AI-Powered Fund Outperforms Market: Phoenix Growth Fund's Playbook
Jack Hu's Phoenix Growth Fund is beating many growth peers by plugging AI directly into its investment process. The team uses Anthropic's Claude to score companies across ~80 factors, from news flow to sector momentum. Boutique Capital handles compliance, administration, and licensing so the investment team can focus on decisions.
How the AI edge works
The model compresses research time. What took Hu 15 minutes per term sheet now takes about one. That speed means more deals screened, quicker no's, and faster allocation to high-conviction ideas-without skipping diligence.
Under the hood, the framework ranks signals like momentum, management quality, cash runway, and competitive position. The output doesn't replace judgment; it sharpens it.
Clear stance on SaaS
Hu expects AI to shake up a lot of SaaS names. His filter: stick to platforms with deeply embedded workflows, regulatory friction, high switching costs, and infrastructure that customers can't live without.
Recent adds reflect that view. Pro Medicus (medical imaging software) and WiseTech Global (logistics software) both sell mission-critical systems with durable adoption hooks.
Portfolio construction: three buckets
- Bucket 1 - Equity capital market deals: New and special situations; current focus on gold, silver, and rare earths names.
- Bucket 2 - Core listed positions (ASX-heavy): The "graduates" from Bucket 1; roughly 35% of the fund.
- Bucket 3 - Alternatives: Options, managed funds, and pre-IPO exposure, including companies preparing to float like SpaceX.
This structure lets Phoenix rotate quickly, scale winners, and keep optionality without concentrating risk in a single theme.
Performance
Phoenix avoided the drawdowns that hit many growth funds. It returned 6% in the three months to January and is up 32% over the past year-22% ahead of its benchmark. The combination of AI-driven screening and a balanced portfolio design is doing the heavy lifting.
What finance teams can take from this
- Codify your investment criteria. If it can't be scored, it can't be scaled.
- Use AI as a first-pass filter to speed up no/yes decisions and widen the funnel.
- Favor businesses with structural moats: embedded workflows, regulatory barriers, and high switching costs.
- Segment the book into clear buckets (opportunistic, core, alternatives) to keep agility and manage risk.
- Outsource low-differentiation functions (compliance/admin) to protect analyst bandwidth.
Further resources
- AI for Finance - methods, tools, and use cases for investment analysis and risk.
AI isn't a silver bullet, but paired with a disciplined process, it compounds edge. Phoenix's results show what that looks like in practice.
Your membership also unlocks:
