An empirical study, Johansen Co-integration test, stationarity, long term relationship
In this article, we will look at the effects of Bitcoin on the American financial market . In fact, digital money can basically be examined in two categories. The first is produced by financial companies and the other is produced by non-financial companies.
As I mentioned in my previous article, credit cards issued by banks can be an example of digital money of financial institutions. The common markets of financial institutions for this type of digital money also serve under the name of internet banking.
As mentioned at the beginning, the effects of bitcoin on the American financial and economic market are studied in this article. In addition, in the analysis, before and after the crisis were examined separately on the basis of the 2008 crisis.
In the study, in which the Bitcoin price was taken as the dependent variable, the variables thought to be explanatory were the ounce gold price, DXY Dollar index, Mortgage index, 10-Year Treasury Rate, Federal Reserve Index USA and 10-Year Bonds.
The model was simply created as follows:
lnBitcoint= b0 + b1 lnGoldt+ b2 lnDollarindext+b3 InMortgageindex+b4 InTreasury rate+b5 10-Year bonds yield+b4 InFederal Reserve Index + et
|LBitcoin 1st Difference||-6.7288||-3.57782||-2.9253||0.0000||-6.9899||-4.1653||-3.5085||0.0000|
|LGold Price||1st Difference||-3.4762||-3.5777||-2.9251||0.0096||-3.3994||-4.1609||-3.5081||0.0432|
The test results are as seen in the table.
Variables were examined on a daily basis.
In the table, it is seen that all of the variables at the level contain unit root. The first differences of the data are stationary with 5% significance level. The Johansen cointegration test is applied to the series which are stationary at the same level as the cointegration test. This study was conducted to find possible cointegration between seven variables, Bitcoin price, Gold price, Dollar index, Mortgage index, Treasury index, Federal reserve index and bond yield, which were found as I(1) and Johansen tests. Thanks to this technique, the dependent variable is Bitcoin price, while the gold price, Dollar index, Mortgage index, Treasury index, Federal reserve index and bond yield are independent variables. According to the results of the test, it was understood that there was cointegration confirming the long-term relationship between the variables.
Finally, we can run a constrained VAR model, which is VECM (vector error correction model), if the variables mean cointegrating, meaning they move together in the long-run association-ship. This model shows the term C (1)>> Error correction…… If C (1) is negative and significant (p value <5%), we can say that there is an Adjustment Speed towards Equilibrium in two issues. There is long-run causality from independent variables to dependent variables.
From Bitcoin price means a long-term relationship with Gold price, Dollar index, Mortgage index, Treasury index, Federal reserve index. Bond yield has a negative relationship with Bitcoin price.
Also read” Digital Currency, Bitcoin (Volume 1)