Short Interest & Thesis

Short Interest & Thesis — Toast, Inc. (TOST)

Bottom line. No official reported short-interest, short-sale-volume, or securities-lending data was staged for TOST, so positioning, crowding, days-to-cover, and borrow pressure cannot be quantified from this run — that absence, not a number, is the headline. There is also no credible public short-seller report, activist short campaign, or accounting allegation specific to Toast in the staged research; the only short-selling material found is generic commentary about other companies. The decision-useful work here is therefore the fundamental bear case, which lives in Toast's own disclosures — payments-revenue concentration, take-rate pressure, ongoing dilution, and a dual-class structure — set against a clean audit and legal record, a company that turned GAAP-profitable, and a stock that has already de-rated roughly 45% from its 2025 high. Short positioning is not, on the evidence available, an independent input to the investment case.

1. Reported positioning — unavailable, not "low"

The staged short-interest feed (FINRA-style official sources) returned no reported short-interest rows, no short-sale-volume rows, no public net-short disclosures, no borrow-pressure rows, and no peer context for TOST. Average daily volume was also not provided by the feed, so even a source-supplied days-to-cover is impossible. The correct institutional reading is absence of data, not a confirmed light short book — these are different things, and the page treats them as such.

Source: staged short-interest feed for TOST — all channels returned zero rows (reported short interest, short-sale volume, net-short disclosures, borrow); not derived from any filing page.

2. Liquidity context — what can and cannot be said about crowding

Crowding is a comparison of short exposure to liquidity, float, and peers. The short-exposure side of that comparison is missing, so TOST cannot be called "crowded" or "uncrowded" on evidence. What is observable is the liquidity denominator: the stock trades roughly 10.9M shares a day (60-day average), about $287M of notional, against 607M Class A and B shares outstanding and a 629M fully diluted count — a deep, liquid tape where, in principle, a short book would be coverable, but there is no measured short book to cover.

Sources: ADV and dollar volume derived from staged daily price data (trailing 60 sessions, as reported); shares outstanding and the 629M fully diluted count per FY2025 10-K, MD&A — Dilution [1]; Class B voting power per Item 1A Risk Factors [2].

Toast discloses a fully diluted share count of 629M against 589M Class A and B shares issued and outstanding at year-end 2025, the gap being options, unvested RSUs, warrants, and charitable-donation reserves [3]. For a name with no measured short interest, the more relevant supply-side overhang is this stock-based-compensation dilution, not borrow.

3. Dilution — the real overhang a bear leans on

Share count is the one "positioning" series that is fully observable, and it tells the cleaner story: shares outstanding have more than doubled since the September 2021 IPO, from ~290M to ~607M, as a high-SBC fintech funded growth with equity. This is the structural dilution overhang — not short selling — that weighs on a per-share thesis.

Source: shares outstanding per reported financials, FY2021–FY2025; FY2025 fully diluted count of 629M per FY2025 10-K, MD&A — Dilution [4].

4. Public short-thesis check — no campaign, but a fundamental bear case

There is no TOST-specific short-seller report, activist short campaign, accounting allegation, restatement, auditor change, or regulatory investigation in the staged research. Forensic web research surfaced only generic short-selling commentary about other names (Hindenburg, Muddy Waters, Andrew Left and similar), with nothing alleging anything about Toast. The company's own record is correspondingly clean: an unqualified audit opinion on both the financial statements and internal control, dated February 18, 2026 [5], and a legal-proceedings disclosure stating Toast is not party to any litigation that, if adverse, would be material [6].

So the bear case is not a forensic one — it is a valuation-and-business-quality case built entirely from Toast's own disclosures. The ledger below separates each thesis strand from what the filing actually says and from the company's offsetting evidence; the right-hand column is the unresolved risk a PM should price.

Source: bear strands mapped to FY2025 10-K disclosures — revenue mix, Results of Operations [7]; competition / pricing pressure, Item 1A [8]; dilution [9]; dual-class [10]; Toast Capital critical audit matter [11].

On revenue quality (strand 1). Financial technology solutions — gross-basis payment transaction fees [12] — were $5,037M of $6,153M total revenue (about 82%) in FY2025 [13], but they carry a far thinner gross margin than the subscription line, so headline revenue overstates the quality of the economics.

Source: FY2025 10-K, MD&A — Results of Operations (revenue by stream) [14]; segment gross margins derived from reported segment financials.

About 82% of revenue is the low-margin payments stream (roughly 23% gross margin) versus the ~72%-margin subscription line — the core of strand 1, and exactly the mix sensitivity a bear presses on.

On take-rate pressure (strand 2). Toast explicitly discloses that well-capitalized competitors offer discounted processing rates and lower fees, that competitive pressure may force it to maintain or lower its processing rates, and that such efforts "have negatively affected, and may continue to negatively affect, our financial performance" [15]. That is the company conceding, in its own risk factors, the mechanism behind the bear's take-rate thesis.

On governance (strand 4). Class B shares carry ten votes each; with 66M Class B shares outstanding, pre-IPO holders control roughly 55% of the vote [16], and Toast warns the dual-class structure may make it ineligible for certain indices, narrowing the passive buyer base [17].

On the one accounting soft spot (strand 5). The sole critical audit matter is the valuation of the Toast Capital contingent liability for expected credit losses, $46M at year-end 2025 — a subjective, macro-sensitive estimate flagged for its complexity, but immaterial in size [18]. It is the closest thing to a forensic angle, and it is small.

5. The bull rebuttal — why the fundamental short is contested

A short thesis on TOST has to fight the fact that the business inflected to real profitability: FY2025 net income of $342M (versus $19M in FY2024) and operating cash flow of $661M [19], on GPV of $195.1B (up 23%) and ARR of $2,047M (up 26%) [20]. Profitability, strong cash generation, a clean audit, and no material litigation are precisely the conditions under which a static short book bleeds — and likely part of why no public short campaign has formed.

Source: FY2025 10-K, MD&A — Liquidity and Capital Resources (net income and operating cash flow) [21]. FY2024 operating cash flow shown per the same statement comparison.

6. Market setup — the de-rating has already happened

Without short-interest data, the tape is the only positioning proxy, and it points the other way from a squeeze setup: TOST fell from a ~$48 high in mid-2025 to roughly $26, a drop of about 45%, with a visible step down after the FY2025 results were filed in February 2026 despite the profitability print. A name that has already halved on fundamentals, with no measured short crowding and a deep, liquid tape, has a muted squeeze setup and an asymmetry driven by fundamentals, not positioning — the catalysts that matter are take-rate and growth, not a borrow-driven unwind.

Source: staged daily price series (monthly samples, trailing 12 months, as reported); not a filing figure.

7. Evidence quality

The most important line on this page is what is not known. The table separates official-but-absent channels from inferred context and the disclosure-grounded thesis work, so a reviewer can see exactly how much weight each input bears.

Source: staged short-interest feed (all channels zero rows) and staged web research for the availability/status column; thesis and record rows per the FY2025 10-K pages cited above [22] [23].