Key Takeaways
- • Tensor Investment, an Asia-based AI-driven proprietary trading firm, launched PaladinAI, its first external product offering targeting institutions and investors in the digital asset managed futures space.
- • PaladinAI delivered a 10-month net return of 116% (January–October 2023) with a maximum drawdown of 15% and a Sharpe ratio of 3.87, positioning it as a market-neutral long-short strategy focused on medium-frequency digital asset futures trading.
- • The strategy leverages Tensor's proprietary AI research platform, powered by 120 GPUs and over 100 deep learning models, enabling real-time signal generation and execution optimization across directional trends, mean reversion, and reversal patterns.
- • Phase 1 fundraising capacity is $55 million, with the product having already secured nearly 50% of the quota from internal capital and partner commitments at launch.
- • Tensor plans to expand its product suite with AI-driven statistical arbitrage strategies while continuing investment in technology infrastructure and talent acquisition, signaling ambitions to scale its asset management operations beyond proprietary trading.
Summary
Tensor Investment, an Asia-based proprietary trading firm specializing in AI-driven strategies, announced the launch of PaladinAI, marking its entry into the institutional asset management market. The product represents the firm's first external offering designed to serve institutional investors and hedge funds seeking exposure to algorithmic trading in digital assets. This move reflects a strategic pivot from pure proprietary trading toward building a scalable asset management business.
PaladinAI is structured as a market-neutral, long-short managed futures strategy classified as "medium-frequency," trading digital asset futures with the largest liquidity pools and highest market capitalizations. The strategy targets directional market inefficiencies across multiple time horizons, capturing both trending alpha and mean reversion dynamics with a focus on superior risk-adjusted returns. Through October 2023, the 10-month track record shows a net return of 116% with a maximum drawdown of 15% and a Sharpe ratio of 3.87—metrics that position the strategy competitively within the managed futures landscape.
The product's infrastructure reflects Tensor's substantial investment in AI capabilities. PaladinAI is built on the firm's proprietary AI research platform, which deploys 120 GPUs and over 100 proprietary deep learning models for signal generation. The firm's tick-level data streaming architecture and real-time deep reinforcement learning-based parameter optimization system provide what Tensor characterizes as a real-time execution edge—a critical differentiator in medium-frequency trading where latency and model precision are paramount.
At launch, PaladinAI has captured nearly 50% of its Phase 1 fundraising target of $55 million, drawing from both internal capital and partner commitments. Tiger Li, Chief Operating Officer at Tensor, indicated that the AI research team is experimenting with additional strategies, particularly AI-driven statistical arbitrage, with plans for future product releases. The firm's stated priorities remain focused on technology and infrastructure development and expanding its talent base, suggesting a long-term strategy to establish Tensor as a leading AI-native multi-strategy asset manager.
PaladinAI is structured as a market-neutral, long-short managed futures strategy classified as "medium-frequency," trading digital asset futures with the largest liquidity pools and highest market capitalizations. The strategy targets directional market inefficiencies across multiple time horizons, capturing both trending alpha and mean reversion dynamics with a focus on superior risk-adjusted returns. Through October 2023, the 10-month track record shows a net return of 116% with a maximum drawdown of 15% and a Sharpe ratio of 3.87—metrics that position the strategy competitively within the managed futures landscape.
The product's infrastructure reflects Tensor's substantial investment in AI capabilities. PaladinAI is built on the firm's proprietary AI research platform, which deploys 120 GPUs and over 100 proprietary deep learning models for signal generation. The firm's tick-level data streaming architecture and real-time deep reinforcement learning-based parameter optimization system provide what Tensor characterizes as a real-time execution edge—a critical differentiator in medium-frequency trading where latency and model precision are paramount.
At launch, PaladinAI has captured nearly 50% of its Phase 1 fundraising target of $55 million, drawing from both internal capital and partner commitments. Tiger Li, Chief Operating Officer at Tensor, indicated that the AI research team is experimenting with additional strategies, particularly AI-driven statistical arbitrage, with plans for future product releases. The firm's stated priorities remain focused on technology and infrastructure development and expanding its talent base, suggesting a long-term strategy to establish Tensor as a leading AI-native multi-strategy asset manager.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.