Source-linked AI summary
Editorial: Understanding Cryptocurrencies
Wolfgang Karl Härdle, Campbell R. Harvey, Raphael C. G. Reule
TL;DR
Cryptocurrency research faces a rapidly developing technology with substantial potential and risks, alongside unusually rich transaction data. The paper explains cryptocurrency and blockchain mechanics, summarizes available data and market behavior, and identifies future research applications. It reports positive co-movement among cryptocurrencies but little evidence of co-movement or cointegration with gold, while emphasizing unresolved risks and measurement challenges.
Problem
Cryptocurrency technology is widely misunderstood, while empirical research remains nascent despite freely available histories of cryptocurrency transactions.
Method
The paper explains cryptocurrency mechanics, describes data sources, reports summary statistics, and identifies possible financial-economics research applications.
Results
The analysis finds positive correlations among cryptocurrencies but little evidence of significant correlation with the S&P 500, gold, or VIX, and no evidence of a cointegrating relationship with gold.
Takeaways & Limitations
Freely available cryptocurrency transaction data provide an extraordinary opportunity for empirical financial-economics research.
Takeaways & Limitations
Cryptocurrency risks are harder to quantify than benefits such as low transaction costs, security, and quick processing.
Abstract
from arXiv · showhide
Cryptocurrency refers to a type of digital asset that uses distributed ledger, or blockchain, technology to enable a secure transaction. Although the technology is widely misunderstood, many central banks are considering launching their own national cryptocurrency. In contrast to most data in financial economics, detailed data on the history of every transaction in the cryptocurrency complex are freely available. Furthermore, empirically-oriented research is only now beginning, presenting an extraordinary research opportunity for academia. We provide some insights into the mechanics of cryptocurrencies, describing summary statistics and focusing on potential future research avenues in financial economics.
1 Introduction
Cryptocurrencies use peer-to-peer blockchain systems that can reduce reliance on traditional financial intermediaries and support fast, low-cost transactions. The paper motivates empirical research by highlighting freely available transaction data while noting that cryptocurrency risks are harder to quantify.
- Motivation: Cryptocurrencies can eliminate financial intermediaries, requiring neither bank accounts nor credit cards for transactions.Their peer-to-peer design could broaden financial inclusion among more than two billion unbanked people.
- Motivation: Blockchain-based systems potentially enable cheap, secure, near-instant transactions outside traditional banking and credit-card infrastructure.They may also support near-real-time micropayments for internet products and services.
- Motivation: Cryptocurrencies can operate as decentralized autonomous organizations whose algorithmic rules govern money supply and whose network integrity replaces trust in individual participants.This structure challenges traditional monetary authorities and central banks.
- Risks: Anonymous transactions in some cryptocurrencies may facilitate illegal activity or broader threats to society and institutions.The passage presents anonymity as a specific risk of the nascent market.
- Risks: Cryptocurrency benefits are readily measurable, whereas quantifying associated risks is less straightforward.The contrast concerns measurable transaction cost, security, and processing speed versus harder-to-measure risks.
- Paper goals: The paper explains cryptocurrency mechanics, details useful data sources, provides summary statistics, and proposes future financial-economics research applications.It emphasizes abundant free transaction data and the early stage of academic research.
2 Cryptocurrencies and Blockchains
Cryptocurrencies are open-source digital assets secured through cryptography and consensus on distributed blockchains. The paper explains how chained hashes support an append-only ledger, how proof-of-work validates blocks, and why cryptocurrency systems encompass diverse uses beyond traditional coins.
- Cryptocurrencies: A cryptocurrency is a digital asset that uses cryptography to secure transactions, regulate additional value units, and verify asset transfers.Different cryptocurrencies vary in consensus mechanism, latency, and cryptographic hashing algorithms.
- Blockchain structure: Blockchains distribute ledger data across participants and use a structure of linked blocks rather than a centralized database.The paper describes blockchain structure as the key distinction from ordinary distributed databases.
- Blockchain structure: Each block’s digest is repeated at the start of the next block, so altering historical data breaks the chain and causes the network to reject the corrupted block.Append-only updates and consensus make rewriting transaction history extremely unlikely.
- Hashing: Cryptographic hashing transforms input data into a digest without revealing the original information and makes finding matching outputs computationally infeasible.Changing even a small input component produces a completely different hash output.
