About the Academy

The Academy exists for a simple reason: there is no shortcut to understanding markets, but there are better ways to practice the work.

A training ground, not a marketing channel.

ABX Capital Partners Academy is structured as an internal-first, education-focused environment. It is designed to help students, early-career professionals, and technically curious people build a durable toolkit around data, code, and markets.

The Academy draws on perspectives from quantitative research, discretionary trading, and engineering— including work carried out within ABX Capital Partners—but it is intentionally separated from any client-facing activity. The goal is clarity, not distribution.

Abstract illustration of structured learning paths

How the Academy is structured

The Academy combines three elements: curriculum tracks, a small research program, and limited-capacity mentorship. Each part is designed to reinforce the others.

Curriculum tracks

Multi-week sequences that cover probability, statistics, market microstructure, systematic thinking, and practical AI tooling for research workflows.

Research notes

Short, transparent write-ups that document data sets, methods, and limitations rather than "calls" or predictions. Many are inspired by questions that arise in the classroom.

Mentorship

Small, time-boxed mentorship pods focused on reading lists, project feedback, and career questions around quantitative and technology-driven roles.

Influences and philosophy.

The Academy's approach is informed by quantitative asset management, venture investing in financial infrastructure, and hands-on engineering experience—similar to the blend of perspectives that drives ABX Capital Partners Ventures.

That means treating markets as complex systems, zooming out to understand incentives and structure, then zooming back in to design concrete experiments, models, and workflows.

The tone is rigorous but practical: enough theory to avoid category errors, enough pragmatism to recognize all the ways real-world data refuses to fit perfectly inside a model.