Lucas Bennett
I research AI semiconductor indices and the rules used to construct investment products around computing power, memory, and AI infrastructure. My work focuses on index methodology, constituent selection, weighting, thematic classification, and the changing composition of AI-related semiconductor benchmarks. I publish on this website because AI investment themes are evolving rapidly. Companies that contribute to AI infrastructure can come from different semiconductor categories, making index construction particularly important when evaluating thematic exposure. My research examines AI semiconductor indices, their constituent companies, weighting structures, rebalancing methods, and exposure to GPUs, ASICs, memory, networking, equipment, and advanced packaging. I also compare different definitions of AI infrastructure and study how changes in technology trends influence index composition. Through my articles, I aim to help readers understand the methodology behind AI semiconductor benchmarks and how index design shapes the exposure provided by related ETFs.
AI semiconductor index analyst covering thematic methodology, constituent selection, weighting, rebalancing, GPUs, ASICs, memory, networking, and packaging.
Lucas Bennett is an AI semiconductor index researcher specializing in the construction and methodology of AI-focused benchmarks. His work examines how companies are classified and weighted within AI computing and semiconductor indices.
Lucas follows exposure to GPUs, ASICs, memory, networking, semiconductor equipment, and advanced packaging. He studies how different definitions of AI infrastructure lead to different index compositions.
Index methodology and rebalancing are central to his research. Changes in constituent selection or weighting can materially alter the exposure of an AI-themed benchmark even when its overall theme remains unchanged.
Through his articles, Lucas explains the methodology behind AI semiconductor indices and helps readers understand how index design influences the structure of related ETF products.
His Articles
AI Infrastructure vs. Memory Indices: The 0.9 Correlation
There's a number that keeps appearing in semiconductor factor models: 0.9. It's the rolling 12-month correlation between AI infrastructure capex (data center buildout, GPU deployments, networking equipment) and memory demand (DRAM bit growth, NAND bit growth, HBM shipments).
Cross-Border Capital Flow Characteristics of Hong Kong and Mainland Semi ETFs
Semiconductor themed ETFs in Greater China sit at the crossroads of domestic industrial policy and global capital markets. Hong Kong-listed semi ETFs and mainland A-share semi ETFs both track the chip story, but they live in very different ecosystems. One operates as an offshore gateway for international money, the other as a domestic vehicle tied closely to onshore investor sentiment and policy guidance. When capital moves between the two, it reveals a lot about how investors view China’s semiconductor ambitions, risk, and opportunity at any given moment.
6-Month Lead Validation of China’s Credit Impulse Indicator on Semi CapEx
Macro linkages rarely move in straight lines, but some signals consistently show up ahead of the cycle. China’s credit impulse is one of them. It has long been treated as a leading indicator for global manufacturing, commodities, and emerging markets. Increasingly, investors are asking a more specific question: does China’s credit impulse also lead semiconductor capital expenditure, and if so, by how much? The hypothesis many macro desks now test is a 6‑month lead—a forward window where changes in credit impulse foreshadow shifts in semi CapEx.