An industry push to harness artificial intelligence (AI) has continued to be embraced by asset managers and lenders, many of which are investing sizable resources into innovations. Nomura Asset Management (NAM) has become the latest group to make inroads in this space, teaming up with one of the leading think-tank and systems integrators in Japan, the Nomura Research Institute.
NAM will be looking to explore natural language processing (NPL) to steer portfolio managers into making improved investment decisions. Powered by AI, sentiment-based trading platforms and other mediums have been deployed to help harness large swaths of data. More so than ever before, investors, market participants, and asset managers are burdened by a sea of news, blogs, social media, and other forecasts.
Nowhere is this more evident than in equity markets, where such data and information is instrumental in trading strategies. As such, NAM has been working with the Nomura Research Institute for several months on a streamlined investment module that utilizes AI to bolster the accuracy of portfolio managers’ investment decision-making.
How to Prepare for CySEC’s New Tiered LeverageGo to article >>
NLP scoring system
Using NLP, both NAM and the Institute attempted to analyze a given pool of information made available to portfolio managers. Subsequently, this consumed data was disaggregated and scored into two groups – the information was deemed positive if it correlated to a rise in company performance or corporate value or negative, suggesting this data was unlikely to improve the company.
Simple tests like this are instrumental in understanding foundational elements of investment decision-making. NLP was also central to a reporting to highlight the shifts in decision-making patterns, i.e. a shift from neutral to overweight or from neutral to underweight. In particular, the language patterns for ‘positive’ and ‘negative’ were then isolated and used as training data for AI.
The final result was the autonomous calculation by AI, noting the similarities between the training data and the targeted materials and consequently scoring whether each piece of information was ‘positive’ or ‘negative’. An important result from the analysis yielded that most matters of information, even text from news websites or blogs, could ultimately be quantitatively scored and used to strengthen the ability of portfolio managers to make investment decisions.
In addition, AI also allowed the quantitative assessment of information, which portfolio managers traditionally view as qualitative. The initial findings are part of a long-term partnership between NAM and the Institute, which will continue to explore other cases where AI can better assist portfolio managers. However, the challenge will remains for both for NAM and other fintechs on how to best harness data streams and information that is more readily available than ever before.