https://www.um.edu.mt/library/oar/handle/123456789/692 Thu, 05 Nov 2026 14:44:23 GMT 2026-11-05T14:44:23Z https://www.um.edu.mt/library/oar/handle/123456789/150787 Title: Sectoral and regional volatility connectedness : the case of CDS spreads and equities Authors: Manicaro, Christian Abstract: This study analyses volatility connectedness at sectoral and regional level within and across the US, UK, EU and Japanese regions between the CDS and equity markets. Analysis is made on 32 sectors and 70 sub sectors within the regions under study with each having 2,479 observations, covering the period between 2008 until June 2017. The sample is divided between crisis and after crisis period and the novel connectedness index by Diebold Yilmaz ( is proposed. The domestic and regional analysis show that connectedness between the two asset classes is in general higher during the crisis period. Although the static Gaussian results for the regional analysis show low levels of connectedness across the board, the dynamic analysis show significant connectedness levels, with level s being predominantly higher during the crisis period, signifying contagion effects also at regional level between the two asset classes. When considering the dynamic volatility connectedness between the two asset classes, equity is the asset class which transmits volatility the most. In the US and EU connectedness between the two asset classes in most sectors is predominantly large during disturbed periods, particularly the 2009 crisis and the EU sovereign crisis. Sun, 01 Jan 2023 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150787 2023-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150739 Title: How machine learning based analysis of financial ratios can assist beginner investors Abstract: This study addresses the gap between the increase in retail investors joining the market, but the financial literacy has not kept pace. This subject is referred to as the information and action gap, where beginners struggle to interpret complex financial data. This research aims to investigate how Explainable Artificial Intelligence can be applied to simplify fundamental stock analysis. A quantitative experimental approach was adopted using a dataset of US companies, and being processed within WEKA, a machine learning platform. Fourteen key financial ratios were analysed using a 2 method, including supervised and unsupervised machine learning. The J48 Decision Tree was applied for supervised classification and achieved an accuracy of 65.71%, and also generated a transparent tree of how its logic classifies the data. Simple K-Means clustering was used as an unsupervised machine learning algorithm that identifies natural groupings within the data, and this machine learning algorithm achieve 80% alignment with the pre-defined company classifications. The findings prove that machine learning models can effectively structure and simplify financial data into interpretable outputs and support user understanding. However, the moderate predictive performance highlights the limitations of the models in real-world investing scenarios. In conclusion, this study demonstrates the potential of explainable machine learning as an educational tool that enhances financial understanding, rather than replacing human decision-making. The results support a shift toward FinTech systems that prioritise transparency, user learning, and informed decision support. Description: B.Sc. (Hons) Bus.& IT(Melit.) Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150739 2026-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150738 Title: How Maltese boutique hotels exploit AI agents to enhance customer experience Abstract: This dissertation examines how decision-makers in Maltese boutique hotels perceive, evaluate, and potentially adopt AI chatbots and AI agents to enhance customer experience across the guest journey. The study addresses a context in which hospitality is becoming increasingly digital, while boutique hotels continue to compete through intimacy, personalisation, trust, and visible human service rather than scale and standardisation. An interpretivist qualitative approach was adopted. The research was exploratory in design and based on eight semi-structured in-depth interviews with boutique hotel decision-makers in Malta. All interviews were analysed using thematic analysis. The study focused on managerial perceptions of AI capability, service suitability across the pre-stay, in-stay, and post-stay journey, operational feasibility, trust, integration, and governance. The findings indicate that Maltese boutique hotels do not reject AI-enabled guest service in principle. Most participants regarded AI as potentially useful when deployed selectively for repetitive, low-emotion, and administratively burdensome communication tasks, particularly routine pre-stay and in-stay enquiries such as FAQs, logistical questions, and multilingual support. However, participants also drew clear boundaries around automating arrival, complaint handling, service recovery, and other emotionally sensitive guest interactions. The findings further show that adoption depends not only on favourable attitudes, but also on PMS and API integration, organisational readiness, governance confidence, data and privacy concerns, and clear human handover protocols. The dissertation concludes that the most credible path for Maltese boutique hotels is carefully governed augmentation rather than replacement. The main implication is that boutique hotels should adopt AI cautiously and selectively, using it to support bounded service tasks while preserving human-led interaction at the trust-intensive moments most central to the boutique experience. Description: B.Sc. (Hons) Bus.& IT(Melit.) Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150738 2026-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150737 Title: Assessing AI and automation readiness for productivity growth in Maltese SMEs Abstract: This study examines the maturity and appetite for adopting artificial intelligence and automation technologies among Maltese small and medium-sized enterprises, with particular attention to their potential to improve productivity and operational efficiency. The research addresses a gap in Malta-specific academic literature by investigating current digital practices, readiness for adoption, and the technological, organisational, and environmental conditions influencing implementation. A qualitatively driven mixed-methods approach was adopted. Primary data were collected through semi-structured interviews with Maltese SME owners, managers, and senior decision makers, together with a structured survey aligned to the EU Digital Maturity Assessment (EU DMA) and the Technology, Organisation, Environment (TOE) framework. The qualitative data were analysed using thematic analysis, while the quantitative findings were used to strengthen and support the overall interpretation of the results. The findings show that Maltese SMEs are not digitally inactive, but are uneven in their maturity, readiness, and depth of AI adoption. Although many SMEs already use some form of digital tools and basic AI applications, readiness for deeper AI integration remains uneven. The results indicate that openness to AI is generally stronger than actual implementation. Key barriers include limited time, resource constraints, skills gaps, fragmented data environments, and uneven access to relevant guidance and support. At the same time, SMEs can identify clear areas where AI and automation could improve productivity, particularly in administrative work, reporting, customer communication, finance-related tasks, workflow coordination, and marketing. Description: B.Sc. (Hons) Bus.& IT(Melit.) Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150737 2026-01-01T00:00:00Z