**3. Empirical analysis**

There is a plethora of information about Bitcoin. The blockchain provides free information about every detail of the market on an hourly basis. To better understand the market size of Bitcoin, market capitalization [9] data are used. Further investigation is carried out to understand the price impact of Bitcoin as a currency against the US Dollar and the economic movement in terms of a commodity, and the following time series variables are analyzed. The sample is taken from Sep 2015 till July 2017 from coinmarketcap [10].


#### **3.1. Regression analysis**

the price and the volume. In public stock exchanges, shares are publicly traded, transactions are recorded and the volumes are displayed. Hence, it becomes easy to calculate the trading volume for such stock exchanges. To better understand this trade volume, let us consider the

Suppose that Samsung only changes hands twice during the day. Let's assume that 20 million shares were bought at \$10 a share; later 30 million shares were purchased at \$10.5 a

There is a plethora of information about Bitcoin. The blockchain provides free information about every detail of the market on an hourly basis. To better understand the market size of Bitcoin, market capitalization [9] data are used. Further investigation is carried out to understand the price impact of Bitcoin as a currency against the US Dollar and the economic movement in terms of a commodity, and the following time series variables are analyzed. The

following example:

**3. Empirical analysis**

The total trading volume = 20 +30 = 50 million shares

**Figure 3.** Ethereum price in USD from Sep 2015 to July 2017.

272 Financial Management from an Emerging Market Perspective

Dollar amount equals = (20X10) + (30X10.5) = \$515 million

So, the total circulation of Samsung Shares is 515 million for that day.

sample is taken from Sep 2015 till July 2017 from coinmarketcap [10].

EXAMPLE:

share.

A regression analysis is used to explain the impact of the variables, market capitalization (CAP), output volume (VOL) and Bitcoin closing price (BTC), and then the relationship of the regression equation to the model is discussed. The best regression equation, based on the analysis of DW (Durbin-Watson), AIC (Akaike **i**nformation **c**riterion) and SC (Schwarz **c**riterion), is fitted (**Table 1**).

From the resultant equation, the t-statistics, r-squared and adjusted r-squared show a strong positive relation and suggest that the regression equation fits well as indicated in **Figure 4**.

#### **3.2. The augmented Dickey-Fuller (ADF) unit-root test**

ADF is used under the three conditions for every time series. The random process includes intercept (c) and trend (t), second includes intercept (c) but no trend (0) and third includes no intercept (0) and trend (t). It was observed that each variable under the Augmented Dickey-Fuller test statistic has a unit root at various lag lengths. This augments the data and the model (**Figure 5**).

#### **3.3. Objectives of the gradient function**

**Table 2** and **Figure 4** allow us to observe the gradients of the objective function for Bitcoin and to help us find a unit root before applying the cointegrating techniques. The result indicates that all the variables, Bitcoin price, market capitalization and volume have a unit root in their levels and are stationary in their first-order differences (**Figure 6** and **Table 3**).


**Table 1.** Regression results.

**Figure 4.** Residual, actual and fitted.

**Figure 5.** BTC estimated value, denoted as BTCF.

#### **3.4. Var cointegration test statistic**

The Johansen **c**ointegration technique is used to check the behavior of the variables. The results obtained are presented in **Table 4**. The cointegration relationships are determined with lag intervals between 1 and 4 with 5% critical values. The unrestricted co-integration rank is applied.


#### **Null hypothesis: D(BTC) has a unit root**

**Table 2.** The ADF test statistic.

#### **3.5. Granger causality tests**

The Granger causality test is used to analyze further the relationship between the three variables. Pairwise tests are carried out in Eviews, and results are shown in **Table 5**. When the lag is 2, the Granger-cause between the variables does not exist. Hence, this proves that the above cause-and-effect relationship is unidirectional and not bidirectional for BTC.

