TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
FirstQFM reported that a quantum reservoir computing system running on Rigetti hardware outperformed the classical models in its evaluation of next-day forecasts for 43 anonymized assets across 600 out-of-sample periods. The company reported 52.02% directional accuracy and 112% higher cumulative gross portfolio profit than the strongest classical benchmark by that measure; independent replication and further detail on the evaluation remain open questions.
FirstQFM said its quantum reservoir computing system, using Rigetti Computing hardware, outperformed the classical models tested in an evaluation of next-day forecasts for 43 anonymized assets. The company reported 52.02% directional accuracy and 112% higher cumulative gross portfolio profit than the strongest classical benchmark by profit across 600 out-of-sample periods; the findings are company-reported results, not evidence here of independent replication.
FirstQFM said its system produced a 51.76% macro-F1 score, also the highest among the models in the comparison. For an equally weighted portfolio of all 43 assets, it reported a 1.22 annualized Sharpe ratio, compared with 1.02 for the classical model with the highest Sharpe ratio. The 112% profit comparison refers to cumulative gross portfolio P&L against the strongest classical benchmark by P&L, not to a net return after costs.
The company said quantum features for the system were generated on Rigetti’s Cepheus-1-108Q system. Its evaluation included classical machine-learning and statistical models, as well as time-series foundation models identified as Chronos-2 from Amazon, Granite TTM-R2 from IBM and Toto 2.0 from Datadog. SW7 Group anonymized the assets and forecasting task before providing the data to FirstQFM.
FirstQFM said the data was divided into chronological training, validation and test periods and that model selection relied only on validation data. It also reported that directional accuracy exceeded 50% for 36 of 43 assets, while gross P&L was positive for 41. A separately described five-asset portfolio, selected using a historical diversification method, had a reported Sharpe ratio of 1.73 and higher directional accuracy, macro-F1 and cumulative gross P&L than the full portfolio.
What the Portfolio Results Show
The results matter because they offer a test of whether quantum computing can contribute to a practical forecasting task, rather than only a laboratory benchmark. If the reported performance holds up under scrutiny and across other periods, it could draw interest from financial firms assessing quantum machine learning for asset forecasting and portfolio research.
The figures do not establish that the system is ready for live trading or that it would produce comparable investor returns. The reported portfolio results are gross of costs, and the supplied material does not establish how they would change after transaction costs, market impact, risk limits or operational constraints. Nor does one evaluation show that a quantum approach will consistently beat classical methods across markets and time periods.
Products Worth Considering
Apple Watch Series 8 GPS, 41mm, Midnight Aluminum Case with Midnight Sport Band (S/M)
Apple Watch Series 12 GPS 46mm Smartwatch
Samsung Odyssey G4 25-inch FHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How FirstQFM Ran the Test
The comparison focused on next-day forecasts across 43 anonymized assets and 600 out-of-sample periods. Chronological splits are relevant to forecasting because they preserve the order of time, while using validation data for model selection separates that process from the final test periods, according to FirstQFM’s description.
FirstQFM released a beta version of its foundation-model-based quantum reservoir system in July 2026. The company describes its underlying technology as proprietary and says it is designed for use in areas including asset management, banking, healthcare, predictive maintenance and energy. Those broader commercial applications are company plans and claims; the results described here concern a financial forecasting evaluation.
The announcement included comments from Rigetti and SW7 Group, both involved in the work. Rigetti supplied the quantum hardware, while SW7 Group provided anonymized data. Their participation helps explain how the test was assembled but does not, by itself, amount to an independent audit of the reported performance.
“Empirical work such as FirstQFM’s helps us probe the capabilities and limits of today’s quantum computers, while helping hardware teams understand which improvements matter most.”
— Subodh Kulkarni, Rigetti Computing chief executive
Limits of the Reported Evidence
The announcement and accompanying material are from FirstQFM; the supplied information does not describe an independent replication or review. The source says detailed methods and results are available in a technical blog, but the announcement alone does not provide enough information to assess all modeling choices, the precise construction of every benchmark or the statistical uncertainty around the reported scores.
It is also unclear from the supplied material how the gross P&L figures account for trading costs, whether the evaluation covers market conditions beyond the reported periods, or how sensitive results are to the selection of assets and portfolio methodology. The report does not establish live performance, net investor returns or a general advantage for quantum systems over classical forecasting. Those questions would require fuller methodological detail and testing beyond the reported evaluation.
What Independent Testing Could Establish
FirstQFM has published a technical blog with its detailed methodology and results, according to the announcement. The next useful step for readers and potential users is to examine those materials for benchmark definitions, portfolio construction, test-period details and treatment of costs. Further evaluation on separate data and periods would help clarify whether the reported advantage persists outside this test.
No independent replication, live-trading result or follow-up milestone was specified in the supplied material. Until further evidence is available, the reported figures should be read as FirstQFM’s results from a defined evaluation, not as a settled comparison of quantum and classical forecasting systems.
Key Questions
What did FirstQFM report?
FirstQFM said its quantum reservoir computing system achieved 52.02% directional accuracy and 51.76% macro-F1 in next-day forecasts for 43 anonymized assets. It also reported 112% higher cumulative gross portfolio P&L than the strongest classical benchmark by profit.
Was the quantum system tested on real hardware?
FirstQFM said the system’s quantum features were generated using Rigetti Computing’s Cepheus-1-108Q hardware. The announcement does not describe this as a live financial trading test.
Does the result prove quantum forecasting is better?
No. It reports an advantage in this particular evaluation, but the supplied material does not establish independent replication or consistent performance across other datasets, periods and market conditions.
Were the reported portfolio profits net of trading costs?
The announcement describes the portfolio P&L as gross. It does not provide net returns after transaction costs or other trading expenses in the supplied material.
What happens next?
FirstQFM says its technical blog provides further methodology and results. Independent testing on separate data, along with clearer reporting on costs and portfolio assumptions, could help determine how broadly the findings apply.
Source: thequantuminsider.com
