Source – https://www.analyticsinsight.net/
Upside AI is one of the first funds in India to use machine learning to make fundamental investing decisions. Today, Upside AI ranks among the top-performing PMS’ in the country delivering 71% cumulative returns since Jul 2019. In Jun 2021, PMS AIF World ranked Upside AI Multicap in the top 10 products of its peer group. On the back of performance and organic referrals, our AUM has grown by 10x over the last year to over INR 65cr with funds from several HNIs, and family offices.
Analytics Insight has engaged in an exclusive interview with Atanuu Agarrwal, Co-founder of Upside AI, to discuss the effects of machine learning in the investment space, specialized services offered by the company, and much more.
Kindly brief us about Upside AI, its specialization, and services that the company offers.
Upside AI is a SEBI-registered portfolio management (PMS) start-up that uses technology to drive investment decisions. Nikhil Hooda, Kanika Agarrwal, and I (Atanuu Agarrwal) founded Upside AI in 2017 with the belief that technology will make better investing decisions than humans over the long term. This is because machines are unbiased, unemotional, and unaffected by market euphoria and panic. It is therefore focused on building products that follow a rules-based, systematic approach to investing.
Upside AI’s first few products focus on equity investing using proprietary machine learning algorithms to understand, recognize, and buy companies that are fundamentally good businesses.
The first step is parsing the P&L, balance sheet, and cash flow data of all companies listed on the NSE. This was a non-trivial exercise, as meticulous data clean-up and standardization. The secret sauce is however the next step where the algorithm uses ML to shortlist companies that are not just fundamentally good businesses but also in-demand stocks. We do this exercise quarterly, to pick a portfolio of 10-25 stocks. We have also developed a framework for corporate governance checks to make sure the numbers fed into the algorithm are kosher.
We are a Portfolio Management Services or PMS, which caters to HNI’s, institutions, and family offices in India and globally. The minimum investment size is INR 50 lakhs.
Today, Upside AI ranks among the top-performing PMS’ in the country delivering 71% cumulative returns since Jul 2019. In Jun 2021, PMS AIF World ranked Upside AI Multicap in the top 10 products of its peer group. On the back of performance and organic referrals, our AUM has grown by 10x over the last year to over INR 65cr with funds from several HNIs, and family offices.
What is the company’s biggest USP that differentiates it from its competitors?
As far as we know, Upside AI is among the first investment managers to combine machine learning and fundamentals-based investing. Our algorithms are completely proprietary and have been fully developed in-house.
In addition, driven by machine learning, our algos are dynamic and are not tethered to a particular style of investing (like momentum, value, etc.) It can adapt to different market conditions and tailor portfolios accordingly. Therefore, it can consistently deliver alpha over the index, which is difficult for human managers who may thrive in some market cycles but struggle in others.
Technology and rules-based investing already dominate mature markets; in fact, the top 5 hedge funds in the world are quant. 5-10 years from now this will play out in India too and we are excited to be at the forefront of this movement.
How is machine learning impacting the investment space?
As I said earlier, ML is pretty dominant in the investment space. I think access to clean longitudinal data makes it an ideal environment for AI/ML-based algorithms.
AI is slowly driving the way investors analyze and implement trading strategies. High-performing computer-managed funds like – ETFs, quant funds, and index funds make around 35% of US public stocks. Notably, the US now has more ETFs listed than stocks. Also, the top five hedge funds in the world are now quant funds collectively managing close to $1trillion. According to Deutsche Bank, 90% of equity-futures trades and 80% of cash-equity trades in world markets are executed by algorithms without any human input
In India, algo trading is mostly associated with high-frequency trading and runs about 45-50% of the cash and derivatives market. ETFs are still in their early days here. Initiatives like Digital India and VC funding are going to further disrupt investment space in India, which so far has been dominated by star human fund managers and large bank-owned AMCs.
Kindly mention some of the major challenges the company has faced to date
The idea was really to step zero. After that, there were several milestones in our journey, including, (i) cleaning up ~20y of financial data, (ii) unbiased and robust back-testing, (iii) real-word beta testing, (iv) frameworks for qualitative issues like corporate governance, and (v) last (but not the least) acceptance of tech-driven by distributors and clients.
It took nearly a year to build out the tech and back-test it; then another year to test it in the live market, before we took out of beta and started offering to third-party distributors and clients in July 2019.
One of our major challenges was that we were pitted against established players in an industry built on trust. Adding to that was the fact that we were all perceived as outsiders since we came from a background in venture capital and technology.
Building a network of investors and distributors virtually from scratch remains one of our biggest challenges even today. Additionally, convincing people to put their trust and money into a newly introduced tech-based product was a mammoth task. However, our aim since day one has been to make investors realize that one must diversify away from human-led investing. Over the due course, on the back of our consistent performance and sustained alpha, clients, wealth managers, and distributors associated with us have come to appreciate that tech-based products should be an integral part of one’s asset allocation strategy.
What are the key trends driving the growth in the portfolio management services space? How do you plan to tap into the market?
I think one of the biggest factors has been the underperformance of mutual funds. Over the last 5 years, the biggest mutual funds have failed to beat their benchmarks. At the same time, performance on PMSs has improved, especially on the back of small- and mid-cap rallies over the last couple of years.
In general, greater penetration of household wealth will also drive growth. Today, with 5-10% penetration we are in the early days of the asset management industry. As savings are diversified away from traditional assets like real estate and gold, and with negative real interest rates, PMSs may see higher allocations from HNIs and family offices.
We are very focused on building a network of large distributors, wealth managers, brokers, and IFAs – we onboarded several distributors over the past year, and are constantly looking for new partnerships. We believe that pure performance marketing does not work and we instead want to focus on long-term community-building driven by organic content and brand-building through online and offline channels.
What are the key milestones and growth numbers achieved since inception? What are your future plans?
We want Upside AI to be a world-class 100% tech-driven asset management company (AMC). We aim to achieve this by building an array of products across asset classes and strategies. Apart from the two live products, we have another couple of products in development, which we plan to launch over the next 6-12 months. We are also working on building an end-to-end digital platform where we want tech to drive the entire chain from client onboarding to sophisticated investment tech to live tracking and reporting via a seamless UI/UX experience for clients. Over the next couple of years, we want to add 1,000 HNIs, family offices, and institutional clients and grow to INR ~1,000cr in AUM.