AI Score
AI Score is a stock rating system. Every company gets a single score from -100 to 100 that summarizes how attractive that business looks based on a large set of signals around value, quality, growth, financial strength, management, risk, and more.
The goal is simple: compress a complex investment picture into a single number you can use to quickly evaluate a company's potential reward and risk. Higher scores point to companies with stronger fundamentals and better outcomes. Lower scores point to weaker businesses, higher risk, overpriced valuations, or less attractive setups.
Sample AI Score
excellent
Reward
Safety
Our study on AI Score
Higher AI Scores correlates with better results
We grouped +10,000 stocks into score buckets and measured their performance over time in a roughly 50-year out-sample period. We used a 2-year holding period for each position, and the chart shows the avg. annual return and win ratio for each bucket.
The correlation is clear: Higher AI scores consistently line up with better results.
The AI score is designed to find companies with great fundamentals that are undervalued. However lower scores does not mean a company is necesarilly bad. A healthy company even at a reasonable price could get a low score if the margin of safety is not significant. In our study we found that even in the worst bucket between -100 and -90, 48% of the investments were still sold with a profit, and overall profitable buckets start appearing at an AI score of -60 to -50.
The available pool of high scoring companies to invest in also highly varies on market conditions, when the market is overvalued there will be fewer or sometimes no companies that receive a +90 AI score.
*Past performance does not guarantee future results. Historical studies are meant to show signal quality and correlation, not promise a specific return.