Insiders Dumped 895 Million Quantum Shares With Almost No One Buying — The Cross-Stock Warning Wall Street Can't Unbundle

The September 22, 2026 tally of Form 4 filings across the four pure-play quantum bellwethers is unambiguous: insiders have dumped895.1 million more shares than they have bought since September 18, 2024. IonQ leads at 454.5 million shares net sold against just 2.25 million shares bought — a sell-to-buy ratio above 200x. D-Wave follows at 331.1 million shares sold against 1,795 shares purchased, the most lopsided ratio in the group at roughly 184,000:1. Rigetti shows 75.7 million shares sold against zero insider buying. Quantum Computing Inc. rounds out the picture with 33.8 million shares sold against zero insider buying. The data was compiled by The Motley Fool from SEC insider-transaction disclosures; the publication framed the headline figure as an \"$895 million warning,\" but the underlying metric is share count, not dollar value — and at recent prices the implied dollar exposure of insider selling is many multiples higher than $895M. What the figure actually measures is something more corrosive: at every name, the people who know the order books, the customer pipelines, and the cash runway best are choosing to be sellers, not buyers.

This is not how healthy, pre-profit growth stories behave. Insider selling in venture-stage tech is normal — early investors and founders need liquidity, RSUs vest, secondary sales are expected. What is not normal is the absence of the counter-signal: any insider with conviction that a sub-scale quantum franchise was about to inflect on bookings, on a CHIPS Act milestone, or on a hyperscaler distribution deal would buy. Across all four names, the buy side is functionally zero — IonQ's $3.32M in trailing insider buying is the only meaningful figure in the group, and even that is dwarfed by IonQ's own net selling inside a single quarter.

Long before the insider data was compiled, the tape itself had already priced the warning. On August 18, 2026 — three trading days before the Treasury's 30-year yield hit a 19-year high and the 10-year hit 4.68%, up roughly 3% in a month — IonQ fell about 6%, D-Wave dropped 6%–7%, and Rigetti fell 5%. The mechanic is structural: quantum valuations rely on deeply discounted long-duration cash flows, so a steepening yield curve hits the entire cohort with disproportionate force. That is why a single macro event — long-end yields — produced a synchronized move across stocks whose technologies are actually competing (trapped-ion vs. superconducting gate-model vs. annealing), not complements.

The August 18 move was a stress test. The September 22 insider data is what the stress test revealed: a sector where the marginal buyer had already left, and where the marginal seller is the insider. The same four names are now 52%–72% below their 52-week highs, yet every single one still commands a market cap in the billions — IonQ at $15.2B, Rigetti at $5.5B, D-Wave at $6.4B, Quantum Computing Inc. at $2.0B — on combined TTM revenue of ~$282M. The price-to-sales ratios (61x for IonQ, 411x for Rigetti, 519x for D-Wave, 207x for QUBT) only make sense inside a multi-year procurement-grade adoption story. The insider data undermines exactly that thesis.

The Q2 2026 print, filed in early August, made the gap between narrative and cash flow impossible to ignore. IonQ reported record revenue of $80.1M, up 287% year-over-year — the strongest top-line print in the cohort — but a GAAP net loss of $1.87 billion driven by a $1.6 billion non-cash mark-to-market charge on warrant liabilities. Rigetti grew revenue 185% YoY to $5.1M but still posted a $28.1M operating loss on $541.3M of cash. D-Wave held revenue essentially flat year-over-year at $3.1M and saw its operating loss more than double, from $26.5M in Q2 2025 to $54.7M in Q2 2026, even as bookings for the first half surged 1,120% year-over-year to $35.5M. Quantum Computing Inc. reported $5.6M of Q2 revenue against an operating loss of $23.0M, and remains a software aggregator whose Qatalyst platform routes workloads to D-Wave, IonQ, and Rigetti QPUs.

The mismatch is the message. D-Wave trades at 519x TTM sales despite revenue going sideways and operating losses widening; Rigetti is at 411x sales with only $5.1M of quarterly revenue; Quantum Computing Inc. is at 207x with no path to operating leverage in sight. Even the best-positioned of the four, IonQ, is at 62x sales — a multiple that prices in roughly 100 quarters of current revenue with no margin. Insider selling at those multiples is not a signal of fraud or operational failure. It is a signal that the people running these companies believe the public-market valuation has decoupled from the operating trajectory they can see.

