At DISCUS we are always on the lookout for articles we believe our readers will find interesting and relevant. This week we came across ‘Big data and investing’, written by Anthony Lawler, Co-head of GAM Systematic.
The article offers insight into how investment managers are using big data and artificial intelligence to enhance the diversification and risk/return potential of portfolios. Read on for key insights…
What is big data?
Today huge and growing volumes of data are being generated. A report by research group SINTEF in 2013 stated that 90% of all the available data in the world had been generated over the previous two years. This massive surge in the amount of data being recorded is what we call ‘big data.’
This big data has significant value for businesses; in an increasingly dynamic and competitive market one of the main challenges for companies is to identify important strategic data and use the most appropriate tools to understand and analyse this data. It allows businesses to understand the preferences of the public and tailor their commercial offerings to be more competitive.
Systematic investing and big data
Analysis of big data also forms an important opportunity for systematic investing. Systematic strategies are able to assess big data using machine learning to improve investment signals across asset classes. This data and evidence focus in systematic investing allows the investment style to avoid the typical behavioural biases present in investing, and also dramatically extends the investment capacity of what is possible from an individual, given machines coded correctly can assess thousands of variables in real time to inform investment decisions, while adhering to strong risk controls. Complex algorithms are able to hone signals from big data and make decisions based on them.
Systematic strategies refer to those where the investment process is rules based. This is the core commonality and simply means the process is written down and adhered to. The investment decisions are taken using rules formulated with clarity and transparency, which means the process is not tainted by cognitive bias, which can cloud human judgement.
It is also repeatable.
These rules are written or codified into algorithms able to then operate over millions of data points to make investment decisions for a portfolio. The algorithms and analyses are all under the close research and supervision of a team of highly experienced quantitative specialists, often with the objective of achieving positive absolute returns with low correlation to traditional asset classes or low correlation to other investment styles.
Computerised investment strategies
Systematic investing is a vibrant arena and a place of real innovation, where researchers are actively studying new data types and new techniques. For example, advances in natural language processing now allowing assessment of written and spoken words as part of systematic investment decisions.
Investors are becoming more at ease with the idea of computerised investment strategies. This is not a revolution but rather it is the evolution of how investment analysis and decisions can be enhanced through the use of algorithms, big data and increasingly also artificial intelligence. Aside from their scalability, quantitative strategies have delivered competitive results in recent years, while some more traditional approaches have suffered by comparison, partly because algorithms are not subject to behavioural bias and cannot be distracted by emotions, unlike humans. We view systematic investing as a positive step in offering clients diversification of approach and the potential to improve their investment outcomes.