Thus motivated, this study will delve into and elucidate the abnormal pricing patterns of DeFi assets and NFTs by investigating the occurrence of bubbles in specific strand times of NFT
i00 1 5
Scope of Content
Two DeFi assets, namely Chainlink [LINK] and Avalanche [AVAX], two NFTs: Decentraland [MANA] and Internet Computer [ICP], 11 other crypto assets including
The article discusses various digital assets, including Ox (ZRX), Ren (REN), Terra (LUNA), Synthetix (SNX), Fantom (FTM), Reserve Rights (RSR), THORChain (RUNE), Theta Network (THETA), Enjin Coin (ENJ), as well as major cryptocurrencies like Bitcoin (BTC) and Ethereum (ETH) It also highlights traditional financial assets, such as Brent oil (OIL), gold (GOLD), key US stock market indices like the S&P 500 (SPX) and Dow Jones (DJI), along with the global stock market index (MSC).
Expected research Contributions cceceessessesesseeeceeeseeecseescseeseeeesceseesseeessesseeesaeeeees 2
This research seeks to improve readers' comprehension of crypto assets within the DeFi and NFT markets by offering detailed insights drawn from numerous trustworthy sources Additionally, it presents recommendations for dependable information and data sources related to cryptocurrencies and financial assets, catering to diverse needs.
This study enhances the current literature by systematically analyzing price bubbles specifically in the NFT and DeFi markets, ultimately concluding that explosive bubble behaviors are present in both sectors.
This study expands on previous research that utilized various time-series econometric frameworks to assess price bubbles in financial markets Notably, it is one of the first empirical investigations to thoroughly identify price explosiveness in the DeFi and NFT markets, comparing these findings to other financial markets using ADF, SADF, and GSADF tests.
The study's findings are crucial for investors, academics, market regulators, and policymakers, as identifying price bubbles can act as an early warning signal This is particularly relevant in NFT and DeFi markets, where continuous trading is significantly affected by news and sudden events, as noted by Lucey et al (2022) and Wang et al (2022).
The thesis is organized into four chapters: Chapter 2 reviews prior research on price bubbles, NFTs, and DeFi markets, outlines models for bubble detection, identifies research gaps, and presents hypothesis development Chapter 3 details the econometric model and describes the data used for testing price bubbles Chapter 4 analyzes empirical results and highlights key findings, while Chapter 5 concludes with essential insights and their practical and policy implications.
THEORETICAL BASIS AND LITERATURE REVIEW
Overview Of NI ẽ ch Tho Tho TH TH HH TH nghệ 4 1 The concept of 1
Recent advancements in blockchain technology have led to the rise of Non-Fungible Tokens (NFTs), a new class of digital assets that have garnered significant attention and scrutiny, especially since early 2021.
NFTs, or Non-Fungible Tokens, are unique digital assets stored on a blockchain ledger, representing items like artwork, video game assets, or music Unlike traditional cryptocurrencies, NFTs are not interchangeable or divisible, as each token has its own unique identifier and metadata, ensuring their irreplaceable nature within the blockchain.
NFTs have emerged as a significant source of revenue in the cryptocurrency market, especially through NFT marketplace development projects Notably, in the initial eight months of 2022, NFT marketplace revenue represented 49% of the total market revenue As of September 2023, the market capitalization of NFTs is estimated to be around [insert current value].
$674 thousand (Coinmarketcap, 2023) Additionally, India leads the world in terms of
NFT holders, with 7% of the country's population involved in NFT ownership On average, about 2% of women worldwide claim ownership of NFTs, compared to 4% of men (Finder, 2022).
The NFT marketplace is projected to generate revenue of $1601 million in
2023, with an average revenue per user in the NFT marketplace reaching $114.80 for the year The revenue is expected to grow at a compound annual rate of 18.5% from
2023 to 2027, reaching an estimated total of $3162 million The number of NFT users is expected to reach 19.31 million by 2027, with an anticipated penetration rate of 0.2% (Statista, 2023)
NFTs are primarily found in the cryptocurrency market, showcasing their potential through successful marketplace sales Nevertheless, the market encounters significant risks, such as environmental concerns and the possibility of speculative bubbles, which could result in growth fluctuations in the future.
2.1.1.2 History of NTFs The creation of NFTs on the blockchain dates back to 2013 with Colored Coins. The concept was simple: each token represented land or different types of metal It is the first kind of NFT with use cases in the real world NFT experiments have multiplied over the years.
The first NFT sale, “Quantum,” was created by Kevin McKoy in 2014 on the Namecoin blockchain That same year, the Counterparty platform enabled the release of the Spell of Genesis game’s initial cards in 2015 In 2016, the trading card game Force of Will ranked fourth in sales in North America, following Magic: The Gathering, Pokemon, and Yu-Gi-Oh The NFT landscape expanded in 2017 with the airdrop of CryptoPunks on Ethereum and the development of Decentraland’s metaverse Additionally, Dapper Labs contributed to the establishment of the NFT standard on Ethereum through the popular CryptoKitties.
In mid-2018, the ERC-721 standard was established on Ethereum, enabling users to mint and utilize NFTs on the blockchain This development led to a surge of NFT projects, including metaverses like The Sandbox and Voxels, artistic marketplaces such as SuperRare and Known Origin, and popular video games like Gods Unchained and Axie Infinity Additionally, specialized NFT blockchains like WAX, Ronin, and Flow have emerged, further expanding the NFT landscape.
In 2021, numerous brands, licenses, and celebrities embraced NFTs, leading to the announcement of major projects and partnerships Notably, Nike acquired the virtual wearable studio RTFKT, while Coca-Cola established a presence in Decentraland.
2.1.1.3 Characteristics of NFTs Non-Fungible Tokens (NFTs) possess several distinctive characteristics that distinguish them from other types of digital assets and cryptocurrencies.
NFTs are unique digital assets that cannot be divided or replicated, ensuring each one is distinct This uniqueness is a core characteristic of NFTs, supported by the immutable nature of blockchain technology, which verifies ownership and grants exclusive rights to the holder Consequently, owning an NFT guarantees possession of a genuine, one-of-a-kind item, with the ownership record consistently reflecting the owner's name.
NFTs can be created, bought, sold, and traded on various marketplaces, as long as they comply with blockchain standards such as ERC-721 or ERC-1155 for Ethereum Built on decentralized blockchain networks, NFTs facilitate peer-to-peer transactions while minimizing the need for intermediaries.
