FirstQFM's Quantum Forecasting System Achieves Major Success Against Classical Models in Financial Evaluation

FirstQFM's Quantum Forecasting Breakthrough

FirstQFM, a leading name in the realm of quantum technology, has recently made headlines with the impressive performance of its foundation-model-based quantum reservoir computing (QRC) system. This system was evaluated during a financial forecasting assessment utilizing Rigetti Computing's advanced Cepheus™-1-108Q system. The results were nothing short of revolutionary, as FirstQFM’s system outshined various classical benchmarks across multiple metrics, including directional accuracy, macro-F1 score, gross portfolio profit and loss (PL), and the annualized Sharpe ratio.

A Detailed Evaluation

The financial forecasting evaluation examined next-day predictions across a diverse range of 43 anonymized assets and 600 out-of-sample periods. The outcomes revealed that FirstQFM's quantum QRC system achieved an impressive 52.02% directional accuracy and a 51.76% macro-F1 score, marking the highest results among all models tested.

In a portfolio equally weighted across all 43 assets, FirstQFM’s forecasting system generated a staggering 112% higher cumulative gross PL compared to the strongest classical benchmark. Additionally, it boasted an annualized Sharpe ratio of 1.22, surpassing the classical benchmark's Sharpe ratio of 1.02. Notably, for individual assets, the system exhibited over 50% directional accuracy for 36 out of 43 assets and produced positive gross PL for the majority, achieving a favorable outcome for 41 of the 43 assets.

Dr. Subodh Kulkarni, CEO of Rigetti Computing, stated that empirical work like that carried out by FirstQFM provides crucial insights into the capabilities and limitations of current quantum computers. This work aids hardware teams in identifying which enhancements are most impactful. As we venture into an era of fault tolerance, there will be a growing need for investment in algorithms, applications, and the essential tools and skills to leverage these advancements effectively, transforming capable machines into valuable assets for industries.

A Comparative Study

The evaluation contrasted FirstQFM's system with classical machine learning models, statistical models, and foundation models including Chronos-2 from Amazon, Granite TTM-R2 developed by IBM, and Toto 2.0 from Datadog. The SW7 Group, contributing by anonymizing assets and the forecasting tasks, facilitated this groundbreaking analysis. The methodology employed chronological training, validation, and testing phases, with model selection solely based on validation data. The features harnessed in FirstQFM’s quantum endeavors were generated using Rigetti’s quantum hardware, demonstrating the seamless integration of cutting-edge technology and strategic foresight.

Moreover, a focused five-asset portfolio utilizing FirstQFM's system—selected through a historical diversification methodology—delivered superior directional accuracy, macro-F1, and cumulative gross PL than the broader 43-asset portfolio, showcasing an annualized Sharpe ratio of 1.73.

The in-depth methodology and comprehensive results of this groundbreaking evaluation can be found in the technical blog provided by FirstQFM at FirstQFM's Technical Blog.

Future Implications

Oleksiy Kondratyev, Head of Risk at SW7 Group, emphasized the significance of developing innovative solutions for complex quant finance challenges. The remarkable results underscore the potential of QRC to formulate viable quantum machine learning solutions that can surpass traditional models.

FirstQFM launched the beta version of its foundation-model-based QRC forecasting system in July 2026. Developed atop proprietary foundation model technology, this system operates at a scale of one billion parameters and caters to various applications across verticals like asset management, banking, healthcare, predictive maintenance, and energy.

Vish Ramakrishnan, CEO of FirstQFM, remarked on the strategic advantages of their foundation model-based QRC solution. By leveraging patented foundation models, they deliver superior performance compared to classical time series models, providing a utility-scale, commercially viable solution that demonstrates tangible value for enterprise clients.

About the Organizations

SW7 Group is a London-based deep technology research firm dedicated to developing innovative AI technologies and their practical applications in fields like finance and medicine. Founded in 2014 by former Goldman Sachs partners, SW7 Group aims to leverage technology transfers across various domains.

FirstQFM, established in Stockholm, specializes in creating machine learning foundation models that enhance both performance and scalability of quantum computing. By crafting models tuned to the nuances of specific quantum processors, FirstQFM paves the way for the creation and implementation of high-performance quantum applications on NISQ (Noisy Intermediate-Scale Quantum) hardware and beyond. The core technology of FirstQFM is protected by granted and pending patents, ensuring a competitive edge in the evolving tech landscape.