- Consensus and applications: Cryptocurrencies use peer networks in which participants hold ledger copies and algorithmically agree which blocks to accept, while some blockchains restrict ledger updates to trusted parties.Bitcoin was the first digital asset based on blockchain technology without backing or intrinsic value.
- Consensus: In bitcoin’s proof-of-work consensus, miners search for a nonce producing the required leading zeros; the network verifies the winning block before adding it.The successful miner receives freshly minted cryptocurrency as a reward.
- Applications: Cryptocurrency concepts extend beyond traditional bitcoin- and ethereum-like assets, and the paper’s seven-category taxonomy is not exhaustive.Examples that do not fit easily include Overlay and Facebook’s Libra.
3 Summary Analysis of Cryptocurrencies
The paper benchmarks selected cryptocurrencies against traditional assets using return, correlation, volatility, distributional, and cointegration analyses over May 1, 2017–June 30, 2019. Cryptocurrencies are highly volatile, positively correlated with one another, weakly related to traditional assets, non-normal in their return distributions, and not cointegrated with gold in this sample.
- Data and assets: The analysis covers BTC, ETH, and XRP alongside the S&P 500, GOLD, and VIX, using cryptocurrency and traditional-asset data sources.Cryptocurrency data come from the CRIX database and Cryptocompare API, while traditional-asset data come from Bloomberg.
- Correlation: Both XRP and ETH are positively correlated with BTC, while no significant correlation appears with the S&P 500, GOLD, or VIX.The analysis uses daily and monthly returns and 250-day rolling windows relative to BTC.
- Volatility: Cryptocurrency volatility is much higher than volatility for GOLD and the S&P 500, making most cryptocurrencies extremely risky stores of value.The comparison uses 100-day rolling-window standard deviations.
- Distributional properties: Cryptocurrency return distributions depart substantially from normality, as indicated by their higher moments and the QQ plots for BTC and GOLD.The paper specifically discusses excess kurtosis and skewness in the return distributions.
- Cointegration: The correlation analysis finds little evidence of co-movement between cryptocurrencies and gold, and cointegration tests find no cointegrating relationship.The paper reports this result for both the Johansen procedure and a method suited to high-dimensional nonstationary time series.
4 Potential Research Areas
The paper identifies research opportunities spanning cryptocurrency network design, sentiment, valuation, trading, and monetary interaction. Existing work uses blockchain activity, social-media text, economic models, and protocol analysis to study these areas, while scalability and fragmented market structure remain important concerns.
- Research scope: Cryptocurrency research opportunities include trading dynamics, pricing, volatility forecasting, network design, sentiment, valuation, monetary systems, institutions, and adoption.The paper frames these as areas for future financial-economics research.
- Network design and markets: More than 200 fragmented, mostly unregulated exchanges and substantial off-chain trading create a complex market structure for studying price formation.The paper notes that exchanges can function more like broker-dealers than traditional exchanges.
- Network design: Blockchain scalability is a major challenge, motivating distributed-ledger designs that target higher transaction throughput.One cited approach appends transactions to a tangle, with each new transaction approving two existing tips.
- Consensus and incentives: Blockchain consensus research examines miner incentives, strategic behavior, forks, transaction fees, and differences between proof-of-work and proof-of-stake.The cited work studies these mechanisms using stochastic games and long-run market-design theory.
- Sentiment and valuation: Blockchain activity and sentiment data provide cryptocurrency-specific inputs for studying market capitalization, returns, and volatility predictability.Examples include GitHub activity, active accounts, comments, cryptocurrency lexica, word embeddings, and natural-language processing.
- Valuation and monetary interaction: Economic models also study bitcoin’s interaction with monetary policy, including a simplified implication that bitcoin prices form a martingale.Other cited work investigates intrinsic value related to computing power and network adoption.
4.2 Monetary systems and financial development
Blockchain-based monetary systems could challenge traditional banking roles, compete with fiat currencies under some conditions, and support central-bank digital currencies. The paper also links these systems to financial inclusion and highlights constraints on monetary authorities’ coordination.
- Monetary systems: Blockchain-based payment systems could challenge the traditional roles that banks have played in the macroeconomy.The paper presents this as a potential effect of new monetary systems.
- Fiat competition: Cryptocurrencies may compete with fiat currencies when a central bank is perceived as weak or untrustworthy.The paper distinguishes this possibility from blockchain technology’s potential use by central banks.
