Forecasting
Improve Model Performance with Forecast Combination
How combining forecasts can improve robustness when individual models are unstable, biased, or sensitive to sample choice.
View article briefArticles
Research-led articles on quantitative methods and analytical systems.
Research-led articles on forecasting, econometrics, machine learning, risk modelling, financial analytics, clustering, model validation, and automated analytics systems.
Research approach
Research-led, implementation-focused
Hurst Analytics draws on academic literature and applied quantitative research to evaluate methods and translate them into practical systems.
Hurst Analytics draws on ongoing research work with academic collaborators in econometrics, empirical finance, forecasting, statistical modelling, and machine learning.
The emphasis is not theory for its own sake. The goal is to understand which methods are defensible, where they fail, and how they can be implemented in practical systems for forecasting, risk measurement, reporting, and decision support.
Consulting work is informed by up-to-date methods used in academic research and industry practice, with attention to validation, implementation limits, and commercial usability.
Articles
Article briefs
Short previews for a research-led article library. Full articles can be added as each piece is written and reviewed.
Risk Modelling
Improving Risk Management with Quantile Forecasting
Why modelling conditional quantiles can be more useful than focusing only on average outcomes in risk-sensitive settings.
View article briefClustering
How Clustering Pipelines Can Improve Analytical Reliability
How clustering workflows can support more stable segmentation, monitoring, and review when group definitions matter.
View article briefRisk Modelling
Distributional Forecasting for Better Risk Measurement
How forecasting the full distribution can support clearer downside analysis, thresholds, and decision rules.
View article briefFinancial Modelling
Forecasting Volatility: From Statistical Models to Practical Risk Tools
A practical view of volatility models, diagnostics, and the reporting outputs that make them useful.
View article briefMachine Learning
Combining Econometric and Machine Learning Models for Forecasting
Where flexible models can complement econometric structure, and how to compare them without losing interpretability.
View article briefModel Validation
Backtesting Forecasting Models Without Fooling Yourself
Common backtesting traps, including leakage, unstable benchmarks, short holdouts, and misleading evaluation windows.
View article briefModel Validation
Model Risk in Forecasting Systems: Why Validation Matters
Why model performance, assumptions, data quality, and operational use all need review before forecasts become routine.
View article briefEconometrics
Asset Pricing Signals, Forecast Combination, and Out-of-Sample Performance
How asset-pricing signals can be evaluated with disciplined out-of-sample testing and combination methods.
View article briefClustering
Clustering Ensembles for More Stable Financial and Operational Groupings
How ensemble approaches can reduce instability when grouping assets, customers, products, or operational units.
View article briefMachine Learning
When Machine Learning Improves Forecasting - and When It Does Not
A restrained look at where machine learning helps forecasting workflows, and where simpler benchmarks still win.
View article briefReporting Automation
From Static Reports to Automated Analytics Systems
How recurring analysis can move from manual reports into repeatable systems with validation, scheduling, and support.
View article briefDiscuss a Project
Share the analytical decision, reporting burden, or forecasting requirement you want to improve.