Thirdly, creators can designate the scarcity of NFTs, determining the number of tokenized copies or editions Scarcity often contributes to the perceived value of NFTs.
NFTs are capable of storing important metadata about the digital items they represent, such as creator information, creation date, and descriptions Additionally, they can feature programmable properties and functionalities via smart contracts, which can include provisions for royalties, enabling creators to receive a percentage of future sales.
NFTs function like physical assets, allowing for buying and selling based on market dynamics such as supply and demand They represent a variety of assets, including tangible items like art and real estate By tokenizing real-world goods, the processes of buying, selling, and trading can be streamlined, which may also lower the risk of fraud (Kaspersky, 2023).
2.1.2 Overview of DeFi assets 2.1.2.1 The concept of DeFi
Decentralized Finance (DeFi) encompasses a variety of financial services and tools that function on blockchain networks, eliminating the need for intermediaries such as banks The main goal of DeFi is to create an inclusive and transparent financial ecosystem that is accessible to anyone with internet access.
DeFi assets are digital assets created and managed within decentralized blockchain networks, operating independently of traditional intermediaries such as banks These assets include cryptocurrencies, stablecoins, and various digital tokens that serve multiple financial functions like lending, borrowing, and trading A key feature of DeFi assets is their ability to facilitate peer-to-peer transactions without third-party involvement Additionally, many DeFi assets utilize smart contracts to automate financial processes, ensuring that transactions are executed according to predetermined rules.
The decentralized finance (DeFi) market experienced a notable shift in investor interest and funding levels in 2022 In 2023, it is projected to generate $16.96 billion in revenue, equating to an average revenue of $2,026 per user This revenue is expected to grow at a compound annual growth rate of 19.6% from 2023 to 2027, potentially reaching $34.7 billion Additionally, the number of DeFi users is forecasted to rise to 9.33 million by 2027, with a penetration rate of 0.11% (Statista, 2023).
A comparison of the above mOd€ÌS - + t3 sEEekekrskekerkreree 19 2.3 A review of related SEUC1©S - - + L1 9 vn nh nh TT TH TH HH HH ngờ 20 2.4 Cu na
The PSY model is a powerful analytical tool that accurately identifies bubble periods in asset markets, determining their start and end times Its agile GSADF approach allows it to adapt to rapidly changing market data, making it effective in diverse market conditions Additionally, the model not only detects bubbles but also serves as an early warning system for potential market collapses and suspicious movements With its reliable risk assessment capabilities, the PSY model is an invaluable resource for analyzing financial market dynamics and trends, making it a key focus of this research.
2.3 A review of related studies Since the introduction of the first cryptocurrency - Bitcoin in 2008, a large number of research studies have examined the different characteristics of cryptocurrencies, especially their hedging properties (Kang et al., 2020; Urquhart et al., 2019; Corbet et al., 2019; Stensas et al., 2019) To be specific, cryptocurrencies are shown to be able to hedge against inflation (Wagenaar, 2022; Smales, 2022), economic policy uncertainty (Vadar et al., 2019; Bouri et a/, 2019; Cheng et al., 2020), geopolitical risk (Tan, 2017; Allen et al., 2018), and the downside risk of stocks (Conlon et al (2020) Bouri et al., 2018), bonds (JuSkaité et al., 2022), gold (Hasan et al., 2022), oil (Heikal et al., 2022), and other commodities (Urom et al., 2021). Recently, the hedging capabilities of cryptocurrencies have been tested during turbulent times such as the COVID-19 pandemic (Lavelle, 2022) or the Ukraine war (Larsson and Johansson, 2022).
Cryptocurrencies, particularly Bitcoin, have been the focus of extensive research regarding their price patterns and trading dynamics Lee et al (2009) highlight that trading volume and price fluctuations in the Bitcoin market are influenced by cross-correlations, providing insights into cryptocurrency market behavior Various scholars have employed different factors to predict Bitcoin prices, including technical analysis elements (Islam et al., 2018) and market psychology influences (Vo et al., 2019) Additionally, some researchers, like Kamps, explore the abnormal behaviors exhibited by cryptocurrency prices.
(2018) and Morgia e¿ al (2021) identify the periods of cryptocurrency pump and dump and specific indicators of an impending surge in market activity preceding a price rise.
Despite the wealth of research on Bitcoin and cryptocurrencies, there is limited literature on DeFi and NFT markets Notably, Corbet et al (2021) explore whether DeFi tokens represent a unique asset class, while Dowling (2022) examines NFT pricing and its correlation with other cryptocurrencies Dowling's findings reveal that NFT price dynamics are distinct from those of traditional cryptocurrencies.
Numerous empirical studies have investigated bubbles across different sectors, notably in financial markets (Ghosh et al., 2021), commodity markets (Figuerola-Ferretti et al., 2015; Caspi et al., 2018), and foreign exchange markets (Hu and Oxley).
2017), and real estate markets (Deng et al., 2017), the examination of bubbles in cryptocurrency markets is still in its infancy.
A bubble in economic terms signifies a deviation from fundamental values, particularly challenging to ascertain in the realm of cryptocurrencies (Cox and Hobson, 2005; Kyrazis et al., 2020) Phillips et al (2015a, 2015b) characterize a bubble by explosive price movements, prompting researchers to develop statistical methods for identification Notably, Diba and Grossman (1988) utilize a unit root test to uncover price explosiveness, while MacDonell (2014) effectively applies the JLS model to forecast the Bitcoin price crash on December 4, 2013 Additionally, Phillips et al (2011) and Phillips et al (2015a) introduce the PWY and PSY methods to enhance the Augmented Dickey-Fuller (ADF) test for bubble detection.
Corbet et al (2018) and Bouri et al (2019) utilize the PSY method to identify bubbles in cryptocurrencies, with Corbet et al concentrating on Bitcoin and Ethereum, while Bouri et al analyze seven different cryptocurrencies Their findings reveal that bubble periods in one cryptocurrency often coincide with bubble occurrences in others.
So far, there appears to be a scant number of research examining bubbles in DeFi and NFT markets or the factors that can predict these bubbles Maouchi et al.
