AI race is no longer a two-player game
Explaining why PPFAS Mutual Fund is avoiding direct exposure to pure-play AI model companies, Rajeev Thakkar said the competitive landscape has changed significantly over the past year.
“From a time early in this space where it seemed like OpenAI and ChatGPT would win all the way, there is now a proliferation of companies and models. OpenAI / ChatGPT, Anthropic / Claude, Gemini, Grok, Llama, DeepSeek, Kimi, GLM, Qwen and many more are fighting for this space.”
Rather than assuming one company will dominate the AI ecosystem, Thakkar believes the market is becoming increasingly complex.
“Increasingly, it looks like the latest and greatest model may not be the only winner and other aspects like sovereignty, access, costs per token and per task, speed, data privacy and so on will matter.”
He added that PPFAS Mutual Fund has deliberately stayed away from investing directly in companies developing foundation AI models.
“Luckily for us, we have no direct exposure to pure play AI model companies like OpenAI or Anthropic.”
Why PPFAS prefers hyperscalers
Rather than investing in companies whose fortunes depend on a single AI model, PPFAS Mutual Fund has chosen to invest in what Thakkar describes as “hyper scalers.”
“Our investee companies can be termed as hyper scalers. These companies are the companies which run these massive data centers and offer cloud computing services and AI computing services to enterprises. Concerns have been raised around the capex of these companies.”
He pointed out that these companies continue to derive substantial revenues from businesses beyond AI.
“Microsoft for example still has enterprise software, Amazon still has e commerce and the traditional cloud business and Google still has its digital advertising and subscription business.”
Addressing concerns over heavy AI-related capital expenditure, Thakkar said the spending is supported by customer demand rather than speculation.
“The new capital expenditure, while large, is backed by increasing demand from their enterprise customers. They are not tied to the success / failure of a single AI model. Is there some risk of over-investment here? Sure, it could happen. But over capacity will not be permanent in nature. It will get consumed in some time as the workloads increase over time. It will result in a slowdown in capex down the road if the demand takes time to scale.”
Staying away from AI chip makers
Thakkar also explained why the fund has not participated in another popular AI investment theme—chip and memory manufacturers.
“These companies have been the biggest beneficiaries of the AI boom. The demand and margins have gone through the roof. It is almost a universal consensus that the current heady days are cyclical in nature, and the profit margins will mean revert. We have missed out on the upside in this space and are unlikely to have comparable downside exposure if there is a cyclical downturn here.”
Investment philosophy remains unchanged
Thakkar concluded by reiterating that PPFAS will continue to invest based on valuations and risk-reward rather than following market narratives or fashionable themes.
“We have not been living under a rock. We do see the media headlines and are aware of Defense, Energy transition, AI, Fintech and other trends / themes in the markets. We will invest where we find attractive, actionable ideas and not invest purely on account of a theme being in fashion.”
Disclaimer: Business Today provides market and personal news for informational purposes only and should not be construed as investment advice. All mutual fund investments are subject to market risks. Readers are encouraged to consult with a qualified financial advisor before making any investment decisions.