Romeonruapeheamlani
Latent Variable Research
Our Faculty
Researchers who work with latent variable methods every day
Each faculty member brings active institutional or research experience in factor extraction, state-space estimation, and return decomposition — not a historical relationship with the field but a current one, sustained through ongoing work with real asset return data and peer-reviewed publication.
Contact usFaculty with active research experience in quantitative return modelling and latent variable estimation
The platform's faculty are drawn from quantitative research roles at asset management
firms, central banks, and research-intensive academic departments. Each brings a specific
area of methodological depth — from Bayesian unobserved component models to high-dimensional
factor structures — and continues to engage with current empirical literature alongside
their platform responsibilities.
Dr. Eleanor Voss, Lead Quantitative Researcher
Lead Quantitative Researcher, Latent Variable Methods
Sydney, Australia — Australian National University affiliation
Dr. Voss holds a doctoral degree in econometrics and spent a decade as a senior quantitative analyst at a major Australian asset management firm before joining the platform as lead researcher. Her work focuses on the application of state-space methods to return decomposition across multi-asset portfolios, and she has published on the identification challenges that arise when latent factor structures are applied to illiquid asset classes. She leads curriculum development for the Practitioner and Research tracks and reviews all capstone submissions.
Core areas of methodological expertise
State-space model specification
Kalman filter estimation and diagnostics
Unobserved component decomposition
Factor model identification theory
Dr. Marcus Okafor, Senior Research Fellow
Senior Research Fellow, Dynamic Factor Models
Melbourne, Australia — University of Melbourne research associate
Dr. Okafor completed his doctoral research at the London School of Economics, focusing on high-dimensional factor models for international equity return data. He subsequently held a research position at a European central bank's financial stability division before relocating to Australia. His platform contributions centre on dynamic factor model architecture and the treatment of regime-switching behaviour in latent factor dynamics. He brings a comparative international perspective to how latent variable methods are applied across different regulatory and market environments.
Core areas of methodological expertise
Dynamic factor model architecture
Regime-switching latent structures
High-dimensional factor extraction
Cross-sectional return variation analysis
Dr. Sian Hartley, Applied Risk Researcher
Applied Risk Researcher, Attribution and Volatility
Brisbane, Australia — Queensland University of Technology affiliation
Dr. Hartley's research sits at the intersection of latent volatility modelling and applied risk attribution, with a particular focus on how unobserved volatility components can inform position-level risk analysis in institutional portfolios. She spent several years as a risk methodology specialist at a regional sovereign wealth research team before moving into full-time research. On the platform, she leads the applied modules in the Practitioner track and developed the framework used for evaluating capstone submissions in the risk attribution segment of the curriculum.
Core areas of methodological expertise
Latent volatility structure modelling
Risk attribution via factor decomposition
Bayesian posterior estimation methods
Drawdown analysis and latent signals
Dr. Tomás Reyes, Econometric Methods Specialist
Econometric Methods Specialist, Foundations and Inference
Perth, Australia — University of Western Australia research fellow
Dr. Reyes specialises in the inferential foundations of latent variable models, with particular attention to the conditions under which maximum likelihood estimation produces reliable factor solutions and the diagnostics that reveal when those conditions are not met. His doctoral work examined rotation indeterminacy in exploratory factor analysis applied to macroeconomic return drivers, and he has since extended this work into the Bayesian estimation setting. He leads the Foundations track curriculum and is responsible for the platform's statistical methodology standards, ensuring that every estimation claim is supported by honest treatment of the underlying assumptions.
Core areas of methodological expertise
Maximum likelihood estimation theory
Covariance structure analysis
Factor rotation and identification
Model selection and diagnostic testing
Recognition and institutional affiliations
Quantitative Research Excellence Recognition
Acknowledged by a leading Australian financial research body for contributions to applied latent variable methodology in institutional settings.
2024
Peer Review Distinction in Factor Research
Faculty research on high-dimensional factor model identification received distinction recognition from a peer-reviewed international quantitative finance journal.
2023
Academic Partnership in Applied Econometrics
Formal research affiliation established with a leading Australian university department specialising in applied econometrics and return modelling.
2022
From the Faculty
What distinguishes this platform's faculty from a standard professional development roster is not credential length but the nature of their ongoing engagement with the field. Each member continues to work with real return data, to submit work for peer review, and to grapple with the estimation problems that practitioners face when latent variable methods meet messy institutional reality. The result is a faculty that does not simply describe methods but can explain, with specificity, where those methods tend to break down and why that matters for the analyst sitting in front of a covariance matrix at the end of a reporting quarter. That kind of embedded, active knowledge is difficult to manufacture and easy to recognise when it is present. It shows in how questions get answered, how ambiguity gets acknowledged, and how the limits of a model get foregrounded rather than quietly omitted. The faculty assembled here holds that standard collectively, and the platform's curriculum reflects it at every stage.
The faculty's collective experience spans asset management, central banking, sovereign wealth research, and academic econometrics — not as historical credentials but as active, ongoing contexts in which latent variable methods are tested against real data. That breadth means the curriculum draws on multiple interpretive traditions simultaneously, which is precisely why the peer-to-peer framing of the platform feels accurate rather than aspirational. When a faculty member describes a failure mode in Kalman filter estimation, they are drawing on a specific encounter with that failure, not summarising a textbook account of it. That distinction is the core of what the platform offers.
Dr. Marcus Okafor
Senior Research Fellow
Rotation indeterminacy is not a footnote in factor analysis — it is the central
interpretive challenge, and practitioners who treat it as a technicality
eventually run into it in the most inconvenient circumstances.
Dr. Tomás Reyes
Econometric Methods Specialist
Risk attribution via latent factors only makes sense if you are honest about
what the model cannot tell you. We design every applied module around that
honesty as a structural requirement, not an afterthought.
Dr. Sian Hartley
Applied Risk Researcher
The gap between what state-space methods can do in theory and what they produce
in practice with real institutional data is exactly the territory this platform
is built to navigate.
Dr. Eleanor Voss
Lead Quantitative Researcher
High-dimensional factor models are powerful and genuinely difficult to specify well. We do not smooth over that difficulty — we treat it as the most important thing for a practitioner to understand.
Dr. Marcus Okafor
Senior Research Fellow
Maximum likelihood estimation gives you a number. Understanding whether that
number is meaningful requires a level of diagnostic rigour that most
professional development resources skip entirely.
Dr. Tomás Reyes
Econometric Methods Specialist
Latent volatility structures are among the most practically consequential outputs of the LVRM framework, and also among the most frequently misinterpreted. That tension is where the most useful teaching happens.
Dr. Sian Hartley
Applied Risk Researcher
Speak with a faculty member before you commit to a track
Each of the three methodology tracks is designed for a specific stage of analytical development,
and the right entry point depends on the depth of your existing foundations in statistical
inference and factor model theory. Faculty members hold periodic open sessions for prospective
participants, where you can discuss your current analytical context, ask methodological
questions directly, and determine which track aligns with the work you are actually doing. These
sessions are not sales conversations — they are substantive exchanges with active researchers
who would rather help you find the right starting point than have you begin at the wrong one.