this study aims to analyze the impact of data selection to train machine learning models and forecast Bitcoin prices. Specifically, we train elastic net regularization models using two datasets with almost identical total observations. One dataset emphasizes years of observations (depth) over total variables, while the second one emphasizes the number of variables (width) over years of data. Our results suggest that the dataset with more extended historical time series and fewer variables provides a lower forecasting error than the dataset with shorter time series and more variables. Our results may be helpful to practitioners looking to identify data selection strategies to train ML-based forecasting models.

We observe the relationship and causality between cryptocurrencies on one, and commodities, currencies, equity indexes and web search results on the other side. We use prices of Bitcoin and Ethereum for cryptocurrencies, prices of crude oil and gold for commodities, Euro-US Dollar, Euro-Swiss Franc exchange rates for currencies, Dow Jones Industrial Average for market index and Google Trends® data as a measure of worldwide web search results for cryptocurrencies of interest. We find that Bitcoin and web search results correlation went from highly positive to low negative during the COVID-19 period. The results of the study show that the price of Bitcoin and Ethereum can be modelled using different combinations of commodities, currencies, indexes and web search results, with web search results and Dow Jones Industrial Average exhibiting best predictive power both concurrently and one day in advance. Our best performing models were able to explain more than 95% and 90% of Bitcoin and Ethereum price variability respectively. We also find strong evidence of web search traffic impacting both Bitcoin and Ethereum prices at all tested lags, as well as some evidence of gold impact on Bitcoin and EUR/CHF impact on Ethereum.

Bankruptcy prediction has been in the focus of research for many years. The benefits of bankruptcy predictive ability are several and possibly beneficial for all business entity stakeholders. This paper has an ultimate goal of revealing most significant financial traits of bankrupt companies as opposed to non-bankrupt companies. The research includes 50 bankrupt companies based in Federation of Bosnia and Herzegovina. They were matched with a random sample of 100 non-bankrupt company-years. Financial ratios of companies used in the sample were derived from their financial statements. Using logistic regression analysis and ANOVA, we were able to construct a several bankruptcy prediction models. The comprehensive model exhibited predictive ability of more than 95%, with high predictive ability of both bankrupt and non-bankrupt companies. The research has shown that bankrupt companies do leave significant financial traits that can be detected prior to official bankruptcy proceeding filing, which can be beneficial for all stakeholders.

Abstract Background: Competition in the banking industry has been an important topic in the scientific literature as researchers tried to assess the level of competition in the banking sector. Objectives: This paper has an aim to investigate the market structure and a long term equilibrium of the banking market in Bosnia and Herzegovina nationwide as well as on its constitutional entities as well as to evaluate the monopoly power of banks during the years 2008-2012. Methods/Approach: The paper is examining the market structure using the most frequently applied measures of concentration k-bank concentration ratio (CRk) and Herfindahl-Hirschman Index (HHI) as well as evaluating the monopoly power of banks by employing Panzar-Rosse “H-statistic”. Results: The empirical results using CRk and HHI show that Bosnia and Herzegovina banking market has a moderately concentrated market with a concentration decreasing trend. The Panzar-Rosse “H-statistic” suggests that banks in Bosnia and Herzegovina operate under monopoly or monopolistic competition depending on the market segment. Conclusions: Banks operating on the banking market in Bosnia and Herzegovina seem to be earning their total and interest revenues under monopoly or perfectly collusive oligopoly.

This article has an aim to assess credit default prediction on the banking market in Bosnia and Herzegovina nationwide as well as on its constitutional entities (Federation of Bosnia and Herzegovina and Republika Srpska). Ability to classify companies info different predefined groups or finding an appropriate tool which would replace human assessment in classifying companies into good and bad buckets has been one of the main interests on risk management researchers for a long time. We investigated the possibility and accuracy of default prediction using traditional statistical methods logistic regression (logit) and multiple discriminant analysis (MDA) and compared their predictive abilities. The results show that the created models have high predictive ability. For logit models, some variables are more influential on the default prediction than the others. Return on assets (ROA) is statistically significant in all four periods prior to default, having very high regression coefficients, or high impact on the model's ability to predict default. Similar results are obtained for MDA models. It is also found that predictive ability differs between logistic regression and multiple discriminant analysis.

Abstract Background: During the last four years, the banking sector in Bosnia and Herzegovina has been facing crisis which has caused the stagnation within the sector. Still, the results within the sector vary to a great extent from bank to bank. Objectives: The efficiency score is assessed for each bank and serves as a basis for further comparisons between banks in the period between 2008 and 2010. Methods: A modified model of Data Envelopment Analysis (DEA) has been used in order to combine several financial indicators simultaneously in a unique efficiency measure. The model provides a rounded judgement on a bank's relative efficiency. Results: Efficiency of individual banks varied throughout the observed period and not all of the banks were a part of the negative banking sector trend induced by the crisis. There is no significant difference between performance of banks in different entities of Bosnia and Herzegovina, and between smaller and larger banks. Conclusions: The results of the study can be used by bank managers to assess the performance of their banks, as observing financial ratios separately can result in a misleading conclusion. The most valuable practical implications of the findings are the provided feasible targets for the three observed years.