Human Judgment Enhanced by Data

Technology accelerates discovery, but discipline and patience guide the decision. We combine machine intelligence with human judgment to act only when conviction is highest.

The Limits of Pure Automation

Fully automated investment strategies — algorithmic trading, quantitative funds, systematic approaches — have a fundamental limitation: they can only capture patterns that existed in historical data. When markets behave in genuinely novel ways, algorithms that were trained on historical data can fail catastrophically.

The best investment decisions require contextual judgment that no algorithm currently possesses. Understanding why a business has a competitive moat, whether management is trustworthy, whether a regulatory change will be temporary or permanent — these assessments require human judgment.

The Limits of Pure Human Judgment

At the same time, pure human judgment has well-documented weaknesses. Cognitive biases — overconfidence, recency bias, anchoring, loss aversion — systematically distort human decision-making. And no human analyst can efficiently process the volume of data that is now available about public companies.

The Combination

We have designed our process to use technology where it has clear advantages — processing large volumes of data, identifying statistical anomalies, back-testing hypotheses — and human judgment where it has clear advantages — assessing qualitative factors, evaluating management, understanding competitive dynamics.

Technology accelerates discovery. It helps us find interesting ideas faster and more comprehensively than we could without it. But every investment decision is ultimately made by humans who have applied rigorous fundamental analysis and who have genuine conviction in the outcome.

We act only when both the quantitative signals and the qualitative assessment align. High conviction requires agreement between machine and mind.