UC Santa Cruz

This dissertation consists of three chapters, which discuss three topics of international finance such as the effect of exchange rate volatility on economic growth, evidence of law of one price, and regulation effect on price differential. Each topic is studied by one of three chapters.

The first chapter empirically tests the effect of country’s domestic and international credit access on effect of exchange rate volatility on growth. In this chapter, I uses annual panel data from 2003 to 2015 for 115 countries. System GMM estimation with two step standard error is used for regression analysis. Country’s financial development level is used to measure country’s domestic credit access, and country’s foreign bank presence level is used to measure country’s international credit access. The empirical finding of this chapter show that countries with higher level of financial development are less likely to be adversely affected by exchange rate volatility, and higher level of foreign bank presence can ameliorate the adverse growth effect of exchange rate volatility. Results of this chapter implies that countries with high levels of credit access can use more flexible exchange rate regime since countries’ high credit access helps to insulate the economy from the adverse effect of exchange rate volatility. For countries with low credit access, exchange rate stability is important for economic growth so that it is preferable to use the fixed exchange rate regime.

The second chapter studies the speed of convergence of Bitcoin across markets globally to find supportive evidence for LOOP. The data used in this chapter includes Bitcoin trading data of 20 exchanges under 14 currencies, which is the most comprehensive Bitcoin dataset used in the literature. Cointegration analysis and Vector Error Correction Model are performed to investigate long run relationship and short run dynamics in Bitcoin markets. Empirical results of this chapter confirm the existence long run equilibrium in Bitcoin market. Half-lives of Bitcoin trading pairs are estimated using VEC model and range from 0.133 days to 6.93 days, which are significantly fast than half lives (range from months to several years) of other assets studied in previous literature. The fast speed of converge in Bitcoin markets implies efficient arbitraging activities, which provide strong evidence to support LOOP. Using estimated half-lives as data, I also identifies border effect in Bitcoin markets.

The third chapter empirically examines the effect of regulation changes on Bitcoin cross country premium using my self-collected daily panel data for 22 Bitcoin exchanges under 11 currencies. To measure Bitcoin regulation changes, I constructs an original country level Bitcoin regulation news index. Web crawler is used to collect Bitcoin regulation related reporting from news websites like CNBC, Reuters and Coindesk etc. Results of this chapter suggest that tighter relative regulation increase level of market segmentation, which induce higher premium. By investigating the effect of each type of news individually, warning news and formal news are found to be main drivers of the result, and formal news have higher estimated effect on premium. In this chapter, I also identify the spillover effect from regulation change in other countries. Country's Bitcoin premium increases as other countries tighten their Bitcoin regulations.