Morgan Stanley Analyst Joseph Moore Predicts Significant Increase in DRAM Prices
Morgan Stanley analyst Joseph Moore stated that after communicating with several data center procurement personnel last week, it was found that the current tight supply of memory has not eased.
Prices for memory products such as DRAM are expected to rise by at least 25% on a comparable basis from Q2 to Q3, higher than previously anticipated. The memory shortage may further worsen in 2027 and 2028, as AI demand heavily consumes DRAM capacity and affects PC and smartphone supply.
AI computing power is driving increased memory procurement, with data center operators becoming high-priced buyers. Memory manufacturers benefit from enhanced pricing power due to supply-demand imbalances, while AI expansion plans face bottlenecks, leading to rising costs in the PC and smartphone supply chains.
Source: Public Information
ABAB AI Insight
Morgan Stanley, where Joseph Moore is located, has previously released multiple reports on the AI supply chain. This judgment, based on feedback from frontline procurement in data centers, continues its past analytical path regarding hardware cycles, including early warnings about GPU shortages.
On the capital front, the demand for AI training and inference is driving funds toward memory expansion projects, allowing memory manufacturers to achieve higher returns on capital expenditure, motivated by the wave of AI capital spending. Meanwhile, downstream AI companies and traditional electronics manufacturers face cost pressures, forcing them to accept price increases or delays.
Similar to the case during the chip shortage of 2021-2022, where soaring memory prices significantly boosted profits for manufacturers like Samsung and SK Hynix, the current memory industry is in an AI-driven phase of supply-demand mismatch expansion.
Structural Judgment Essentially, this is a transfer of pricing power: the explosion of AI will shift the pricing power of key memories like DRAM from downstream terminal manufacturers to upstream suppliers. This mechanism arises from demand growth far exceeding the capacity construction cycle, leading capital to pay a premium for limited resources and accelerating industry concentration.
ABAB News · Law of Cognition
The hotter AI gets, the more bottlenecks appear before profits.
Pricing power belongs to the segment that controls scarce physical resources.
When demand explodes, the upstream of the supply chain always smiles first.