Market Volatility and Efficiency Within and Across Cryptocurrency Composite Indexes.pdf

(2511 KB) Pobierz
Market volatility and efficiency within and across
cryptocurrency composite indexes
Dimitrios Koutsoupakis
∗
Department of Economics, National & Kapodistrian University of Athens
†
26
th
August, 2020
Abstract
The study of peer groups as Initial Offerings of diverse cryptocurrencies
continue to expand is important for matching investors’ tastes and pref-
erences in relation to volatility. This paper poses the question whether
stylized facts traditionally found in daily returns differ between individual
cryptocurrencies and comparable indexes. Against this background, we use
daily data frequency of 57 cryptocurrencies throughout their entire trading
history and construct 7 benchmark indexes that track market cap by crypto-
asset class aiming to draw inferences about the presence of effects namely
symmetric time-varying, risk premium, leverage and calendar. The findings
are useful as guidance for examining level of risk more evenly by employing
the framework of composite crypto-indexes.
Keywords
Cryptocurrencies, volatility clustering, risk, alternative investments
JEL Classification
C22, G14, G15
Declaration of Conflicting Interest: The Author declares that there is no conflict of
interest.
∗
†
Email: dimitrios.koutsoupakis@econ.uoa.gr
Address: 1, Sofokleous, 10559 Athens, Greece.
1
Electronic copy available at: https://ssrn.com/abstract=3700285
1
Introduction
The continuous advancement in financial technology has transformed financial markets
at large to exchange values at ways not previously available and, in turn enhance the
potential to revisit core values of industrialized societies including capital formation
(Wales, 2015). Such ways have been Initial Offerings of cryptocurrencies available via
the internet. It is essential for investment professionals, regulators and the general pub-
lic to fully grasp the risk implication trading a particular crypto-asset class for that all
such cryptocurrencies may not actually be alike. The increased demand, therefore, for
insights for understanding the fragmented alternative investments market of cryptocur-
rencies due to the complexity of the plethora cryptocurrencies regarding their diverse
supply schedule design has led to a recent expansion of the literature. In traditional fi-
nancial markets, peer-groups constituted to better monitor trading performance among
homogeneous asset-classes, thus within and across benchmark indexes. As more and
more cryptocurrencies are issued, composite indexes formed by closely comparable (in
terms of the monetary arrangement followed) cryptocurrencies is a new and increas-
ingly accepted means of evaluating risk exposure in the related literature. This paper is
motivated by this new research agenda.
While many empirical studies have been conducted to delve into various aspects of
risk-return relationship of cryptcurrencies (Gandal et al., 2018), this topic as far from
being exhausted as a research area. Lately, it has been identified the need to expand
literature by monitoring benchmarks and indexes (Trimborn & H¨rdle, 2018) so as to
a
allocate cryptocurrencies on the basis of similarities in their respective economy protocols
into market segments similar to equity or commodity indexes in financial markets.
The paper contributes to an evaluation of the historical performance of cryptocur-
rencies conditional on composite indexes that track the market cap of their respective
crypto-asset class. We use daily data frequency of 57 cryptocurrencies throughout their
entire trading history. Cryptocurrencies allocated to peer-groups with common charac-
teristics from the supply-side (how they are constructed based on their economy protocol)
should verify that cryptocurrencies are not alike while showing evidence of competition
within these sub-markets. This methodology highlights the need to choose indexes that
represent this alternative market well to enable the general public being informed about
risk diversity among cryptocurrencies.
Closer look at the definitions and properties of cryptocurrencies allows us to identify
2
Electronic copy available at: https://ssrn.com/abstract=3700285
important differences. This can be useful for investors with different tastes in their
allocation decision among assets and indexes. In this work, we examine the dominant
market player (Bitcoin) as a distinct asset class compared to other cryptocurrency asset
classes as we draw a sample in order to construct different indexes and market cap
portfolios. These categories would be the next:
1.
Bitcoin,
the market pioneer and leader stands for a stand-alone index.
2.
Altcoins,
thus alternative to the Bitcoin blueprint for that they are openly re-
garded as clones only to feature different parameter values in the protocol i.e. dif-
ferent block time for clearing transactions, supply function, and issuance scheme.
3.
Altchains,
thus alternative type of the Bitcoin blockchain blueprint. These assets
create more ecosystems for that their main role is becoming a platform for the
development and execution of various smart contract (called dApps) on top of
their network.
4.
Algorithmic Stable
follow crawling pegs exchange rate arrangements as they
fluctuate over accepted small bands. Their goal it to peg the decentralized cryp-
tocurrency (called Algorithmic Stable) to another anchor (usually traditional cur-
rency at parity) through stability mechanisms governed by holders of a centralized
cryptocurrency (called Smart Token, see below for more). Their reserves are ei-
ther over-collaterized (allowing trading on margin via Collaterized Debt Position
derivatives) or non-collaterized.
