside it, and the same position can be quoted against two different share counts to sound twice as dramatic.
Decomposition steps: pull the Short interest (SI) series and its settlement dates; check the issuer's own filings for share-loan or convertible financing structures (a Convertible-arb hedge block inside SI is mechanical, not directional); net it out and recompute the month-over-month change in the DIRECTIONAL residue; then state Days to cover (DTC) against the move day's actual volume, not a quiet-tape average. Finally, reconcile the "percent of float" claims — shares outstanding versus true free float can differ by multiples, and both versions will circulate simultaneously. None of this needs paid data: the filings are public record.
05Short volume cannot show covering
our own analysisThere is a daily, public number with "short" in its name, and on big days everyone reads it as proof that shorts covered or doubled down. It can prove neither. It only counts short-marked SELL orders — a short seller buying back looks like any other buyer.
Structural, not a data-quality issue: buy-to-cover executions are unmarked in daily Short volume files, so covering is invisible by construction. A collapsing short-volume ratio on an up day says the marginal seller stopped dominating — nothing more. The only figure that can settle how much covering actually happened is the next Short interest (SI) settlement print, weeks later; until that date, every confident covering claim on either side is storytelling. Write the settlement date into the analysis so the question has a scheduled answer.
06What the options market already charged
our own analysisBy the time a move is public, the options market has already charged for the excitement. The morning-after call buyer pays the highest volatility prices of the whole cycle, needs the stock to repeat its entire move again just to break even on a far-out strike, and fights time from the first minute.
Translate the chain before touching it: state Implied volatility (IV) as the daily move it implies; compute the Breakeven (options) as a required percentage move with a deadline and say it out loud; prefer the longest-dated cheapest-vol expiry if anything at all; wait out the opening auction's worst prices; and size so that losing the entire premium is genuinely acceptable — premium at risk is always the full premium. Post-event vol compression can halve a call on an unchanged underlying, which is how right-on-direction still loses. Heavy same-day call flow (0DTE (zero days to expiry)) plus dealer Gamma (dealer hedging) hedging explains the SPEED of these moves — it does not make chasing them a plan.
07Kill triggers — decide what proves you wrong first
our own analysisDecide what would prove you wrong BEFORE you pay — and rank those triggers by how fast they do damage. The fastest killer for a just-spiked small cap is the company selling new shares into the strength; the quietest is the crowd arriving late while the price stalls.
The standing list, fastest first: a Dilution event or a loaded ATM offering (at-the-market) program going live (watch the issuer's SEC feed directly — the registration history tells you if the machinery exists); option premiums deflating faster than the stock rises; a strong open that fades back through the prior high; borrow staying cheap when your thesis needed it to tighten; social chatter peaking while price goes sideways — the crowd arriving late is the exit liquidity. Each trigger is checkable, timestamped, and assigned BEFORE entry; an analysis that cannot name its kill conditions is a mood, not a method.