CHAPTER FOUR

PRESENTATION AND ANALYSIS OF RESULTS

This chapter presents the results of the estimation of the model specified in chapter three.

This is necessary in order to test the hypotheses as stated in chapter one of this work. We would

look at the economic, econometric, statistical criterions coupled with the descriptive statistics of

all the variables used in this work. This chapter will also present its post estimation tests.

4.1 PRE-ESTIMATION TEST RESULTS

Before the estimation of the OLS model, the variables in the data set were standardized because

they existed at different scales. For instance, Bitcoin price (BTC) is in thousands of dollars,

Volume of BTC traded (VBTC) is in millions of dollar while Exchange Rate (EXR), Inflation

(INFL), Interest Rate (INTR), Gross Domestic Product (GDP) and Inflation Rate (INFL). To

correct this the variables were log transformed.

4.1.1 UNIT ROOT TEST RESULTS

TABLE 4.1.1: KWIATKOWSKI-PHILLIPS-SCHMIDT-SHIN TEST STATISTICRESULTS

KPSS
T-
KPSS T-
STATI
STATIST
KPSS T- KPSS 5% STICS KPSS 5% KPSS 5% ORDER OF
ICS AT
VARIABLE STATISTIC CRITICAL AT CRITICAL CRITICA INTEGRATI
SECOND
S AT LEVEL VALUE FIRST VALUE L VALUE ON
DIFFER
DIFFE
ENCE
RENC
E
EXR 0.500323 -2.90485 -2.63784 -2.90485 -6.99274 -2.90692 I(2)
BTC -1.31685 -2.99806 -3.16257 -2.90485 - - I(1)
VBTC -2.42145 -2.90485 -3.5599 -2.90485 - - I(1)
INFL 0.053498 -2.90485 -2.78231 -2.90485 -8.06876 -2.90552 I(2)
INTR -0.374774 -2.90921 -2.27428 -2.90921 -8.63058 -2.90921 I(2)
TOP -1.352840 -1.94567 -3.26714 -3.00486 - - I(1)

The table above shows the unit root test results using the Augmented Dickey-Fuller test.

The Augmented Dickey-Fuller test result can be interpreted using either the t-statistic or the

pvalue.

The Hypothesis tested is;

H0: The Variables are non-Stationary (Unit Root).

H1: The Variables are Stationary (No Unit Root).

Decision Rule

Reject the null Hypothesis if the ADF t-statistic in absolute terms is greater than the ADF

5% critical value or the p-value is less than or equal to 0.05 level of significance. From the table

above, since the ADF t-statistic (in absolute terms) of all the variables is greater than their

corresponding ADF 5% critical value at first difference, therefore we reject the null hypothesis

and conclude that all the variables are stationary at first difference and therefore integrated at

first order.

This requires us to perform a cointegration test to test if the variables in the model have a

long run relationship.

4.1.2 COINTEGRATION TEST RESULTS

TABLE 4.1.2: ENGLE-GRANGER COINTEGRATION TEST RESULT

VARIABLE ADF T- ADF 5% ORDER OF

STATISTICS CRITICAL INTEGRATION
AT LEVEL VALUE

RESIDUAL -3.522967 -2.904848 I(0)

The table above shows the results of the Engle-Granger Two Steps procedure which is

obtained by performing an Augmented Dickey-Fuller test on the residuals of the model.

The Hypothesis tested is;

H0: The variables in the model are not cointegrated.

H1: The variables in the model are cointegrated.

Reject the null Hypothesis if the ADF t-statistic in absolute terms is greater than the ADF

5% critical value or the p-value is less than or equal to 0.05 level of significance. From the table

above, since the ADF t-statistic (in absolute terms) of the residual is greater than the

corresponding ADF 5% critical value at levels, therefore we reject the null hypothesis and

conclude that the residual is stationary at level and therefore integrated at order one. That is, the

variables are cointegrated.

4.1.3 GRANGER CAUSALITY TEST

The result of the Granger causality model specified in chapter three is tabulated in table

4.4.1 and analyzed below. However, note that the hypotheses we are working with are given as:
H01: EXR does not Granger cause VBTC.

H02: VBTC does not Granger cause EXR.

H03: BTC does not Granger cause EXR.

H04: EXR does not Granger cause BTC.

We are concerned with determining the direction of causality among EXR, BTC and VBC.

