The aim of this research was to determine the persistence of profit in an emerging banking sector of the Republic of Croatia. Most developing countries have experienced common changes within restructuring of the banking system and therefore, this issue has become crucial, especially after comparing poor empirical findings in these countries to the findings of developed countries. However, research related to this issue is non-existent in the Croatian banking sector. Moreover, among the few studies that were carried out on the territory of the Central and Eastern European banking markets, none of the studies have analysed the persistence of profit in terms of the Markov Chain stochastic process. In addition to the profit persistence analysis, authors defined and estimated a model that would enable the identification of the profitability determinants of Croatian banks. In this sense, the model incorporated three groups of profitability determinants: bank-specific, industry-specific and macroeconomic. The variables with a statistically significant impact on the profitability of banks were identified using a dynamic panel model, while the application of the Markov Chains stochastic process revealed that profit persistence was less likely to occur in banks with higher profit.

This paper investigates whether a specific type of a recurrent neural network, in particular Jordan neural network (JNN), captures the expected inflation better than commonly used feedforward neural networks and traditional parametric time-series models. It also considers competing survey-based and model-based expected inflation towards ex-post actual inflation to find whose predictions are more accurate; predictions from survey respondents or forecasting modelers. Further, it proposes neural network modelling strategy when dealing with nonstationary time-series which exhibit long-memory property and nonlinear dependence with respect to lagged inputs and exogenous inputs as well. Following this strategy, overfitting problem was reduced until no improvement in forecasting accuracy of expected inflation is achieved. The main finding is that JNN predicts inflation in euro zone quite accurately within forecasting horizon of 2 years. Regarding rational expectation principle we have found a set of demand-pull and cost-push inflation characteristics as exogenous inputs which helps in reducing overfitting problem of recurrent neural network even more. The sample includes euro zone aggregated monthly observations from January 2000 to December 2019. The results also confirm that inflation expectations obtained from JNN are consistent with Survey of professional forecasters (SPF), and thus, monetary policy makers can use JNN as a complementary tool in shortcomings of other inflation expectations measures.

Background: Since high-frequency data have become available, an unbiased volatility estimator, i.e. realized variance (RV) can be computed. Commonly used models for RV forecasting suffer from strong persistence with a high sensitivity to the returns distribution assumption and they use only daily returns. Objectives: The main objective is measurement and forecasting of RV. Two approaches are compared: Heterogeneous AutoRegressive model (HAR-RV) and Feedforward Neural Networks (FNNs). Even though HAR-RV-type models describe RV stylized facts very well, they ignore its nonlinear behaviour. Therefore, FNN-HAR-type models are developed. Methods/Approach: Firstly, an optimal sampling frequency with application to the DAX index is chosen. Secondly, in and out of sample predictions within HAR models and FNNs are compared using RMSE, AIC, the Wald test and the DM test. Weights of FNN-HAR-type models are estimated using the BP algorithm. Results: The optimal sampling frequency of RV is 10 minutes. Within HAR-type models, HAR-RV-J has better, but not significant, forecasting performances, while FNN-HAR-J and FNNLHAR- J have significantly better predictive accuracy in comparison to the FNN-HAR model. Conclusions: Compared to the traditional ones, FNN-HAR-type models are better in capturing nonlinear behaviour of RV. FNN-HAR-type models have better accuracy compared to traditional HAR-type models, but only on the sample data, whereas their out-of-sample predictive accuracy is approximately equal.

The Purpose – An evaluation of 197 hotel websites in Croatia was conducted to determine whether hotels in this economically lagging transition country exploit the potentials of the Internet as a marketing tool. Design/Methodology/Approach – A team of 30 trained assistants evaluated the websites using an amended modified Balanced Scorecard (mBSC) approach from the user-friendliness, site attractiveness, marketing effectiveness, and F&B informativeness perspectives. MANOVA was used to test main and interaction effects of hotel size, quality rating, and location on four website performance perspectives. Findings – Although hoteliers recognize the importance of online presence, most are not effectively using websites from the user-friendliness, marketing effectiveness, and F&B perspectives. Hotel quality rating, location, and their two-way interaction are significant in explaining differences in site attractiveness and marketing effectiveness. Hotel size does not explain the differences among the four website perspectives. Originality – In lieu of the rapidly evolving IT and consumer trends, this study provides hoteliers with an updated website assessment tool that can serve as a point of comparison against contemporary e-marketing approaches. Moreover, since tourism website evaluation is an under researched topic in transition economies, this article makes a contribution in the needed direction. The study ends with a discussion of results and implications for hoteliers, policy makers, and scholars.