Asset manager Franklin Templeton argued that in a phase of expanding agentic artificial intelligence (AI), it is hard to capture all investment opportunities with stocks alone, and holding cryptocurrencies and altcoins may be necessary.

On July 22 local time, blockchain outlet The Defiant reported that Sandy Kaul (샌디 콜), Franklin Templeton’s head of digital assets and innovation, said that if on-chain transactions by autonomous AI agents increase, tokens issued by the relevant networks could become a core asset in investment portfolios.

Kaul said the market’s current way of investing in AI could differ from investing in agentic AI. She said most investors have so far sought to benefit from AI growth by investing in shares of AI-related companies and linked sectors, but questioned whether the same approach would apply to agentic AI.

She said the key logic lies in payment structure. Kaul argued that existing card and bank payment networks are not suited to handling machine-to-machine (M2M) micropayments. She said a typical card payment carries average fees of 2 to 3 percent and a fixed cost of about $0.30, while AI agent payments are made at an average level of around $0.001 for 1 second of compute time or a single data query.

She added that to capture the value of decentralised networks and businesses operating on them, investors need to buy the cryptocurrencies and altcoins issued by those entities. She also said such assets are likely to become core portfolio holdings for investors seeking to capture emerging agentic AI opportunities.

She also presented external forecasts on market size and adoption speed. Kaul cited estimates that the agentic commerce market could reach $3 trillion to $5 trillion by 2030, and said that in 2028, 38 percent of all organisations would have AI agents as team members working alongside humans.

Examples cited included Coinbase’s x402 payment standard and machine-to-machine payment protocols by Stripe and Visa. The subsequent transfer of x402 to the Linux Foundation shows that a structure is being put in place in earnest for software to make payments directly to other software without human intervention.

Against this backdrop, Franklin Templeton’s argument focused on the idea that the means of capturing value in agentic AI may not remain limited to big tech stocks. How widely actual AI agent payments spread on-chain, and which network tokens become established as fee assets, are expected to be key points to watch.