Essays in international finance:Contribution to explain the home bias puzzle in the presence of cryptocurrency and behavior finance

Abstract

The main theme of this thesis lies in our attempt to contribute to explaining the Home Bias Puzzle (HBP) observed in international financial markets in the presence of Cryptocurrencies, within the framework of FinTech and more specifically in the context of Decentralized Finance (DeFi). Through a meticulous review of recent works on the question of international portfolio diversification, encompassing physico-financial assets such as Cryptocurrencies, technology stocks, classical stocks, currencies, commodities, and oil, we examined the issues of arbitrage and the strategy of choosing investment in domestic assets and/or choosing investment in international portfolio diversification. To empirically test our central issue, we validated four essays formulated as hypotheses: In the first essay on Efficiency and Volatility, we examined, through time series modeling, the impact of integrating cryptocurrencies into the investor's portfolio to verify our first hypothesis, namely the transmission of volatility shocks induced by this asset. The use of ARCH and GARCH modeling shows that the coefficients associated with them are close to unity, thus indicating a permanent effect of shocks on conditional variance. However, during the COVID-19 pandemic, Bitcoin was considered a safe haven asset. Also, relying on the econometric results of EGARCH and TARCH models, similar to Wang (2021), we show the existence of an excessive leverage effect on the volatility of future returns for Bitcoin (+26.50%) and Dogecoin (+65.07%). Furthermore, our study shows the absence of leverage for the other cryptocurrencies in the sample. Our second essay aims to validate the second hypothesis borrowed from industrial economics on the Integration-Segmentation-Diversification (ISD) triptych of asset portfolios, i.e., the relationship between goods and services markets and the capital market. To verify this hypothesis, we used a VAR (Vector AutoRegressive) modeling to analyze the causal time relationship between economic variables (real sphere) and financial variables (financial sphere) through standard tests (AIC) in the first stage and (SC) in the second stage. The results obtained show that price variations in the developed markets of the sample do not follow a common long-term trend. In this context, there would likely be an opportunity for diversification among developed markets, a product of financial liberalization (Attig.N. and al. (2023)). Thirdly, the empirical test of the existence of a Home Bias, our third essay and hypothesis were conducted over the period 2006-2021, with 640 observations. The determinants of the Home Bias Puzzle (HBP) were divided into seven panels: governance variables, macroeconomic variables, market size and microstructure variables, information asymmetry, familiarity and geography, Foreign Trade, and finally geopolitical variables. The econometric results we obtained are consistent with previous findings (Garg, Karmakar, M. and Paul, S., (2023); Lee, J. Lee, K. and Oh, F.D. (2023)). Finally, the last essay, reflecting hypothesis four on the relationship between Cryptocurrencies, Portfolio Diversification, and Behavioral Finance, highlights the superiority of the W. Sharpe (1964) performance index compared to other naive portfolio diversification strategies derived from the Mean-Variance approach by H. Markowitz (1952). Our results corroborate those obtained by Hachicha F., and al. (2023).