Grace Investment Machine raises $20mn for capital markets AI

Grace Investment Machine, a Hong Kong-based artificial intelligence investing startup, has raised $20 million in Series A funding to develop agentic systems for capital markets, Finextra reported on July 10. The round was co-led by a US venture capital firm and Hony Capital, with participation from IDG Capital and existing investor Monolith Capital.

GIM is building AI systems intended to move beyond support for investment research. According to the company, its approach, which it calls a “Visionary Machine”, is designed to generate, test and refine investment hypotheses using market data, feedback loops and groups of coordinated AI agents.

The company says capital markets provide a rich environment for machine learning because investment ideas can be converted into trades or other market actions, and those actions can then be assessed through observable outcomes. In GIM’s framing, that creates a closed loop in which an AI system proposes a hypothesis, tests it against market conditions and incorporates the results into later reasoning.

Agentic AI refers to systems built to pursue defined goals through sequences of actions, rather than only responding to single prompts. In a capital markets setting, GIM says that means coordinating agents that can work across tasks such as idea generation, signal validation and the refinement of investment logic.

The startup is also developing foundation models tailored to capital-market environments. Foundation models are general-purpose AI models trained on broad datasets and adapted for specific tasks. GIM’s stated focus is on models that can operate within the data-heavy and feedback-rich conditions of financial markets.

Its multi-agent systems are being designed to produce, assess and improve investment signals across coordinated reasoning layers, according to the company. The structure suggests a workflow in which different AI agents handle distinct parts of the investment process, while the system evaluates whether signals hold up under market feedback.

“We believe investment AI is moving from information assistance to autonomous hypothesis generation and testing,” Jiahao Xu, chief executive of GIM, said. “GIM is building systems that can reason across market data, evaluate signals through feedback, and improve over time in real-world capital markets.”

The financing brings together venture and strategic capital from investors with exposure to technology and financial markets. GIM did not frame the product as a replacement for investment professionals in the reported announcement, instead presenting its systems as a new layer of AI-driven reasoning for capital-market use cases.

This story draws on original reporting from Finextra Research.