Scottish Financial Risk Academy Covid-19 Projects
The Scottish Financial Risk Academy has several coronavirus-related projects underway that would benefit from further assistance from SFE members.
Any assistance from members or quants who wish to join this national effort that will ultimately be coordinated through the Royal Society would be greatly appreciated.
You may either form a team and submit your proposal for research to: https://epcced.github.io/ramp/
Alternatively, you may participate in existing project efforts underway. Examples of these projects are outlined below. If you would like more information on the projects and would like to support by offering expertise please contact SFRA Academic Director Gareth Peters, firstname.lastname@example.org
- Heriot-Watt and Chinese Academy of Science are building network dynamic models for epidemic-economic integrated impact models of the Covid-19 epidemic and stress testing components on the macroeconomy for understanding paths to recovery from recession/depression that may arise.
- economists and asset managers required or experts on capital stress testing frameworks.
- University College London-Heriot Watt-JP Morgan teams are building stochastic population epidemic models based on compartment models with self-excitation features and population mixing dynamics with age and gender structures.
The goals are to determine and provide advice from modelling on how likely it will be for epidemic to spread through different age structures and gender as a second wave of infection and the time-scale/severity of this.
- experts on network modelling and epidemics (from say past PhD or insurance side etc for morbidity and mortality modelling in life insurance).
- Experts on epidemic mixing in populations in city and urban environments for influenza.
- A team at Heriot Watt is working on sentiment analysis of news reports and medical reports to produce Natural Language Processing (NLP) summary of information for perception and response to this crisis.
Details of this global challenge call: https://pages.semanticscholar.org/coronavirus-research
and CORD-19 challenge.
- Machine Learning and Natural Language Processing (NLP) experts required
- There is a worldwide call for assistance in NLP experts to process data from medical journals and news media to understand causal risk drivers of this epidemic -- there are many NLP experts in banks and we could use some right now.
- Heriot-Watt, University of Technology Sydney (UTS) Australia and University College London
A team is starting to explore Insurance linked securities such as Pandemic bonds and how to structure these in future that will not be index-based (and not payout or payout far too late as is the case with current World Bank Bonds) but rather have triggers and pricing kernels which reflect the epidemic risk factors.
- Insurance and reinsurance experts required in particular expertise in pandemic and mass mortality bonds from epidemics.
- A Heriot-Watt team is working on a range of time-series models for forecasting possible loads on hospitals daily from new infections. This would be the input into network stochastic queuing models such as Jackson networks and beyond. This would have particular value in regard to decision making for triage and decision making on efforts to send people to different hospitals and optimal allocations of treatments. Understanding network queuing models on how to manage triage at hospitals and ventilator management
- service sector experts who have worked on networks queuing models for customer service, stochastic queuing model experts
A team at the Credit Research Centre at the University of Edinburgh is working with a fintech to look at changes in individual expenditure patterns following the COVID-19 outbreak and the effects of the pandemic on sources of income. This is part of a project to understand and predict paths to financial distress.
- To support a wider project on the effects of the pandemic on incomes and expenditures the CRC would like to gain access to anonymised bank transactional data.
For this project, please contact Professor Jonathan Crook, University of Edinburgh, email@example.com