Workshop on Risk Management and Predictive Analytics in Financial and Actuarial Mathematics, January 2025

This workshop, supported by the London Mathematical Society, aims to bring together researchers at all career stages who are interested in finance and actuarial mathematics. A particular focus will be on sentiment analysis with applications in risk assessment, financial modelling in jump detection, and predictive analytics in coherent multipopulation mortality modelling. We recognise the value of diversity in our discipline and welcome all attendees, with a special invitation to PhD students. The event is hybrid, and you can join online with Meeting ID: 322 126 705 504 (Passcode: Jb3MH6i4).

Speakers

• Jia Shao (University of Birmingham)
• Paresh Date (Brunel University London)
• Bo Wang (University of Leicester)
• Maggie Chen (Cardiff University)

Schedule

13:30 - 14:15, Jia Shao, Leveraging sentiment analysis to improve customer satisfaction in UK banks
14:15 - 15:00, Paresh Date, ESG news-enhanced volatility prediction
15:00 - 15:30, Coffee Break
15:30 - 16:15, Bo Wang, Multipopulation mortality modelling and forecasting: the weighted multivariate functional principal component approaches
16:15 - 17:00, Maggie Chen, Can machine learning models better volatility forecasting? A combined method
17:00 - 18:00, Social reception to celebrate the lectureship

Abstracts

Leveraging sentiment analysis via text mining to improve customer satisfaction in UK banks (Shao)

Abstract: This study examines the role of online customer reviews through text mining and
sentiment analysis to improve customer satisfaction across various services within the UK
banking sector. Additionally, the study analyses sentiment trends over a five-year period.
Both positive and negative sentiments provide valuable insights. The results indicate a high
prevalence of negative sentiments related to customer service and communication, with
HSBC and Santander receiving 90.8% and 89.7% negative feedback, respectively, compared
to Tesco Bank’s 66.8%. Key areas for improvement include HSBC’s credit card services and
call center efficiency, which experienced increased negative feedback during the COVID-19
pandemic. The findings also demonstrate that DistilBERT excelled in categorizing reviews,
while the SVM model, when combined with customer ratings, achieved 96% accuracy in
sentiment analysis. Online customer reviews become more informative when categorised by
service sector. To enhance customer satisfaction, bank managers should pay attention to both
positive and negative reviews, and track trends over time.

ESG news-enhanced volatility prediction (Date)

Abstract: We study predictive ability of news for stock price crashes for a set of governance
failure events across world markets, over the past 15 years. From a database which provides a
number of open-source news items daily for each company tagged as positive, negative or
neutral, we construct a simple and easily explainable extension of GARCH model which uses
positive and negative news sentiment as exogenous inputs. We demonstrate that this
significantly enhances ability to predict an increase in volatility due to governance failure at a
company. The broad objectives of the work, which is still ongoing, are (i)  to provide hard
quantitative evidence for good corporate behaviour (as evidenced in a large number of
positive news items) generally leading to good stock price performance (as evidenced by
'buy' or related investment signal); (ii) to combine E, S and G sentiment scores for
standardised ESG reporting, with a peer-reviewed and open source methodology.

Coherent multipopulation mortality modelling and forecasting: a multivariate functional principal component approach (Wang)

Abstract: Human mortality patterns and trajectories in closely related populations are likely
linked together and share similarities. It is always desirable to model them simultaneously
while taking their heterogeneity into account. When a mortality model is applied to each
population separately, they tend to result in divergent forecasts of life expectancy in the long
term. We introduce a method for joint and coherent mortality modelling and forecasting of
multiple subpopulations using the multivariate functional principal component analysis
techniques, which ensures the non-divergent forecasting in the long run when several
subpopulation groups have similar socio-economic conditions or common biological
characteristics. We demonstrate the proposed methods by using sex-specific mortality data,
and the forecast performances are compared with several existing models, including the
independent functional data model and the Product-Ratio model.

Can machine learning models better volatility forecasting? A combined method (Chen)

Abstract: Volatility forecasting for Bitcoin constantly gains attention due to the increased in- vestment interest and the high riskiness of cryptocurrencies. The traditional forecast- ing models, such as the GARCH family models, are widely adopted. However, there should be careful consideration whether whether they can capture extreme market shocks and associated extreme risk (e.g. contagion). Hence, we fit several GARCH models and the EGARCH shows the best goodness of fit. We further take its his- torical volatility observations for an automated forecasting solution,using the Long short-term memory (LSTM) neural network to take predictions. Our results show clear improvement in volatility forecasting regarding both the model’s in-sample and out-of-sample accuracy. More importantly, the LSTM can optimize information intake through the short- and long-memory states. Overall, our new LSTM neural network model is more robust in reflecting to market shocks and regime changes.

Participation

Participation is open to all university-level staff and students. All participants of the workshop are expected to adhere to the School of Mathematics Code of Conduct. The School of Mathematics has established a programme to offer funded child-care services to visiting researchers. To take advantage of this opportunity, it is recommended that you contact the organiser at the earliest convenience to ensure proper arrangements are made.

Funding

We acknowledge funding from the London Mathematical Society via Celebrating New Appointments (Scheme 9), and School of Mathematics, University of Birmingham.