[Paper Review] The Bitcoin price formation: Beyond the fundamental sources

Jamal Bouoiyour, Refk Selmi|arXiv (Cornell University)|Jul 5, 2017

Blockchain Technology Applications and Security12 references31 citations

TL;DR

This study employs Bayesian quantile regression to analyze Bitcoin price formation across the full conditional distribution, identifying distinct drivers under varying market regimes. It reveals that trade use and geopolitical uncertainty (e.g., China’s slowdown, Brexit) drive price appreciation during market upswings, while macro-financial factors like hash rate and gold prices dominate during downturns.

ABSTRACT

Much significant research has been done to investigate various facets of the link between Bitcoin price and its fundamental sources. This study goes beyond by looking into least to most influential factors-across the fundamental, macroeconomic, financial, speculative and technical determinants as well as the 2016 events-which drove the value of Bitcoin in times of economic and geopolitical chaos. We use a Bayesian quantile regression to inspect how the structure of dependence of Bitcoin price and its determinants varies across the entire conditional distribution of Bitcoin price movements. In doing so, three groups of determinants were derived. The use of Bitcoin in trade and the uncertainty surrounding China's deepening slowdown, Brexit and India's demonetization were found to be the most potential contributors of Bitcoin price when the market is improving. The intense anxiety over Donald Trump being the president of United States was shown to be a positive determinant pushing up the price of Bitcoin when the market is functioning around the normal mode. The velocity of bitcoins in circulation, the gold price, the Venezuelan currency demonetization and the hash rate were found to be the fundamentals influencing the Bitcoin price when the market is heading into decline.

Motivation & Objective

• To investigate the determinants of Bitcoin price beyond traditional fundamental factors.
• To examine how the influence of various determinants—fundamental, macroeconomic, financial, speculative, and technical—varies across different market conditions.
• To assess the role of specific geopolitical and economic events (e.g., Brexit, Trump’s election, demonetization) in shaping Bitcoin price dynamics.
• To identify regime-specific drivers of Bitcoin price using a non-linear, distributional approach.
• To provide a comprehensive framework for understanding price formation in decentralized digital assets under economic and political stress.

Proposed method

• Application of Bayesian quantile regression to model the conditional distribution of Bitcoin price returns.
• Incorporation of diverse determinants: fundamental (e.g., hash rate), macroeconomic (e.g., gold price), financial (e.g., volatility), speculative (e.g., investor sentiment), and technical (e.g., velocity of money).
• Inclusion of event dummies for major geopolitical and economic shocks (e.g., 2016 U.S. election, India’s demonetization, Venezuela’s currency reform).
• Estimation of regression coefficients at multiple quantiles (e.g., 0.1, 0.5, 0.9) to capture regime-dependent effects.
• Use of Markov Chain Monte Carlo (MCMC) methods for posterior inference under Bayesian framework.
• Model validation through diagnostic checks and comparison of predictive performance across quantiles.

Experimental results

Research questions

• RQ1 How do the determinants of Bitcoin price vary across different points in the conditional distribution of returns?
• RQ2 Which factors—fundamental, macroeconomic, speculative, or technical—most strongly influence Bitcoin price during market upswings versus downturns?
• RQ3 To what extent do geopolitical and economic events such as Brexit, Trump’s election, or demonetization events drive Bitcoin price movements?
• RQ4 How does the influence of speculative sentiment and investor anxiety manifest in Bitcoin price formation during stable versus stressed market conditions?
• RQ5 What role do monetary and technological fundamentals (e.g., hash rate, velocity of transactions) play in shaping Bitcoin price during bear markets?

Key findings

• The use of Bitcoin in trade and uncertainty surrounding China’s economic slowdown were the most influential positive drivers during market upswings.
• Geopolitical events such as Brexit and India’s demonetization significantly contributed to upward price pressure during periods of market improvement.
• Anxiety over Donald Trump’s presidency acted as a positive determinant, particularly when the market operated near its median or normal mode.
• During market downturns, the velocity of bitcoins in circulation, gold price, Venezuelan currency demonetization, and hash rate emerged as key fundamental drivers.
• The study identifies three distinct groups of determinants based on their influence across different quantiles of the return distribution.
• The Bayesian quantile regression model successfully captured the non-linear, regime-dependent structure of Bitcoin price dependence on its determinants.