Sector Context Before Single-Stock Conviction
Before concluding a company is mispriced, rule out the simpler explanation: everything in its industry is moving together.
May 7, 2026
Research
Our thinking, published openly — how we analyse companies, how we read the market, and what we believe investors most often get wrong.
Before concluding a company is mispriced, rule out the simpler explanation: everything in its industry is moving together.
May 7, 2026

Research that always produces a recommendation is not research. The discipline to conclude 'we don't know' is a feature.
August 9, 2023

You can be right about most of your names and still lose money. How much you own is usually the more consequential decision.
February 17, 2021

Screens are excellent at narrowing a universe and useless at judgement. Knowing the boundary is most of the skill.
November 12, 2018

A 300-page filing has maybe 20 pages that change your mind. Here is a practical order of operations for finding them.
March 8, 2016

The five filters every company clears before it earns a place in our research coverage — and why most names fail at filter two.
September 15, 2014