Statistical modelling and applications for sustainable-development goals
Bibliographic record
Abstract
In 2015, all United Nations member states adopted the Agenda 2030, which consists of 17 Sustainable Development Goals to create a better world in terms of social, economic, and environmental development by 2030. Numerous countries and governments have initiated measures to achieve these goals, including curbing traditional energy consumption, advancing renewable energy development, cutting carbon emissions, and addressing climate risks. In this thesis, statistical modelling techniques are developed to examine economic and financial challenges encountered in the pursuit of sustainable development.\nThe first study in this thesis explored the dynamic link between Canada's economic growth and renewable energy use using a panel autoregressive distributed lag model framed within the neoclassical production function. We incorporated an OECD-based indicator to identify Canada's economic phases and used the Pooled Mean Group method to analyse long-term and short-term correlations. The findings show a unidirectional causality going from renewable energy to economic growth only during expansion phases in the short-term, highlighting the need for policies that recognise the nonlinear connection between renewable energy and economic growth.\nIn our second study, we analysed the US stock market's reaction to both types of physical climate risks (chronic and acute) as well as transition climate risk. Using a multivariate hidden Markov model, we employ two climate variables and one sentiment variable to build an indicator for the chronic risk. Acute and transition risks are deemed to be reflected in the natural disaster and policy news, respectively. We evaluated their effects on stock returns using an event study methodology. Our results indicate that some sectors are more susceptible to climate risks than others. Firms with lower environmental scores face greater exposure, affecting their stock returns negatively. This implies that enhancing environmental performance can boost a company's financial resilience against climate risks.\nWe modelled the Emission Allowance price dynamics in the EU Emission Trading System in our third study. Capitalising on prior studies, we integrated a Markov-switching mechanism into four stochastic models. Parameters were estimated using change of reference probability measures alongside the EM algorithm. The fitting accuracy and usefulness of our model were assessed through an out-of-sample forecasting and the pricing of European-style call options. Notably, the Markov-switching Geometric Brownian motion model surpassed both the non-Markov switching and other Markov-switching stochastic models in in-sample and out-of-sample performance.\nThe fourth study in this research work investigated the influence of green bond issuance on an issuer's environmental performance, utilising an interrupted time series with a control group. This study also looked into how companies' financial characteristics and specific green bond data can enhance an issuer's environmental performance, using both the Random Forest and Generalised Additive models. Our findings indicate that the environmental performance of most issuers improves following the issuance of green bonds. Additionally, we discovered that certain company's characteristics as well as the specific features of the green bonds play significant roles in determining the efficacy of green bonds in enhancing a company's environmental performance. These findings are valuable for investors when selecting green bonds.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.010 | 0.050 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.004 | 0.005 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.012 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".