Pulse
Multi-asset regime-aware quantitative research
PULSE is an independent research architecture for sophisticated professional and institutional participants, providing structured, non-directive analysis of evolving market regimes, factor behaviour, and model-derived updates, with current coverage of the S&P500 and Gold.
Research Scope
PULSE evaluates each defined market series under the premise that market dynamics evolve across distinct regimes rather than conforming to a single, uniform environment. Regime classification provides the basis for examining variations in momentum, volatility, liquidity, participation, and other relevant market characteristics.
Modelling
The PULSE research process comprises sequential stages, such as:
• data ingestion and series construction
• market-state and regime classification
• momentum and factor decomposition
• probabilistic state-transition analysis
• distributional assessment
• out-of-sample validation
Each stage is evaluated for methodological consistency across changing market conditions.
Output
PULSE produces structured research outputs describing market conditions, exposure states, and quantitative evidence. Currently applied to the S&P500 and Gold, the methodology is designed for broader application across financial markets.
Projected Metrics (as of July 2026)
SR 2.67 | DD -0.01 | t 2.64
Higher-order moments are incorporated into the statistical inference framework, with a 95% confidence level defined as the target threshold for walk-forward validation.
-
PULSE separates data preparation, factor construction, optimisation, statistical validation, and research distribution into distinct analytical stages. This structured workflow enhances methodological transparency, mitigates overfitting risk, and enables each stage to be independently evaluated before contributing to the final research output.
-
Large-scale parameter evaluation and repeated walk-forward validation are conducted using the University of York’s VIKING high-performance computing facility, enabling extensive empirical testing across millions of parameter combinations and varying market conditions. This computational framework supports reproducible research and the identification of statistically persistent model behaviour.
Performance & Live Updates
Performance plots for each market covered, alongside examples of the live, time-stamped model updates delivered to subscribers, including statistical classification, reference market observations, and projected metrics for independent interpretation.
