The problem with reading about quantum
If you ask a venture capitalist what quantum computing is in 2026, you get one of two answers. The first is breathless: a generational technology that will break encryption, cure cancer, and produce trillion-dollar companies inside a decade. The second is dismissive: a science project that has been a decade away from useful for twenty years and will probably stay that way. Both are wrong in approximately the same ways, and the gap between them is where money actually gets made or lost.
The problem is that the two camps are reading different inputs. The bulls are reading press releases and stock charts. The bears are reading academic critiques of specific advantage claims. Neither is reading the thing that matters most: the small set of variables that actually determine which of these companies becomes a real business. That set is short, it is mostly knowable, and it is mostly ignored.
This essay is the framework I use to evaluate every quantum company. It has four layers. Skim it once and you have a vocabulary that survives the next press cycle. Read it twice and you can tell a real quantum business from a marketing budget with a cryostat in the lobby.
Layer one: separate the four products in a single field
"Quantum" in 2026 is four different commercial categories with four different time horizons. Conflating them is the single biggest analytical mistake in the public discourse.
1. Quantum-safe cryptography (PQC). Already revenue.
NIST standardized ML-KEM, ML-DSA, and SLH-DSA in August 2024. Apple, Cloudflare, Google, Mozilla, and AWS have already migrated meaningful portions of their TLS handshakes to hybrid post-quantum schemes. As of late 2025, roughly 52% of inbound TLS traffic to Cloudflare uses a post-quantum hybrid handshake — the largest deployment of new cryptography to consumers in two decades. McKinsey's mid case puts cumulative PQC services spend at $40-50B through 2035.
None of this requires a working quantum computer. The driver is harvest-now-decrypt-later — adversaries are presumed to be capturing encrypted traffic today and betting on future quantum computers to decrypt it. Any data with a secrecy lifetime longer than the time to a cryptographically-relevant quantum computer is exposed. PQC is the only quantum category where the customer doesn't need to believe in quantum computers to buy. The migration tracker shows what is shipped, what is committed, and — the more interesting list — who is silent.
2. Quantum sensing and PNT. Near revenue.
Atomic clocks, optically-pumped magnetometers, quantum gravimeters, and atom-interferometer inertial sensors are commercial products today. Royal Navy submarines use a Q-CTRL quantum clock. Cerca Magnetics ships optically-pumped magnetometers for MEG brain imaging. Quantum gravimeter prototypes are deployed in oil & gas surveys and UK civil engineering. The combined market is probably $1-2B annually today, with defense (AUKUS Pillar 2) and healthcare (MEG) as the two scaling verticals. The technology risk is much lower than quantum computing because each product is a single-purpose device, not a general-purpose computer.
3. Quantum networking and QKD. Niche revenue, growing.
Quantum key distribution is operational on China's national backbone, in EuroQCI pilot links, between Swiss banks, and across Korean telco networks. The market is small ($1-2B annually) but it is one of the few categories where actual quantum hardware (single-photon sources, detectors, polarization analyzers) generates real recurring revenue. The NSA explicitly prefers PQC over QKD for U.S. national-security systems, citing engineering challenges (dedicated fiber, distance limits, authenticated classical channel requirements). For non-NSS contexts and defense-in-depth postures, QKD remains viable. The long-term direction is device-independent QKD certified by Bell-inequality violation — the same physics our Bell test demo illustrates.
4. Quantum computing (the gate model). Pre-revenue at scale.
This is the category every quantum stock and venture round is about. The category where Quantinuum just filed its S-1 at $20B against $30.9M of trailing revenue. The category where IBM, Google, Microsoft, and PsiQuantum are competing for FTQC leadership. And — being honest — the category where 95% of the marketing budget is spent.
The point of separating these four products is that each requires a different valuation lens. PQC vendors trade on services revenue and customer migration timelines. Sensing companies trade on hardware unit economics and defense procurement cycles. QKD vendors trade on government infrastructure spend. Gate-model quantum computing trades on a 2029-2033 fault-tolerance milestone discount rate. A portfolio that doesn't separate them is mispricing every name in the basket.
Layer two: read the only benchmark that matters
Inside the gate-model category, every company will tell you about its qubit count, its gate fidelity, its quantum volume, its algorithmic qubits, its connectivity, its roadmap. Most of this is real, most of it is hype, and exactly one number summarizes the real signal: has any independent technical body looked at the architecture and said yes?
DARPA's Quantum Benchmarking Initiative is the single most useful filter the public has on this category. The program runs an IV&V team described by its founder as "almost certainly the world's best quantum computing test and evaluation team." That team has now scored ~20 companies, advanced 11 to Stage B, and selected two — Microsoft and PsiQuantum — for the deeper US2QC final phase. They have no equity exposure and no commercial conflict. Their answer is the closest thing to truth available outside an NDA.
Stage A advancement is a $1M technical exam. Stage B advancement is $15M and a 12-month engineering audit. US2QC final phase is the architectural blessing. Each step up is a real signal. Each step down is a real signal too — Rigetti participated in Stage A and did not advance to Stage B. The market still trades Rigetti at ~890× sales. The gap between those two facts is the kind of inefficiency this sector produces, and the kind of inefficiency you can underwrite against.
The DARPA QBI tracker reads the program continuously. The short version: if a company is not in QBI Stage B or US2QC final, treat their roadmap as marketing until proven otherwise. If they are, the roadmap is at least credible enough that a non-conflicted government technical team thinks it can work.
Layer three: the four-dimensional read
Once you have separated the four products and indexed against DARPA validation, the remaining work is portfolio construction. The framework I publish in the Ledger Score scores every meaningful company on four dimensions, each 0–100.
