The economics of BitCoin price formation

 The inefficiency of Bitcoin

Introduction:

The existing literature on Bitcoin focuses on the main topics related to this new cryptocurrency: the economic, legal and security aspects. Researcher Urquhart  had completed the literature on Bitcoin by studying the efficiency of its market. Following a set of tests, Urquhart  reveals that Bitcoin returns are significantly inefficient over the entire study period. For this reason, he divides the period into two subsamples. He notes that some tests indicate that the Bitcoin market is inefficient in the first period. However, its market could be evolved into an efficient market in the second period. Indeed, the efficient market hypothesis is one of the crucial topics of finance, developed by Fama in 1970. An efficient market means that the prices of Goods and Services reflect all the information available on the market. Fama distinguished three forms of efficiency. The form most examined in the review was the weak form of efficient where the market does not display the information necessary to predict future asset returns.

Objective:

In this article we will 

• We study the informational efficiency of Bitcoin 
• We employ a battery of tests and find evidence of market inefficiency 
• We find that some tests indicate market efficiency in the latter period 

Problem: 

This article aims to  find answer to this question:

Is the Bitcoin market efficient over time?

Literature review:

• Cheach and Fry (2015) argue that if Bitcoin were a real unit of account or a store of value, it would therefore not display high volatility leading to speculative bubbles and crashes. 
• Dwyer (2015) finds that the volatility of Bitcoin is higher than that of gold and foreign currencies.
• Cheung et al (2015) show the existence of bubbles, in the short term, on the Bitcoin market. the most important is the bursting of Mt.Gox. 
• Brière et al (2015) show that Bitcoin offers diversification advantages for investors.
• Dyhrberg (2016a, 2016b) indicates that Bitcoin has similar hedging capabilities compared to gold and the dollar, and therefore it can be used in risk management.

Data and methodology:

• David Urquhart (2016) collects data from site www.bitcoinaverage.com , which represents the first aggregated Bitcoin price index that aggregates the rates of all Bitcoin exchanges available in the world and provides a volume-weighted average Bitcoin price. The data shows the daily closing prices for Bitcoin in USD from August 1, 2010 to July 31, 2016.

• Urquhart (2016) examines the efficiency of the Bitcoin market over any sampling period by calculating the Bitcoin yield :
  
  

With : 

R_t: The Bitcoin yield, ln⁡(P_t ) and ln⁡(P_(t-1) ): represent the logarithmic function of Bitcoin prices in period t and t-1.

•  And then, he divides the sample into two subsamples in order to determine if the level of efficiency varies over time. Therefore, the total sampling period is from August 1, 2010 to July 31, 2016, and the two subperiods are from August 1, 2010 to July 31, 2013 and from August 1, 2013 to July 31, 2016.

To analyze the efficiency of Bitcoin over time, Urquhart uses a set of tests to present adequate results and determine the dynamics of Bitcoin prices :

•   Firstly, Urquhart examine the autocorrelation of returns which are assessed via the Ljung-Box (Ljung and Box 1978).

• Secondly, the runs test (Wald and Wolowitz 1940) and the Bartels test (Bartels 1982) are employed to determine whether returns are independent.

• Thirdly, he employs the variance ratio test (Lo and MacKinlay 1988). A problem with this test is the choice of parameters q and p, so we adopt the automatic variance test (AVR) of Choi (1999) where they are determined automatically using a data dependent procedure. Urquhart utilize the wild-bootstrapped AVR test of Kim (2009b), which greatly improves the small sample properties of the AVR test.

•  Fourth, he uses the BDS test (Brock et al 1996), which is a popular nonparametric test for the serial dependence of stock returns.

• Finally, the rescaled Hurst exponent (R/S Hurst) for long memory of stock returns is employed

Results:

According to this research Urquhart finds that the informational efficiency of Bitcoin can be rejected. That is, the total sampling period indicates a significant inefficiency of Bitcoin.

Urquhart thus finds, according to the tests used, that the Bitcoin market is inefficient in the first subperiod. However, the Bitcoin market seems to become less inefficient in the second subperiod. This means that its market could be evolved into an efficient market over time as more and more investors know and trade this cryptocurrency.

Conclusion:

Bitcoin has received much attention in the media and by investors in recent years, although there remains skepticism and a lack of understanding of this cryptocurrency. In this article the researchers add a value to Bitcoin background by studying the market efficiency of Bitcoin.  

Through a battery of robust tests, evidence reveals that returns are significantly inefficient over their full sample, but when they split the sample into two subsample periods, they find that some tests indicate that Bitcoin is efficient in the latter period.  Therefore they conclude that Bitcoin in an inefficient market but may be in the process of moving towards an efficient market.