Financial Sentiment Scorer
General NLP models fail in finance because words like "liability" or "tax" are neutral in a 10-K but negative in conversational English.
Modern quant funds have moved beyond simple dictionary approaches (like the Loughran-McDonald dictionary) to context-aware models like FinBERT. These models don't just count words; they measure executive evasion and uncertainty.
Read our full research on NLP in Earnings Calls to understand how the alpha in unstructured text is evolving.