Browsing by Author "Lawuobahsumo, Kokulo Kpai"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Modelling Cryptocurrency Market Risk: Do Macroeconomic Indicators and Financial Data Explain the Risk within the Cryptocurrency Market?
    (Università della Calabria, 2025-01-29) Lawuobahsumo, Kokulo Kpai; Piluso, Fabio; Algieri, Bernardina; Leccadito, Arturo
    This dissertation explores the intricate relationship between cryptocurrency market risk and an array of macroeconomic indicators and financial data. The central inquiry revolves around whether traditional financial metrics and broader economic factors can elucidate the underlying risk dynamics within the cryptocurrency market. By integrating insights from financial economics, econometrics, and risk management, this study aims to provide a nuanced understanding of how external economic variables influence cryptocurrency return. Chapter 2 aims to investigate calendar effects in the cryptocurrency market. We consider the day-of-the-week, the month-of-the-year, quarter-of-the-year, the US Holidays, and Weekend calendar anomalies for the leading cryptocurrencies: Bitcoin, Dash, Dogecoin, Litecoin, Ripple, and Stellar. Our study employs the Autoregressive Conditional Density model with dummy variables to scrutinize these calendar effects. We find anomalies in the mean, variance, skewness, and kurtosis for these cryptocurrencies' returns. Our result suggests that the cryptocurrency market in some periods tends to violate the Efficient Market Hypothesis. Chapter 3 aims to jointly predict conditional quantiles and tail expectations for the returns of the most popular cryptocurrencies (Bitcoin, Ethereum, Ripple, Dogecoin and Litecoin) using financial and macroeconomic indicators as explanatory variables. We adopt a Monotone Composite Quantile Regression Neural Network (MCQRNN) model to make one- and five-steps-ahead predictions of Value-at-Risk (VaR) and Expected Shortfall (ES) based on a rolling window and compare the performance of our model against the Historical simulation and the standard ARMA(1,1)-GARCH(1,1) model used as benchmarks. The superior set of models is then chosen by backtesting VaR and ES using a Model Confidence Set procedure. Our results show that the MCQRNN performs better than both benchmark models for jointly predicting VaR and ES when considering daily data. Models with the implied volatility index, treasury yield spread and inflation expectations sharpen the extreme return predictions. The results are consistent for the two risk measures at the 1\% and 5\% level both, in the case of a long and short position and for all cryptocurrencies. Chapter 4 use a robust measure of non-linear dependence, the Gerber cross-correlation statistic, to study the cross-dependence between the returns on Bitcoin and a set of commodities, namely wheat, gold, platinum and crude oil WTI. The Gerber statistic enables us to obtain a more robust co-movement measure since it is neither affected by extremely large nor small movements that characterise financial time series; thus, it strips out noise from the data and allows us to capture effective co-movements between series when the movements are “substantial”. Focusing on the period 2014--2022, we construct the bootstrapped confidence intervals for the Gerber statistic and test the null that all the Gerber cross-correlations up to lag kmax are zero. Our results indicate a low degree of dependence between Bitcoin and commodities prices, both when we consider contemporaneous correlation and when we employ correlations between current Bitcoin and lagged (one day, one week, or one month) commodities returns. Chapter 5 proposes a novel framework leveraging an asymmetric Student-t distribution for asset returns enhanced with correlations governed by Generalized Autoregressive Score dynamics. We incorporate explanatory variables to examine their impact on correlations. Empirical analysis using cryptocurrency (Bitcoin and Ripple) and traditional financial market (S\&P 500, NASDAQ, VIX, and WTI) data reveals that the asymmetric Student-$t$ model consistently outperforms competing models, as it effectively captures asymmetry and heavy tails.