Contrarian Corpus
short seller research note initial thesis
2025-04-01 · 19 pages

D-Wave Quantum Inc QBTS

D-Wave's quantum annealing is a commercial dead end, its 'hybrid' solvers are almost entirely classical, and at 57x 2026E revenue the stock will collapse as reality sets in.

N 4 Narrative
V 3 Visual
C 2 Craft
Source URL unavailable

Thesis

Kerrisdale is short D-Wave Quantum (QBTS), a $2.3 billion quantum annealing company whose stock has surged over 600% on next-gen computing hype despite never generating more than $9 million in annual recurring revenue. Academic studies and customer interviews (Airbus, BASF, GlaxoSmithKline, Unisys, Amazon) consistently show D-Wave's annealer has no proven scaling advantage over classical solvers like Gurobi or CPLEX, while former engineers describe its 'hybrid' offering as 'almost entirely classical' with quantum as marketing veneer. D-Wave's 2021 pivot to gate-model quantum has produced no architecture papers or benchmarks three years later, and its recent 'quantum supremacy' claim rests on a toy spin-glass lattice with no industrial relevance. At 57x inflated 2026E consensus revenue (152x Kerrisdale's estimate), the valuation will unwind as customers defect and physics catches up with the narrative.

SCQA

Situation

D-Wave Quantum (QBTS) is a $2.3bn publicly traded quantum annealing company whose stock rallied 600% on broader quantum-computing enthusiasm, despite pursuing a niche annealing architecture distinct from gate-model leaders IBM and Google.

Complication

Quantum annealing has no proven scaling advantage over classical solvers, D-Wave's 'hybrid' offering is almost entirely classical per former engineers, its gate-model pivot is stalled with zero benchmarks, and customers in key verticals report zero benefit.

Resolution

Investors should short QBTS — Kerrisdale discloses an active short position — betting the stock unwinds as customers defect, physics reasserts itself, and the gap between commercial reality and quantum hype narrative closes.

Reward

Shares trade at 57x 2026E consensus revenue (152x Kerrisdale's estimate) on $9m of ARR with stalled growth and persistent cash burn; a rerating to defensible fundamentals implies very large downside.

The three reasons

  1. 1

    Quantum annealing shows no scaling advantage over classical solvers in peer-reviewed studies

  2. 2

    D-Wave's 'hybrid' solvers are 'almost entirely classical' per former insiders

  3. 3

    QBTS trades at 57x 2026E revenue on just $9m ARR with stalled customer growth

KPIs cited

Stock performance
Up over 600% since gate-model quantum hype reignited last year
Market capitalization
$2,311m (307m fully-diluted shares at $7.53)
Annual recurring revenue
Never more than $9m; 2024A total revenue $9m
EV / Sales multiple
229x 2024A, 92x 2025E, 152x 2026E, 137x 2027E
Consensus 2026E revenue multiple
57x consensus (152x Kerrisdale's estimate)
Revenue growth y/y
22% 2023A, 1% 2024A, 150% 2025E, -39% 2026E per Kerrisdale forecast
Free cash flow
Adj. EBITDA less capex losses of ($49m)–($71m) from 2022A through 2027E
Pattison Food Group scheduling efficiency
Reported 80% improvement was vs. an Excel baseline, not a modern classical solver

Pattern membership

Where this document fits across the library's 12 rhetorical / structural patterns.

Precedents cited

  • Microsoft's overhyped Majorana qubit announcement
  • D-Wave's prior 'carefully selected benchmark' speedup claims that failed under scrutiny

Notable slides (6)

Notes

Prose-style short research note (not a slide deck) — Kerrisdale's standard format: cover with executive-summary paragraphs, financial summary table on p.4, extensive verbatim quotes from former D-Wave engineers and customers (Airbus, BASF, GSK, Unisys, Amazon Bracket) forming the rhetorical core. Quotes used as primary-source evidence rather than CEO-vs-CEO contradictions. No named individual villain — critique aimed at the company and its marketing narrative. No stake size disclosed, only that Kerrisdale 'stands to realize gains' if the stock falls. No explicit target price — downside is implied via peer multiples and Kerrisdale's own 152x estimate vs consensus 57x.