Entropic approximate learning for financial decision-making in the small data regime

Financial decision-making problems based on relatively few observations and several explanatory variables can be problematic for the common machine learning (ML) tools, since they cannot efficiently discriminate the relevant information. To investigate the challenges of this “small data” regime, we employ several state-of-the-art ML methods for predicting whether three selected stocks from the Swiss Market Index will outperform the market, by using, as classification features, a set of commonly used technical indicators. We show that the recently introduced entropic Scalable Probabilistic Approximation (eSPA) algorithm significantly surpasses its competitors in both prediction accuracy and computational cost. We then discuss the interpretability of the employed ML methods and suggest some statistically derived heuristics to select the most appropriate and parsimonious financial decision-making candidate model.     

Perceptions of the learning context and learning approaches: Implications for quality learning outcomes in accounting

This study analyses learning approaches, course perceptions and learning outcomes of a group of second year accounting students at an Australian university using qualitative data analysis techniques. The research method involves the development of a series of matrices linking types of motives and strategies used by students in their study, together with their perceptions of the learning context associated with learning outcomes. The study focuses on assessing the links between learning approaches and a qualitative assessment of students' conceptual understanding of aspects of financial accounting studied at the undergraduate level. The results confirm how individual differences in the perceptions of the learning context relate to study motives and strategies. The findings show how different forms of memorisation relate to study strategies and how the completion of accounting tasks link to students' perceptions of course requirements. There was also some evidence that, in terms of learning outcomes, students with sophisticated levels of understanding of concepts, tended to have consistent deep and achieving approaches to learning. This result was compared with students' academic performance as a measure of learning outcome. Discrepancies between these two measures of learning outcome are highlighted in the conclusions. The findings strengthen the case for further investigation of the use of measures other than academic performance in examining relationships between learning approaches and learning outcomes.     

The budgetary interview: Intentional learning for students in governmental and non-profit accounting

Learning-to-learn skills are critical to the future success of accounting students. This paper reports on a budgetary interview exercise that helps students develop as intentional learners. Students select a government or non-profit agency to investigate, arrange an interview with an agency official to discuss the budgetary process, write a technical paper on what was learned, and record their reflections on the experience. The budgetary interview exercise was implemented with undergraduate students in governmental and non-profit accounting courses over four academic years (one course per year). Effectiveness of the exercise was assessed via content analysis of student papers and reflections. Results indicate the exercise was highly effective in helping students develop intentional learning skills. Furthermore, students successfully connected classroom material to “real-world” practice, and most students reflected on potential careers in governmental or non-profit sectors. Appendices provide materials that instructors may use to implement this exercise.     

Bridging knowing and learning: a suggestion for accounting education

This teaching note describes an exercise which appears to prompt students to reflect on their learning. The exercise is based on five pairs of words concerned with knowing and learning: (1) learn/study, (2) educate/train, (3) know/understand, (4) teach/tutor, and (5) student/pupil. The exercise may appear trivial, even frivolous, but when successful, it can compel students to reflect, often for the first time, on the meaning and purpose of their educational experience. Frequently, students make explicit previously unarticulated views and feelings about the process and content of their educational programmes.     

Machine learning for quantitative finance: fast derivative pricing,hedging and fitting

In this paper, we show how we can deploy machine learning techniques in the context of traditional quant problems. We illustrate that for many classical problems, we can arrive at speed-ups of several orders of magnitude by deploying machine learning techniques based on Gaussian process regression. The price we have to pay for this extra speed is some loss of accuracy. However, we show that this reduced accuracy is often well within reasonable limits and hence very acceptable from a practical point of view. The concrete examples concern fitting and estimation. In the fitting context, we fit sophisticated Greek profiles and summarize implied volatility surfaces. In the estimation context, we reduce computation times for the calculation of vanilla option values under advanced models, the pricing of American options and the pricing of exotic options under models beyond the Black–Scholes setting.     

Changing management roles in the Scottish NHS: implications for management learning and development

The UK government's austerity means that public service providers, such as the National Health Service (NHS), are looking for efficiencies from service integration and collaborative working. This paper highlights how NHS (Scotland) management is coping with these changes and how the role of the manager and the nature of management development is being transformed. New forms and processes of management learning and development are needed for collaborative partnership working in multi-agency public service environments.     

Reasoning about safety management policy in everyday terms: a pilot study in citizen engagement for the UK railway industry

This paper reports on a pilot study in citizen engagement which formed part of a broader stakeholder engagement and consultation programme addressing safety decision‐making for UK rail industry activities. In addition to developing tools to support engagement initiatives, the study was concerned specifically with investigating everyday lay notions of what is a ‘reasonable’ basis for establishing safety. In view of the technical complexity of this issue, the exercise therefore presented an important methodological challenge: how to ‘translate’ specialised economic and legal issues in such a way that lay citizens were able to grasp, and reason about, these issues in an informed and considered way. The engagement exercise worked well in terms of its capacity to promote such a process of informed consideration, and in being ‘user friendly’ for participants. Despite the exercise involving a relatively small number of discussion group meetings, the quality and depth of the evidence collected allows some cautious provisional conclusions to be drawn regarding lay sensibilities concerning certain technical aspects of rail safety management.     

Forecasting for regulatory credit loss derived from the COVID-19 pandemic: A machine learning approach

The economic onslaught of the COVID-19 pandemic has compromised the risk management of financial institutions. The consequences related to such an unprecedented situation are difficult to foresee with certainty using traditional methods. The regulatory credit loss attached to defaulted mortgages, so-called expected loss best estimate (ELBE), is forecasted using a machine learning technique. The projection of two ELBEs for 2022 and their comparison are presented. One accounts for the outbreak's impact, and the other presumes the nonexistence of the pandemic. Then, it is concluded that the referred crisis surely adversely affects said high-risk portfolios. The proposed method has excellent performance and may serve to estimate future expected and unexpected losses amidst any event of extraordinary magnitude.     

Managing group processes in transdisciplinary future studies: How to facilitate social learning and capacity building for self-organised action towards sustainable urban development?

Challenges for future urban development are complex and characterised by ambiguous problem definitions or unclear, conflicting and dynamically changing goals. Transdisciplinary research promises new ways of dealing with uncertainty and complexity by including non-academic actors into the research process and fostering social learning for better and more effective research. Depending on the level of participation and the number and heterogeneity of actors involved, appropriate designs for group processes but also associated skills are essential. In this article, we scrutinise the dynamics of groups to better understand how to effectively promote social learning and capacity building for selforganised action beyond project enc. Based on experiences of a participatory scenario planning process in the city of Korneuburg and substantiated with theories on groups and their development, we conclude with five propositions emphasising researchers’ responsibility in processes of societal change, the role of external facilitators, the scope and time needed for group building, the acknowledgement of various phases of group processes as well as requirements for social learning.