5.
Stable Tokens
follow the currency board exchange rate arrangement and are
privately (by firms) issued following the acceptance of another anchor (traditional
currency or commodities) which is kept as collateral reserve.
6.
Smart Tokens,
are not related to products or services in the real economy and
either act as governors for algorithmic stable cryptocurrencies or simply as tokens
that allow the participation in a virtual applications.
7.
Utility Tokens
are privately issued (by firms) as pre-payments for the payment
of a particular product or service in the real economy (offered by the originator).
All said, Stablcoins (Algorithmic Stable and Stable Tokens) aim at fluctuating at parity
(1:1) with another traditional currency while all others independently float. As the
3
Electronic copy available at: https://ssrn.com/abstract=3700285
cryptocurrency matures in size and value the need to classify these assets in indexes that
follow common patterns alike traditional regulated (stock and foreign exchange) markets
will become greater. At the same time, evidence of market anomalies in cryptocurencies
and constructed indexes hint at possible return predictability which is inconsistent with
the efficient market hypothesis. The remaining paper is organized as follows. Section
2 reviews literature, section 3 cites preliminary statistics and explains the methodology
while sections 4 provide the estimation results. At the end, section 5 discusses and
extends while section 6 concludes.
2
Literature
The fundamental concepts of efficient market hypothesis were first systematically studied
by Samuelson (1965), then formulated to their current framework that distinguishes
between
weak-form efficiency, semi-weak form efficiency
and
strong form efficiency
by
Malkiel & Fama (1970). The market efficiency hypothesis has been refined over the years
by Fama (1991) who methodologically relates the three forms with
predictability tests,
event studies
and
private information test
respectively.
From the empirical investigation standpoint of view, the Efficient Market Hypothe-
sis is greatly evaluated under two frameworks. The first examines the assets’ time-series
behavior over time either in relation to the announcement of information or stand-
alone, thus looking at their own trading path. The goal is to identify patterns of non-
randomness in returns. In the cryptocurrency market, there are no announcement of
information for that supply is usually pre-determined or fixed while these assets are not
associated with the future economic benefits such as receipt of dividends and participa-
tion in the corporate governance of an entity.
The second strives to identify empirical (market) anomalies that violate the bold
assumption that no-one can beat the market. On the research question of efficient market
hypothesis in cryptocurrency markets, the findings of current literature are inconclusive
on account of different modeling techniques and periods selected. Caporale et al. (2018)
identify persistence (positive correlation between past and future values) of changing
degree over time that can allow trading strategies to gain abnormal profits. Kaiser (2019)
examines seasonality effects in key trading variables (volume, spread, volatility) for ten
cryptocurrencies to conclude that the weak-form market efficiency cannot be rejected.
4
Electronic copy available at: https://ssrn.com/abstract=3700285
An event study for macroeconomics news is found in Al-Khazali et al. (2018) while the
size-effect is explored in Shen et al. (2019) who use a three-factor pricing model. In a
first attempt to investigate cryptocurrencies in peer-groups, Trimborn & H¨rdle (2018)
a
highlight the difficult of constructing index constituents to represent market segments
while they built an active management portfolio able to reduce tracking errors.
In this paper, we take a broader perspective to examine the cryptocurrency market
as a whole. Consequently, we test the validity of stylized facts with regards to volatility
and efficiency in daily returns of cryptocurrencies as individual assets and as part of
composite indexes that track the market cap of their respective crypto-asset class.
3
Preliminaries
This section is meant to provide a concise overview of the methodology employed and
thee main statistical features of the assets in question prior to the main analysis by
reporting graphs and figures.
3.1
Data & descriptive statistics
The asset price of cryptocurrencies, that is their exchange rate in terms of another
currency either traditional (such as the US dollar or the Euro) or alternative (other
cryptocurrencies) is determined in numerous private digital exchanges, thus not in a
single exchange similar to equities or commodities. It can be, therefore, said that from
an institutional design perspective the cryptocurrency alternative market is attached to
the traditional foreign exchange market which is “in operation twenty-four hours a day,
seven days a week, and is the closest analogue to the concept of a continuous time global
marketplace” (Bollerslev & Domowitz, 1993).
The data used are for the period from April 28, 2013 until July 29, 2019, thus
account for almost the entire cryptocurrency market trading history which had started
a few three years ago, yet with considerably low volume and capitalization.
1
In the next
figure, the lines display the market capitalization (market cap indexes at levels) evolution
for the selected sample of 57 cryptocurrencies segregated into groups. The market cap
indexes is the summation of individual cryptocurrency market cap corresponding to each
1
The data were obtained from coinmarketcap.com
5
Electronic copy available at: https://ssrn.com/abstract=3700285

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin