01
Observe
Begin with the phenomenon and competing explanations.
About
Disciplined observation without the theater of certainty
XG Capital Strategies is a private applied research laboratory. It works on problems where the structure that matters is buried in noise and the cost of being confidently wrong is real, in markets, in operating decisions, and in the systems built to support both.
Research philosophy
The promise is not certainty.
It is disciplined observation, explicit uncertainty, inspectable reasoning, and a serious attempt to distinguish durable structure from persuasive noise.
Most of what the laboratory runs is not published. The programs, instruments, and infrastructure behind this site are larger than the surface it exposes, and that is deliberate: work appears here only where the question, the evidence, and the boundary can all be stated. What is public is not a summary of the laboratory. It is the part of it that can be inspected.
We preserve questions, failed paths, boundaries, and nulls because a research record loses value when uncertainty is edited out.
Method / Durable sequence
02
Measure
Make source, method, state, and limitations inspectable.
03
Challenge
Seek disconfirming evidence and preserve what fails.
04
Record
Turn supported observations into durable research or applied artifacts.
Who runs the laboratory
Two people, and no one else yet.
XG Capital Strategies is deliberately staff-light. The research is built and run by the two people below. Where work is done by someone else, it is named as theirs.
-
Xander Gray
Founder
Built the research this laboratory publishes and the validation infrastructure used to test it: the pre-registration gates, the null registry, and the forward evaluation track. Author of the Ebb and Flow framework, published here as a research program rather than a product.
LinkedIn , Xander Gray (opens in a new tab) -
Alexander Jerome (Jerry) Leon
Co-founder
Designs and maintains the infrastructure the research runs on: NixOS, Proxmox, bare-metal virtualization, and the compute environments used for training, backtesting, and evaluation. Responsible for making the research systems reproducible and fault-tolerant.
LinkedIn , Alexander Jerome (Jerry) Leon (opens in a new tab)
Contact