ILXQuant Large Language Model Value Chain Basket Index

Fixed-basket index of 19 companies with genuine economic exposure to large language models -- the firms training frontier models, the silicon they train on, the data that feeds them and the software surfaces that monetise them. Layers covered are frontier model developers (Microsoft, Alphabet, Meta, Amazon, Alibaba, Tencent, Baidu, NAVER, SoftBank), training and inference silicon (NVIDIA, Broadcom, AMD, TSMC), data and retrieval infrastructure (Snowflake, MongoDB), data labelling and reinforcement learning from human feedback (Innodata, Appen), and application surfaces (Palantir, ServiceNow). Deliberately global -- the second frontier-model cluster is Chinese and Korea and Japan host sovereign-model programmes that Western thematic baskets largely omit. Rebalances quarterly to equal weight.

Index identifier: ILXLLM
Category: Basket
Universe: Fixed basket -- 19 large language model value chain companies (equal weight)
Currency: USD
Inception: 2020-10-05
Rebalancing: Quarterly (first Monday of March, June, September, December)
Index fee: 0.60% p.a.
Reference benchmark: Technology SPDR (XLK)

Performance and risk

All figures below are measured over a trailing window of up to 1,260 trading days (approximately five years), not since index inception.

Performance and risk — 2021-08-06 to 2026-08-13 00:00:00
Annualised return	+24.76%
Sharpe ratio	0.80
Calmar ratio	0.46
Volatility	30.85%
Maximum drawdown	-54.24%
Return, 1 month	—
Return, 3 months	—
Return, 1 year	—
Return, year to date	+27.24%

Full performance history, drawdown analysis, rolling risk metrics and portfolio breakdowns for this index are available in the ILX Quant platform.

Customized Indexing and Advanced Analytics

ILX Quant is a research platform for systematic equity strategies. It supports a wide range of approaches, including factor-based, thematic, alternative and regional strategies, enabled by scalable, high-speed development and a high degree of customization.

Methodology

Rapid Index Construction

Create and iterate systematic equity strategies using natural language instructions, without writing code. From factor-based models to complex, multi-layered strategies, strategies can be defined, adjusted and rebuilt in seconds. All strategies remain fully rules-based, transparent and consistently validated, with ongoing maintenance and rebalancing handled automatically.

Flexible Data Integration

Incorporate a wide range of data sources, including alternative datasets, regional universes, custom signals and external inputs. Each strategy can be precisely tailored to reflect specific ideas, constraints or market views, without limitations on structure or scope.

Scalable Analytics and Exploration

Every strategy is supported by a comprehensive analytics and risk framework, including performance breakdowns, drawdowns, rolling metrics and risk attribution. Strategies can be tested, compared and refined at scale, enabling a continuous and systematic research process.

Our Approach

Factor-Based

Systematic exposure to established risk premia such as value, momentum, quality and low volatility, capturing persistent drivers of returns through transparent, rules-based frameworks.

Thematic and Signal-Based

Targeted exposure to long-term trends including AI, clean energy and digital assets, structured to reflect evolving market themes within clearly defined strategy frameworks.

Sector and Industry

Focused strategies across sectors and industries, enabling precise and structured allocation to specific segments of the equity market.

Alternative Strategies

Advanced approaches including long-short strategies, AI-assisted systematic models, and other non-traditional strategy constructions — all rules-based, expanding beyond traditional long-only exposure.

© ILX Quant

info@ilxquant.de