By Sudharson Gunasekaran 

Algorithms written for investment management today are achieving a level of sophistication that was previously unimaginable. In a relatively short period of time, rapid growth has occurred for usage in market intelligence, sentiment analysis, trading, smart routing, and transactions. Among this expansion, the use of robo-advisors—the algorithm-based digital platforms that offer financial planning advice with minimal or no human supervision—remains the industry’s most controversial option. Advocates argue that robo-advisors work best with traditional investments and do not threaten the need for an intellectual human element required for more complex investment planning. Opponents counter that human advisors can never be replicated with technology, even for traditional investment planning, largely due to the emotional complexity and empathy required for thoughtful decision-making. The reality is that it’s unlikely human involvement will become obsolete, and yet algorithmic investing is not expected to slow any time soon. How can banks ensure trust and confidence among their customers given this dynamic? The answer is a comprehensive pairing of both man and machine working toward a common goal.

The evolving state of algorithmic investing

The roots of digital and “algo” trading can be traced back to the 1960s when the New York Stock Exchange (NYSE) began applying computer data processing technologies to its market operations. High-speed data networks captured and disseminated trading data and market information, greatly increasing market efficiency. Following the 1976 advent of the Designated Order Turnaround (DOT) system which allowed brokers to route orders directly to specialists on the floor, electronically transmitted orders became the norm, with human oversight required. Nearly two decades later in 1984, the NYSE launched its SuperDot system, a program that electronically delivered orders to the trading post from a broker’s office and then followed up with an execution report within seconds.

Nearly 40 years later, the presence of technology and algorithmic investing has grown out of a simple, rule-based approach into one that has the ability to identify market fluctuations through the integration of artificial intelligence (AI) and machine learning programs. These technologies help apply big data to provide appropriate guidance on when to sell or buy stocks that are ideal for specific portfolios. Perhaps most notable is the work of robo-advisors that offer customers investment advice and automated portfolio management based on personal preferences and financial goals. Introduced in 2010, these advisors seemingly work independently of human intervention and are expected to be managing as much as $16 trillion in assets by 2025, according to a report by Deloitte. Regardless of the technologies available now and in the future, society can’t rely on this type of technology to perform the work and provide the guidance that humans are capable of conducting.

The ongoing role of humans

Continued advancements in technology into the “smart” domain have also helped the majority of people to become more trusting of integrating technology into essentially every aspect of personal and professional life, including the management of money. A survey of approximately 15,000 people conducted in 2022 by the Edelman Trust Barometer, found that the tech sector earned a trust score of 74 percent, a four-point jump from 2021. While most industries registered a rise in trust in recent years, the tech industry scored better than any other sector, according to the report—including healthcare and social media, the latter of which dropped two points to a score of just 44 percent. 

AI and algorithms are penetrating the healthcare industry in more substantial ways with influential regulatory oversight by the U.S. Food & Drug Administration and the Centers for Medicare & Medicaid Services. When it comes to preparing financially for their future and retirement, people are quite trusting of technology. For a population that increasingly becomes less likely to earn a pension, reliance on others to help secure a financial future has been the norm for some time. According to a study released by the National Institute on Retirement Security in 2020, a mere 6.8 percent of Americans ages 60 and older who work less than 30 hours per week receive money from Social Security, pensions, and workplace retirement savings such as a 401(k). Meanwhile, 40.2 percent of older Americans receive income through Social Security only and 14.9 percent receive no income from a pension, savings, or Social Security. Is there a line to draw between trust when it comes to humans and machines making financial investments? Would more people be inclined to go the route of a robo-advisor if it cost them less money to manage their investments? At the very least, there is a difference between what machines can legitimately handle and what human employees can do for their clients when customers make their considerations. 

For instance, human financial advisors are still needed for the management of more complex portfolios, estate planning, tax handling, and assisting customers in establishing their long-term investment goals. Banks and their customers can rely on robo-advisors to manage portfolios reaching the $20,000-$30,000 range and they can advise investors on a few assets based on their individual goals. Robo-advisors, however, are not a reliable source for creating well-diversified portfolios. Traditional financial planners also continue to play a much greater role in building and maintaining customer relationships, namely through having the ability to provide “good faith” investment advice. Humans are also more capable of reviewing balance sheets and assessing company behaviors that could lead to finding “silver lining” stock profits in the long term, whereas robo-advisors can only act upon straightforward data in the present tense. Humans are also more valuable in assessing risk and can call upon their emotional intelligence and intuition to make suggestions from their “gut” that come from what real-life experiences have taught them.

Robo-advisors are most useful for retail investors with small-to-medium-sized investments that are concentrated primarily on exchange-traded and index funds. Machines are also useful for creating short-term profit opportunities when the appropriate data is available. They can also react to market changes more quickly, can potentially predict the movement of a stock price in the short term, and are less expensive to “employ.”

Future trends: Humans and machines investing together 

Given the benefits that both humans and algorithmic trading offer to customers and the industry, it’s likely that increasingly banks will seek ways to integrate elements of technology with one-on-one business relationships in an attempt to provide customers with the ideal guidance for their financial decisions. Much like it was realized long ago that human oversight was required during the early days of automated NYSE investment technologies, humans will continue to collaborate with machines moving forward, even when it might seem as if the machines are doing “all of the work.” At their baseline, algorithmic trading systems are developed by feeding instructions and training the system by using the vast amount of data that is available to them to execute investments and trades on their own. Consider this utilization of data: robo-advisors can handle more data in terms of pure volume, but humans continue to play a significant part in defining data quality and refining the data because of their ability to better understand the overall financial landscape. Human intervention allows machines to be equipped with the best information to make more sophisticated decisions that are most applicable to institutions and their customers.

Working in tandem, robo-advisors are potentially best utilized through their ability to recommend an ideal portfolio among the seemingly infinite number of combinations using data collected from investors such as an investor’s risk appetite, expectations of return, the investor’s employment industry, and duration of investment time before retirement. Through this approach, robo-advisors can quickly compare market changes with investment opportunities to construct attractive and lucrative portfolios to recommend to investors. Those investors will still retain complete control over their investments and have the ability to liquidate at any time.

Algorithmic trading, robo-advisors, and AI overall will likely always rely on human involvement for risk management and mitigation. AI needs human intervention to be trained appropriately and to consolidate big data. Algorithmic trading can be partial, in particular requiring hedge fund managers to override a system to avoid automated trading during volatile and illiquid market conditions, for example. Humans are also necessary for managing edge cases where machines cannot understand certain factors, such as explaining arbitrage opportunities once they are identified.

Further, it’s the role of humans to regulate the registration and usage of robo-advisors, as well as other types of AI technology, just as human advisors are regulated to ensure the secure movement of assets and transactions. According to the U.S. Securities and Exchange Commission (SEC), investment advisers may be primarily regulated by the SEC or by one or more state securities authorities. The SEC issued a set of observations and advice in 2021 to those advisers providing electronic investment services in an effort to facilitate guidelines for those managing the estimated 200 robo-advisors in the United States today. Increasingly, traditional asset management firms are beginning to launch robo-advisor businesses independently. While regulations over machine trading and advisory interest will continue to grow, both will serve different segments in the future based on the role of humans.

About the Author: 

Sudharson Gunasekaran is a highly skilled professional working as a senior software engineer with a leading investment management firm. He has more than 12 years of experience providing software engineering solutions to the banking and finance industry and more than 16 years of experience in software engineering. For more information, email [email protected].