Heartisans

Heartisans

Founded in 2014 by data scientist and quantitative investor Timothy Li, Heartisans Limited operates from Hong Kong as a dedicated systematic fund. The firm emerged during a period when machine-learning applications were still nascent within Asia's hedge fund community, giving it a rare, multi-year head start on the region's algorithmic learning curve. Li structured Heartisans to operate less as a discretionary platform with quant overlays and more as a pure-play AI-native manager, training models on market microstructure data across multiple regional exchanges. The firm runs liquid equity market-neutral and directional strategies focused primarily on developed and emerging Asian markets, including Japan, Australia, Hong Kong, and mainland China share classes. Heartisans' core intellectual property sits in its proprietary deep-learning models, which process order-book dynamics and alternative data signals to forecast short-to-medium-term price movements. While the firm has not publicly disclosed its full track record, its concentrated emphasis on applying machine learning at every layer of the research stack sets it apart from quant funds that rely on traditional statistical factor models. Heartisans maintains a lean operational profile typical of early-stage systematic managers, with investment and engineering talent concentrated in its Hong Kong headquarters. The firm has not disclosed external capital raises or headcount changes, though its durable presence across multiple market cycles suggests a stable capital base. No adjacent vehicles, philanthropic structures, or co-investment club affiliations have been publicly reported, reflecting a tightly held, strategy-first organizational design. Structurally, Heartisans represents a small but distinct cohort of Asia-based AI-native hedge funds that compete for talent and capital primarily against larger global quants such as Two Sigma or WorldQuant. Its differentiator lies in a region-specific model-building culture: training models on fragmented Asian market data with local market-access infrastructure. That focus creates a narrower, deeper moat than a multi-region algorithmic generalist could likely maintain with equivalent resources.