TABLE 4.4.1: THE GRANGER CAUSALITY MODEL

Null Hypothesis Probability Decision Conclusion

H01: EXR does not Granger Reject H0 Causality
cause VBTC. 0.0682
H02: VBTC does not Reject H0 Causality
Granger cause EXR. 0.6986
H03: BTC does not Granger Reject H0 Causality
cause EXR. 0.9493
H04: EXR does not Granger Reject H0 Causality
cause BTC. 0.0873

Decision Rule: If the probability is less than 0.05, we reject the hypothesis. Otherwise, we do not

reject. Therefore, we conclude that there is no causal relationship between volume of bicoin

traded, bitcoin price and exchange rate as their pvalue is greater than 0.05 in the table above.

TABLE 4.2: DEPENDENT VARIABLE: EXR

VARIABLE COEFFICIENTS STANDARD T-STATISTICS P-VALUE

ERROR
C 1171.803 558.4082 2.098471 0.0398
BTC 68.69114 31.70600 2.166503 0.0340
VBTC -244.3182 145.4049 -1.680261 0.0978
 INFL 2.802111 0.859472 3.260268 0.0018
INTR -11.93449 5.013676 -2.380387 0.0203
TOP -11.66427 54.25807 -0.214978 0.8305

Note: Standard Error, T-Statistics and Pvalue reported above are Newey-West

Heteroskedasticity and Autocorrelation corrected

The regression result above will be interpreted using the following Criteria:

4.3 ECONOMIC CRITERIA:

4.3.1 APRIORI EXPECTATION

TABLE 4.3: APRIORI EXPECTATIONS

VARIABLE APRIORI OUTPUT REMARKS

SIGNS
EXPECTATIONS

BTC Positive Positive(+) Conform

Conform
VBTC Negative Negative(-)

INFL Positive Positive (+) Conform

Conform
INTR Negative Negative(-)

TOP Negative Negative (-) Does not conform

COEFFICIENT INTERPRETATION

 BITCOIN PRICE (BTC)

The coefficient of BTC is 68.7 which show that, all other variables in the model held

constant, a percentage increase in bitcoin price leads to 69 unit increase in Exchange Rate. This

shows a positive relationship between BTC and EXR

 VOLUIME OF BITCOIN TRADED (VBTC)

The coefficient of VBTC is -244.3 which shows that, all other variables in the model held

constant, a percentage increase in VBTC leads to about 244 decrease in exchange rate. This

shows a negative relationship between VBTC and EXR.

 INFLATION (INFL)

The coefficient of INFL is 2.8 which shows that, all other variables in the model held

constant, a percentage increase in the value of inflation leads to about 2.8 increase in exchange

rate. This shows a positive relationship between INFL and EXR.

 INTEREST RATE (INTR)

The coefficient of INTR is -11.9 which shows that, all other variables held constant in the

model, a unit increase in interest rate leads to a 11.9 decrease in exchange rate. This INTR shows

that have a negative relationship with EXR.

 TRADE OPENESS (TOP)

The coefficient of TOP is -11.7 which shows that, on the average, all other variables held

constant in the model, a unit increase in trade openness leads to about 11.7 decrease in exchange

rate. This shows that TOP has a positive relationship with EXR.

4.4 STATISTICAL CRITERIA:

The first order test results include:

T-test: the individual statistical significance of the variable in the result can be checked using

either the t-statistic (if t-statistic is > or = |2|) or the p-value (if it is < or = 0.05 level of

significance).

TABLE 4.4: T-TEST USING THE P-VALUES OF THE VARIABLES

VARIABLE P-VALUE DECISION AT 5% STATISTICAL

SIGNIFICANT

C 0.0398 Statistically significant

Statistically significant
BTC 0.0340
Statistically significant
VBTC 0.0978
Statistically significant
INFL 0.0018

INTR 0.0203 Statistically significant

TOP 0.8305 Not statistically significant

F-test: the overall significance of the model can be checked using the probability value of the F-

statistic which should be less than or equal to 0.05 level of significance. Looking at the

regression result above, the value is 0.0000, which shows that the model is statistically

significant.

