This thesis is based on three empirical essays in financial development and economic growth. The first essay, investigated in the third chapter, the effect of financial development on economic growth in the context of Saudi Arabia, an oil-rich economy. In doing so, the study distinguishes between the effects of financial development on the oil and non-oil sectors of the economy. The Autoregressive Distributed Lag (ARDL) bounds test methodology is applied to yearly data over the period 1968 to 2010. The finding of this study is that financial development has a positive impact on the growth of the non-oil sector. In contrast, its impact on the oil-sector growth and total GDP growth is either negative or insignificant. This suggests that the relationship between financial development and growth may be fundamentally different in resource-dominated economies. The second essay revisited, in the fourth chapter, the relationship between financial development and economic growth in a panel of 52 middle-income countries over the 1980-2008 period. Using pooled mean group estimations in a dynamic heterogeneous panel setting, we show that there is an inverted U-shaped relationship between finance and growth in the long-run. In the short run, the relationship is insignificant. This suggests that too much finance can exert a negative influence on growth in middle-income countries. The finding of a non-monotonic effect of financial development on growth is confirmed by estimating a dynamic panel threshold model. The third essay empirically explores cross-country evidence of the effects of financial development shocks on economic growth. It employs a Global Vector Autoregressive (GVAR) model, which allows us to capture the dynamics of this relationship in a multi-country setting, and connects countries through bilateral international trade. Given the progressive role that Brazil, Russia, India, China and South Africa (BRICS) play in the world economic arena, this essay focuses on whether financial development in one BRICS member state affects economic growth in the other BRICS. To this end, the study finds empirical evidence that credit to the private sector has a positive spillover effect on growth in some of the BRICS countries. However, the results imply that the current level of financial integration among the BRICS countries is still not mature enough to spur economic growth for all the BRICS members.

This thesis examines the exchange rate pass-through into consumer prices in Central and Eastern Europe. The study is based on quarterly data of 12 countries from 2003 to 2013. Estimations are conducted using heterogeneous panel cointegration methods, namely the mean group and the pooled mean group estimators. Fixed effects are used as a reference. The thesis provides short- run and long-run estimates of the exchange rate pass-through for the individual countries and for the region as a whole. Based on the results, we conclude that the exchange rate pass-through is highly variable across Central and Eastern Europe. We find that there is no clear distinction between the pass-through rates in euro area countries, EU countries not using the euro and non- EU countries. Further, we find that the generally accepted concept of higher exchange rate pass- though in developing countries does not hold in this region. JEL Classification C23, E31, E52, F31 Keywords exchange rate pass-through, pooled mean group, mean group, heterogeneous panel cointegration Author's e-mail bara.mirkova@centrum.cz Supervisor's e-mail roman.horvath@gmail.com

This dissertation comprises two topics. In Chapter 1, I explore the short- and long-run relation between insurance development and economic growth for 40 countries between 1981 and 2010. Applying a pooled mean group estimation, I find that life and nonlife insurance have different short- and long-run effects on the growth. On a full sample analysis, life insurance exerts a significantly positive long-run effect on the growth, while its short-run effect is not significant. Nonlife insurance, in contrast, has a significantly positive short-run growth effect but no long-run effect. On a reduced sample analysis, the observation on life insurance is qualitatively similar, but the growth effect of nonlife insurance is no longer significant both in short and long run, suggesting that specific countries drive the overall effect in the full sample. The results pass a battery of robustness tests. The analysis on individual countries reveals that the short-run effect and adjustment speed toward the long-run equilibrium varies across countries. I also analyze if the level of income and insurance development makes any difference on the growth effect of insurance. The results show that the growth effect of life insurance is significant in non-high income countries and countries with low level of life insurance development, while the effect is not significant both for life and nonlife insurance in high income countries. In Chapter 2, I employ the dynamic panel threshold model to investigate how institutional environments shape the impact of insurance development on economic growth. I conduct four hypotheses for possible intermediate effects of institutional environments on insurance-growth nexus: quasi-institution positivity, quasi-institution negativity, quasi-institution duality, and quasi-institution neutrality. I use multiple measures related to political, economic, and legal environments to evaluate the soundness of institutional environments. Empirical results show that the quasi-institution negativity hypothesis is supported for life insurance because the observation is consistent across all institution-related measures. The results in nonlife insurance are not as uniform as those in life insurance. The quasi-institution positivity, negativity, and neutrality are respectively supported in different institutional measures, and the coefficients in most cases are significant only at a marginal significance level. The overall findings suggest that a sound institutional environment does not necessarily benefit the growth effect of life insurance, but an unhealthy one does deter it and that the effect depends on specific measure in the case of nonlife insurance. In Chapter 3 I briefly introduce some directions for further research.