Technology (30% weight). Best published 2-qubit gate fidelity, number of logical qubits demonstrated, and architecture credibility. Quantinuum currently scores 92 on this dimension — 99.921% fidelity on Helios, 48 logical qubits live, the most credible Apollo 2029 fault-tolerance commitment. Rigetti scores 52. The 40-point gap is the structural read.
Capital (20% weight). Runway, valuation momentum, market access. This is where mega-cap parents (Google, Microsoft, IBM) get full marks regardless of unit economics, and where Rigetti — for all its strategic weakness — gets credit for $590M of cash. Capital is the dimension that buys time. In a sector where time-to-FTQC is the binary variable, having time is non-trivial.
Commercial (30% weight). TTM revenue, growth, customer mix, concentration risk. This is where the public cohort sorts. IonQ at $130M, growing 200%+, with diverse customers, scores 82. Quantinuum at $30.9M with 60% RIKEN concentration scores 64 — strong revenue, structural concentration penalty. Rigetti at $7M with declining revenue scores 18. The commercial score is the dimension that most often diverges from stock price, and the divergence is usually the trade.
Government validation (20% weight). DARPA QBI stage, US2QC, allied programs, named contracts. PsiQuantum and Microsoft score 95 and 92 here — US2QC final phase is the deepest validation available. Rigetti at 25 reflects Stage A elimination. This dimension is the closest thing to an outside opinion the sector produces.
The composite score is opinionated by construction. The point of publishing the methodology is so you can disagree with the weights, see the inputs, and reach your own number. The methodology page documents every rubric.
Layer four: the five questions that survive every cycle
When a quantum announcement crosses the wire — a new chip, a new round, an IPO filing, a benchmark result — there are five questions that strip the news down to the part that actually moves the model.
- Which of the four product categories does this affect? A photonic chip announcement affects FTQC and possibly networking; it does not affect PQC. A NIST standard affects PQC; it does not affect FTQC roadmaps. Match the news to the category before pricing it in.
- Does it move the DARPA validation signal? If the news is not in or adjacent to QBI / US2QC / NIST / DOE QIS programs, it is marketing until proven otherwise. If it is, weight it heavily.
- Which Ledger Score dimension does it change, and by how much? A new gate-fidelity record affects technology, possibly capital (valuation step-up). It does not affect commercial. A new customer affects commercial, possibly capital. A new round affects capital. Map the news to the dimension before assuming systemic impact.
- What is the read-through for the other names? Most quantum news affects the cohort directionally. A negative quarter from Rigetti is positive for IonQ and Quantinuum because it reduces the available capital pool to a structurally weaker peer. A positive Stage C advance for PsiQuantum is positive for the FTQC-2029 cohort thesis broadly. The read-through is where most of the trade lives.
- What is the calendar implication? Has the news pulled forward a milestone (Apollo 2029, Starling 2029, Stage C decisions Q4 2026), pushed it back, or revealed a new milestone? Quantum stocks discount future dates aggressively — a six-month pull-forward is a meaningful multiple expansion event.
Run any quantum announcement through those five questions and the part of the news that matters survives. The rest is press-release texture.
Where the framework leads in 2026
Apply the framework to the current cohort and a few conclusions emerge.
The IPO matters. Quantinuum's S-1 pricing in June 2026 is the single most consequential pricing event for the entire public cohort. A $20B price either drags IonQ, D-Wave, Rigetti, and the rest up by re-rating the comp set, or becomes the ceiling everyone trades against. The Ledger Score composite suggests QNT should price between IonQ's implied multiple and Rigetti's — somewhere around $12-18B. The IPO watch tracks pricing updates daily.
The Q4 2026 DARPA decision is the next discrete event. Stage C advancement from QBI Stage B will further consolidate the credibility hierarchy. Companies that advance gain Ledger Score points (and probably valuation step-ups); companies that do not, lose ground. The market is currently not pricing this discretely.
The 2029 FTQC convergence is the long-cycle bet. IBM, Quantinuum, and Xanadu have all publicly targeted 2029 fault-tolerance milestones. Three independent architectures converging on the same year is the most credible long-term signal in the sector. The bear case is none of them hit. The bull case is two of three do, in which case the public cohort is meaningfully under-priced today. The Ledger Score framework lets you bet on the convergence without picking the winner.
The misclassification trades are still open. The market continues to price Rigetti and Arqit as quantum-computing exposure when they are, respectively, a sub-scale gate-model vendor and a quantum-safe cryptography services company. Both trade at multiples that reflect category confusion. The Ledger Score is structured specifically to surface these misclassifications.
Why this site exists
The Quantum Insider does breadth. Quantum Computing Report does technical depth. Pitchbook does private-company data behind a paywall. Nobody does a published, opinionated, financially-literate synthesis with a transparent scoring framework. I built Quantum Ledger because that was the resource I wanted and couldn't find.
Every page on the site is meant to defend the framework above. The Ledger Score ranks every company. The earnings tracker updates the commercial score every quarter. The QBI tracker updates the government score after every stage announcement. The IPO watch tracks the swing event. The PQC migration tracker covers the first-product category. The companies section is the per-name due diligence. The primer and glossary are the vocabulary.
If the framework holds up to a year of news, the site becomes useful. If it doesn't, I update it publicly. The methodology page is the contract. Disagree productively, disagree often, push back on the weights. The site improves when the framework is pressure-tested.
Connor Reuter runs Quantum Ledger. He is an investor at Caruso Ventures, a Single Family Office, and writes about quantum sector intelligence for a living. Comments and disagreements welcome at the address on the about page.