The Coefficient of Regression (R 2): with the R-squared value being 0.994055, the result revealed

that the regressors have very high explanatory power. It shows that about 99.4% of the variations

in EXR is accounted for by the regressors. The model is a good fit for the data.
The Adjusted R2: this performs the same function as the R2 but penalizes based on the number

of variables included in the model. The adjusted R 2 value 0.993591 which differs a little from the

R2 value showing that the regression is parsimonious.

4.5 ECONOMETRIC CRITERIA:

Here, it is necessary to conduct the second order test, to know if the model fulfilled some

of the assumptions of OLS. The tests results include:

4.5.1 NORMALITY TEST

TABLE 4.5: THE NORMALITY TEST RESULT

JACQUE-BERA STATISTIC 4.463384

PROBABILITY VALUE 0.107347

The Hypothesis tested is

H0: The residuals are normally distributed.

H1: The residuals are not normally distributed.

Decision Rule

Reject H0 if the p-value of Jarque-Bera is less than or equal to 0.05 level of significance.

From the table above, the p-value of the Jarque-Bera statistic is 0.107347 which is greater than

0.05 level of significance. Therefore, we fail to reject H 0 and conclude that the residuals are

normally distributed.
4.5.2. AUTOCORRELATION TEST

Using the Breusch-Godfrey Serial Correlation LM test as shown below:

TABLE 4.6: THE BREUSCH-GODFREY AUTOCORRELATION TEST RESULT

Obs*R-squared 66.32133

Prob. Chi-Square(2) 0.0000

The Hypothesis tested is:

HO: There is no autocorrelation.

H1: There is autocorrelation.

Decision Rule

Reject the null hypothesis if the p-value of the observed chi-square is less than or equal

to 0.05 level of significance. The table above shows that the probability of the chi-square

distribution is 0.0000 which is less than 0.05 significance level. Therefore, we reject the null

hypothesis and conclude that there is presence of autocorrelation in the residual.

4.5.3: HETEROSCEDASTICITY TEST

TABLE 4.7: BREUSCH-PAGAN-GODFREY TEST RESULT

Obs*R-squared 22.03366
0.0005
Prob. Chi-Square(5)

The Hypothesis tested is

H0: There is no heteroscedasticity (homoscedastic)

H1: There is heteroscedasticity

Decision Rule

Reject the null hypothesis if the p-value of the chi-square is less than 0.05. The table

above shows that the probability of the chi-square distribution is 0.0005 which is less than 0.05

significance level. Therefore, we reject the null hypothesis and conclude that the residuals are

heteroscedastic.

4.5.4 MULTICOLLINEARITY TEST CORRELATION MATRIX

TABLE 4.8: CORRELATION MATRIX OF THE VARIABLES

EXR BTC VBTC INFL INTR TOP

EXR 1.0000 0.7780 0.6040 0.9710 -0.9894 0.9619
BTC 0.7780 1.0000 0.8990 0.6670 -0.7797 0.8564
VBTC 0.6040 0.8990 1.0000 0.4722 -0.6352 0.7332
INFL 0.9710 0.6670 0.4722 1.0000 -0.9455 0.8853
INTR -0.9894 -0.7797 -0.6352 -0.9455 1.0000 -0.9814
TOP 0.9619 0.8564 0.7332 0.8853 -0.9814 1.0000

Decision Rule

If the correlation coefficient is greater than 0.8, we conclude that there is presence of

multicollinearity. The table above shows the correlation matrix which depicts the relationship

among the various variables used for this analysis. Multicollinearity exists if the independent

variables are correlated with each other up to 0.80. From the table above, we can say that there is

presence of multicollinearity. The problem of multicollinearity can be solved by doing nothing.

4.6 EVALUATION OF RESEARCH HYPOTHESES

Here, the hypotheses stated in chapter one will be duly evaluated. They are:

H0: There is no significant relationship between cryptocurrency price and value of fiat money in
Nigeria.

H0: There is no significant relationship between volume of cryptocurrency trade and value of fiat
money in Nigeria.

H0: There is no significant causality between cryptocurrency price and value of fiat money in
Nigeria.

H0: There is no significant causality between cryptocurrency volume and value of fiat currency in
Nigeria.

HYPOTHESIS ONE

Using the p-value to check for statistical significance, we reject the null hypothesis if the pvalue
is less than or equal to 0.05 using the 5% level of significance. Since the p-value for BTC
(0.0340) is less than 0.5, we reject the null hypothesis and conclude that there is a significant
relationship cryptocurrency price and value of fiat money in Nigeria.

