Short Interest & Thesis
Short Interest & Thesis
Bottom Line
Reported short interest is material and trending up — 39.5M shares (12.1% of float) as of the May 15, 2026 settlement, more than 1.5x the share count short six months earlier — but the position is not crowded relative to liquidity (2.4 days to cover against ~$2.7B ADV) and the borrow is general-collateral (~30 bps APR, ample lendable supply). The decision-useful signal is therefore not a squeeze setup; it is a credible, multi-source short thesis that has now drawn a fresh wave of bearish positioning after Strategy broke its "never sell bitcoin" pledge on June 1 and after the mNAV premium that anchored four years of accretive ATM issuance compressed to roughly 1.0x. Evidence quality is mixed: the staged official-source pull returned no FINRA rows for this ticker, so the position figures here come from third-party aggregators that re-publish the official NASDAQ semi-monthly settlement — directionally reliable, but two to three weeks lagged.
Source classification. All position figures below are reported short interest (official semi-monthly NASDAQ settlement, re-published by third parties). Borrow fee and lendable-supply rows are third-party securities-lending indicators, not exchange data. Allegations from short sellers and class-action plaintiffs are claims, not findings. FINRA daily short-sale volume is trading flow and is shown only as tape context — never as a stand-in for outstanding short interest.
Reported Position Trend
The reported short-interest position has roughly doubled since the September 2025 settlement, while the price collapsed and float was diluted further. The 12.1% short-float reading at May 15, 2026 is the highest sustained level in the available history, though not a record (Feb 13, 2026 briefly printed 14.4%).
Shares Short (M, May 15 2026)
Short % of Float
Days to Cover
Notional Short ($B)
The position roughly doubled (21M → 40M shares) over eight months as the bitcoin premium compressed and the four-year "never-sell" pledge frayed. The May 15 reading shows a small pullback (-3.0%) from the April 30 print — the first sequential decline of the cycle — suggesting some shorts began covering into the late-April rally before the June 1 BTC sale re-engaged the trade.
Crowding vs. Liquidity
The percent-of-float number reads "elevated" in isolation but MSTR is one of the most liquid US equities — 20-day ADV is ~17.4M shares ($2.7B/day), so even the highest-ever 40.8M short position cleared at about 2 days to cover. There is no liquidity-driven squeeze setup here.
Crowding read. 12.1% short-float is high in absolute terms — peer software comp Cadence sits near 2.0% — but with $2.7B/day of liquidity and 2.4 days to cover, the position is not "crowded" in the institutional sense. A 20% short-covering rally would clear in under half a session of normal volume. The squeeze tail risk is real because of vol and reflexivity, not because the borrow side cannot exit.
Borrow Pressure
The borrow indicator data are third-party (not exchange-published) and only partial windows are available, but every available window points the same direction: MSTR is general collateral. Borrow fees in early April 2026 sat in a 29-43 bps APR band, lendable supply was 5–8 million shares intraday, and a 50.7% off-exchange short-volume ratio (FINRA dark-pool tape) indicates that locating shares to short was not a constraint.
The clean read: no borrow squeeze, no hard-to-borrow regime, no locate friction — at least through mid-April. Subsequent intraday windows are not in the public dataset, so a borrow tightening into the late-May BTC sale would not be visible here. Given the institutional liquidity, a sudden HTB shift is unlikely without a corporate action or a sharp drop in float.
Public Short Thesis Ledger
The short thesis on MSTR is credible and multi-sourced — capital-structure / wrapper-premium critique, not GAAP accounting fraud. The ledger below separates allegation, evidence, company response, and unresolved risk.
Chanos closed his short before the worst leg down. Kynikos exited the MSTR short on November 10, 2025 after a ~50% drawdown, calling the premium thesis played out. The position has roughly doubled since — different cohort, different capital base, different catalyst. Treating the current short cohort as a continuation of Chanos's thesis is incorrect; the current cohort is positioned for closed-loop preferred-dividend stress, not premium compression alone.
Disclosed Institutional Short Positions
The US does not have UK/EU-style holder-level net-short disclosure, so the only institutional shorts visible are those that elect to disclose via 13F (rare, voluntary, lagged). The two MSTR shorts visible in the most recent 13F cycle are below — this is not an exhaustive holder map, just the publicly searchable subset.
Read carefully. HSBC and PEAK6 appearing on a 13F short list almost always reflects dealer-book or options-market-maker hedging, not a directional fundamental short. The 12.1% short-float reading is being driven primarily by non-disclosed hedge-fund positions; the named disclosures are not the marginal seller.
Market Setup Around the Short
Three pieces of evidence link the elevated short interest to the current tape and price action:
Peer Context
A clean apples-to-apples short-interest comparison is not available in staged data for the full peer set. The single peer comparison point referenced by MarketBeat — Cadence Design Systems at 1.98% — sits in the software sector but is not an economic comp (it is a stable EDA franchise, not a treasury vehicle). The more relevant comparison is to other bitcoin-balance-sheet / mining vehicles; that comparison is not staged here and would need to be sourced directly from each issuer's reported short interest. Marking as partial in the JSON output.
Evidence Quality
What Would Change The Read
Decision-useful upgrades. A documented borrow-fee spike above 100-150 bps APR or a sharp drop in lendable supply would shift the read from "thesis-driven short" to "stress-driven short" and create a real squeeze tail. A second public short-seller report from a forensic shop (Hindenburg, Muddy Waters, Kerrisdale, etc.) — distinct from the existing capital-structure critiques — would warrant escalation. STRC trading below $90 with a concurrent jump in days-to-cover would tighten the closed-loop thesis the current short cohort is trading.
Decision-useful de-escalation. A successful preferred raise at par with the dividend rate unchanged would weaken the closed-loop thesis. A short-interest decline of two consecutive settlements while bitcoin is flat-to-down would suggest the marginal short has run out of conviction. Dismissal of the Levi & Korsinsky class action without a material settlement would remove one ledger item.
How This Affects The Investment Case
For a PM weighing a long: short-interest data alone does not change sizing — the position is not crowded and the borrow is cheap, so a structural squeeze is not the asymmetric risk. The short thesis itself does change the case, because it has now been partially validated by management's own June 1 action, and the next mNAV reading will determine whether the ATM accretion engine that funds the strategy is still working. For a PM weighing a short: the trade is uncrowded by squeeze metrics but priced — the rate of change in short interest is more useful than the level, and the right entry is on rallies that re-test mNAV premium rather than on tape that already reflects the bear case.