Our Methodology
We built this engine to answer one question: Who is actually beating the market? Here is exactly how we calculate the numbers.
1. Data Collection (The Scrape)
We don't manually watch videos. We run automated scripts that scrape YouTube transcripts from the top finance channels daily. We use Natural Language Processing (NLP) to detect specific keywords like "Buy," "Load up," "Sell," and identify ticker symbols (e.g., $TSLA, $NVDA) in the context of a prediction.
2. The "Entry" Price
Accuracy requires fairness. Since most viewers watch videos hours after they are posted, we assume an investor cannot buy at the exact moment of upload.
Our Rule: We log the "Entry Price" as the Open Price of the next trading day following the video upload.
3. Benchmarking (The Alpha)
Returns are meaningless without context. If a YouTuber is up 5% but the S&P 500 is up 10% in the same timeframe, they actually lost you money.
We calculate "True Alpha" by subtracting the S&P 500 (SPY) return from the YouTuber's stock return over the same holding period.