International Journal of Blockchains and Cryptocurrencies (4 papers in press)

Regular Issues

• AI and machine learning in cryptocurrency volatility prediction: insights and perspectives  
  by Rifa Atrous 
  Abstract: This paper aims to investigate cryptocurrency volatility prediction by integrating insights from econometrics and artificial intelligence (AI). This systematic literature review analyses studies published between January 2020 and July 2025 by comparing four categories of models: econometric, machine learning (ML), deep learning (DL), and hybrid approaches. The results highlight that DL architectures, notably long short-term memory (LSTM) and gated recurrent unit (GRU), along with hybrid GARCH-LSTM models, consistently outperform traditional benchmarks while enhancing risk-return trade-offs. Nevertheless, econometric models from the GARCH-family remain advantageous in terms of interpretability and robustness under specific conditions. The main shortcomings identified concern the limited interpretability of AI-based models and the concentration of analyses on Bitcoin and Ethereum at the expense of other digital assets. We recommend strengthening the interpretability of machine learning and hybrid models. These findings provide valuable insights for researchers and practitioners in a turbulent economic and geopolitical environment.
  Keywords: cryptocurrencies; volatility; artificial intelligence; prediction; GARCH; generalised autoregressive conditional heteroscedasticity.
  DOI: 10.1504/IJBC.2026.10075565
   
  
• Blockchain and smart contracts for transparent public finance in climate resilience  
  by Sarvesh Chand 
  Abstract: Climate resilience activities in vulnerable regions often confront problems relating to the financial opacity, corruption, and the unreliable verification of outcomes. The paper proposes a blockchain system to strengthen transparency and traceability in climate adaptation funding, particularly for green infrastructure. Using smart contracts and IoT-verified geospatial data, the system assures the real-time monitoring of performance metrics and releases funds securely upon assessed performance. A hybrid PoW/PoS consensus mechanism is suggested to enable scalability while remaining energy friendly toward resource-constrained geographies. Pilot implementation among small countries is scheduled to assess the adoption and governance. The goal is for this model to restore public faith and speed climate resilience activities in places where the traditional setup fails.
  Keywords: blockchain; climate finance; smart contracts; public governance; geospatial verification; IoT; Internet of Things.
  DOI: 10.1504/IJBC.2026.10075790
   
  
• Important considerations for designing and issuing a central bank digital currency (CBDC) in Africa  
  by Peterson K. Ozili 
  Abstract: This study examines the important considerations that African central banks need to take into account when designing and issuing a central bank digital currency. The article begins by stating the possible motivations for issuing a central bank digital currency in African countries. Thereafter, it discusses the important development, design, technology, policy, legal, payment system safety and resilience, stakeholder engagement, political, and human resources considerations to consider when issuing a CBDC in African countries. It also presents case studies on the important considerations that were taken into account when issuing the eNaira, eRupee, DCash, SandDollar and eCedi CBDC.
  Keywords: CBDC; central bank digital currency; retail CBDC; wholesale CBDC; digital payment; privacy; renumeration; consideration; financial stability; monetary policy; Africa; cryptocurrency.
  DOI: 10.1504/IJBC.2026.10075885
   
  
• Designing a deep learning model for bitcoin price trend prediction in dynamic markets  
  by Sachin Kumar, Gagan Tiwari 
  Abstract: The market for cryptocurrencies, definitely Bitcoin, is dishonourable for its life-threatening unpredictability and unpredictability. A deep learning-based background for forecasting Cryptocurrency responsibility arrangements over innumerable time epochs and market circumstances is accessible in this work. The prototypical seeks to expansion accuracy in approximating for both short-term and long-term alternations in prices by manipulating cutting-edge DL algorithms and providing insights into various bazaar scenarios like bear markets as well as bull bazaars. The approach addresses the exertion of adjusting to rapid market fluctuations, which is decisive for merchants and predictors in finance. In this paper, we present a deep learning-based model for calculating Bitcoin price trends that is intended for self-motivated market environments. To improve predicting stability, the model incorporates temporal educational architecture and multiple feature combining of data. Experimental consequences show that the projected method achieves 94.2% recognition accuracy, reduces the error rate by 5.8%, and outperforms baseline LSTM and CNN models across numerous evaluation metrics. These consequences establish the usefulness of the proposed construction in highly unpredictable cryptocurrency markets.
  Keywords: bitcoin price prediction; deep learning; LSTM; long short-term memory; CNN; convolutional neural network; dynamic market adaptation; cryptocurrency.
  DOI: 10.1504/IJBC.2026.10076137