The recent uptick in US inflation has introduced two new sources of concern: rising costs associated with artificial intelligence and increasing energy expenses. A key question is to what extent these factors are reflected within the core Consumer Price Index (CPI).
China Securities Co., Ltd. (SSE: 601066) has released a research report analyzing this dynamic. The report finds that while AI has a stronger downstream transmission effect, oil prices currently exert a larger direct pull on inflation. The core CPI has increased by approximately 0.2% year-on-year this year, with AI and energy costs combined contributing nearly 0.5%, effectively accounting for the entirety of the inflationary pressure. Other components within the index have generally shown a decelerating trend.
Initial Analysis of Cost Increases
Energy prices have seen a significant rise, primarily driven by higher costs for crude oil and electricity. In contrast, the weighted increase in AI-related costs has been relatively modest, which presents a divergence from common perception. Although storage prices have surged, their weight in the overall cost structure is limited. The hardware component, which carries the largest weight, has continued to trade at low levels, tempering the overall AI cost increase.
Transmission to Downstream Prices
In terms of overall impact, the transmission coefficient for AI costs is notably high at 4.6%, far exceeding energy's 0.5%. Regarding the magnitude of the pull on inflation, the core CPI's 0.2% year-on-year increase can be attributed to an additional 0.39% from AI and 0.08% from energy, summing to 0.47%. This suggests these two factors explain the entire inflation uptick, with other components decelerating, potentially indicating negative feedback from rising costs.
Structurally, AI's influence is diffusing more broadly across downstream sectors, particularly within services like business, finance, and education. This diffusion is likely linked to increased corporate investment in digitalization, suggesting its inflationary impact could be more persistent. Energy's impact, however, remains more concentrated in transportation sectors with high exposure to crude oil costs, making it relatively narrower and more characteristic of a one-off shock.
A historical comparison indicates that the current cycle's transmission of AI costs is weaker than in past periods, while energy's transmission is in line with historical averages. This points to underlying inflationary momentum that is not particularly strong, possibly related to subdued consumer demand.
Outlook and Key Factors
Looking ahead to the second half of the year under a baseline scenario, the contribution from oil prices is expected to decline but remain significant due to the very low base in Q4 of last year. Meanwhile, the contribution from AI could nearly double from current levels. The combined influence of both factors is projected to remain at an elevated level.
Upside risks to this outlook primarily involve oil prices exceeding expectations or a broader-than-expected penetration of AI, leading to price increases in more industries. Downside risks mainly stem from negative demand feedback causing further declines in other CPI components.
For a substantive decline in year-on-year inflation to materialize, the market may need to see one or more of the following developments: a more pronounced drop in oil prices, core goods returning to deflationary territory, or the shift to a higher base for oil price comparisons in the coming year.
Comments