摘要
虛擬貨幣市場中占比最大的比特幣，近幾年學術界和產業界對於它在財務經濟或環保領域的研究方面持續不斷。本研究發現在財務經濟領域方面，效率市場假說(Efficient Market Hypothesis, EMH)和噪音交易者理論通常被廣泛的驗證於股票市場，卻鮮少驗證於虛擬貨幣市場。此外，在環保領域方面，對於商品期貨中的原油、二氧化碳和比特幣這三者之間的關係則尚未有文獻的探討。 本研究主要在探究不同時間軸，效率市場假說應用於比特幣市場能否成立？在此期間比特幣的波動性是屬於經濟基本面還是投資者情緒的表現？以及應用新方法(分位數中介分析)探討原油、比特幣和二氧化碳這三者之間是否有因果關係？因此本研究將針對三個研究主題進行深入研究與探討，主題一為「經濟基本面和投資者情緒對比特幣波動性的影響」，主題二為「商品期貨與比特幣之間相關性分析」，主題三為「原油和比特幣對二氧化碳排放之影響」。 針對主題一，本研究是基於效率市場假說和噪音交易者理論，分別建構了經濟基本面和投資者情緒的模型，可以從不同的理論角度闡明比特幣的波動性。就主題二，本研究以時間軸建構非預期事件前後商品期貨和比特幣波動之間的向量誤差修正模型(Vector Error Correction Model, VECM)，以探討商品期貨波動對比特幣波動的短長期影響效果，以及是否合乎效率市場假說？就主題三，本研究建構原油、比特幣和二氧化碳的分位數中介分析模型，藉由比特幣的中介效果探討三者間之因果關係。 本研究採用Augmented Dickey-Fuller (ADF)單根檢定、Johansen共整合檢定、VECM、Chow test、分位數中介分析和環境庫茲涅茨曲線(environmental Kuznets curve, EKC)理論，探討投資者情緒、商品期貨與比特幣之間的關係。研究結果顯示，短期是投資者情緒模型變數，長期是經濟基本面模型變數顯著影響比特幣的波動性，這意味著這些變數可以作為衡量比特幣波動性的指標。同時也顯示，非預期事件前多數商品期貨波動顯著影響比特幣波動，但是影響力微小；然而在非預期事件中少數商品期貨波動依然顯著影響比特幣波動，但是影響力卻非常大，此結果驗證了非預期事件下效率市場假說依然成立。本研究的新方法發現比特幣在原油對二氧化碳排放的影響中是有部分中介效果，以及原油與二氧化碳排放之間的關聯包含直接和間接影響，而且二氧化碳排放量分佈越高，原油的正向影響就越大。EKC非線性分析結果顯示，在二氧化碳排放分佈的任何百分位數上，比特幣與二氧化碳排放之間存在非線性U形關係。
並列摘要
In the virtual currency market, Bitcoin, which has the largest market share, has been the subject of research in the academic and industrial sectors in terms of its financial and economic implications, as well as its environmental impact. This study finds that in the field of financial economics, the Efficient Market Hypothesis (EMH) and the Noise Trader Theory are commonly validated in stock markets but have been rarely examined in the virtual currency market. Additionally, in the field of environmental sustainability, there is currently no literature exploring the relationships between commodity futures, such as crude oil, carbon dioxide, and Bitcoin. This study aims to investigate the applicability of the EMH to the Bitcoin market across different time horizons. During different periods, is the volatility of Bitcoin driven by economic fundamentals or investor sentiment? Additionally, the study explores the causal relationships among crude oil, Bitcoin, and carbon dioxide using a new method (quantile-mediated analysis). Therefore, this study conducts in-depth research and analysis on three main topics: (1) “Examines the impact of economic fundamentals and investor sentiment on Bitcoin volatility”, (2) “Analyzes the correlation between commodity futures and Bitcoin”, and (3) “Investigates the influence of crude oil and Bitcoin on carbon dioxide emissions”. For the first topic 1, this study constructs models based on the EMH and noise trader theory to elucidate Bitcoin's volatility from different theoretical perspectives related to economic fundamentals and investor sentiment. For the second topic 2, this study constructs a Vector Error Correction Model (VECM) based on the timeline to investigate the short and long-term effects of commodity futures volatility on Bitcoin volatility before and after unexpected events. This helps determine whether it conforms to the EMH. For the third topic 3, this study constructs a quantile mediation analysis model for crude oil, Bitcoin, and carbon dioxide, exploring the causal relationships among the three variables through Bitcoin's mediating effect. This study employs Augmented Dickey-Fuller (ADF) unit root tests, Johansen cointegration tests, VECM, Chow tests, quantile-mediated analysis, and the Environmental Kuznets Curve (EKC) theory to investigate the relationships among investor sentiment, commodity futures, and Bitcoin. The results indicate that in the short term, variables of investor sentiment models and in the long term, variables of economic fundamentals models significantly affect the volatility of Bitcoin. This implies that these variables can be used as indicators to measure the volatility of Bitcoin. This suggests that these variables can serve as indicators for measuring Bitcoin volatility. Furthermore, the study reveals that the majority of unexpected events in commodity futures have a significant but small influence on Bitcoin volatility. However, a few commodity futures exhibit a substantial impact on Bitcoin volatility during unexpected events, thus confirming the validity of the EMH in such circumstances. The novel approach in this study discovers partial mediation effects of Bitcoin in the relationship between crude oil and carbon dioxide emissions. It also reveals that the association between crude oil and carbon dioxide emissions involves both direct and indirect effects, and the positive influence of crude oil increases with higher levels of carbon dioxide emissions. The non-linear analysis of the EKC indicates a non-linear U-shaped relationship between Bitcoin and carbon dioxide emissions at any percentile of the carbon dioxide emission distribution.