#### **3.6. Impulse response function**

**3.4. Var cointegration test statistic**

**Figure 5.** BTC estimated value, denoted as BTCF.

0 400 800 1,200 1,600 2,000 2,400 2,800

3,200


**Figure 4.** Residual, actual and fitted.


0

200

400

600

274 Financial Management from an Emerging Market Perspective

III

The Johansen **c**ointegration technique is used to check the behavior of the variables. The results obtained are presented in **Table 4**. The cointegration relationships are determined with lag intervals between 1 and 4 with 5% critical values. The unrestricted co-integration rank is applied.

III IV I II III IV I II III 2015 2016 2017

Residual Actual Fitted

Mean Absolute Error 22.27751 Mean Abs. Percent Error 2.296855

Theil Inequality Coefficient 0.022783 Bias Proportion 0.000000 Variance Proportion 0.001420 Covariance Proportion 0.998580

Forecast sample: 9/05/2015 7/22/2017 Included obser vations: 687 Root Mean Squared Error 47.24789

0 500 1,000 1,500 2,000 2,500 3,000 3,500

Forecast: BTCF Actual: BTC

IV I II III IV I II III

BTCF ± 2 S.E.

2015 2016 2017

To categorize the dynamic structure of Bitcoin, the Monte Carlo simulation is applied through the impulse response functions in the model. It shows how shocks to any one variable filter through the model can affect every other variable and eventually feed back to the original variable itself.

Gradients of the Objective Function

**Figure 6.** Gradients of the objective function.


**Table 3.** Gradients of the objective function.




**Figure 6.** Gradients of the objective function.



0

400

800


0E+00

2E+13

4E+13

III IV I II III IV I II III 2015 2016 2017

CAP

III IV I II III IV I II III 2015 2016 2017

C

III IV I II III IV I II III 2015 2016 2017

Gradients of the Objective Function VOL



0E+00

1E+12

2E+12

276 Financial Management from an Emerging Market Perspective


Trace test indicates 2 cointegrating eqn(s) at the 0.05 level; Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.05 level.

\* Rejection of the hypothesis at the 0.05 level.

\*\*MacKinnon-Haug-Michelis (1999) *p*-values [12].

**Table 4.** Johansen cointegration test.


**Table 5.** Pairwise Granger Causality Tests with lags 2.

Monte Carlo [13] simulation is best employed through possible random movements in the model. This perfectly fits the scenario of Bitcoin. There are two components to a stock's price movements: drift, which is a constant directional movement, and a random input, representing market volatility. By analyzing historical price data, the drift, standard deviation,

**Figure 7.** Impulse response functions—Monte Carlo simulation.

variance and average price movement for a security is determined. These are the building blocks of a Monte Carlo simulation. The horizontal spool in **Figure 7** delegates trace periods of the response function, and the vertical spool delegates responses of dependent variables to independent variables.

#### **4. Bitcoin and the future economy**

In an era where having a smart phone is highly more likely than having a Bank account, Bitcoin is here to stay. The architecture of money that we live in today brings the borrower and lender together through a financial intermediary, usually a Bank. One of the parties usually conforms to the rules and regulations put forth by the financial institution making it the most sought-after financial product the banks have to offer. People with limited education and a profile not matching to open a bank account can open a Bitcoin account in under 30 seconds.

Bitcoin is Gold 2.0 because Bitcoin is great as a store of value.

As of June 2017, the total value of all cryptocurrency in circulation is now [14] almost \$100 bn.

This is almost double of what was at the beginning of this year. The value of the fiat currency issued by the Federal Reserve Bank of \$1.4 trillion cannot be undermined, but an alternative is underway albeit faster than predicted. Cryptocurrencies are borderless, decentralized, and capable of replacing money in just about any transaction. Without any physical infrastructure, Bitcoin has paved its way into the world economy because of ease of use. The "nostrings" attached idea has appealed to the masses. It is positively one of the most innovative inventions ever since the Internet came. It is Internet money and a digital currency; to stop this, we would have to stop the Internet. This is the power of Bitcoin.

From pizza to Porsche, hotel bookings and even goat, [15] you can buy just about anything by using Bitcoin.