Research from 2022 highlights digital phenomena and the emergence of bubbles in DeFi and NFTs during the COVID-19 pandemic Specifically, studies have employed the Hot NFT Asset Proxy to detect price bubbles and manage portfolios (Vidal-Tomás, 2022; Wang et al., 2022) This work enriches the cryptocurrency literature by examining both the short-term and long-term performance of the evolving NFT and DeFi markets, positioning them as diversification tools for crypto portfolios, reminiscent of the ICO bubble that impacted the cryptocurrency market in 2017 (Momtaz, 2021).
Current research on DeFi and NFTs primarily focuses on their innovative ability to connect market participants via smart contracts, effectively bypassing traditional centralized intermediaries like banks and brokers (Harvey et al., 2021; Schọr, 2020), as well as the factors influencing their profitability (Lucey et al., 2022; Wang, 2022) Yousafa and Yarovaya (2022) indicate weak static returns and volatile spreads among NFTs, Bitcoin, DeFi assets, and select financial markets Additionally, studies on NFT price mechanisms (Aharon & Demir, 2021; Dowling, 2022a; Vidal-Tomás, 2022) and average NET domain prices (Aharon & Demir, 2021; Gubareva et al., 2022) have emerged However, the infrequent trading and quality issues surrounding NFTs complicate the creation of a comprehensive NFT price composite index, resulting in contentious findings in research such as Dowling (2022b) and Vidal-Tomas (2022).
2.4 Research gap Through the process of referring to existing research papers, it can be seen that there have been some studies on the connection between financial assets, but there are some limitations that still exist in the existing literature.
There is a notable lack of literature examining volatility changes preceding asset bubbles in over-the-counter finance markets compared to centralized financial markets Additionally, limited research has focused on the volatility connectedness between NFTs and DeFi assets within the cryptocurrency market, especially concerning price explosiveness (Maouchi et al., 2022; Wang et al., 2022).
Secondly, while rational models have attempted to analyze and justify the price movements of cryptocurrencies, they have encountered challenges in accurately
The prominent Bitcoin equilibrium model by Biais et al (2020) aims to calibrate and rationalize the dynamics of Bitcoin's price However, despite these calibration efforts, a significant portion of Bitcoin's price volatility remains unexplained, largely due to variations in fundamental factors like transaction costs and the advantages of using Bitcoin.
The differing risk factors between traditional financial markets and cryptocurrencies add complexity to understanding cryptocurrency price movements, making it difficult to effectively apply conventional financial models to these digital assets.
This research aims to enhance the understanding of bubble detection in cryptocurrency markets by focusing on Non-Fungible Tokens (NFTs) and Decentralized Finance (DeFi) sectors, addressing existing gaps in the literature.
RESEARCH METHODOLLUOOYY G5 G G5 99 990 09 0996909 996 24 3.1 Research method - The PSY method - - - 2 5c SE 2E 331 **EE+EE+EESEEsrsrrrrerrrrrrrrerre 24
This study analyzes asset price bubbles in the NFT and DeFi markets, alongside key financial assets like oil, gold, and stock indices The focus is on identifying bubble-like patterns using the PSY methodology by Phillips et al (2015), which enhances the supremum augmented Dickey-Fuller test This method allows for the detection of multiple bubble phases, including their start and end dates.
The method proposed by Phillips et al (2015) for identifying and timestamping bubbles is based on tests for an explosive root, with the null hypothesis defined as a random walk that includes a minimally impactful trend component.
Where: a, =dT" with n > 0.5 The alternative hypothesis posits a process with a moderately explosive root, as defined by Phillips and Magdalinos (2007):
The PSY method encompasses an iterative estimation of the regression model: p1,
As well as the calculation of the t-statistic (Augmented Dickey-Fuller):
In the analysis, the sub and superscripts À and À denote the specific sample portions used for estimating the coefficients, ensuring they fall within the range of 0 < A 1* A < 1 For example, the term a, signifies the constant term in the equation.
24 in the regression estimated between LTA a and LTA, where L - | denotes the floor function.
The PSY methodology employs the Supremum Augmented Dickey-Fuller (SADF) test to effectively identify multiple bubble episodes This process involves expanding the window size, Le, from its initial value, ro, which represents the smallest sample window, to 1, the largest window size in the recursive procedure In the SADF test, the starting point, A v 1S, is set at 0, while the subsample endpoints, À„, range from ro to 1 The SADF test is notably resilient against multiple disruptions.
The Generalized Supremum Augmented Dickey-Fuller (GSADF) procedure is applied recursively, allowing for a variable window width within a specified range This approach incorporates additional segments of the overall sample, enhancing flexibility in detecting multiple bubbles In GSADF, the starting point, denoted as d>, ranges from 0 to À, with À 2 € [A 0, enabling a more comprehensive analysis of financial bubbles.
1], and the conclusion of the subsamples is set at À„ spanning from A, to 1 The
GSADF is defined as follows:
The null hypothesis is discarded when the GSADF statistic exceeds the critical threshold, leading to the estimation of the bubble period.
=i A¿ BSADF (A) > cơ 12 fo inf A„ me 2" af 0) cv, 3 (12) where BSADF(r,) for r,€ lrạ, 1] is the backward sup ADF statistic defined as:
25 and cơ ` is the 100 - (1 — B,„)% critical value of the sup ADF statistic based B
Table 3.1 A summary of the test’s null and alternative hypotheses
Test Null hypothesis Alternative hypothesis ADF Unit root Explosive process
SADF Unit root Single periodically collapsing bubble period GSADF Unit root Multiple periodically collapsing bubbles
The PSY procedure is frequently used to assess the explosive behavior of a price-fundamental ratio When the null hypothesis of a unit root is rejected, it indicates that the variable y demonstrates explosive behavior If this variable y is linked to an economic fundamental, the detection of explosive behavior implies the existence of a bubble.
This study uses the daily closing price of two DeFi tokens (Chainlink [LINK] and Avalanche [AVAX]) and two NFTs (Decentraland [MANA] and Internet Computer [ICP]).
The selection of these four cryptocurrencies is based on their market capitalization rankings and the availability of data, particularly the length of their time series This includes two DeFi coins and two NFTs, both of which are relatively new in their respective markets Additionally, it is noted that ICP began trading on May 11.
