GROUP 4
How do U.S. Inflation rates affect Bitcoin
Price?
Brandon Farrel 13301910025
Ferdyanto
John Arnold Matthew 133
Matthew Jeremiah Natawidjaja
TABLE OF CONTENTS
1.1 INTRODUCTION
1.1.1 Bitcoin
1.1.2 Inflation
1.1.3 Inflation Hedging
1.2 Research Question
1.3 Research Objective
2. LITERATURE REVIEW
3. METHODOLOGY
4. RESULTS
5. CONCLUSION
6. RECOMMENDATION
CHAPTER 1
INTRODUCTION
1.1 Introduction
Cryptocurrency refers to digital tokens that enable people to make payments directly to each other
through an online system (Reserve Bank of Australia, 2022). Specifically, cryptocurrency refers to a type
of digital asset that uses distributed ledger, or blockchain technology to enable a secure transaction
(Hardle et al., 2019). The rise of cryptocurrency in the last decade poses a serious threat to many
traditional functions in finance as cryptocurrency embraces peer-to-peer mechanisms which effectively
eliminate the “middle man”. Cryptocurrency could potentially tap people who aren’t able to have bank
accounts otherwise known as the “unbanked”. With a smartphone and the internet, the potential exists for
a revolution in financial inclusion given that over two billion people are unbanked. (GlobalFindex, 2017;
World Bank, 2017). The growth of cryptocurrency technology, therefore, poses a challenge to traditional
monetary authorities and central banks with currently over 2000 such coins and tokens and the volume is
still growing every day. Central banks realized this issue and have initiated their own national
cryptocurrency initiatives (Bech and Garratt, 2017).
1.1.1 Bitcoin
Bitcoin can be considered the father of all cryptocurrencies as it can be considered the first and
this research is still the Cryptocurrency with the highest market capitulation. Bitcoin is an online
communication method that makes it easier to use virtual currencies and make payments online. The rules
governing Bitcoin were developed by engineers without any apparent input from lawyers or regulators.
(Böhme et al, 2015). By acting as money and a means of payment independent of any one person, group,
or entity, cryptocurrencies like Bitcoin eliminate the need for third parties to get involved in financial
transactions. It is available for purchase on several exchanges and is given to blockchain miners as
compensation for their efforts in verifying transactions.
Since then, it has grown to be the most well-known cryptocurrency worldwide. Numerous other
cryptocurrencies have been developed as a result of their popularity. These rivals either try to displace it
as a means of payment or are employed in other blockchains and cutting-edge financial technologies as
utility or security tokens.
If you've been following the cryptocurrency market, you've probably observed that as Bitcoin's
price falls, other cryptocurrency prices (often referred to as altcoins) follow. The opposite is also true:
when the price of bitcoin rises, we expect altcoins to follow suit. This is as Bitcoin has the largest market
capitulation. If we draw a connection with stock markets, the IHSG (Index Saham Gabungan) will crash
just because the stock price of Bank BCA falls significantly, therefore creating negative sentiments,
leading investors to pull their investment from the market as a whole.
1.1.2 Inflation
Inflation refers to the loss of a currency's relative buying power over time. The increase in the
average price level of a basket of certain goods and services over time in an economy would provide a
quantitative measure of the rate at which the decrease in buying power occurs. The increase in pricing,
frequently stated as a percentage, implies that a unit of money buys less than it did previously. The loss of
buying power affects the general cost of living for the general public, resulting in a slowing of economic
growth. Mainly there are 3 types; Demand-pull, Cost-push, and Built-in Inflation.
1.1.3 Inflation Hedging
Over time, inflation has caused people’s income to gradually lose real value to commodities that
said income is used to purchase. Hence, people have always searched for ways to lessen that devaluation
through inflation hedging. Inflation hedges are an investment that is perceived as preserving the
diminished buying power of a currency as a result of its value being lost due to rising prices as a result of
inflation or macroeconomic growth. During inflationary cycles, the investments are expected to hold or
increase in value.
1.2 Research Question
How do the U.S. Inflation rates affect Bitcoin Price and hence Other Cryptocurrencies?
