Dancing with data: The Simpson’s Paradox in VC Investing

By Ashita Gupta | 02 May 2024

Assur, the ancient Mesopotamian city, was famous for the ingenuity of its merchants, among other things. The city was the cradle of notable inventions and fascinating business practices, leading the charge for which were its merchants. One among the many timeless customs of those merchants was investing in various trading routes and dealing with multiple commodities, from grains to textiles, precious metals, and exotic spices. After all, Mother Nature or a baby human could play havoc anytime, and this was probably the only way to hedge their risks against any calamity – famine, wars, floods, et al.

This ancient wisdom found modern roots in the pioneering work of Charles Darwin - father of Theory of Evolution and Harry Markowitz, who took a quantum leap to shape the idea of diversification, and heralded the era of modern portfolio theory. Their studies formed the basis for portfolio managers to appreciate the value of diversified portfolios to manage risks and returns.

When fund managers started constructing diversified portfolios, data defied. How could this ancient theory that evolved over centuries of human wisdom and flourishing trade defy the numbers when juxtaposed over Venture Capitalists (VC) investing. Blame the Assur merchants? Not really.

Simpson’s Paradox, coined after British statistician and civil servant Edward H. Simpson in 1951, describes an intriguing phenomenon: that data interpretation is not always straightforward. It reveals that when the data of multiple groups is combined, the aggregate results can contradict the individual data performance of those group. In some cases, the expected trends may disappear altogether while in others, they are reversed. What does it imply for VCs?

Imagine a VC fund that has invested in two sectors: Technology and Healthcare. Both the sectors have shown impressive Return on Investment (ROI) individually, say X and Y respectively. However, when one combines their individual performance into a single diversified portfolio, the aggregate return might be less than either of the individual portfolios i.e., Z < X & Y. This is counterintuitive, isn’t it? One might expect that the overall average ROI of the fund would be a weighted average of the individual sector ROIs. But here comes the paradox. Prick!

Bursting a myth?

A possible explanation of this anomaly is the composition and size of the investment portfolio, which are often overlooked when benchmarking individual returns. It also leads to negative correlation between sectors that were not evident when the portfolios were separate. In a scenario where technology companies thrive due to positive industry news and healthcare companies are in the doldrums due to shifts in industry behaviour and investor sentiment. As a result, the combined portfolio returns might be dampened due to offsetting movements. Guess the Mesopotamians did not rummage through multiple bone-chilling terrains and arduous trade routes for the love of sailing!

Diversification – the only free lunch in finance

Several studies have proven that diversification outperforms specialised portfolios as it hedges risk, it is crucial for fund managers to be aware of Simpson's Paradox and understand the underlying composition and interaction of the fund's assets with each other. This paradox highlights the importance of analysing investment performance at a granular level, considering subgroups or sectors separately, while considering factors such as risk profiles, growth potential, and market conditions while not relying solely on aggregate figures.

A fund manager can decide their own thesis as long as they are true to their positioning and manage sectorial concentration better i.e., diversify enough. Like we do at Merak.

At Merak, we embrace this principle while maintaining our expertise in the B2B sector. We understand the value of portfolio diversification and are cognizant of paradoxes playing out in real life. Our strategy is to take a sector-agnostic bet, and this reflects in our investment spectrum - from deep-tech startups such as RACE Energy, LightspeedAI to tech-enabled businesses such as Progcap, 4baseCare, Medpay, among others.

For fund managers (aka Assur merchants) seeking to build a robust startup portfolio, do reach out to us if you have any ideas on innovative diversification strategies or methods to defy portfolio paradoxes. While we are skilled at decrypting cuneiform, you may simply write to us at investments@merakventures.com.