The structural bull case for the cohort has shifted from \"software margin in a new market,\" not from \"operating leverage on revenue.\" On September 8, 2026, the U.S. Department of Commerce finalized a definitive agreement with D-Wave for up to $100M in CHIPS and Science Act funding, structured as an R&D award that will support next-generation systems including a planned 100,000-qubit machine. Rigetti and Quantinuum each locked in parallel $100M CHIPS Act awards on the same date. These awards fund U.S. quantum foundries — the Department named IBM and GlobalFoundries as quantum-foundry partners — and they fund the seven U.S. quantum hardware companies including D-Wave and Rigetti.

The most consequential vertical move in the sector is IonQ's July 31, 2026 closing of its acquisition of SkyWater Technology, the largest U.S.-headquartered semiconductor foundry. The deal was framed by IonQ as \"securing a fully scalable supply chain domestically\" for its trapped-ion roadmap — IonQ had previously relied on a third-party foundry for advanced packaging of its quantum chips. With SkyWater inside the consolidated entity, IonQ has effectively removed a foundry-bottleneck risk that had constrained the cadence of its IonQ Tempo deployments. For Rigetti, which builds superconducting qubits in its own Fab-1 facility but uses external atomic-layer-etch equipment from Oxford Instruments, the CHIPS Act funding de-risks the capex line on the next-generation84-qubit and 108-qubit systems.

The cross-stock warning has a clear spillover: the hyperscaler cloud platforms — Microsoft Azure Quantum, Amazon AWS Braket, and Alphabet Cloud Marketplace — distribute access to all three QPU makers' hardware as a managed service. Their quantum revenue is bundled inside multi-billion-dollar cloud contracts and never separately disclosed. The implication is that the marginal institutional buyer of the underlying quantum hardware exposure is increasingly the hyperscaler ecosystem, not the public equity holder. The pure-plays are the optionality wrapper, not the cash-flow engine. When insiders across all four pure-plays are net selling simultaneously, it is because they are reading the same hyperscaler-roadmap signal: the public-market wrapper is rich, but the procurement-grade thesis that justifies the wrapper is still years from monetization.

The second spillover runs into AI infrastructure. NVIDIA NVQLink now integrates with Rigetti's control systems, positioning Rigetti as a quantum-classical co-processor inside NVIDIA's CUDA-Q hybrid stack. That integration is the most concrete industrial proof-of-concept in the cohort — a path from superconducting qubit to GPU-accelerated inference — but it does not move the P&L in 2026. The 2026 revenue impact at Rigetti is still measured in single-digit millions per quarter. The fact that insiders are selling despite this partnership is the warning's second layer: industrial progress is real, but the public-market multiple has already priced a much larger revenue trajectory than the partnerships can deliver inside a1–3 year window.

The next 90 days will be decided by three catalysts, in this order of importance. First, the Q3 2026 prints: D-Wave, IonQ, Rigetti, and Quantum Computing Inc. are all expected to report Q3 in early November, and any sequential contraction in bookings — particularly after D-Wave's 1,120% H1 bookings surge — will be read as a direct falsification of the procurement-grade thesis. Second, the long end of the Treasury curve: the August 18 selloff showed that quantum stocks are the highest-beta exposure to long-end yields. A retest of the 30-year high near 4.80% would re-trigger the cross-stock move. Third, the buy-side: any insider Form 4 with a purchase above $1M at any of the four names would meaningfully reset the warning. As of the September 22 data set, insider buying is functionally zero — the asymmetry is the signal.

Over a 1- to 3-year horizon, the cohort splits into two trajectories. The industrial trajectory is constructive: IonQ has the trapped-ion performance lead and now owns the foundry; Rigetti is integrated into the NVIDIA CUDA-Q stack; D-Wave is the only public pure-play in the annealing niche with a real customer base and a CHIPS Act line item for a 100,000-qubit system. The financial trajectory is more fragile. At current TTM revenue, every name needs roughly 10x revenue growth to justify a 50x P/S multiple — and even then, the market cap would only hold if growth does not decelerate. Insider selling at this scale is the people closest to the operating data voting against the financial trajectory. A re-rating is the base case over1–3 years; the question is whether industrial progress (CHIPS-funded100,000-qubit systems, hyperscaler distribution at scale) arrives fast enough to defend the market caps that exist today.