This thesis improves the understanding of cryptocurrencies as financial assets by examining the Bitcoin reaction to high-frequency news compared to Forex, explores the role of news within financial markets, quantifies the economic value of a novel investment strategy which times financial noise able to manage price noise-risk, and assesses the extent to which climate change physical and transition risks are incorporated into asset prices. The thesis consists of three essays. The first essay "News sentiment in the cryptocurrency market: An empirical comparison with Forex" considers high frequency intra-day data to investigate the influence of unscheduled currency and Bitcoin news on the returns, volume and volatility of the cryptocurrency Bitcoin and traditional currencies over the period from January 2012 to November 2018. Results show that Bitcoin behaves differently to traditional currencies. Fiat currencies typically experience a decrease in returns after negative news arrivals and an increase in returns following positive news whereas Bitcoin reacts positively to both positive and negative news. This suggests investor enthusiasm for Bitcoin irrespective of the sentiment of the news. This phenomenon exacerbates during bubble periods. Conversely, cryptocurrency cyber-attack news and fraud news dampen this effect, decreasing Bitcoin returns and volatility. The second essay "The economic value of financial noise timing" proposes a dynamic noise-timing strategy which exploits the temporary dependence in noise traders' beliefs. Decomposing prices of the portfolio assets (stocks, bonds, gold, and cryptocurrencies) into permanent and noise components, we assess the economic value of a dynamic investment strategy which times the noise component. Our results show that risk averse and short horizon investors would be willing to pay a positive annual performance fee of between 314 and 940 basis points to switch from an ex-ante static investment strategy to a noise timing strategy. Our findings are robust to comparisons with other benchmark strategies, such as the volatility timing, and different periods of heightened volatility, including the Covid-19 period. The third essay "Transition versus physical climate risk pricing in euro area financial markets: A text-based approach" prices climate change risks in equity markets within a Fama-French five factor model. We build two novel vocabularies on physical and transition climate risks, and we construct a Physical Risk Index and a Transition Risk Index comparing them to a corpus of news over the period 2015-2019 using the cosine-similarity approach. Climate news are found to carry relevant information especially for brown firms, with transition risk appearing to be more concerning for investors. Returns of low environmental and ESG scores firms negatively relate to both shocks to physical and transition risk, whereas returns of high Greenhouse Gas emissions levels and intensity firms further decline with transition risk news. While investors appear to penalise high climate risk exposure, there is no evidence of an increase in returns of less exposed firms.
Date of Award | 1 Aug 2021 |
---|
Original language | English |
---|
Awarding Institution | - The University of Manchester
|
---|
Supervisor | Stuart Hyde (Supervisor) & Sarah Zhang (Supervisor) |
---|
- Economic value
- Portfolio allocation
- Climate risk
- Physical risk
- Pricing of climate risk
- Textual analysis
- Transition risk
- Kalman filter
- Noise risk
- Noise
- Foreign Exchange
- News sentiment
- Bitcoin
- Cryptocurrency
- Digital currencies
- Noise trading
Essays in empirical finance: News sentiment in cryptocurrency, the value of noise timing, and the pricing of climate change risks
Rognone, L. (Author). 1 Aug 2021
Student thesis: Phd