Cette thèse étudie comment le taux de change effectif réel (TCER) et ses mesures associées (volatilité du TCER et désalignement du TCER) affectent la croissance de la productivité totale des facteurs (CPTF). Elle analyse également les canaux par lesquels le TCER et ses mesures associées agissent sur la productivité totale des facteurs (PTF). La première partie étudie comment le TCER lui-Même, d'une part, et la volatilité du TCER, d'autre part, influencent la productivité. Une analyse du lien entre le niveau du TCER et la PTF dans le chapitre 1 indique qu'une appréciation de taux de change cause une augmentation de la PTF. Mais cet impact est également non- inéaire: en-Dessous du seuil, le TCER influence négativement la productivité tandis qu'au-Dessus du seuil il agit positivement. Les résultats du chapitre 2 illustrent que la volatilité du TCER affecte négativement la CPTF. Nous avons également constaté que la volatilité du TCER agit sur PTF selon le niveau du développement financier. Pour les pays modérément financièrement développés, la volatilité du TCER réagit négativement sur la productivité et n'a aucun effet sur la productivité pour les niveaux très bas et très élevés du développement financier. La deuxième partie examine les canaux par lesquels le TCER et ses mesures associées influencent la productivité. Les résultats du chapitre 3 illustrent que la volatilité du TCER a un impact négatif élevé sur l'investissement. Ces résultats sont robustes dans les pays à faible revenu et les pays à revenu moyens, et en employant une mesure alternative de volatilité du TCER. Le chapitre 4 montre que le désalignement du taux de change réel et la volatilité du taux de change réel affectent négativement les exportations. Il démontre également que la volatilité du taux de change réel est plus nocive aux exportations que le désalignement. Ces résultats sont corroborés par des résultats sur des sous-Échantillons de pays à bas revenu et à revenu moyen. / This dissertation investigates how the real effective exchange rate (REER) and its associated asurements (REER volatility and REER misalignment) affect total factor productivity growth (TFPG). It also analyzes the channels through which the REER and its associated measurements act on total factor productivity (TFP). The first part studies how the REER itself, on the one hand, and the REER volatility, on the other hand, influence productivity. An analysis of the link between the level of REER and TFP in chapter 1 reveals that an exchange rate appreciation causes an increase of TFP. But this impact is also nonlinear: below the threshold, real exchange rate influences negatively productivity while above the threshold it acts positively. The results of chapter 2 illustrate that REER volatility affects negatively TFPG. We also found that REER volatility acts on TFP according to the level of financial development. For moderately financially developed countries, REER volatility reacts negatively on productivity and has no effect on productivity for very low and very high levels of financial development. The second part examines the channels through which the REER and its associated measurements influence productivity. The results of chapter 3 illustrate that the exchange rate volatility has a strong negative impact on investment. This outcome is robust in low income and middle income countries, and by using an alternative measurement of exchange rate volatility. Chapter 4 show that both real exchange rate misalignment and real exchange rate volatility affect negatively exports. It also demonstrates that real exchange rate volatility is more harmful to exports than misalignment. These outcomes are corroborated by estimations on subsamples of Low- ncome and Middle-Income countries

Cette thèse étudie comment le taux de change effectif réel (TCER) et ses mesures associées (volatilité du TCER et désalignement du TCER) affectent la croissance de la productivité totale des facteurs (CPTF). Elle analyse également les canaux par lesquels le TCER et ses mesures associées agissent sur la productivité totale des facteurs (PTF). La première partie étudie comment le TCER lui-même, d'une part, et la volatilité du TCER, d'autre part, influencent la productivité. Une analyse du lien entre le niveau du TCER et la PTF dans le chapitre 1 indique qu'une appréciation de taux de change cause une augmentation de la PTF. Mais cet impact est également non-linéaire: en-dessous du seuil, le TCER influence négativement la productivité tandis qu'au-dessus du seuil il agit positivement. Les résultats du chapitre 2 illustrent que la volatilité du TCER affecte négativement la CPTF. Nous avons également constaté que la volatilité du TCER agit sur PTF selon le niveau du développement financier. Pour les pays modérément financièrement développés, la volatilité du TCER réagit négativement sur la productivité et n'a aucun effet sur la productivité pour les niveaux très bas et très élevés du développement financier. La deuxième partie examine les canaux par lesquels le TCER et ses mesures associées influencent la productivité. Les résultats du chapitre 3 illustrent que la volatilité du TCER a un impact négatif élevé sur l'investissement. Ces résultats sont robustes dans les pays à faible revenu et les pays à revenu moyens, et en employant une mesure alternative de volatilité du TCER. Le chapitre 4 montre que le désalignement du taux de change réel et la volatilité du taux de change réel affectent négativement les exportations. Il démontre également que la volatilité du taux de change réel est plus nocive aux exportations que le désalignement. Ces résultats sont corroborés par des résultats sur des sous-échantillons de pays à bas revenu et à revenu moyen.