Could the next Warren Buffet be an AI tool? That’s what Kanika Agarrwal, co-founder and chief investment officer, Upside AI believes. The Mumbai-based Upside AI uses machine learning to make better investing decisions than humans over the long term. They have recently opened up their flagship PMS (portfolio management services) product to retail investors like you and me, to allow technology to diversify investing away from humans.
We caught up with Agarrwal to understand how human biases occur in investing, how to avoid them, and the future of AI in investing.
What are the common human biases witnessed by you while investing?
Most biases that human investors have come from the inability to accept that they are wrong. This keeps them from learning from the past and keeps our eternal wheel of “history repeats itself” going. But you can break down some of the common ones situationally:
Confirmation bias: Looking for information that confirms your existing thought process/ ‘hunch”’. The correct approach is objectively examining data first and then coming to a conclusion, instead of working backwards from conclusion to data.
Loss aversion: This is our tendency to avoid losses instead of obtaining gains – as a result, your money is tied up in losers that you are unable to exit. A Nobel Prize-winning study by economist Daniel Kahneman and psychologist Amos Tversky found that human beings feel the psychological and emotional impact of a loss twice as much as that of an equal gain.
Hindsight Bias: Investors often think all past events are “obvious” because hindsight is 20-20. This stops them from learning from history. Banker and investor Sir John Templeton put it aptly: “The four most dangerous words in investing are ‘this time it’s different’.”
Self-attribution Bias: When you attribute successful picks to your own ability and bad outcomes to external factors. This makes you overconfident and starts disguising luck as skill.
Trend chasing bias: When you forget that “past performance is not an indicator of future results”. Don’t flock to the “hottest stock” or the newest trend – make a set of rules to follow in the market and do it consistently.
Is technology playing a bigger role in investing now?
At least 60% of the US markets are run by technology today. The world’s top funds are all technology-driven. The investing industry as a whole is facing a revolution as technology becomes front and centre. We believe technology will make better investing decisions than humans since machines are unbiased and unemotional decision-makers.
They do not react to market euphoria or panic and react purely to data. We will also see this change in India over the next decade. Today in India, technology is doing the basics, screeners, technical analysis, high-frequency trading, etc. But the ‘holy grail’ is learning true investing, which is the area we are focused on solving.
Why should we consider machine learning-based (ML) fundamental investing as opposed to traditional fund managers?
In general, any conventional wisdom will tell you to diversify. Diversification does not just mean owning 10 stocks instead of five. It also means, diversifying the approaches you invest behind. We are of the view that all investors should also diversify away from human experts and have part of their allocation to technology. Further, our research of Indian fund managers has shown that over time, most fund managers revert to average performance.
This is not a new trend. We have studied this in the Indian context between the years 2013-2020 and from our research, we found that irrespective of whether they were the best (top 25%) or worst managers (bottom 25%) in 2013, over the eight-year period, PMS managers consistently reverted to the mean.
What is your take on the digital transformation within the financial services industry?
This is on-going and inevitable. We have been ‘transforming’ for the last 100 years, and this remains true today. While in the past it was moving from paper to calculators to computers, today it will be moving to machine learning and AI.
Tech is changing how products are distributed, as more customers start buying financial products online (just like they buy a T-shirt on Amazon now) – this is true for loans, credit cards, insurance, mutual funds, and stocks. Companies making these products are also being disrupted – HDFC/ICICI will be disrupted by neo banks, NBFCs by new-age lenders like Lending Kart, brokers by Zerodha, and mutual funds by Upside AI. Products built in deep tech are simply more efficient and responsive to a customer better than the old way of doing things.
Do AI/ML-based or Robo advisory have a future in India?
It is inevitable that technology will drive investing. You can see the disruption in the mutual funds space where distribution is moving online with users (including first-time investors) using platforms like PayTM Money, Groww, etc.
The next big wave will be the transformation of actual investing products. Over the last three years, investors have seen a lack of alpha [often considered the active return on investment] while paying high fees to human experts. While all markets have cycles, a prolonged period of underperformance is leading to a shift into ETFs, tech-based products like Upside AI and Smallcase.
Any trends that you are tracking for 2021 in the wealth management space?
Wealth managers should be recommending technology to be a part of asset allocation in order to diversify away from human fund managers. We see this as a big trend for the next year as tech increasingly becomes a core part of portfolios.
Further, the regulators are doing more for consumer protection making changes to rules on fees and disclosures. All of this means that our investment is set to get cheaper and more transparent in 2021 – all of this will be made possible by the use of technology in investment advice, distribution, and execution.
All images: Courtesy Getty