Contextual NLP – redefining the Natural Language Processing landscape Natural Language Processing is one of the fastest growing areas of Machine Learning / AI and is used by both quant and discretionary investors. Traditional NLP structures language to derive sentiment signals and commonly uses bagof- words / dictionaries. These models have a number of limitations in practice. Contextual NLP is different in that it moves away from these methods, taking in a whole section of text to understand context: this is the machine acting more like a human. Using company newswires data, we create systematic stock portfolio strategies with impressive backtest results. We obtain Sharpe Ratios between 0.7 and 2.9 across a variety of factors. After reasonable transaction costs of 10bps the Sharpe Ratios are still up to 0.8 trading in GDM: the returns & overall risk-adjusted performance argument appears very compelling.