Studying the patterns and long‐run dynamics in cryptocurrency prices

• Date: 01 July 2020
• Author: Mathew Abraham
• Published date: 01 July 2020
• DOI: http://doi.org/10.1002/jcaf.22427

BLIND PEER REVIEW

Studying the patterns and long-run dynamics in

cryptocurrency prices

Mathew Abraham

Accounting and Finance, Whitireia

Community Polytechnic Faculty of

Business, New Zealand

Correspondence

Mathew Abraham, Accounting and

Finance, Whitireia Community

Polytechnic Faculty of Business,

New Zealand.

Email: matthew.abraham@whitireia.ac.nz

Abstract

This study analyses the price movements of a select sample of cryptocurrencies

and examines whether they are cointegrated and predictable using machine

learning algorithm and Johansen Test. The study used daily historical trading

data of 76 cryptocurrencies sourced from different cryptocurrency exchanges.

A sub-sample of six cryptocurrencies were chosen for the cointegration and

machine learning analysis based on their market share, attractiveness to the

investors and availability of data for the full sample period. The data records

starting from April 29, 2013 to February 7, 2019 were considered for the study.

An error correction model was estimated to investigate both the long-run and

short-run dynamics between the cryptocurrency prices. The evidence from the

error correction model estimates shows that there is a long-run association

between the prices of crypto currencies. The machine learning algorithm

involving neural networks (multilayer perception) was used to comprehend

the data patterns in the cryptocurrency price series, and the results show that

the model fits well in identifying and predicting the data patterns. The study

also examines the possible value drivers of cryptocurrencies by estimating a

linear regression with a set of covariates, which include the cryptocurrency

demand and supply interaction variables and financial variables such as the

NZX/S&P 50 index and exchange rates. The linear model estimates confirm

that cryptocurrency market fundamentals have an important impact on cryp-

tocurrency prices; however, they do not support the prediction that financial

fundamentals are the major value drivers of cryptocurrencies.

1|INTRODUCTION

Cryptocurrency has received much attention recently,

and the first cryptocurrency, Bitcoin, introduced in 2009,

represents over 81% of the total market of

cryptocurrencies (Chan, Chu, Nadarajah, & Osterrieder,

2017). With the UK government considering paying out

research grants in Bitcoin and an increasing number of

IT companies stockpiling Bitcoin to defend against

ransomware, the interest in cryptocurrencies has

acquired a global dimension. Many Chinese investors

now see Bitcoin as an investment opportunity, and the

US Federal Reserve is currently investing its resources in

the study of block chain and distributed ledger technolo-

gies in the finance industry.

Cryptocurrencies are “minted”through “mining”in

which computer network participants (users who provide

their computing power) record payments into a public

ledger (block chain), and in return they receive transac-

tion fees and newly minted cryptocurrencies

(e.g. Bitcoin). Cryptocurrency is based on cryptography,

which provides many advantages over traditional

Received: 26 September 2019 Accepted: 26 October 2019

DOI: 10.1002/jcaf.22427

98 © 2019 Wiley Periodicals, Inc. J Corp Acct Fin. 2020;31:98–113.wileyonlinelibrary.com/journal/jcaf

payment methods (such as Visa and MasterCard) includ-

ing high liquidity, lower transaction costs, and anonym-

ity. Cryptography is used to facilitate and record

transactions in cryptocurrencies on a set of electronic led-

gers (databases of financial accounts), and

cryptocurrencies have no tangible existence since they

are merely electronic records that keep track of transac-

tions. The two main advantages of cryptography are to

ensure that only an appropriate “address holder”can

“spend”the funds attributed to an address, and people

cannot fraudulently tamper with their cryptocurrency

balances. Since 2009, numerous cryptocurrencies have

been developed, and in November 2018, there were 2503

cryptocurrencies in existence (Coinrivet.Com, 2019).

