According to the standard neoclassical growth theory, capital should flow from countries with more capital (rich countries) to countries with less capital (poor countries). Starting with Solow growth model country with more capital per worker should export capital (savings) to country with less capital per worker until the equalization of capital per worker. In the first era of globalization rich countries were indeed «bankers of the world» (Schularick, 2006, p. 340) but in the second era international capital flows changed dramatically. In his paper «Why doesn‘t capital flow from rich to poor countries?» Lucas (1990) refers to inconsistency between neoclassical theory and capital flows in USA and India. The Lucas paradox becomes one more puzzle of international macroeconomics and finances together with the Feldstein-Horioka puzzle, the home bias puzzle and the risk sharing puzzle. The most important question is why the neoclassical growth model doesn‘t exactly solves the direction of capital flows as well as the amount of capital. Capital flows and its drivers could explain the growth in the long run, the convergence of countries and consequently could be the solution to inequality and poverty in the world. Lucas (1990) presented three possible solution to the paradox and although triggered an avalanche of research on capital flows even now, 30 years after, there are no standardized tools or data to accurately determine and predict capital flows or drivers of those flows. Most of the work on Lucas paradox is based on global capital flows and analysis are based on net capital flows, data driven from surplus and deficit of the current account. According to newer research gross capital flows are higher and more volatile while the net capital flows remained the same in the past decade. While the research of capital flows on a global scale is represented in many papers, work on Lucas paradox in European union is rear and the results are inconsistent. With very high financial integration, significantly larger capital flow, monetary union and high indebtedness of the member states, some European Union countries are experiencing strong uphill capital flows especially in years of crisis. The doctoral thesis focuses on studying the existence of the Lucas paradox and the allocation puzzle in EU member states and the overall objective of this work is to research the direction, stability, procyclicality and dynamics of capital flows. The specific research objectives of this work based on theoretical and empirical insights are: 1. To define differences in stability, dynamics and the procyclicality of net and gross capital flows. 2. To determine the importance of studying gross capital flows in research of Lucas paradox and its dynamic correspondence. 3. To analyze and determine the direction of capital and the dynamic correspondence in the member states of the EU. 4. To identify differences in dynamics and allocation of capital between the countries of the euro area and the EU, as well as the core and periphery. 5. To describe and systematize the reasons of Lucas paradox, and to analyze the fundamental reasons for the uphill capital flows in the EU. 6. To analyze the impact of the financial crisis on the direction of net and gross capital flows in the countries of the European Union. Based on research objectives following main hypothesis, with seven auxiliary hypotheses, are specified: H 1) In times of globalization, financial liberalization, and especially the crisis, the capital flows and dynamic correspondence can be determined by studying gross and not by net capital flows. H 2) There is a Lucas paradox in the countries of the European Union. H 3) There is an allocation puzzle in the countries of the European Union. The empirical analysis of Lucas paradox and its dynamic correspondence is conducted using panel analysis and linear regression, following the methods of previous works, on data for 28 EU member states from 1995. to 2018. A shift in research is done by comparative analysis of gross and net capital flows. The conclusion of research on the first hypothesis is that, regarding the procyclycli and the volatility of the capital flows in the EU, the empirical analysis of the Lucas paradox and the allocation puzzle should be conducted using gross instead of net capital flows. Net capital flows are more volatile in all samples while the gross capital flows are more volatile in countries with higher income. Gross inflows and outflows are moving simultaneously, and high correlation of gross flows indicates procyclicality. The research on the second hypothesis confirms the existence of the Lucas paradox in EU member states, with the highest uphill capital flows in countries with GDP per capita above average. The Lucas paradox is confirmed on all three gross capital composition, while net capital flows are according to the theory. Also, the fundamental assumptions are not the solution to the reverse capital flow in the EU. The existence of the allocation puzzle is not confirmed and the capital is allocated in countries with higher economic growth. Understanding the existence of capital flows in the European Union outside the theoretical framework should direct further efforts towards finding the reasons for such flows and their possible solution.