In 2021, the cryptocurrency market experienced significant fluctuations, particularly with the launch of the Internet Computer Protocol (ICP) and a market capitalization exceeding $1,000,000, coinciding with the COVID-19 pandemic Research by Maouchi et al (2021), Shilov & Zubarev (2021), and Shu et al (2021) highlights a speculative surge in cryptocurrency values, especially Bitcoin, during the first half of 2021 This suggests that ICP, along with decentralized finance (DeFi) assets and non-fungible tokens (NFTs), faced heightened risks of price bubbles during this tumultuous period.
For an in-depth analysis of the timeframe and characteristics of cryptocurrency price series, along with their descriptive statistics, please consult Tables 3.2 and 3.3, which present all asset prices in USD.
Table 3.2 Overview of crypto assets under examination.
Series Ticker Type Issue Date Rank Max S T.S % of C.S # of pairs
The rank of a cryptocurrency is based on its overall market capitalization, with each asset assigned a position relative to all other cryptocurrencies and within specific categories For example, Decentraland is ranked 61st among all crypto assets and 9th within the NFT category The term "Max S." represents the maximum number of coins that will ever exist, while "T.S." indicates the total circulating supply at present The "% of C.S." reflects the percentage of the circulating supply compared to the total supply, and "# of pairs" denotes the total number of cryptocurrency and fiat currency combinations available for trading each asset.
Table 3.3 Descriptive statistics of the crypto assets under examination.
Series From To Mean Max Min High Low Market No of
Note: The Price and Market Capitalization (Market Cap.) are as of 5th September
In 2023, the maximum and minimum prices analyzed are represented by Max P and Min P., with notable all-time highs and lows for various cryptocurrencies For MANA, the all-time high occurred on November 25, 2021, while the all-time low was recorded on October 14, 2017 ICP reached its highest price on May 10, 2021, but lacks a recorded low due to insufficient data AVAX peaked on November 21, 2021, and hit its lowest price on December 31, 2020 Lastly, LINK's all-time high was also on May 10, 2021, with its lowest price being on September 23, 2017.
This study analyzes the daily closing prices of two major cryptocurrencies, Bitcoin (BTC) and Ethereum (ETH), due to Bitcoin's dominance in the market and Ethereum's crucial role in decentralized finance (DeFi) and non-fungible token (NFT) protocols A significant portion of smart contracts in these sectors operates on the Ethereum blockchain Additionally, a selection of notable NFT and DeFi assets, chosen based on market capitalization and data availability, was examined for potential price bubbles The selected DeFi assets include 0x (ZRX), Ren (REN), Terra (LUNA), Synthetix (SNX), Fantom (FTM), Reserve Rights (RSR), and THORChain (RUNE).
This study analyzes the NFT market by examining Theta Network (THETA) and Enjin Coin (ENJ), while also considering traditional financial assets such as Brent oil (OIL), gold (GOLD), and two key indices of the US stock market, the S&P 500 (SPX) and Dow Jones (DJI), along with the global stock market index (MSCT).
This study utilizes data sourced from two distinct platforms: daily closing prices of cryptocurrencies obtained from Coinmarketcap.com and prices for gold, Brent oil, the US stock index, and the MSCI World Index gathered from Investing.com.
EMPIRICAL RESULTS AND DISCUSSION <- << <<<<<< 29 4.1 Summary statistics aăăẳiỪỒằẳŸỪỒỦỒ®d
Results of PSY model . - - + 1n nh nh TT TT TH TH HH Hà HH TT ch nà 32
To investigate the presence of price bubbles and evaluate the confidence in our results, we utilize the PSY model, incorporating three key tests: the Augmented Dickey-Fuller (ADF), Supremum Augmented Dickey-Fuller (SADF), and Generalized Supremum Augmented Dickey-Fuller (GSADF) The PSY model relies on the ADF test for statistical calculations and employs a Monte Carlo simulation to determine the corresponding confidence levels.
This article analyzes the outcomes of the ADF, SADF, and GSADF tests to identify potential price bubbles in the NFT and DeFi markets Critical values for these tests are computed through Monte Carlo simulations, assuming the null hypothesis of a unit root Additionally, finite sample critical values are derived from a Monte Carlo simulation with 2,000 bootstrap replications, and the minimum window size is established accordingly.
A market is classified as experiencing a bubble when the price at the end of a minimum trading day exceeds the initial price, following the prescribed criterion of A 07 0.01 +, where T represents the sample length Conversely, it is categorized as a crash if the ending price falls below the starting price within the same period.
Table 4.2 Results of PSY test.
Note: The asterisks (*) denote the level of ADF, SADF, and GSADF test for test Statistics of all financial assets: *, **, ***, indicating critical values at 90%, 95%, and 99% levels (* 0.9 < tstat < 0.95; ** 0.95 < tstat < 0.99; *** tstat > 0.99), respectively.
Table 4.2 presents the Augmented Dickey-Fuller (ADF), Sup Augmented Dickey-Fuller (SADF), and Generalized Sup Augmented Dickey-Fuller (GSADF) test statistics alongside their critical values for various crypto and financial assets The SADF and GSADF tests reveal that two test statistics nearly surpass the critical values at the 90%, 95%, and 99% confidence levels, strongly supporting the notion that the NFT and DeFi markets exhibit explosive sub-periods and potential price bubbles This finding also applies to the cryptocurrencies Bitcoin (BTC) and Ethereum (ETH) In contrast, traditional financial markets show minimal variation, with the SADF and GSADF test statistics for US stock market indexes and gold prices remaining unchanged compared to the aforementioned crypto assets Notably, the SADF test for the world stock market index indicates test statistics greater than 90% but less than 95%, while the SADF test for Brent oil prices reveals test statistics lower than all others.
33 critical values This may suggest that Brent oil prices and world stock market indexes have few unique periodic bubble collapse periods.