HYPOTHESIS TWO

Using the p-value to check for statistical significance, we reject the null hypothesis if the pvalue
is less than or equal to 0.05 using the 5% level of significance. Since the p-value for VBTC
(0.0978) is greater than 0.5, we fail to reject the null hypothesis and conclude that there is no
significant relationship volume of cryptocurrency trade and value of fiat money in Nigeria.

HYPOTHESIS THREE

Using the p-value to check for statistical significance, we reject the null hypothesis if the pvalue
is less than or equal to 0.05 using the 5% level of significance. Since the pvalue from the granger
causality test was greater than 0.5, we fail to reject the null hypothesis and conclude that there is
no significant causality between cryptocurrency price and value of fiat money in Nigeria.

HYPOTHESIS FOUR

Using the p-value to check for statistical significance, we reject the null hypothesis if the pvalue
is less than or equal to 0.05 using the 5% level of significance. Since the pvalue from the granger
causality test was greater than 0.5, we fail to reject the null hypothesis and conclude that there is
no significant causality between cryptocurrency volume and value of fiat currency in Nigeria.

CHAPTER FIVE
 SUMMARY, POLICY RECOMMENDATIONS AND CONCLUSION

5.1 SUMMARY

The aim of this research is to investigate the impact of cryptocurrency on traditional fiat
money in Nigeria. To achieve the aim of this research work, the study employed quarterly data
from January, 2018 to December, 2023 from Central Bank of Nigeria (CBN) Statistical Bulletin,
World Development Indicators and Yahoo Finance.

The study adopted Augmented Dickey Fuller unit root test to check for stationarity and

found that all the variables are stationary at first difference which further led us to use Engle-

Granger Two Step Procedure for cointegration test. The study found the existence of a long run

relationship between the variables. The study employed Ordinary Least Square (OLS) technique

for empirical analysis. The model did not meet the no heteroscedasticity and no autocorrelation

assumptions which led to an estimation of a Newey West Heteroscedasticity and Autocorrelation

Corrected Standard Errors.

The result of the estimation showed that while cryptocurrency trading has a significant

effect on value of fiat money in Nigeria. Furher analysis showed that while cryptocurrency price

showed a positive relationship with fiat money, the volume of cryptocurrency traded was shown

to have a negative relationship with the value of fiat money. This is because high trading

volumes decrease liquidity in foreign exchange market, making it more difficult for buyers and

sellers to transact, and causing the exchange rate to depreciate hence the positive relationship

with volume of cryptocurrency traded. However, if cryptocurrency price increases, Nigerian

traders holding cryptocurrency may want to exchange their cryptocurrency assets for fiat money

in order to realize profits hence the negative relationship with value of cryptocurrency price. The

Granger Causality Test was also employed to test for the existence of causal relationship
between cryptocurrency trading in Nigeria. The result showed that there was no causal

relationship between the variables.

5.2 POLICY RECOMMENDATIONS

1. Policy makers should closely monitor cryptocurrency price to anticipate potential changes in

exchange rate.

2. Investors should diversify their portfolio too hedge against the risk from both the

cryptocurrency and foreign exchange market.

3.Policy makers should make concerted efforts to regulate the cryptocurrency market as it has

been shown to affect other financial markets too.

4. Regulators should also implement regular market surveillance to detect potential risks and take

proactive measures to mitigate them

5.3 CONCLUSION

The study adopted the Ordinary Least Square (OLS) technique to empirically investigate the

the impact of cryptocurrency on traditional fiat money in Nigeria . The results of the study revealed

that cryptocurrency has a strong significant effect on the value of traditional fiat money in

Nigeria. The study also discovered that the price of cryptocurrency has a positive relationship

with the value of traditional fiat money while the volume of cryptocurrency has a negative

relationship. This was attributed to the role of fiat money in cryptocurrency transaction. The

study concluded that cryptocurrency stands as an alternate to fiat money, however, it is not a

perfect substitute as it is not generally acceptable and fiat money are still employed in making
cryptocurrency trading successful. The study recommended that the government should regulate

and conduct cryptocurrency market surveillance to detect potential risk. The study also

recommended that investors should diversify their portfolio to other markets to hedge against the

risk in both the cryptocurrency and the foreign exchange market.