To enable more proactive management of the underlying sources of operational risks in financial institutions, this pre-registered study seeks to improve traditional qualitative approaches to causal factors analysis. A Bayesian network-based approach is used to leverage both incident and operations data to model the probability of operational loss events. The approach is applied and empirically tested in a case study on an Australian insurance company. The outputs from the model go beyond simply identifying key risk drivers to offer risk managers a deeper under-standing of how causal factors influence risk. Insights into the collective effects of causal factors, their relative importance and critical thresholds strategically inform more efficient and effective mitigation decisions, ultimately enhancing firm performance and value.

PurposeMachine learning (ML), and deep learning in particular, is gaining traction across a myriad of real-life applications. Portfolio management is no exception. This paper provides a systematic literature review of deep learning applications for portfolio management. The findings are likely to be valuable for industry practitioners and researchers alike, experimenting with novel portfolio management approaches and furthering investment management practice.Design/methodology/approachThis review follows the guidance and methodology of Linnenluecke et al. (2020), Massaro et al. (2016) and Fisch and Block (2018) to first identify relevant literature based on an appropriately developed search phrase, filter the resultant set of publications and present descriptive and analytical findings of the research itself and its metadata.FindingsThe authors find a strong dominance of reinforcement learning algorithms applied to the field, given their through-time portfolio management capabilities. Other well-known deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN) and its derivatives, have shown to be well-suited for time-series forecasting. Most recently, the number of papers published in the field has been increasing, potentially driven by computational advances, hardware accessibility and data availability. The review shows several promising applications and identifies future research opportunities, including better balance on the risk-reward spectrum, novel ways to reduce data dimensionality and pre-process the inputs, stronger focus on direct weights generation, novel deep learning architectures and consistent data choices.Originality/valueSeveral systematic reviews have been conducted with a broader focus of ML applications in finance. However, to the best of the authors' knowledge, this is the first review to focus on deep learning architectures and their applications in the investment portfolio management problem. The review also presents a novel universal taxonomy of models used.

Operational risks are increasingly prevalent and complex to manage in organisations, culminating in substantial financial and non-financial costs. Given the inefficiencies and biases of traditional manual, static and qualitative risk management practices, research has progressed to using data analytics to objectively and dynamically manage risks. However, the variety of operational risks, techniques and objectives researched is not well mapped across industries. This paper thoroughly reviews the emerging research area applying data analytics to operational risk management (ORM) within financial services (FS) and energy and natural resources (ENR). A systematic literature search resulted in 2,538 publications, from which detailed bibliometric and content analyses are performed on 191 studies of relevance. The literature is classified using a novel multi-layered framework, informing critical analyses of the analytics techniques and data employed. Five core themes emerge, relevant to practitioners, researchers, educators and students across any sector: risk identification, causal factors, risk quantification, risk prediction and risk decision-making. Generally, ENR studies focus on identifying causal factors and predicting specific incidents, whereas FS applications are more mature surrounding risk quantification. To conclude, the comprehensive review reveals areas where further research is needed to advance ORM within and beyond FS and ENR, in pursuit of improved decision-making.

The purpose of this work is to compare predictive performance of neural networks trained using the relatively novel technique of training single hidden layer feedforward neural networks (SFNN), called Extreme Learning Machine (ELM), with commonly used backpropagation-trained recurrent neural networks (RNN) as applied to the task of financial market prediction. Evaluated on a set of large capitalisation stocks on the Australian market, specifically the components of the ASX20, ELM-trained SFNNs showed superior performance over RNNs for individual stock price prediction. While this conclusion of efficacy holds generally, long short-term memory (LSTM) RNNs were found to outperform for a small subset of stocks. Subsequent analysis identified several areas of performance deviations which we highlight as potentially fruitful areas for further research and performance improvement.

This paper provides an examination of term structure models in the Australian bond market. Specifically, we examine the comparative ability of various models to forecast at the short, medium and long ends of the yield curve. Overall, we find that model performance varies along the yield curve. Out-of-sample pricing tests show that most of the term structure models underprice a bond at the short and medium ends of the term structure and generally overprice bonds at the long end Further, the level of mispricing is related to time-to-maturity, coupon payments and interest rate volatility. The results have implications for bond pricing in relatively illiquid markets like Australia's.

There is substantial argument that political risk is an important and increasing influence on international portfolio allocation decisions. The purpose of this paper is to investigate the relation between political risk and stock returns within the context of emerging markets. The issue is examined using a framework that controls for global and local return influences. Consistent with the paper's predictions, the findings reveal that political risk is important in explaining return variation in individual emerging markets, particularly in the Pacific Basin, but not in developed markets. At an aggregate portfolio level, supportive evidence is found of a positive relation between political risk and ex-post returns in emerging markets that is robust to alternative risk measures, and more prevalent during the 1990s.

Emerging stock markets have been identified as being at least partially segmented from global capital markets. As a consequence, it has been argued that local factors rather than global factors are the primary source of equity return variation in these markets. This paper seeks to address the question of whether local macroeconomic variables have explanatory power over stock returns in emerging markets. Moderate evidence is found to support this contention. Furthermore, using a principal components approach, two types of commonality in returns are examined. Evidence is found that supports commonality in the factors that drive return variation across emerging markets. A test is also conducted for identical sensitivity to a common set of extracted factors. While little evidence of common sensitivities is found when emerging markets are considered collectively, considerable commonality is found at the regional level. These results have implications for international investors as they suggest that the benefits from diversification are enhanced when the allocation of funds is spread across, rather than within, regions.