The rest of the paper is organized as follows: Section 2 reviews the literature and develops a hypothesis; Section 3 describes data; Section 4 explains the methodology; Section 5 presents results; and Section 6 concludes.
2. Literature Review and Hypotheses
Bitcoin appeared in 2009 as the first cryptocurrency. It was created by an anonymous internet group operating under the pseudonym Satoshi Nakamoto and was initially introduced as an alternative to conventional currencies. Bitcoin is the most important of all virtual currencies because it held an 89% share of all virtual currency market capitalization as of December 2016 (Bariviera et al., 2017). Bitcoin prices over time are substantially more volatile than conventional currency. Blau (2017) suggests that the volatility of Bitcoin prices is double the average of the 51 regular currencies under their study from July 2010 to June 2014. Through correlation testing, Blau (2017) concludes that Bitcoin returns were unrelated to speculative trading. Differently, Cheah and Fry (2015) show that Bitcoin price exhibits speculative bubbles and the fundamental value of Bitcoin is zero.
Although it is a non-traditional currency Bitcoin has similarities with traditional currencies. Yermack (2013) evaluates the validity of Bitcoin as a currency against the three required functions of a currency. He states that although Bitcoin satisfies the function as a medium of exchange, it cannot be a store of value or a unit of account, which are two of the three attributes required when being considered a currency by economists. Dyhrberg (2016b) investigates whether Bitcoin more resembles a commodity or a currency. The author finds that Bitcoin returns have a significant positive reaction to the US Federal Funds rate, similar to the US dollar. Bitcoin is also found to provide risk-management capabilities against Dollar-Pound and Dollar-Euro exchange rates, similar to what researchers found in gold. Dyhrberg (2015b) concludes that Bitcoin can be classified as something in between the US dollar and gold as a commodity asset. Luther and Salter (2017) analyze the increase in Bitcoin app downloads after Cyprus bailout announcement. While Bitcoin app downloads increased in both the US and Cyprus after the bailout announcement, interest was greater in the US, possibly suggesting that Bitcoin is not replacing the currencies of countries with troubled banks. The other line of research analyzes Bitcoin’s hedging ability as an asset. Dyhrberg (2016a) used daily data from 2010 to 2015 to test Bitcoin’s hedging ability against some UKrelated assets including USD-EUR Exchange rates, USD-GBP Exchange rates, and the FTSE. The results showed that Bitcoin was uncorrelated to the FTSE Index in both lagged and contemporaneous returns, and is uncorrelated with Dollar-Euro and Dollar-Sterling contemporaneous returns, but is positively correlated to the lagged exchange rate returns. These indicate that Bitcoin could be a weak hedge against UK assets.
Bouri et al. (2017) hypothesized that Bitcoin could hedge negative movements in world stock indices (S&P 500, FTSE 100, DAX 30, Nikkei 225, Shanghai A-Share, MSCI, Bond, US dollar, and Commodity), gold prices, and oil prices. Using daily and weekly price index data from mid-2011 to end of 2015, their results indicated that Bitcoin had the ability to hedge against the Nikkei, the MSCI Pacific and commodity index as their coefficients were negatively correlated and significant. However, the significance of these coefficients faded when examining the weekly data. This is an example of how data frequency may change the hedging ability of Bitcoin.
Bouri et al. (2017) and Dyhrberg (2016a) demonstrate that Bitcoin possesses riskmanagement abilities against regional assets related to UK and Asia. Bouri et al. (2017) caution that the diversification ability of Bitcoin is not constant over time and future studies on the timevarying nature of these risk-management abilities are necessary. We will use more up-to-date data to investigate Bitcoin’s daily, weekly, and monthly hedging abilities against five major stock indices, i.e., the S&P 500, Nikkei, Shanghai A-Share, the TSX Composite Index, and the Euro Index. Two of these five indices (TSX and Euro index) haven not been studies by other research. Longer horizon investors would more likely require monthly data trends rather than daily or weekly analyses.
Bouri et al. (2017) and Dyhrberg (2016a) find that Bitcoin can hedge Asian and UK assets, but the hedging relationship fades with weekly data. Therefore, we formulate our hypotheses as follows:
H1: Bitcoin can hedge and diversify against certain assets among S&P 500, Nikkei, Shanghai A-Share, TSX Composite Index, and Euro Index.
H2: The hedging and diversification abilities of Bitcoin differ under different data frequencies.
3. Data
Bitcoin (BTC) daily price data from October 2010 (the earliest data available) to October 2017 was retrieved from the Coindesk Price Index (2017). We choose five indices, S&P 500 (GSPC), Nikkei (N225), Shanghai A-Share (SSE Composite), TSX Composite Index (GSPTSE) and Euro Index (STOXX50E), which represent different regions whose currencies top the share of Bitcoin trade. Bitcoin trading against the Chinese Yuan used to account for most of Bitcoin trading volume until China started to clamp down on digital currency exchanges in early 2017 and eventually banned the trading of Bitcoin in September of 2017. Japan’s Yen then took over as the regulators in Japan adopted digital currency-friendly rules. The US Dollar and Euro are also among the top five most active Bitcoin trading currencies (Russo and Migliozzi, 2017). Daily, weekly and monthly prices were sourced from Yahoo. Our sample consists of 1,828 observations for daily frequency, 366 observations for weekly frequency, and 85 observations for monthly frequency for all assets under study.
To be continued
No comments:
Post a Comment