Evolution Of The 'Humachine'

In 1997, the IBM computer, Deep Blue, beat chess world Champion Garry Kasparov in New York in what was that time thought of as a blow to chess professionals. Nineteen years later, the game - thought to be facing a threat from machines working on pure reason without the handicap of human emotions - is far from dead. It is, in fact, thriving more than ever.

Will the same thing happen in the field of personal finance too? The recent launches of robo advisory platforms, which offer financial advice based on automated, algorithm-driven services with little to no human supervision, raise a similar "Man Vs Machine" question. Can machines work independently of humans?

A Solid Foundation

The Modern Portfolio Theory, proposed by Noble Laureate Harry Markowitz, gave a framework that combined risk and return to build efficient portfolios. Other frameworks such as Capital Asset Pricing Model helped define risk-adjusted returns better. Though these developments occurred a few decades back, they provide a solid foundation to investing strategies. There is an increased understanding that good advice is not just about returns but also the risk one takes in achieving those returns. Most robo advisors are founded on these core frameworks that provide robustness to the advice.

Behavioural Finance: Research in this area suggests that our decisions can be laced with biases that can lead to poor investing choices. Robo advisory, being data driven, takes away many biases by suggesting actions that are driven by a rational decision-making engine. Take, for example, the simple principles of rebalancing or exiting poor investments. It is known that following an asset allocation strategy and regular re-balancing for a target makes people buy assets at their lows and sell them at their highs. However, this strategy is difficult to implement as investors rarely sell while markets are doing well. And vice versa, investors rarely buy while markets are falling. However, many studies have shown that aligning portfolio to a target allocation at regular intervals reduces risk and enhances returns. Other biases like loss aversion makes investors delay exiting poor investments. Robo advisors take action clinically and systematically and reduce the play of emotions in investing.

Quantitative Finance: Many sophisticated robo advisors consider the risk profile of the client, his/her financial goals and the risk for each goal to define an asset allocation strategy, embed market factors like size in the equity portfolio and even choose the duration of the fixed income portfolio. It is common for robo advisors to discount a large number of client-specific profile variables and use millions of simulations and optimisation techniques to arrive at an investment strategy. This brings in more accuracy and rigour to the advice.

Challenges for Robo Advisors

For robo advisors, knowing their clients well enough is a challenge. Unlike a good relationship manager or an investment advisor who can understand both said and unsaid needs of clients, robo advisers work only on the inputs collected. There is a need to sharpen understanding of clients. There are also no regulatory standards. It is also known that the theoretical frameworks do not work in all market conditions. Another limitation of robo advisors is ability to give assurance during volatile market movements.

It is estimated that the robo-advisory market worldwide will grow to $2.2 trillion by 2020 as against the current estimate of over $230 billion. As the preference for fully automated advice grows, it is likely that growth will come to robo advisors who offer a healthy mix of offline and online interactions with clients. This will bring not just the rigour required to the process of financial advice but also the assurance of a human touch.

The writer is MD & CEO of ICICI Securities