- Central-bank digital currencies: Blockchain technology could improve central-bank operations and provide a platform for issuing central-bank digital currencies.The Venezuelan petro is cited as an early example of a government-issued digital currency.
- Monetary coordination: Nonregulated cryptocurrencies can restrict monetary authorities’ coordination capabilities when different transaction types coexist.This conclusion comes from a model of monetary equilibrium and competing transaction systems.
- Financial development: Cryptocurrencies may increase financial inclusion and economic activity in emerging markets, where bank-account ownership remains limited.The paper cites sub-Saharan Africa, where 34% of adults had a bank account in 2014.
4.3 Institutions
Cryptocurrency trading is organized across many fragmented exchanges with uneven liquidity, controversial volume measures, and persistent price differences. These institutional features create both arbitrage opportunities and constraints from settlement risk.
- Exchange structure: Hundreds of exchanges operate worldwide, each with distinct traits, while reported trading volume remains controversial.Coinmarketcap lists 218 exchanges, and a 2019 SEC filing argues that 95% of bitcoin trading volume is fake.
- Exchange structure: Price discrepancies arise from fragmented venues, uneven regulation, sentiment-driven prices, and market inefficiencies.Some apparent discrepancies outside the ten exchanges with credible volume may not represent real prices.
- Arbitrage: Arbitrage opportunities can persist for days or weeks across cryptocurrency–fiat markets despite substantial trading volumes.Spreads are much smaller for cryptocurrency pairs, suggesting that cross-border fiat controls contribute to the opportunities; order flow also helps explain price differences.
- Arbitrage: Exchange fragmentation supports profitable short-run arbitrage and high-frequency trading bots, although China later banned all cryptocurrency exchanges.Algorithms were used to identify mispricings across numerous Chinese exchanges before the 2017 ban.
- Settlement: Settlement protocols limit arbitrage because random waiting times expose traders to price risk.Theoretical arbitrage boundaries increase with expected latency, latency uncertainty, spot volatility, and risk aversion.
4.4 Adoption, price discovery and high-frequency data
Research on cryptocurrency adoption, price discovery, and high-frequency markets exploits publicly available user and transaction data. Findings address platform adoption, order informativeness, manipulation, market efficiency, and the effects of futures trading.
- Adoption: Token price appreciation can induce platform adoption, causing tokens to capitalize future user adoption and generally reducing user-base volatility.This fundamentals-based dynamic pricing model incorporates user-base externalities and endogenous adoption.
- Adoption: Bitcoin offers a unique opportunity to study technology adoption alongside publicly available micro-level user-to-user transaction and interaction data.The framework connects cryptocurrency adoption and pricing to broader information-technology adoption themes.
- Price discovery: Order informativeness generally rises with order aggressiveness but reverses in the outer layers of the bitcoin order book.Liquidity appears to migrate outward in response to information asymmetry.
- Market dynamics: Evidence supports price manipulation as a source of substantial distortive effects in cryptocurrency markets.The analysis examines whether tether supply or investor demand drove price effects involving bitcoin and other cryptocurrencies.
- Market dynamics: Bitcoin markets have been described as more efficient, but the effects of high-frequency trading, continuous trading, and increased liquidity remain unresolved.It remains to be seen whether greater liquidity will reduce previously observed harsh price swings.
- Futures and high-frequency data: Bitcoin futures expand institutional access, permit fiat settlement, and potentially boost liquidity, with CME volume approximately matching Binance.Almost all futures volume is reported in the CME contract.
4.5 Index construction
Cryptocurrency index construction is difficult because markets operate continuously across hundreds of venues with highly uneven liquidity. Researchers must resolve price selection, volatility measurement, and the treatment of forks.
- Market coverage: Cryptocurrency indices aggregate assets traded 24/7 across hundreds of venues, unlike traditional indices based on narrower trading intervals and fewer venues.Examples include WorldCoinIndex, CoinMarketCap, CryptoCompare, CCI30, and CME CF Cryptocurrency Indices.
- Price measurement: There is no agreed bitcoin spot price because liquidity varies widely and even CME and CBOE futures use different price-data sources.Price manipulation on some exchanges contributed to the SEC blocking cryptocurrency ETF creation.
- Volatility measurement: An implied volatility proxy such as VCRIX can capture cryptocurrency-market expectations when derivatives are unavailable for most cryptocurrencies.Kim et al. use relationships between VIX and underlying-asset volatility to guide proxy selection.
