| dc.description.abstract | This thesis examines whether the unconventional monetary policy measures pursued by the Federal Reserve, the Bank of England, the Bank of Japan, and the European Central Bank since the Global Financial Crisis (GFC) are associated with an appetite for cryptocurrency as a resistance movement against the perceived domination of institutional authority, reflecting a lack of trust in central banks. Previous studies have not found evidence of a relationship due to their exclusive focus on cryptocurrency price effects (Benigno & Rosa, 2023; Fama, Fumagalli & Lucarelli, 2019; Ly'csa et al., 2020; Nguyen et al., 2019; Pyo & Lee, 2020; Vidal-Tom's & Ibañez, 2018) which have been shown to be heavily manipulated (Gandal, 2018; Griffin & Shams, 2020; Hu et al., 2020; Peterson, 2021; Twomey & Mann, 2020). This research extends beyond cryptocurrency price effects to the underlying cryptocurrency activity metrics, specifically for the largest cryptocurrency, bitcoin, and utilises a novel approach to capturing monetary policy media attention, extracted from a unique corpus of Twitter data. In its capacity as one of the world's leading media platforms, Twitter activity data provide an overview of how news is being perceived, in terms of both the level of attention and the type of reaction that it generates. As a media platform, it is particularly relevant in this context as cryptocurrency users tend to receive their news from social media rather than directly from traditional media outlets (Anser et al., 2020; Garcia, 2014). The thesis follows a three-paper format, comprising three related studies. The first study investigates the linkage between two types of unconventional monetary policy, Quantitative Easing (QE) and Negative Interest Rate Policy (NIRP), and the cryptocurrency, bitcoin. Using ARDL-OLS modelling, the findings demonstrate a significantly positive relationship between QE and NIRP media attention and bitcoin-related activity, and, consistent with previous literature, no association with bitcoin price returns. The second study focuses exclusively on the relationship between Quantitative Easing (QE) and bitcoin, to investigate the scope of central bank QE policy association with a selection of bitcoin-related activity metrics. I employ a novel event study approach using dynamic ARMAX modelling of controlled interrupted time series and show that QE media attention linked to the Federal Reserve (Fed) has the most significant association with bitcoin-related activity, with the ECB and the Bank of England exerting intermittent significance. To reflect the evolvement of Fed QE policy, I distil the study period into five distinct phases, namely (i) Expansionary (ii) Tapering (iii) Stable (iv) Quantitative Tightening (QT) and (v) Crises (of short-term US dollar liquidity and the Covid-19 pandemic). I find that the Fed had significant association with bitcoin-related activity across all phases, including, surprisingly, the Tapering and QT periods, which suggests that it is not the direction of QE policy that is of most relevance, but rather the perceived manipulation of monetary supply generally, emblematic of institutional distrust stemming from the centralised control of the fiat monetary system. My final study delves deeper into the public discourse surrounding QE on Twitter, using supervised Machine Learning to perform a sentiment analysis. This mixed-method study comprises an initial qualitative inquiry to identify the sentiment polarities and underlying themes, which are subsequently used to train a classification model. Comparing the various algorithm classifiers of Logistic Regression, Na've Bayes, and Support Vector Machines, Binary Logistic Regression with a TF-IDF vectoriser is selected due to its highest accuracy score. I identify a predominance of negative sentiment towards QE, comprising five emergent themes, namely `anti-government', `anti-central bank', `inequality', 'ineffectiveness', and `inflation'. Furthermore, I test if the sentiments of inequality, ineffectiveness and inflationary concerns are associated with an anti-establishment conviction. These negative tweet sentiments indicate that the GFC and the ensuing unconventional policy responses precipitated a distrust amongst the public towards central banks, providing evidence for increased appetite for cryptocurrency as a form of resistance movement. This dissertation makes a theoretical contribution to Social Exchange Theory, where I posit that the increased appetite for cryptocurrency associated with unconventional monetary policy since the GFC stems from a lack of trust in the central banks, and I identify three meaningful trust determinants that support the public-institutional exchange partnership, where there exists a power imbalance. From an empirical perspective, the Twitter data employed in the study represent two wholly unique datasets, obtained directly from the Twitter API, and which are applied both as an attention proxy, and to offer rare insights into the sentiment response of social media users to QE. Undoubtedly, the advent of cryptocurrencies, combined with the reduced utility of physical cash since the Covid-19 pandemic, has paved the way for central banks to adopt their own central bank digital currencies (CBDC), and the results of this research will be important to policymakers in the consideration of a CBDC design. Finally, referencing the extensive history of the financial system via the Schools of Economic Thought, I discuss future avenues for the evolution of bitcoin and propose some recommendations for central banks to regain and rebuild public trust. | en |
