ADS Capstone Chronicles Revised

Figure 7 Bitcoin Price Change Stationarity Test

4.7.2 Stationarity Test of Bitcoin Price Change To assess the stationarity of Bitcoin’s daily percentage price changes, the Augmented Dickey-Fuller (ADF) test is employed. This test specifically looks for a unit root in the time series, indicating non-stationarity. If the test statistic is significantly negative, it suggests that the series does not have a unit root and is instead stationary. The stationarity tests, particularly the ADF test applied to Bitcoin and other cryptocurrencies like Ethereum and Solana, substantiate the reliability of historical price data for forecasting purposes. By confirming these financial time series are stationary, we lay a robust statistical foundation for predictive modeling, which is essential for investors and analysts engaged in the volatile cryptocurrency market. Interpretation The results from the ADF test are quite definitive for Bitcoin, showing a test statistic of -33.42 and a p-value of 0.0000. This very low p-value suggests the probability of the Bitcoin price series containing a unit root is almost zero, confirming its stationarity. This implies that the historical price data for Bitcoin has been consistent over time, making it reliable for future statistical forecasting and modeling. 4.7.3 ADF Test Results and

4.7.4 Implications for Other

Cryptocurrencies This stationarity testing approach is not limited to Bitcoin but extends to the entire spectrum of cryptocurrencies in our study, including ADA (Cardano), SOL (Solana), MATIC, LTC (Litecoin), ETH (Ethereum), UNI (Uniswap), LINK (Chainlink), COMP (Compound), BTC (Bitcoin), and AVAX (Avalanche). The ADF tests on these cryptocurrencies uniformly yielded p-values of 0.00, conclusively indicating that all are stationary. This consistent result across diverse digital assets underpins the robustness of our analytical approach and enhances the reliability of using historical data for predictive modeling in the cryptocurrency market. With stationarity confirmed, the price series of these cryptocurrencies are deemed suitable for further time series analyses and forecasting models, which are crucial for informed trading and investment decisions. 4.8 EDA Conclusion The exploratory data analysis (EDA) for ‘The Guardians of the Crypto’ platform provided critical insights into the structure and behavior of cryptocurrency data. Initial inspection and cleaning of the dataset ensured its integrity by addressing missing values, removing duplicates, and correcting data types.

79