The investment profession is changing by leaps and bounds. That’s not just due to industry specific dynamics; it’s also because of the technological evolution under way that is having an impact on industries all over. In May 2019, CFA Institute published a report, “Investment Professional of the Future,” based on surveys from its members, candidates, and industry experts. The conclusions were hardly surprising, although the pace of change has been remarkably fast. The main themes in the report emphasise how the roles of investment professionals, along with their core skill sets and career trajectories, will evolve over the next 5 to 10 years. The disruptions created by COVID-19 have only accelerated the rate of change as the need to digitize workplaces has become urgent.

Some interesting conclusions stand out in the report. Eighty-nine percent of industry leaders agree that the investment professional will have to evolve multiple times over the course of their career, and 43 percent of professionals believe their jobs will be quite different than today in five to 10 years. Most importantly, the report acknowledges the importance of artificial intelligence (AI) as a dominant, evolutionary force in the industry. It coins the “AI+HI” (human intelligence) combination as the inevitable future, where recent advances in machine learning, data science, and computational prowess will supplement human intelligence.

Evolving role of the investment professional

The report’s findings imply that the investment professional of the future needs proficiency beyond investment analysis and portfolio management; they must both have a better understanding of technology and be programmed to adapt and innovate. These so-called “T-shaped skills” are ranked as the most important by investment industry leaders, according to a May 2019 CFA Institute report titled Investment Professional of the Future: Changing Roles, Skills, and Organizational Cultures. Professionals with these skills are subject-matter experts, quickly adapt to changing environments, and can work across organizational silos.

A growing number of research analysts, for example, are now integrating various sources of alternative data in their analysis to gain an information edge. Doing so would require the analyst to acquire the data, clean it, and prepare it for analysis; perhaps by integrating it into a Python notebook to produce analytics and forecasts with a strong understanding of the economic fundamentals of the investment. Alternative data such as consumer transactions, cargo movements, real-estate construction activity, and other wide-ranging datasets, are widely being used by the buy-side community.

So, it is imperative that the CFA Charterholder of the future have a high technical proficiency in a broad set of skills, in AI and data science especially. As most CFA candidates would agree, the content of the 2020 CFA Curriculum reflects the broadening of required subject matter, and the increased number of hours now needed to master it. CFA Institute has adapted its curriculum to the new realities of our industry by adding important learning outcomes focused on fintech, big data, and machine learning.

Evolving CFA curriculum

Financial technology, or fintech, is a newer term for a gradual evolution the entire financial ecosystem has been witnessing, not just in investment management but in other sectors of financial services, in which technology is used to improve, and to automate, the delivery and use of financial services. Payment technologies, retail and commercial banking, and borrowing and lending have all seen technical advancements from incumbents and newer players, and these advancements have reduced the cost of services and increased market penetration to various customer segments. Robo-advisors and free-to-trade platforms such as Robinhood are some examples that exemplify this trend. The fintech section of the CFA curriculum introduces candidates to the general industry and its terminology with emphasis on big data, artificial intelligence and machine learning. It covers some popular techniques and their uses, such as natural language processing (NLP), algorithmic trading, blockchain, and distributed ledger technology (DLT).

The emphasis on big data and machine learning underscore the rapid improvements in computational sophistication and decision-making capabilities powered by reliable data in recent years. According to IBM, 90 percent of the data in the world today has been created in the last two years. As the amount of data increases exponentially, so does its utility. “Data is the new oil” has become a common metaphor, although some would argue that data is much more important than oil. Firms are increasingly looking to capture and monetize their data assets, irrespective of their industry, so it is imperative that investment professionals understand how to harness the power of big data.

Big data

The 2020 CFA Curriculum ensures that candidates are prepared to lead big data projects by providing a breakdown of all the stages of a typical project and an understanding of how to manage each of the stages. Data comes in all shapes and sizes, from unstructured to structured, and usually requires a lot of preprocessing. Just like oil, which needs to be refined to be fit for use, data may need to be cleaned and transformed to be useful for making inferences; this is called data wrangling. Data wrangling is followed by data exploration, which involves analyzing the wrangled data to uncover trends, correlations, biases, and so on. This is the point where data visualization plays an important role.

Data preprocessing methods depend on the type of data being used and the specific use case. Text-based data requires different tools and techniques than time series data, for example. The sophistication of these methods, and the importance of doing it right to get an information edge, has introduced data engineers and data scientists to the investment process lifecycle. The investment professional of the future needs to understand the overall process and be able to effectively communicate with these participants.

Machine learning

Clean, good-quality data is the raw ingredient for machine learning models, which are typically used to make inferences about the future. Just like the data they consume, machine learning models come in different flavours, both supervised and unsupervised. These models can also be combined sequentially to get the desired outcome. The 2020 CFA Curriculum prepares candidates to understand the differences in and applicability of supervised, unsupervised, and deep learning models, which is probably the most technical content included as part of the new curriculum.

Candidates are not only provided with a solid foundation by delving into fundamental techniques, but more importantly, being taught which situations the techniques are most suited for. Similarly, candidates should understand the variance-bias trade-off, as well as be able to recognize and deal with the issue of overfitting (also known as the data mining bias in traditional statistics). Candidates must master these analytical concepts as some may continue in their career using software applications and pre-built code libraries to implement these models in practice.

“As the amount of data increases exponentially, so does its utility. “Data is the new oil” has become a common metaphor, although some would argue that data is much more important than oil.”

Looking ahead

Just like traditional statistical analysis, data science and machine learning are also both an art and a science. Both require a deep understanding of the underlying data and the context of the problem at hand. Biases in data can distort processes, and need to be handled with care. The tools and techniques needed require a certain level of technical competency to understand their advantages and shortcomings, and to determine their usefulness in different situations.

In addition, the regulatory, ethical, and professional rules that apply to the investment industry often create an aversion to black box algorithms that produce a desired outcome without any economic intuition. Explainability and interpretability of machine learning models are, therefore, gaining traction in both academia and industry. Suffice it to say that decision makers should not simply rely on model predictions, and should apply their own objective judgment based on all the available information. From that perspective, big data and machine learning are just an additional set of tools in the investment professional’s toolkit.

This emphasis on learning in the 2020 CFA Curriculum harmonizes the investment professional’s analytical skills, industry knowledge, and technical capabilities into a cohesive yet adaptable career trajectory. A broad knowledge base also equips professionals with a mindset that embraces innovations in seeking solutions for new problems as they arise. What we can be sure of is that the evolution of the industry will continue, especially when it comes to technological advancements.

Achieving the CFA Charter is no easy feat. The increased complexity, breadth, and depth of the curriculum may keep candidates up at night, but it is meant to set them up for success in an industry where change is the only constant. Rest assured, the CFA Curriculum lives up to its standard of a professional designation by preparing the candidates for the realities of a dynamic industry. A few hours of sleep are a very reasonable price to pay for that.