Research

Ten projects, five themes

Each project below is led by a named principal investigator from the Department of Econometrics and Business Statistics.

Theme

Financial and neuro-econometrics

Volatility is the common object. The same class of conditional-variance models is applied to market returns and to cortical signal, on the argument that clustering in the second moment carries information the mean equation discards.

2 projects

Volatility clustering in cortical dynamics: conditional heteroskedasticity as a group marker in the alcoholism EEG cohort

Principal investigators Nazirul Hazim A. Khalim · Erniel B. Barrios

Tests whether conditional heteroskedasticity in electroencephalographic recordings distinguishes the alcoholic and control arms of a public cohort, treating volatility clustering in cortical signal as a candidate group marker rather than as noise to be filtered out.

170Mobservation rows

111hypotheses tested

B = 200bootstrap replications

UCI COGAEEG cohort

Non-parametric volatility testing
Bootstrap inference
Signal processing

Active

AI News Shocks and Cross-Sector Volatility between AI and Electricity

Principal investigators Nazirul Hazim A. Khalim · Akram Hasanov

Measures how news shocks about artificial intelligence transmit into volatility across AI-exposed equities and the electricity sector, and establishes which direction the transmission runs in.

Event study
EGARCH
Volatility connectedness

Active

Theme

Climate, agriculture and environment

Production under climate stress, estimated at the frontier. Work in this theme asks what extreme weather does to the technology available to producers, and what it does to how far below that technology they actually operate.

3 projects

Agricultural Production and Greenhouse Gas Emission: Dynamic Stochastic Frontier Model with Extreme Weather Indicators

Principal investigators Erniel B. Barrios · Lee How Chinh

Estimates a dynamic stochastic frontier for agricultural production with greenhouse gas emission treated as a joint outcome, and admits extreme weather indicators into the inefficiency component.

Stochastic frontier
Dynamic panel
Climate indicators

Active

Climate-production capacity and environmental disclosure in Chinese agricultural firms: Frontier Shifts and Efficiency Gaps

Principal investigators Lee How Chinh · Erniel B. Barrios

Separates two things environmental disclosure by Chinese agricultural firms might be associated with: movement of the production frontier itself, and movement of the firm relative to it.

Stochastic frontier
Firm-level panel
Environmental disclosure

Active

Climate change, agricultural productivity and food security: global evidence on the transmission channel

Principal investigator Hashem Salarzadeh Jenatabadi

Traces the channel running from climate change to food security through agricultural productivity, using cross-country evidence to establish where in that chain the effect is concentrated.

Structural modelling
Cross-country panel
Food security

Active

Theme

Information, narrative and disinformation

Quantitative computational social science on public discourse. The work combines large-corpus text modelling with structural estimation, so that claims about narrative and credibility are testable rather than interpretive.

2 projects

Monash Change Narrative Lab

Principal investigator Nazirul Hazim A. Khalim

A standing programme measuring how narratives of change are formed, carried and revised across public discourse, using topic modelling over large multilingual corpora alongside structural estimation of the results.

Topic modelling
Narrative measurement
Multilingual corpora

Active

Disinformation Equilibrium – Truth-Bearing Capacity: Competition, Credibility, and the Supply of Verified Information

Principal investigator Nazirul Hazim A. Khalim

Models the conditions under which a competitive information market supplies verified content, and asks what competition and credibility do to the equilibrium quantity of truth-bearing information.

330,000+posts and comments

80+political actors

BERTopic + STMtext pipeline

Computational social science
Structural topic models
PLS-SEM

Active

Theme

Computational heritage and culture

Methods built for biological descent, applied to material culture. The question is whether design traditions can be related to one another by inference rather than by connoisseurship.

1 project

Multi-Model Computational Phylogeny of Southeast Asian Textiles

Principal investigators Nazirul Hazim A. Khalim · Lee How Chinh · Rahman Rom

Infers relationships among Southeast Asian textile traditions from motif and construction characters under several competing phylogenetic models, and compares what those models disagree about.

Phylogenetic inference
Image analysis
Cultural heritage

Active

Theme

Household finance and well-being

Panel evidence on how households absorb shocks and on what moves quality of life over time, with model explanations that a policy reader can act on.

2 projects

Saving Composition, Digital Payments and Household Financial Resilience in Developing Economies

Principal investigator Nur Syazwani Mazlan

Asks whether what households hold their savings in, and whether they have adopted digital payments, changes how well they absorb financial shocks in developing economies.

Household finance
Financial inclusion
Panel data

Active

What Shapes Well-Being Over Time? Uncovering the Drivers of Quality of Life and Mental Health Through Longitudinal Modelling and Explainable AI

Principal investigators Mogana Darshini Ganggayah · Shin Zhu Sim

Identifies which factors move quality of life and mental health across repeated observation of the same individuals, pairing longitudinal models with explainable AI so that the drivers can be read off the fitted model directly.

Longitudinal modelling
Explainable AI
Health analytics

Active

Capability

Methods and platforms

Estimators, pipelines and tooling that carry across the portfolio rather than belonging to any single project.

Time series and volatility

Time-varying parameter VAR and connectedness measurement
EGARCH and related conditional-variance models
Non-parametric volatility testing
Rolling-window structural break detection
High-frequency series at five-minute resolution

Computational statistics

Bootstrap inference and resampling designs
Monte Carlo study design at scale
Stochastic frontier and efficiency estimation
PLS-SEM with HTMT and bootstrap confidence intervals
Bayesian estimation by MCMC, including DSGE calibration

Text, image and machine learning

BERTopic and structural topic models
Transformer-based classification over multilingual corpora
Image and whole-slide analysis pipelines
Explainable AI for model-driven policy reading
On-premise inference and fine-tuning of open-weight models

Named tooling

EconBench

A benchmark of large language model accuracy on econometric estimation tasks, with a companion dataset release. Run locally across multiple open-weight models, which is the only way the inference volume is affordable.

ESTIMAND-OS

An applied-econometrics research platform for large-scale multiverse and Monte Carlo computation, supporting methodological work across the portfolio.

DSGE Explorer

A browser-based front end for calibrating and simulating dynamic stochastic general equilibrium models, built for teaching and for dissemination of results.

ASEAN Disinformation Index

A measurement instrument for regional information integrity, built from large social media corpora using topic modelling and structural estimation.

Replication

Where the underlying data permit, our published work ships with a full replication package in R or Python. Requests for materials go through the contact page.

Research with us

We supervise higher-degree research candidates, host visiting researchers, and develop joint proposals with partners who bring a problem and data.

Get in touch Meet the team