Why Data Science Is a Key Skill for Investment Professionals

The technology-enabled democratization of data has made developed public markets more efficient, with investors now rarely able to profit from asymmetric information  – though notable exceptions remain in private and emerging markets. 

“Everybody has more or less the same information. Gone are the days when an experienced muni bond salesman could sit in an ivory tower and make decisions that move the market,” observed Sri Krishnamurthy, CEO & Founder, QuantUniversity. 

“The question then becomes, what are you going to do with that information?”

In search of the answer, the investment industry was an early adopter of data science to improve competitiveness and drive better decisions. After all, data provides the foundation for every investment strategy. Whether you are looking at long-term fundamentals, short-term momentum or a complex quantitative model, understanding the data is key to spotting the opportunity. 

As advances in machine learning and artificial intelligence open up new productivity gains, the successful investment firms of the future will be those that incorporate new technologies into everything they do. In order to do that, their entire workforces will need at least a basic familiarity with data science.

Data has become so critical to investment management that all professionals have a stake in dealing with it – just as all employees of a bank need to participate in managing risk, even though there is a dedicated department for that function, argues Augustine Backer, Vice President, Lead Analyst of Investment Portfolio at Wells Fargo. 

“They should all be able to understand where the data came from, how it has been transformed, and whether anything has changed in the transformation process,” he said. 

Data serves three main purposes in the investment industry, noted Krishnamurthy. It helps managers understand what happened in the past, so they can avoid pitfalls or spot opportunities based on current developments. It is used to generate summaries to make the sheer volume of data more easily digestible, and trends easier to discern. And it can be leveraged to develop models that can attempt to predict future outcomes. 

The third purpose holds particular promise. Whereas only a small minority of investment professionals describe themselves as proficient in artificial intelligence (AI) and machine learning, the tide has clearly turned: the majority of investment professionals are pursuing or planning to pursue skills development in AI and machine learning, according to a 2022 CFA Institute survey.

When asset owners are vetting an investment manager these days, they inevitably ask, “what is your secret sauce?” More often than not, the answer is “data,” said Krishnamurthy. “Everybody should be able to talk about it,” he said.

Indeed, given investment professionals are often better positioned to determine the relevance of data to their specific business lines than data scientists, instilling data science skills across the entire organization has potential to yield considerably deeper and more valuable insights. After all, given that only a fleetingly small percentage of data is considered “useful,” experience is crucial in identifying and exploiting it.

While investment professionals don’t need to be expert data scientists, “they should at least be aware of the general topics so they can communicate effectively with the data science team,” said Sree Mallikarjun, Chief Scientist, Head of AI Innovation at Reorg. For instance, it would be helpful if they could indicate which type of forecasting model they require or specify how robust it should be for their investment needs.  

“A key benefit of the investment folks having a bit more knowledge of these things is that they are able to make better judgement calls about what projects to work on, and how long to allow for them,” said Mark Ainsworth, an insight and data science expert who previously headed data insights at Schroders.

“When you don't have that interface working, the data science team can end up being asked to work on a project that's doomed to failure, trying to fit a model to some dataset that is just all noise and very little signal. Or if the business doesn't have its own intuition that the project will take three months to do in a rigorous, robust and sustainable way, and just asks the data scientists to create a one-month version of it. Lo and behold, after a year, it will complain it keeps breaking,” he said.

A grasp of data science also opens a lot of doors for investment professionals, paving the way for them to participate actively in algorithmic trading strategies, optimizing, time-series analysis and backtesting of models and strategies, said Mallikarjun.

“And it will be more fun for them because now that they know a little bit about it, they can ask more interesting questions. That will also make the data science team more excited about working with them. If someone does not know much about data science, it becomes much more of a difficult, stilted conversation,” he said.

There is another critical reason all investments professionals need to acquaint themselves with data science: to help make sure the models built using it work as they should.

“These models are complex,” said Julia Bonafede, Co-Founder of Rosetta Analytics. “And if only certain parts of the organization have the ability to monitor them, it becomes very difficult get the outcome you want. There has to be a broad general knowledge of how the models should work, so that you can ask the right questions.”

On a more everyday level, across every industry, practically any employee can make use of small data and basic analytics to improve their team’s performance.

By learning the basics, investment professionals “can build small quantitative models on Excel to maybe forecast prices and find some risk indicators. It enables them to do some groundwork and not have to depend on other people to do even basic tasks,” said Mallikarjun.

Moreover, as the line between finance and technology becomes increasingly blurred, financial institutions will need to instill data skills throughout their teams to compete with data-savvy tech firms. Data is so deeply embedded in everything that Google does, for example, that the company does not even have a Chief Data or Analytics Officer. Rather, it has a Chief Decision Officer whose role is to bridge departments within the firm.

To remain competitive in an ever more data-driven future, the investment industry needs to cultivate a more direct and effective dialogue between its business and data science teams through a greater familiarity with each other’s domains.

“Right now, the conventional way is often to get a middleman, who has engineering training and is also a CFA charterholder, enabling them to translate between the two sides,” said Alan Lok, Director, Curriculum Development at CFA Institute. “But eventually, to move things along, it will always be better for the two sides to communicate directly.”

That will enable the industry’s data science endeavors to begin with key investment questions rather than trying to figure out use cases for data and metrics that data scientists have already built. In other words, finding solutions for problems, rather than vice versa.

As data scientists and fundamental analysts learn to speak the same language, their bond will grow deeper. That will help ensure investment firms can maintain their edge by realising the full potential of the rapid expansion of new data sources and data science techniques.

View more Data & Technology stories

Want to learn more about data science?

The Data Science for Investment Professionals Certificate teaches you the key skills you need to become the investment professional of tomorrow.

Learn More