- Fork treatment: Forks create unresolved index-design questions about when new cryptocurrencies should enter an index and whether smaller-capitalization forks should be included.The issue also affects single-currency futures contracts.
4.6 Portfolio diversification
Cryptocurrency diversification research compares digital assets with gold and evaluates risk-based portfolio methods. Findings suggest potential defensive relevance for bitcoin, while traditional risk-based portfolios do not significantly improve investment performance under cryptocurrency volatility.
- Bitcoin and gold: Bitcoin and gold share finite supply and mining requirements, although gold additionally has uses in jewelry, art, electronics, and medicine.The limited supply and market acceptance of digital gold suggest a potentially similar role for cryptocurrencies.
- Bitcoin and gold: Bitcoin’s correlation of extreme returns with US and European equities increases during stock-market drawdowns and decreases during booms.Gkillas and Longin conclude that bitcoin can play an important asset-management role with results similar to gold.
- Portfolio construction: Traditional equal-risk-contribution, minimum-variance, and minimum-CVaR portfolios do not significantly boost investment performance under cryptocurrency volatility.The finding is attributed to the volatility structure of cryptocurrencies.
4.7 Bubbles
Research on cryptocurrency bubbles finds substantial volatility and bubble-like behavior, but evidence about whether bitcoin prices constitute a bubble remains inconclusive. The section also situates ICOs and blockchain-based credit networks as important adjacent research areas.
- Bubble evidence: Bitcoin returns are highly volatile, exhibit large kurtosis, and show negative skewness, making a price bubble plausible but not conclusive.The cited study analyzes daily bitcoin returns from January 2015 through March 2018.
- Bubble evidence: Research variously characterizes cryptocurrencies as exhibiting bubble-like behavior or as requiring extended bubble tests that account for time-varying volatility.The literature therefore uses multiple empirical approaches to diagnose cryptocurrency bubbles.
- Bubble evidence: The LPPLS confidence indicator fails to provide effective warnings during large short-term bitcoin price fluctuations, especially for positive bubbles.This limits its reliability as an early-warning diagnostic under sharp price movements.
- Alternative financing: 329 of 2027 listed ICOs failed, corresponding to 16.23%.The failure rate is reported for ICOs listed on tokendata.io.
- Alternative financing: Blockchain-based distributed credit networks are being studied as alternatives in which users transfer credit on demand and extend credit according to trust.The passage describes these systems as distributed systems of trust between users.
4.9 The role of energy in consensus mechanisms
Consensus mechanisms have differing energy implications, with bitcoin’s proof-of-work approach drawing environmental concern. Research links cryptocurrency mining to energy-sector performance and identifies further environmental questions for blockchain research.
- Energy costs: Bitcoin uses an energy-intensive consensus method, raising environmental concerns amid coal-dependent mining in China.Mining pools can further exacerbate energy consumption in proof-of-work blockchains.
- Energy costs: Cryptocurrency mining’s annual electricity consumption is growing, and its carbon production likely exceeds that of Portugal.The passage emphasizes that mining-related energy use has become environmentally consequential.
- Research implications: Continued cryptocurrency energy use affects the performance of the energy sector, highlighting the need to assess the environmental impacts of cryptocurrency growth.The cited research examines effects on energy markets, utilities companies, and green ETFs.
- Research implications: Blockchain technology also creates environment-related research opportunities beyond measuring mining electricity consumption.The section points to environmental applications and questions associated with blockchain technology generally.
5 Closing Remarks
Cryptocurrencies remain a promising but poorly understood financial innovation, with substantial confusion about their concepts and valuation. The paper responds by clarifying blockchain and cryptocurrency diversity while emphasizing broad opportunities for future research.
- Closing remarks: Cryptocurrencies are intriguing financial innovations, but considerable confusion surrounds their underlying concepts and valuation approaches.The closing remarks frame clarification as necessary for understanding the field.
- Closing remarks: The paper aims to provide a high-level understanding of blockchain technology behind cryptocurrencies.This is identified as the authors’ first goal.
- Closing remarks: Cryptocurrencies comprise multiple classes, including tokens representing traditional assets, utilities such as computational power, and fiat currencies.The authors caution against treating cryptocurrency as synonymous with bitcoin.
- Closing remarks: Research opportunities extend beyond the 2018 bubble in the most liquid cryptocurrencies to a broader new field in finance and economics.The authors describe this field as having much remaining work.