Sánchez López, Julián Fernando, Julián Fernando Sánchez López
abstract
This doctoral thesis presents an in-depth investigation into the optimal liquidation of cryptocurrency assets and the analysis of market dynamics under the Weak Efficient Market Hypothesis (WEMH) across various cryptocurrencies. The first paper of the thesis focuses on the liquidation of Binance Coin (BNB), analyzing temporary and permanent price impacts using Limit Order Book (LOB) data from Binance exchange. It introduces linear and quadratic models for Permanent Price Impact (PPI) and derives optimal liquidation strategies through closed-form solutions. The study finds that the quadratic PPI model notably outperforms linear models in capturing permanent price impacts in financial trading. The second paper extends this investigation by applying finite differences and optimal policy iteration to numerically solve the liquidation problem under different scenarios of price impact estimation. It characterizes optimal liquidation policies based on various parametrizations and compares their performance with naive and common strategies, highlighting the importance of the inventory's functional form in determining revenue-maximizing policies. The third paper diverges to examine market dynamics, introducing a novel methodology for identifying epochs of upward trends, downward trends, and mean reversion using statistical techniques. It seeks to evaluate market efficiency within these periods under the Weak Efficient Market Hypothesis (WEMH) through methods including the Hurst index, Shannon entropy, and autocorrelation tests. The analysis concludes that cryptocurrency markets do not uniformly adhere to WEMH principles. While certain periods and frequencies display efficiency, others exhibit predictability and inefficiency, highlighting the complex and fluctuating nature of market efficiency across different cryptocurrencies and market conditions. Overall, this thesis contributes to the field by providing nuanced insights into asset liquidation strategies in the context of cryptocurrency trading and advancing the understanding of market efficiency dynamics in these novel financial markets.
publication date
July 29, 2024 4:38 PM
Research
keywords
Analytical solutions
Autocorrelation
Binance Coin (BNB)
Cryptocurrency
Cryptocurrency Liquidation
Cryptocurrency Trading
Financial Trading Models
Finite Difference Methods
High-Frequency Data
Hurst Index
Limit Order Book (LOB) Data
Market Efficiency
Market Scenarios (Underestimation, Overestimation, Average Estimation)
