Macro Strategist Warning: LLM Token Spending Index Reverses Course, AI Hardware and Data Center Cycle May Come to an End

According to OnChain Analytics, macro strategist Andreas Steno Larsen recently issued a warning regarding the latest downtrend in the Silicon Data LLM Token Spending Index. He noted that the LLM Token pricing movement is a key barometer determining the overall AI infrastructure cycle. If the Token price continues to weaken, spanning from memory chips to broader hardware and data center construction, this investment cycle may be on the brink of ending.

The Silicon Data LLM Token Spending Index is an expenditure-weighted market indicator designed to track the average payment price per million Tokens in the entire market. As mainstream service providers commonly bill based on Token consumption, this data directly links software-side actual usage to underlying GPU power, DRAM memory, and data center construction, making it a crucial reference for evaluating the AI hardware capital expenditure cycle.

Data shows that the Silicon Data LLM Token Index experienced a significant climb at the beginning of 2026, only to peak and retreat by the end of May. The index's decline is closely tied to a shift in enterprise attitudes toward AI expenditure. Due to early lack of cost control, some tech giants internally even engaged in blind competition known as 'Tokenmaxxing,' where employees excessively consumed Tokens to boost internal productivity rankings. As hefty bills arrived, companies realized that the high Token expenditure did not yield proportionate business returns. Currently, major players like Microsoft and Amazon have begun tightening internal AI tool usage limits, and in some cases, halting related projects altogether. The shift from 'blind stacking' to 'financial rationality' within enterprises directly resulted in the stagnation and softening of the Token Spending Index.