"Scientific research" Almanac

More titles

• 

  APPROACHES AND MODELS FOR MEASURING AND MANAGING COMMERCIAL BANKS LIQUIDITY

  The problems related to liquidity management and maintaining optimal levels of liquid assets of banks have become very topical and have stirred the interest of both the regulatory bodies and researchers in the field of banking and the general public. The pertinence of these problems arises from the importance of banks for the economic development ... The problems related to liquidity management and maintaining optimal levels of liquid assets of banks have become very topical and have stirred the interest of both the regulatory bodies and researchers in the field of banking and the general public. The pertinence of these problems arises from the importance of banks for the economic development of the national economies on the one hand, and from the possibility for transformation of banking problems into sectoral, national and international financial crises, on the other. The integration and the high level of interdependence of the economies worldwide are the key factors for accumulation of covert negative effects, which may trigger crises under certain conditions.

• 

  CUSTOMER RESPONSE PREDICTIVE ANALYSIS USING UPLIFT MODELLING

  Traditional customer response models are based on the understanding that marketing campaign effectiveness could be measured by predicting response probabilities of each customer. Such approach seems to be very reasonable and in most cases it gives better results than random targeting. A fundamental problem of response modelling is that it does not ... Traditional customer response models are based on the understanding that marketing campaign effectiveness could be measured by predicting response probabilities of each customer. Such approach seems to be very reasonable and in most cases it gives better results than random targeting. A fundamental problem of response modelling is that it does not separate the impact of a campaign from other stimuli and spontaneous responses because response models predict the wrong thing. They model and predict response probabilities, not the change in those probabilities as a result of the given marketing intervention. Following the understanding that traditional response models do not give optimal marketing decisions, the goal of this study is to define, estimate and apply on a real data base uplift model for predicting the change in customer behaviour resulting from marketing action. The specific tasks relate to studying the opportunities of a relatively new and less familiar methodology for identifying hesitant and persuadable customers. The study illustrates the application of the direct approach for uplift modeling and demonstrates its significant advantages over targeting based on uplift models over traditional response models.

• 

  STUDY ON THE APPLICABILITY OF INVESTMENT AND PORTFOLIO MANAGEMENT MODELS IN EMERGING MARKETS

  The study aims to test the applicability of three basic investment and portfolio management models in emerging markets in Central and Eastern Europe – the EVA model, RI model and ETF α+ model. Our thesis statement is that the models established in developed economies are applicable to emerging markets but should be modelled according to the ... The study aims to test the applicability of three basic investment and portfolio management models in emerging markets in Central and Eastern Europe – the EVA model, RI model and ETF α+ model. Our thesis statement is that the models established in developed economies are applicable to emerging markets but should be modelled according to the specifics of these markets. Our task was to create modified models that are applicable in the East-European emerging markets. First we tested the economic value-added model (EVA), but we met a number of difficulties in applying it in Eastern Europe, mainly due to some differences in the accounting standards. Therefore, we developed the much simpler RI model, in which with only two modifications we were able to prove its applicability in our markets. Based on this result we then made key changes in the EVA model and as a result we found that both indicators – EVA and RI – can be used as criteria for implementing investment and portfolio management policies. In the final stage of the project we successfully tested the possibility of generating alpha returns in duplicating the WIG220 index, achieving a higher return than the market return.