1.3 Research Objective
Our research looks to work out whether Bitcoin and other cryptocurrencies can be deemed as an
effective hedge against inflation and if so to what degree that is or isn’t so. This comes as there is a
continuous debate on which commodities or financial instruments can be used as a means to hedge
against inflation as recent studies have found gold, which has historically been a popular means of
inflation hedging, is not a perfect hedge against inflation. This is as its prices, which can vary greatly from
year to year due to a variety of factors, are also subject to inflation-adjusted returns. In actuality, realized
inflation has not been a major factor in the variation in the nominal and real returns of gold over the last 1,
5, 10, 15, and 20-year investment horizons (Harvey, Campbell R, and Claude B. Erb, "The Golden
Dilemma)
CHAPTER 2
LITERATURE REVIEW
There is various literature that analyzes the relationship between Bitcoin or another
cryptocurrency with inflation. Felez-Vinas et al. (2021) argue that Bitcoin is often referred to as “digital
gold” and like gold, it must have similar properties. Bitcoin has similar properties which are decentralized
production and low correlation with traditional assets (Dyhrberg, 2016; Corbet et al., 2018; Baur et al.,
2018; Bouri et al., 2020). Similar to Gold, Bitcoin offers no explicit yield and has limited usage as a
medium of directly exchanging goods and services. However, unlike gold, with a finite supply and lower
supply growth than gold, Bitcoin is seen as a more attractive alternative store of value that can’t be
devalued by the monetary policy decisions of central banks. These characteristics and limits to Bitcoin
supply provide a potential hedge against inflation and explain the increased adoption of Bitcoin in
countries with a history of hyperinflation such as Argentina and Venezuela (Kenneth and Haesly, 2016;
Moreno, 2016). Essentially, the characteristics and properties of Bitcoin which is similar to gold has led to
the increase in adoption of Bitcoin as it has gained relevance in society as an alternative long-term digital
store of value with similar anti-inflationary characteristics. In addition, Blau et al. (2021) and Choi and
Shin (2021) empirically demonstrate that Bitcoin returns are positively correlated with future inflation
expectations, which Conlon et al. (2021) show was a particularly acute issue during the Covid-19 crisis
period.
Choi and Shin (2021) notes that during the height of the Covid-19 pandemic, many investors
considered Bitcoin as an inflation hedge. Indeed Bitcoin like gold can hedge against inflation and should
be taken into consideration as an alternative to gold. However, unlike gold, Bitcoin isn't able to hedge
against uncertainty and macroeconomic shocks.
Matkovskyy and Jalan (2020) are eager to find out the hedging capabilities, effectiveness and
returns of Bitcoin assets against inflation using the quantile-on-quantile model. Gold has attracted
significant attention as a hedging tool after the 2008 crisis (Lucey, 2011). The results however are
somewhat contradictory and inconsistent. Conlon et al. (2018) show that gold can be used as an inflation
hedge, while others point out lack of correlation and weak link between gold and inflation (Baur, 2011;
Erb & Harvey, 2013; Batten,Ciner, Lucey, 2014; etc.). In recent years, as Bitcoin grew in popularity,
many have speculated the inflation hedging capabilities. However, results and empirical evidence are still
mixed (Baur et al., 2018; Guesmi et al., 2019). Some researchers found that Bitcoin has inflation-hedging
capabilities. Dyhrberg (2016) shows that the hedging properties of Bitcoin and gold are similar, Bouri et
al.(2017b) find Bitcoin to be a hedge against global uncertainty at short investment horizons and in bullish
regimes only, Sebastião& Godinho (2019) document hedging-efficacy of CBOE bitcoin futures for the
bitcoin and other cryptocurrencies’ price risk, Chan, Le & Wu (2019) document Bitcoin hedging
properties against theEuro-Index, Shanghai A-Share, S&P 500, Nikkei, and the TSX, Bouri et al.(2017b)
find Bitcoin to be a hedge against global uncertainty at short investment horizons and in bullish regimes
only, and Urquhart& Zhang (2019) found that Bitcoin acts as an intraday hedge for certain currencies.
On the other hand, many researchers found that Bitcoin has little power and capabilities to hedge
against inflation. Bouri et al.(2017a) found limited evidence of Bitcoin as a hedging tool, Corbetet
al.(2018) found Bitcoin futures an ineffective hedge, and Wu et al.(2019) showed that neither gold, nor
Bitcoin can serve as a strong hedge against economic policy uncertainty.