Cryptocurrencies have certain unique characteristics

that separate them from conventional currencies issued

by Central Banks. First, cryptocurrencies are fiat curren-

cies and they are intrinsically worthless since they do not

have an underlying value based on consumption. They

are still in the process of establishing their market share

by building credibility among potential users as a

medium of exchange and an investment option. Second,

being digital currencies, cryptocurrencies are prone to

cyber-attacks, which can destabilize the cryptocurrency

systems triggering volatile price responses. For instance,

in 2014 there was a cyber-attack on MtGox (Bitcoins

Exchange) leading to its collapse and loss of 850,000

Bitcoins. Third, Lee (2018) found that changes of positive

and negative news in media about cryptocurrencies gen-

erated high price cycles due to attention-driven behavior

of cryptocurrency investors. Potential investors may pre-

fer investment choices under the attention of news media

because they reduce search costs and trigger high price

responses. This may be the reason why most of the

cryptocurrencies exhibit high volatility in the market.

New Zealand is yet to develop a defined regulatory

environment for cryptocurrencies, and the Financial

Market Authority (FMA) has so far released a few “com-

ments”about cryptocurrencies. The Reserve Bank of

New Zealand (RBNZ) released an “analytical notes”

about cryptocurrencies in 2017, and for tax purposes,

IRD treats cryptocurrency as property, not currency

meaning the foreign currency gain or loss provisions do

not apply.

1

In 2017, the Australian Government declared

that cryptocurrencies were legal and Bitcoin (and

cryptocurrencies that shared its characteristics) should be

treated as property subject to Capital Gains Tax (CGT).

In 2018, the Australian Transaction Reports and Analysis

Centre (AUSTRAC) implemented new crypto regulations

requiring exchanges operating in Australia to register

with AUSTRAC, identify and verify users, maintain

records, and comply with reporting regulations.

Unregistered exchanges would face criminal charges and

financial penalties (Reserve Bank of Australia, 2014).

The main motivations for undertaking the present

study are as follows. First, although several studies have

examined trending in cryptocurrencies, very few studies

have used the machine learning algorithm for forecasting

trends in cryptocurrencies. The present study used the

machine learning algorithm (neural network approach)

to examine the data patterns in cryptocurrency prices.

The neural network framework was executed through a

multilayer perceptron (MLP) method. Unlike the previ-

ous studies (Abdalkafor, 2018; Moller, 2018), which

employed small samples, the present study employed a

large sample of 76 cryptocurrencies to examine the trad-

ing data patterns. Second, after dealing with stationarity

and cointegration, the study analyzed both the short-

run and long-run dynamics in the price behavior of

cryptocurrencies. Third, most of the studies on

cryptocurrencies have so far dealt with the issuance and

trading and there is a dearth of research about consider-

ing cryptocurrency as a serious alternative investment

option. The present study is an attempt to examine this

aspect.

The results show that the cryptocurrency prices are

cointegrated, and they have a long-run relationship. The

machine learning solution indicates that it is possible to

identify and predict the cryptocurrency price patterns.

The OLS estimates show that the main value drivers of

cryptocurrency prices are trading volume, market capital-

ization, block size, generated coins, and block count.

However, the variables of exchange rates (AUS$/US$, NZ

$/US$), the stock market index (NZX/S&P 50), and the

consumer price index do not show much influence in

determining the values of cryptocurrencies as predicted.

The remainder of the article is organized as follows.

Section 2 discusses the main objectives of the study,

which is followed by a brief review of the relevant litera-

ture and hypothesis development in Section 3. Section 4

describes the data, the research methodology, and the

sampling frame. Section 5 discusses the empirical results,

and finally Section 6 concludes the study.

2|OBJECTIVES OF THE STUDY

The study aims to examine the price movements of differ-

ent cryptocurrencies and analyze both the short-run and

the long-run dynamics and trending of cryptocurrency

prices using Johansen cointegrating test and machine

learning algorithm (neural-network approach). In this

framework, the study seeks to answer the following

research questions.

ABRAHAM 99