Recently, we have been getting a lot of questions about the role of Artificial Intelligence in stock investing. It is easy to see the appeal. You enter a prompt into a chatbot, and within seconds, you receive a confident, highly readable stock recommendation.
To be completely honest, I use AI tools like Gemini and Claude quite a bit in my day-to-day work, especially for research and brainstorming ideas for our blog. But using them so frequently is exactly why I see a fundamental problem with this trend in investing.
Whenever I put these language models to the test and ask for stock ideas or deep-dive data, two things usually happen. First, the financial data is often off, sometimes just slightly but other times by a mile. Second, I am almost always fed the same massive, well-known mega-caps that everyone else is already talking about. This completely contradicts the Obermatt method of finding hidden value.
Filtering the Noise to Find the Signal
When you look closely at AI-powered stock analysis, it is essentially advanced pattern recognition. These language models scrape and recycle the same opinions and narratives that human analysts have been publishing. They are exceptionally good at echoing the current market hype, but they do not do the actual math.
They amplify the noise instead of filtering it out to find the real signal. And in investing, there is a massive difference between a well-written story and financial reality.
For investors, relying on AI introduces several hidden risks. First is the "black box" problem. If an algorithm tells you a stock is a strong buy, you cannot easily audit how it arrived at that conclusion. Without a clear, consistent set of rules driving the recommendation, you have to take the result on blind faith.
Furthermore, language models are designed to predict the next plausible word in a sentence, not to calculate valuations. They can, and do, invent financial metrics or misinterpret complex corporate balance sheets—a phenomenon the tech industry calls "hallucination." Because they are trained on existing market commentary, they suffer from the exact same cognitive traps as human investors, particularly recency bias and momentum chasing.
The Obermatt Approach: Indexing the Facts
At Obermatt, our philosophy has always been the exact opposite. We do not predict the future, and we do not listen to market narratives. We measure the present.
Every week, our proprietary algorithm processes the actual financial data of thousands of companies across sectors and market caps. We look strictly at the fundamentals: profitability, growth, and valuation.
The core of our method is the peer group comparison. It allows us to find the signal in the noise. We benchmark each company solely against its direct competitors. A pharmaceutical company is never compared to a software firm; a small-cap is never stacked against a global blue chip. We systematically evaluate companies facing the same business cycles and regulatory environments. It is a strict, apples-to-apples comparison.
The output of this indexing is the Obermatt Rank. It is a simple percentile from 1 to 100 showing exactly where a stock stands among its true peers. We break this down into clear dimensions, including Value, Growth, Safety, and Sentiment, to provide our comprehensive 360° View.
If a company has a Value Rank of 87, it simply means its valuation is mathematically better than 87% of its direct competitors. You do not need a finance degree to understand that, and we keep it simple on purpose. Complexity is often where weak analysis hides. A single, hard number based on indexed performance leaves no room for subjective opinions or hidden narratives.
Because our system is purely mathematical and operates without emotion, it is immune to earnings-call euphoria or media panics. When markets are overheated and traders are chasing a "story" stock, our methodology reveals the actual risk-reward profile.
We know this works because we have tested it systematically. Looking back at eight years of our Top 10 lists across 32 western markets, our combined rank picks beat their index more than half the time, delivering an average alpha of 2.7%. This is not a theoretical backtest. It is a rigorous review of 417 published lists and 3'503 stocks. It works because it measures actual financial delivery, not analyst expectations.
Where AI Fits In
This does not mean we are anti-technology. We believe AI has a practical place in the future of our platform. We are currently exploring AI-powered features designed to help you navigate the Obermatt website, find insights faster, and track your investments more efficiently. However, we will only release these tools after they have been tested and we are sure they are solid and genuinely useful to you. Stay tuned for more.