4.3 Discussion Table 4.3 Date-stamping bubble periods.
Series Min duration bubbles Bubble periods Duration Peak
5th February 2021 - 17 13th February 22nd February 2021 2021
3rd March 2021 - MANA 12th May 2021 70 13th March 2021
12th January 2021 - 21st January 2021 9 16th January 2021
17th August 2021 - 9 19th May 2019 - 26th August 2021 29th May 2019
19th May 2019 - 29th May 2019 10 27th May 2019
17th June 2019 - 15th July 2019 28 3rd July 2019
16th July 2020 - 24th Tuly 2020 8 19th July 2020
6th February 2021 - 1 17th February 27th February 2021 2021
15th April 2021 - 23rd April 2021 8 19th April 2021
4th May 2021 - 20th May 2021 16 13th May 2021
4th February 2021 - 2 13th February 16th February 2021 2021
22nd June 2019 - 7th July 2019 15 2nd July 2019
REN 10th August 2020 - 24 18th August 2020
8th February 2021 - 15 19th February 23rd February 2021 2021
23rd May 2019 - 13th June 2019 21 Ist June 2019
3rd January 2021 - 51 12th February 23rd February 2021 2021
27th February 2021 - 12th March 2021 13 3rd March 2021
15th August 2020 - 26th August 2020 li 20th August 2020
2nd February 2021 - 26 24th February 28th February 2021 2021
Ist March 2021- 11th March 2021 10 1st March 2021 FTM
Ist September 2021 - 13 9th September 14th September 2021 2021
6th October 2021 - 41 26th October 16th November 2021 2021
13th August 2020 - RSR 20th August 2020 7 16th August 2020
15th March 2021 - 24th March 2021 9 21st March 2021
Sth August 2020 - 3rd 29 14th August 2020
13th January 2021 - 21st January 2021 8 18th January 2021
22nd January 2021 - 24th March 2021 61 lst February 2021
26th March 2021 - 19th May 2021 54 18th April 2021
Ist March 2021 - 19th ENJ 1 April 2021 49 15th March 2021
31st January 2011 - 28 11th February 28th February 2011 2011
2 15th May 2011 May 2011 9 5th May 201
21st May 2011 - 13th June 2011 23 9th June 2011
25th January 2013 - 12th April 2013 77 10th April 2013
4th November 2013 - 33 29th November 7th December 2013 2013
12th June 2016 - 21st Tune 2016 9 16th June 2016
27th December 2016 - BTC 16 6th January 2017 10 4th January 2017
2nd May 2017 - 14th July 2017 73 10th June 2017
20th July 2017 - 14th 56 1st September September 2017 2017
31st December 2017 - 11th January 2018 11 5th January 2018
21st June 2019 - 30th 9 26th June 2019 June 2019
16th December 2020 - 152 21st February 17th May 2021 2021
11th October 2021 - 38 20th October 18th November 2021 2021
2nd February 2016 - 14 11th February 16th February 2016 2016
22nd February 2016 - 19th March 2016 26 6th March 2016
19th March 2017 - 3rd April 2017 15 24th March 2017
26th April 2017 - 26th June 2017 61 12th June 2017
11th December 2017 - 1 19th December 22nd December 2017 2017
22nd January 2021 - 34 19th February 25th February 2021 2021
3rd March 2021 - 24th March 2021 21 13th March 2021
28th October 2021 - 19 8th November 16th November 2021 2021
16th March 2020 - OIL 1 25th March 2020 7 18th March 2020
7th August 2019 - 19th August 2019 8 15th August 2019
27th July 2020 - 11th August 2020 11 6th August 2020
4th January 2018 - ond February 2018 20 26th January 2018
16th March 2020 - 25th March 2020 7 23rd March 2020
28th November 2017 DJI 7 3 - 2nd February 2018 45 26th January 2018
16th March 2020 - 25th March 2020 7 23rd March 2020
5th January 2018 - Ist February 2018 19 26th January 2018
12th March 2020 - 25th March 2020 9 23th March 2020
The table outlines the frequency of identified bubble days across various crypto and financial assets, along with the total count of detected bubble days on a daily basis "Peak" indicates the day with the highest bubble intensity, while "Minimum duration" denotes the smallest permissible window size, calculated using the criterion of ro = 0.01 + 3, where T represents the sample length, requiring a minimum of one trading day.
Obs BD Pct HM ABM Weight ABM
Note: Bubble days (abbreviated as “BD”) refers to the aggregate count of days when the crypto asset was in a state of experiencing a bubble The “Percentage of bubble days ”
The "Pct" represents the ratio of total bubble days to the overall number of observations for each token The "Magnitude of a bubble" indicates the percentage increase from the lowest to the highest prices during a specific bubble period The highest magnitude (HM) reflects the most significant price fluctuation across all bubbles for each token, while the average bubble magnitude (ABM) signifies the mean price fluctuation across all bubbles experienced Additionally, the "Weighted ABM" is calculated as the ratio of bubble magnitude to the total number of bubble days.
An analysis of individual assets reveals that most reviewed crypto and financial assets experienced notable price surges, with the exception of ICP.
The analysis of the DeFi and NFT markets reveals that the DeFi market experiences a higher frequency of price bubbles compared to the NFT market This difference is attributed to the continuous trading of DeFi coins, while NFT prices fluctuate only when the underlying assets, such as antiques or clothing, are bought and sold Notably, Internet Computer (ICP), a prominent player in the NFT sector, has shown no signs of price bubbles during the study period In contrast, LINK has been identified as the DeFi asset with the most frequent price bubbles, occurring seven times within a 124-day timeframe.
In the summer of 2020, specifically from July to September, a notable bubble emerged in the LINK - DeFi asset, coinciding with the initial rise of DeFi tokens, commonly referred to as the "DeFi summer 2020."
Maouchi et al (2021) reveal that the price bubbles of AVAX and LINK in the DeFi market correspond with key events in the wider cryptocurrency landscape, including the launch of Libra on June 18, 2019, the cryptocurrency bull market from October 2020 to May 2021, and the market's recovery from July 2021 to November 2021 These results align with earlier studies on cryptocurrency price bubbles, as noted by Corbet et al (2018), Yao and Li (2021), and Li et al (2021).
The period of early 2021 can be characterized as "NFT Spring," mirroring the earlier "DeFi Summer" of 2020, driven by significant milestones such as PieDAO DEFI++ reaching a market cap of $2.34 million and NFT sales soaring to $2.5 billion in the first half of the year This surge included a staggering 315% month-on-month increase in total NFT sales volume and the Bored Ape Yacht Club achieving an extraordinary 58,118% ROI These developments attracted a wave of investors to the burgeoning DeFi and NFT markets, raising concerns about potential speculative bubbles Notably, a substantial bubble was identified in the MANA NFT asset at the start of 2021, followed by a brief cooldown from February 23 to March 2, during which the cryptocurrency bubble deflated Subsequently, MANA experienced a significant price surge from March 3 to mid-May, as detailed in Table 4.3.