Based on their research, they found six key results. Firstly, the Bitcoin return-inflation
relationship is asymmetric for different market states and levels of inflation, Bitcoin return-realized
inflation relationship depends upon the magnitude of inflation shocks and Bitcoin market states (bullish
vs bearish), the Bitcoin return-unexpected inflation relationship depends more on the magnitude of
inflation shocks than Bitcoin investor sentiment, the Bitcoin can be considered a macro hedge against
realized inflation in bullish euro, GBP and JPY markets, the GBP and JPY Bitcoin offer significant ‘safe
haven’ characteristics by offering very high returns during periods of very high inflation and the USD
Bitcoin performs worse with rising levels of inflation. Their empirical research supports the existing
evidence on the asymmetric return-inflation which points out the similarities between Bitcoin and energy
and industrial stocks. In addition, their results are consistent with the asymmetric return-inflation
relationship across different market states and different levels of inflation. This has been documented for
other assets such as equities (Kim and Ryoo, 2011), real estate (Simpson et al. 2007), and gold (Wang,
Lee, and Thi, 2011)
CHAPTER 3
METHODOLOGY
3.1 Methodology
This paper conducted research by applying multiple tests including Linear regression,
Breusch-Pagan Test, and Durbin-Watson Test. The linear regression model uses a straight line to illustrate
relationships between variables, by looking for the value of the regression coefficient(s) that minimizes
the overall model error and locating the line that matches the data. Based on the types of linear regression,
the research is done based on simple linear regression as the model only has one independent variable.
The Breusch-Pagan-Godfrey Test, commonly known as the Breusch-Pagan test, is a test used to detect
whether or not a regression model has heteroscedasticity. The term "heteroscedasticity" refers to a
different spread. The test is done by creating a null and alternate hypothesis, continued by accepting or
rejecting the hypothesis by comparing values of the p-value and the standard significance level of 0.05.
The null hypothesis will be rejected if the p-value is lower than the significance level and conclude with
the alternate hypothesis with the presence of heteroscedasticity. Lastly is the Durbin-Watson Test, which
tests for autocorrelation. Values of the test will always be between 0 and 4. A value of 2 would mean no
autocorrelation is found within the data, values between 0 and less than 2 represent a positive
autocorrelation, and values between 2 and 4 represent negative autocorrelation. The test also has the rule
of thumb to show that values ranging between 1.5 and 2.5 are relatively normal.
Hypothesis
Bitcoin is an effective hedge against inflation during bull market, and ineffective during
H0 bear market
H1 Bitcoin is an effective hedge against inflation during bull and bear market
Variables used within the paper are inflation rate in the United States as the independent variable
and Bitcoin price in USD as the dependent variable, the data is taken monthly from May 2019 to May
2022. The Bitcoin price for the data is taken from Bitstamp based on their monthly closing price.
Whereas, inflation rate of the United States is taken from Bitstamp. Inflation rate refers to the change in
price level's value and expressed in form of percentage. Usually, inflation rate is measured from the value
of CPI (Consumer Price Index) using the equation:
Percent Inflation Rate = (Final CPI Index Value/Initial CPI Value) x 100
The research is done by conducting multiple tests based on different classifications of data
applying the 3 types of tests with the regression model, Bitcoin Price = b0 + b1*U.S Inflation Rate. The
same data of inflation rate is tested onto 3 types of market trends in bitcoin prices; full, bear, and bull.
Full refers to the whole data of bitcoin prices that combines both bear and bull, bear or bearish refers to
the declining trend of prices, and bull or bullish is when the prices of the market experiences an increase
in value. The multiple tests are done based on the findings of Matkovskyy and Jalan in 2020 that mentions
Bitcoin to be an effective hedge against inflation during bullish trends of the Bitcoin price.
CHAPTER 4
RESULTS
Figure 4.1 Bitcoin VS U.S Inflation Chart
According to the graph above, as the inflation rate remains stable under the 2.5% area, Bitcoin
price consolidates and then turns bullish starting from October 2020. Then, Bitcoin then peaked at
$58,782.58 in March 2021, with an inflation rate of 2.6%. However, a sudden hike in the inflation rate happened in
April, almost doubling at 4.2%. This is then followed by a drop in Bitcoin price to $35,037.23 in June 2021. Then
the inflation rate stables at around 5-6%, which lead the Bitcoin price to another peak at $61,359.44 in October
2021. The inflation rate then rallies above the 6.5% level in November 2021, which is then followed by declining
Bitcoin prices over the next months, entering the bear market.
4.1 Full
Figure 4.1.1 Regression
Based on the regression results, the estimated regression line equation is:
Bitcoin Price = 9566 + 493650*U.S Inflation
Looking at the y-intercept (b0), Bitcoin Price will be $9566 if the inflation rate is 0. Furthermore,
looking at the slope (b1), Bitcoin Price will increase by $493650 if the inflation rate increases by
100% in other words, Bitcoin Price will increase by $4936.50 if the inflation rate rises by 1%.
This means that Bitcoin and the Inflation rate have a positive relationship. Therefore:
Bitcoin Price = 9566 + 493650*1%
Looking at the Adjusted R Squared, the value is 0.487. This means that 48.7% of the variability is
successfully explained by the model. Moreover, the F-Statistics is 35.17 between 1 and 35 Degrees
of Freedom, which means the variable U.S Inflation is good to predict the outcome of Bitcoin Price. In
Addition, the p-value from the model is 0.0000009527, which means that the parameters are
significant to the model since it is less than 5%.
Figure 4.1.2 Breusch-Pagan Test
Since the B-P Test shows a p-value of more than 5% (77.21%), the null hypothesis is not rejected.
This indicates that there is not enough evidence that heteroscedasticity exists. Thus, there is
homoscedasticity.
Figure 4.1.3 Durbin Watson Test
Looking at the D-W Test, the results show a value that is less than 2 and the p-value is less than
5%. This means that we reject the null hypothesis so autocorrelation exists.
4.2 Bull
Figure 4.2.1 Regression
Based on the regression results, the estimated regression line equation is:
Bitcoin Price = 12738 + 474486*U.S Inflation
Looking at the y-intercept (b0), Bitcoin Price will be $12738 if the inflation rate is 0.
Furthermore, looking at the slope (b1), Bitcoin Price will increase by $474486 if the inflation rate
increases by 100% in other words, Bitcoin Price will increase by $4744.86 if the inflation rate
rises by 1%. Therefore:
Bitcoin Price = 12738 + 474486*1%
Looking at the Adjusted R Squared, the value is 0.4283. This means that 42.83% of the variability
is successfully explained by the model. Moreover, the F-Statistics is 15.24 between 1 and 18
Degrees of Freedom, which means the variable U.S Inflation is good to predict the outcome of Bitcoin
Price. In Addition, the p-value from the model is 0.001042, which means that the parameters are
significant to the model since it is less than 5%.
Figure 4.1.2 Breusch-Pagan Test
Since the B-P Test shows a p-value of more than 5% (91.49%), the null hypothesis is not rejected.
This indicates that there is not enough evidence that heteroscedasticity exists. Thus, there is
homoscedasticity.
Figure 4.1.3 Durbin Watson Test
Looking at the D-W Test, the results show a value that is less than 2 and the p-value is less than
5%. This means that we reject the null hypothesis so autocorrelation exists.
4.3 Bear
Figure 4.3.1 Regression
Based on the regression results, the estimated regression line equation is:
Bitcoin Price = 2518 + 588732*U.S Inflation
Looking at the y-intercept (b0), Bitcoin Price will be $2518 if the inflation rate is 0. Furthermore,
looking at the slope (b1), Bitcoin Price will increase by $588732 if the inflation rate increases by
100% in other words, Bitcoin Price will increase by $5887.32 if the inflation rate rises by 1%.
Therefore:
Bitcoin Price = 2518 + 588732*1%
Looking at the Adjusted R Squared, the value is 0.5959. This means that 59.59% of the variability
is successfully explained by the model. Moreover, the F-Statistics is 24.59 between 1 and 15
Degrees of Freedom, which means the variable U.S Inflation is good to predict the outcome of Bitcoin
Price. In Addition, the p-value from the model is 0.0001715, which means that the parameters are
significant to the model since it is less than 5%.
Figure 4.1.2 Breusch-Pagan Test
Since the B-P Test shows a p-value of more than 5% (114.78%), the null hypothesis is not
rejected. This indicates that there is not enough evidence that heteroscedasticity exists. Thus,
there is homoscedasticity.
Figure 4.1.3 Durbin Watson Test
Looking at the D-W Test, the results show a value that is less than 2 and the p-value is less than
5%. This means that we reject the null hypothesis so autocorrelation exists.
CHAPTER 5
DISCUSSIONS
Full Bull Bear
Regression Significant Significant Significant
Heteroscedasticity Test Homoscedasticity Homoscedasticity Homoscedasticity
Autocorrelation Autocorrelated Autocorrelated Autocorrelated
Table 5.1 Test Results
Based on the regression result, we can say that inflation has a positive relationship with Bitcoin
Price regardless of the trend cycle. This means that Bitcoin is an effective hedge against inflation.
Looking at the P-Value, all of the regression results are significant. This means we reject h0. Furthermore,
F statistics show that using the inflation rate as the variable is beneficial to predict the outcome.
Looking at the BP Test for all trends, the p-value is more than 0.05. This means that we fail to
reject h0, which means that there is not enough evidence that heteroscedasticity exists. However, looking
at the DW Test for all trends, the p-value is less than 0.05, which rejects the null hypothesis so it indicates
that autocorrelation exists.
The analysis shows that we reject the null hypothesis:
Bitcoin is an effective hedge against inflation during the bull market, and ineffective
H0 during the bear market
H1 Bitcoin is an effective hedge against inflation during bull and bear market
We reject the null hypothesis because our analysis shows that Bitcoin Price will increase as
inflation increases regardless of the trend, which makes Bitcoin an effective hedge against inflation.
CHAPTER 6
CONCLUSIONS
6.1 Conclusion
In conclusion, the null hypothesis is rejected because Bitcoin Price goes up as the U.S Inflation
rate goes up regardless of the trend cycle. Our analysis shows that a 1% increase in inflation will increase
the Bitcoin Price by $4936.50, $4744.86, and $5887.32 by Full trend (Bull+Bear), bull trend, and bear
trend respectively, and these results are significant. The variable U.S Inflation is beneficial for the model
due to its high F-Value in all trend cycles. There is no heteroscedasticity, however, autocorrelation exists
for all trend cycles.
6.1 Recommendations
Since there is autocorrelation found in our analysis, measures to fix this are needed. One way to
fix autocorrelation is by maintaining the model specification as-is while enlarging the confidence intervals
around the regression coefficients to accommodate for the model's non-autocorrelated error assumption
being broken. Heteroscedasticity and autocorrelation (HAC) robust standard errors, such as those
developed by Newey and West(1987), can be used for this.
Furthermore, Bitcoin prices are also relative to market cycles and other economic variables such
as interest rates. According to research done by Eli Noam (2021), interest rate clearly has an effect on
Bitcoin price. So, to improve the model, adding these variables to the model would be beneficial.
Additionally, this research could be improved by comparing the effectiveness of each inflation
hedging instrument to one another. Eg. The returns against inflation of Gold investment to Bitcoin
investment.
REFERENCES
Chen, J. (2022, January 4). What is an inflation hedge? Investopedia. Retrieved July 11, 2022, from
https://www.investopedia.com/terms/i/inflation-hedge.asp
Fernando, J. (2022, June 28). Inflation. Investopedia. Retrieved July 11, 2022, from
https://www.investopedia.com/terms/i/inflation.asp
Kenton, W. (2021, August 4). Understanding the durbin watson statistic. Investopedia. Retrieved
July 11, 2022, from https://www.investopedia.com/terms/d/durbin-watson-statistic.asp
Poras, E. M. (2018, July 18). R linear regression tutorial: LM function in R with code examples.
DataCamp. Retrieved July 11, 2022, from https://www.datacamp.com/tutorial/linear-regression-R
Zach. (2020, December 31). The breusch-pagan test: Definition & Example. Statology. Retrieved
July 11, 2022, from https://www.statology.org/breusch-pagan-test/
Choi, S., & Shin, J. (2021, August 16). Bitcoin: An inflation hedge but not a safe haven. Finance
Research Letters. Retrieved July 11, 2022, from
https://www.sciencedirect.com/science/article/pii/S1544612321003810
Dee, J., Li, L., & Zheng, Z. (2013, March). Is gold a hedge or safe haven? evidence from inflation
and stock market. Retrieved July 11, 2022, from
https://www.researchgate.net/publication/302931124_Is_gold_a_hedge_or_safe_Haven_Evidence_
from_inflation_and_stock_market
Härdle, W. K., Harvey, C. R., & Reule, R. C. G. (2019, August 3). Understanding cryptocurrencies
. Retrieved July 10, 2022, from https://ies.keio.ac.jp/upload/20191125econo_Wolfbang_wp.pdf
Newey, Whitney K; West, Kenneth D (1987). "A Simple, Positive Semi-definite, Heteroskedasticity
and Autocorrelation Consistent Covariance Matrix". Econometrica 55 (3): 703–708.
doi:10.2307/1913610
Reserve Bank of Australia. (n.d.). Cryptocurrencies. Retrieved July 11, 2022, from
https://www.rba.gov.au/education/resources/explainers/pdf/cryptocurrencies.pdf