In our analysis, we applied a real-time price bubble date-stamping strategy to both NFT and DeFi markets, with results detailed in Table 4.4 Figures 4.3, 4.4, and 4.5 depict the trends of NFT and DeFi assets during the tested periods using ADF tests against critical value sequences The findings reveal that DeFi markets exhibit significantly larger price bubbles than NFT markets, especially when comparing assets like AVAX and LINK to MANA and ICP This difference may be attributed to the greater volatility and uncertainty inherent in cryptocurrency markets, as highlighted by Urquhart (2016) and Lucey et al (2022) The average price bubble rate in the DeFi market ranges from 3% to 5%, while MANA has a bubble rate of 4.68% and ICP shows no bubble presence.
MANA and ICP exhibit distinct characteristics due to their differing asset natures and issuance timelines Research on cryptocurrency price bubbles, including studies by Corbet et al (2018) and Maouchi et al., highlights these variations in market behavior.
In 2021, the DeFi and NFT markets exhibited significantly high bubble magnitudes, highlighting an increased occurrence and intensity of price volatility This observation reinforces the idea of inefficiencies in the pricing mechanisms of these markets, consistent with the research conducted by Dowling (2022a).
Figure 4.3 Evolution of MANA, period from Eigure 4.4 Evolution of LINK, period from
17th September 2017 to 5th September 2023 24th September 2017 to 5th September 2023.
Figure 4.5 Evolution of AVAX, period from 22nd September 2020 to 5th September 2023. o4
The DeFi market has exhibited notable price bubbles for various crypto assets such as ZRX, REN, SNX, FTM, RSR, and RUNE, as illustrated in Figures 4.6 to 4.11, with LUNA being an exception These bubbles primarily emerged during two key periods: the "DeFi summer 2020" and from mid-2021 onwards, both of which saw significant price surges for DeFi coins and tokens.
Figure 4.6 Evolution of ZRX, period from
16th August 2017 to Sth September 2023.
Figure 4.8 Evolution of SNX, period from
14th March 2018 to 5th September 2023.
Figure 4.10 Evolution of RSR, period from
24th May 2019 to 5th September 2023.
Figure 4.12 Evolution of THETA, period from
17th January 2018 to 5th September 2023.
Figure 4.7 Evolution of REN, period from 21st February 2018 to 5th September 2023.
Figure 4.9 Evolution of FTM, period from 30th October 2018 to 5th September 2023.
Figure 4.11 Evolution of RUNE, period from
23rd July 2019 to 5th September 2023.
Figure 4.13 Evolution of ENJ, period from 1st November 2017 to 5th September 2023.
Examining the co-explosivity across financial aSS€fS .- - c5 sec s+xssserrrrrrres 45
The analysis reveals the presence of bubbles in 15 cryptocurrency assets and 5 additional financial assets, highlighting three significant co-explosivity bubble periods: (i) from 2017 to 2018, and (ii) during the DeFi summer.
2020, and (iii) the 2021 bubble Additionally, we can observe that the COVID-19 pandemic plays a significant role in exacerbating the latter two series of bubbles.
At the end of 2017 and the beginning of 2018, there was a significant surge in prices for ZRX, BTC, ETH, DJI, SPX, and MSCI, indicating an explosive market trend This phenomenon aligns with existing research on bubble detection in cryptocurrencies, as highlighted by Kyriazis et al (2020) Notably, this period witnessed the highest magnitudes of market bubbles.
Between September 15, 2017, and December 30, 2017, 45 instances of Bitcoin price fluctuations were detected, while Ethereum showed significant activity from April 26, 2017, to June 26, 2017, and again from late December 2017 to mid-February 2018 Additionally, the DeFi coin 0x (ZRX) reached its peak bubble period between December 23, 2017, and January 15, 2018.
The "DeFi Summer 2020" marked a significant rise in market capitalization for decentralized finance (DeFi), largely driven by the launch of Compound's COMP token liquidity mining program in May 2020 This event is considered the beginning of decentralized lending applications, paving the way for the popularization of liquidity mining and yield farming strategies These practices encourage investors to strategically reallocate their crypto assets across various decentralized lending pools and platforms to maximize their returns.
Between mid-2020 and May 2021, all 15 analyzed crypto assets and 5 other financial assets exhibited signs of significant price volatility, with Bitcoin, Ethereum, Decentraland, ChainLink, Terra, Synthetix, THORChain, Fantom, and THETA showing more pronounced bubble durations Notably, February 2021 marked the only month where all 15 crypto assets experienced bubbles simultaneously During this period, major companies and institutional investors, such as Tesla, Mastercard, and Bank of New York Mellon, publicly demonstrated interest in cryptocurrency investments These observations indicate common driving factors affecting both crypto and traditional financial markets, suggesting the need for further research, which aligns with Dowling's (2022b) findings on NFTs.
CONCLUUSIONN .o 5 (5-5 5< HH HH Họ 0008 50 47 5.1 Summary of key findITðS . - c1 22118911191 111 1111 911 11119111 1 91H ng ng hy 47 5.2 x⁄Ui00)9ì xi9 11 47 5.3 Research limitations and directions for future sfudÌIes - - +: s+ss+sx+sseessxss 48 REEFERENGCES 5-5 <5 5 Họ HH Họ TH HT II 1000090 50
This study provides crucial insights into the emergence and timing of price bubbles driven by the rise of NFT and DeFi assets We have identified several bubbles across 15 crypto assets and 5 additional financial assets, highlighting three significant bubble periods: the 2017-2018 phase, the DeFi summer of 2020, and the bubbles of 2021.
The detection of price bubbles in the NFT and DeFi markets can be effectively achieved using ADF, SADF, and GSADF tests Both markets exhibit significant speculative characteristics, with DeFi markets displaying notably larger bubble magnitudes than their NFT counterparts Additionally, DeFi bubbles are not only more frequent but also exhibit greater average explosive magnitudes compared to those found in NFT markets.
The relationship between market hype and uncertainty in the cryptocurrency sector is notably significant, particularly concerning market bubbles Furthermore, there are identifiable phases during which bubbles are absent, suggesting that these markets may possess inherent value.
The swift expansion of NFTs and DeFi can potentially lead to spillover effects on other cryptocurrencies and broader financial markets which raise concerns.
5.2 Policy implications This study suggests several important policy considerations that policymakers should consider.
Implementing a system to identify and timestamp bubbles in the DeFi and NFT markets can provide early warnings for policymakers to assess the need for market interventions or regulatory changes, which is particularly crucial in China, where cryptocurrency trading and mining are banned, yet controlled uses of blockchain and NFTs are being explored (Wang et al., 2022) Additionally, policymakers should identify valuable applications for DeFi and NFT assets and create tailored regulations to avoid overly broad measures that could hinder market potential, especially in emerging economies (Zhao et al., 2021).
DeFi Tokens and NFTs are high-risk assets that can experience significant price bubbles, leading to potential substantial losses This risk is particularly heightened in developing countries, where financial markets may lack stability.
Investors in developed and alternative markets are increasingly drawn to high-risk, high-reward speculative ventures like cryptocurrencies, P2P lending, and NFTs, as traditional investment options remain limited Historical trends indicate that financial crises often stem from the bursting of asset bubbles, highlighting the need for policymakers to prioritize financial literacy This focus is crucial as trading becomes more accessible to a broader audience, ensuring that investors are better equipped to navigate these volatile markets.
Market cycles feature both significant bubbles and calmer phases, with maturing markets expected to stabilize price dynamics, leading to fewer bubbles and improved market efficiency This evolving environment offers lucrative profit opportunities for investors willing to take risks.
Policymakers must assess financial incentives and practices such as liquidity mining and yield farming, especially in light of the growing financialization of NFTs and NFT mining These activities can generate unsustainable returns fueled by significant new capital inflows, which may distort investor expectations and exacerbate market bubbles.
5.3 Research limitations and directions for future studies This study still has certain limitations.
This study focuses on select cryptocurrencies within the DeFi and NFT markets, specifically Decentraland, Internet Computer, Avalanche, and Chainlink, due to constraints in time and resources, rather than analyzing all available options.
The effectiveness of the PSY model is hindered by the limited quality of data available in the cryptocurrency market, where historical data is often biased or uncertain This lack of reliable data can restrict the model's ability to accurately identify specific bubbles, potentially leading to missed opportunities in bubble detection.
Future research should focus on the maturation of cryptocurrency, DeFi, and NFT markets, particularly through the evaluation of price mechanism efficiency Additionally, investigating the transmission channels of risks among these markets presents a valuable opportunity, as they often experience synchronized price bubbles Notably, fluctuations in cryptocurrency markets can significantly trigger bubble formations in both DeFi and NFT sectors.
Recent studies by Karim et al (2022), Yousaf and Yarovaya (2022a, 2022b), and Wang (2022) have explored the interconnectedness of risk spillover across various markets, highlighting the importance of examining the median transmission effects between them.
REFERENCES Aloosh, A., Ouzan, S., & Shahzad, S J H (2022) Bubbles across meme stocks and cryptocurrencies Finance Research Letters, 49, 103155.
Akkus, H T., Gursoy, S., Dogan, M., & Demir, A B (2022) Metaverse and metaverse cryptocurrencies (meta coins): Bubbles or future? Journal of Economics Finance and Accounting, 9(1), 22-29.
Baur, D G., Hong, K., & Lee, A D (2018) Bitcoin: Medium of exchange or speculative assets? Journal of International Financial Markets, Institutions and
Bianchetti, M., Ricci, C., & Scaringi, M (2018) Are cryptocurrencies real financial bubbles? Evidence from quantitative analyses Evidence from Quantitative Analyses (February 24, 2018) A version of this paper was published in Risk, 26.
Bollerslev, T (1986) Generalized autoregressive conditional heteroskedasticity. Journal of econometrics, 31(3), 307-327.
Bouri, E., Gupta, R., Lau, C K M., Roubaud, D., & Wang, S (2018) Bitcoin and global financial stress: A copula-based approach to dependence and causality in the quantiles The Quarterly Review of Economics and Finance, 69, 297-307.
Carol, A (2001) Market models: a guide to financial data analysis ED John Wiley, Nueva York.
Caspi, L, & Graham, M (2018) Testing for bubbles in stock markets with irregular dividend distribution Finance Research Letters, 26, 89-94.
Cheng, H P., & Yen, K C (2020) The relationship between the economic policy uncertainty and the cryptocurrency market Finance Research Letters, 35, 101308.
Conlon, T., Corbet, S., & McGee, R J (2020) Are cryptocurrencies a safe haven for equity markets? An international perspective from the COVID-19 pandemic. Research in International Business and Finance, 54, 101248.
Costa, F J M (2017) Forecasting volatility using GARCH models (Doctoral dissertation, Universidade do Minho (Portugal).
Dowling, M (2022a) Fertile LAND: Pricing non-fungible tokens Finance Research Letters, 44, 102096.
Dowling, M (2022b) Is non-fungible token pricing driven by cryptocurrencies? Finance Research Letters, 44, 102097.
Engle, R F (1982) Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation Econometrica: Journal of the econometric society, 987-1007.
Enoksen, F A., Landsnes, C J., LušivJanská, K., & Molnar, P (2020). Understanding risk of bubbles in cryptocurrencies Journal of Economic Behavior & Organization, 176, 129-144.
Geuder, J., Kinateder, H., & Wagner, N F (2019) Cryptocurrencies as financial bubbles: The case of Bitcoin Finance Research Letters, 31.
Ghosh, B., Papathanasiou, S., Ramchandani, N., & Kenourgios, D (2021). Diagnosis and prediction of IIGPS’countries bubble crashes during BREXIT. Mathematics, 9(9), 1003.
Hall, A., & Virrantaus, K (2016) Visualizing the workings of agent-based models: Diagrams as a tool for communication and knowledge acquisition Computers, Environment and Urban Systems, 58, 1-11.
Hanappi, H (2017) Agent-based modeling History, essence, future PSL Quarterly Review, 70(283), 449-472.
Harvey, D I., Leybourne, S J., & Zu, Y (2020) Sign-based unit root tests for explosive financial bubbles in the presence of deterministically time-varying volatility. Econometric Theory, 36(1), 122-169.
Hasan, M B., Hassan, M K., Karim, Z A., & Rashid, M M (2022) Exploring the hedge and safe haven properties of cryptocurrency in policy uncertainty Finance Research Letters, 46, 102272.
Heikal, M., Saragih, M M S., Ilham, R N., Khaddafi, M., & Rusydi, R.
(2022) Effect of World Oil Prices on Cryptocurrency Return Journal of Accounting Research, Utility Finance and Digital Assets, 1(1), 61-68.
Hollekim, E R., & Raa, J (2018) Exploration of bubble properties in cryptocurrencies: a hybrid-study with quantitative models for crash estimation supplemented with industry experts (Master's thesis).
Hu, Y., & Oxley, L (2018) Do 18th century ‘bubbles’ survive the scrutiny of 21st century time series econometrics? Economics Letters, 162, 131-134.
Jensen, J R., von Wachter, V, & Ross, O (2021) An introduction to decentralized finance (defi) Complex Systems Informatics and Modeling Quarterly,
Juškaitẻ, L., & Gudelyté-Zilinskiené, L (2022) Investigation of the feasibility of including different cryptocurrencies in the investment portfolio for its diversification Business, Management and Economics Engineering, 20(1), 172-188.
Kang, S H., Yoon, S M., Bekiros, S., & Uddin, G S (2020) Bitcoin as hedge or safe haven: evidence from stock, currency, bond and derivatives markets. Computational Economics, 56, 529-545.
Karim, S., Lucey, B M., Naeem, M A., & Uddin, G S (2022) Examining the interrelatedness of NFTs, DeFi tokens and cryptocurrencies Finance Research Letters,
In their 2022 study, Larsson and Johansson analyze the investment potential of Bitcoin, Ethereum, and Gold as hedging tools during periods of stock market volatility, specifically focusing on the OMX30 index amid the COVID-19 pandemic and the Russian invasion of Ukraine Their research highlights the safe-haven properties and diversification benefits of these cryptocurrencies, providing a Swedish perspective on alternative assets during economic crises.
Lavelle, B., Yamamoto, K N., & Kinnen, M (2022) Cryptocurrencies, correlations, and COVID-19: diversifiers, hedge, or safe haven? Review of Integrative Business and Economics Research, 11(2), 25-35.
MacDonell, A (2014) Popping the Bitcoin bubble: An application of log-periodic power law modeling to digital currency University of Notre Dame working paper, 1-33.
Maouchi, Y., Charfeddine, L., & El Montasser, G (2022) Understanding digital bubbles amidst the COVID-19 pandemic: Evidence from DeFi and NFTs Finance Research Letters, 47, 102584.
Ozili, P K (2022) Decentralized finance research and developments around the world Journal of Banking and Financial Technology, 6(2), 117-133.
Phillips, P C., Shi, S., & Yu, J (2015) Testing for multiple bubbles: Historical episodes of exuberance and collapse in the S&P 500 International Economic Review, 56(4), 1043-1078.
Phillips, P C., Shi, S., & Yu, J (2015) Testing for multiple bubbles: Limit theory of real-time detectors International Economic Review, 56(4), 1079-1134.
Phillips, P C., & Shi, S (2020) Real time monitoring of asset markets: Bubbles and crises In Handbook of statistics (Vol 42, pp 61-80) Elsevier.
Podobnik, B., Horvatic, D., Petersen, A M., & Stanley, H E (2009). Cross-correlations between volume change and price change Proceedings of the National Academy of Sciences, 106(52), 22079-22084.
Rokosz, R (2018) Evaluating Asset Bubbles within Cryptocurrencies using the LPPL Model (Doctoral dissertation, Duke University Durham).
Schueffel, P (2021) DeFi: Decentralized Finance-An Introduction and
Overview Journal of Innovation Management, 9(3), I-XI. Širca, Š J., & Omladié, M (2016) The JLS model with ARMA/GARCH errors ARS Mathematica Contemporanea, 13(1), 63-79.
Smales, L A (2022) Cryptocurrency as an alternative inflation hedge?. Available at SSRN 3883123.
Stensas, A., Nygaard, M F., Kyaw, K., & Treepongkaruna, S (2019) Can Bitcoin be a diversifier, hedge or safe haven tool? Cogent Economics & Finance, 7(1),
Umar, Z., Abrar, A., Zaremba, A., Teplova, T., & Vo, X V (2022) The return and volatility connectedness of NFT segments and media coverage: fresh evidence based on news about the COVID-19 pandemic Finance Research Letters, 49, 103031.
Urom, C., Abid, I., Guesmi, K., & Chevallier, J (2020) Quantile spillovers and dependence between Bitcoin, equities and strategic commodities Economic Modelling, 93, 230-258.
Urquhart, A., & Zhang, H (2019) Is Bitcoin a hedge or safe haven for currencies? An intraday analysis International Review of Financial Analysis, 63,
Velankar, S., Valecha, S., & Maji, S (2018, February) Bitcoin price prediction using machine learning In 20/8 20th International Conference on Advanced Communication Technology (ICACT) (pp 144-147) IEEE.
Vo, A D., Nguyen, Q P., & Ock, C Y (2019) Sentiment analysis of news for effective cryptocurrency price prediction International Journal of Knowledge Engineering, 5(2), 47-52.
Wagenaar, L P (2022) Are cryptocurrencies good hedges against inflation? (Bachelor's thesis, University of Twente).
Wanfeng, Y (2011) Identification and forecasts of financial bubbles (Doctoral dissertation, PhD Thesis Diss ETH).
Wang, L., Ahn, K., Kim, C., & Ha, C (2018, June) Agent-based models in financial market studies In Journal of Physics: Conference Series (Vol 1039, No 1, p.
Wang, M C., Chang, T., & Min, J (2022) Revisit stock price bubbles in the COVID-19 period: Further evidence from Taiwan’s and Mainland China’s tourism industries Tourism Economics, 28(4), 951-960.
Wang, Y., Horky, F., Baals, L J., Lucey, B M., & Vigne, S A (2022) Bubbles all the way down? Detecting and date-stamping bubble behaviours in NFT and DeFi markets Journal of Chinese Economic and Business Studies, 20(4), 415-436.
Wang, Y., Lucey, B., Vigne, S A., & Yarovaya, L (2022) An index of cryptocurrency environmental attention (ICEA) China Finance Review International, 12(3), 378-414.
Yousaf, I., & Yarovaya, L (2022a) The relationship between trading volume, volatility and returns of non-fungible tokens: evidence from a quantile approach. Finance Research Letters, 50, 103175.
Yousaf, I., & Yarovaya, L (2022b) Static and dynamic connectedness between NFTs, Defi and other assets: Portfolio implication Global Finance Journal, 53, 100719.