Proposal for a Comprehensive (Crypto) Asset Taxonomy
Thomas Ankenbrand, Denis Bieri, Roland Cortivo, Johannes Hoehener, Thomas Hardjono
PProposal for a Comprehensive(Crypto) Asset Taxonomy
Thomas Ankenbrand & Denis Bieri, Lucerne University ofApplied Sciences and Arts;Roland Cortivo & Johannes Hoehener, Swisscom AG;Thomas Hardjono, MIT Connection ScienceJuly 24, 2020
Abstract —Developments in the distributed ledger technology haveled to new types of assets with a broad range of purposes.Although some classification frameworks for common instru-ments from traditional finance and some for these new, so-called cryptographic assets already exist and are used, a holisticapproach to integrate both worlds is missing. The present paper fills this research gap by identifying 14 attributes, each of which isassigned different characteristics, that can be used to classify alltypes of assets in a structured manner. Our proposed taxonomy,which is an extension of existing classification frameworks,summarises these findings in a morphological box and is testedfor practicability by classifying exemplary assets like cash andbitcoin. The final classification framework can help to ensure thatthe various stakeholders, such as investors or supervisors, have aconsistent view of the different types of assets, and in particularof their characteristics, and also helps to establish standardisedterminology. I. I
NTRODUCTION
Since the inception of the Bitcoin network in the year 2009,the space for cryptographic assets has developed rapidly.The continuing technological innovation in the underlyingdistributed ledger technology could consequently lead to anincreasing transformation of traditional financial markets intocrypto-based markets. Although different asset classificationframeworks exist for both worlds, a holistic approach mergingboth traditional finance and the crypto economy is still lacking.This poses a challenge to the various stakeholders such asinvestors or regulators in retaining an overview of existingassets of different types and, in particular, of their design andindividual characteristics. In order to fill this lack of research,we propose a taxonomy for the systematic classification of alltypes of assets, be it of physical, digital or tokenised nature.II. L
ITERATURE R EVIEW
The characteristics and properties of the most common typesof financial instruments such as stocks, bonds, and derivativeshave been the subject of research for some time, not onlyin the academia, but also in the industry. Therefore, a widerange of publications exist that deal with the functioning ofthese different instruments in a structured way.One framework defining the structure and format for theclassification of financial instruments (CFI) was first proposedby the International Organization for Standardization (ISO)in the year 1997. The last revised version of the framework is Our research is part of the FinTech programme supported by Finnova,Inventx, Swiss Bankers Prepaid Services, and Swisscom. called ISO 10962:2019 and was published by ISO in 2019. Itseeks to provide a standard for identifying the type of financialinstrument and its main high-level features in the form ofspecific codes consisting of six alphabetical characters, andshould thus help to standardise country and institution-specificterminology in relation to financial instruments [1]. The firstcharacter of the CFI code indicates the main category offinancial instruments. These include equities, collectiveinvestment vehicles, debt instruments, different types ofderivatives, and others. The second character of the CFI codeindicates multiple subclasses in a given main category, calledgroups. Equities, for example, are divided into the groupscommon/ordinary shares, preferred/preference shares, andcommon/ordinary convertible shares, among other groups.The last four characters of the CFI code define the specificattributes of a financial instrument and depend on the groupto which the asset is allocated. For financial instrumentsin the group common/ordinary shares from the “equities”main category, relevant attributes include voting rights,ownership, payment status, and form. These attributes comewith predefined possible values that determine the final codeof a financial instrument [1]. For other groups such as bondsfrom the “debt instruments” category, alternative attributes,e.g., the type of interest or guarantee, are of relevance.A second framework for classifying financial instruments isproposed by Brammertz and Mendelowitz [2]. Their so-calledACTUS taxonomy is based on the specific nature of financialcontracts and in particular on their cash flow profiles and seeksto create a global standard for the consistent representationof financial instruments. It distinguishes between financialcontracts, which in turn are split into the subcategoriesof basic contracts and combined/derivatives contracts onthe one hand, and credit enhancement on the other. Basiccontracts consist of fixed income and index-based products,whereas combined/derivative contracts comprise symmetricfinancial products, options, and securitisation products.The second main category of the ACTUS taxonomy, i.e.,credit enhancement, includes guarantee contracts, collateralcontracts, margining contracts, and repurchase agreements.The standard is implemented on the SolitX platform witha technical API layer and DLT adapter for transactionsystems and accounting, and in the AnalytX architecture forrisk management analysis, simulations, asset and liabilitymanagement, and business planning [3].The standards proposed by [1] and [2] show that sophisticatedclassification frameworks for traditional financial assets exist,which are used in practice. For cryptographic assets, onthe contrary, the characteristics of many tokens in variousrespects, for example in terms of regulation, utility orvaluation, were and are still largely ambiguous and hard tomeasure. Several initiatives from governments, the academia,and the industry have sought to reduce these uncertainties For a detailed description of each category, see [1]. a r X i v : . [ q -f i n . GN ] J u l y systematically structuring the hundreds of existing tokensbased on predefined criteria.The Swiss Financial Market Supervisory Authority (FINMA),for example, issued guidelines for enquiries regarding theregulatory framework for initial coin offerings in early 2018,in which it distinguishes between three types of tokens, i.e.,payment tokens, utility tokens, and asset tokens, based onthe underlying economic purpose [4]. Whether a particulartoken is a financial instrument and thus would be subjectto certain laws and regulations depends on its economicfunction and the rights associated with it. Other jurisdictions,such as the European Union, Israel, Malta, and the UnitedKingdom, follow a similar classification approach, althoughtheir terminologies differ to some extent [5]. Additionally,some jurisdictions follow the approach that the three maintypes of tokens are not necessarily mutually exclusive. Rather,there are also hybrid forms that share characteristics of twoor three main types. Accordingly, particular cryptographicassets could thus, for example, have certain characteristics ofboth payment and utility tokens.In April 2019, the U.S. Securities and Exchange Commission(SEC) through its strategic hub for financial innovation,FinHub, published guidelines to determine whether a digitalasset, which may be a cryptographic asset, is an investmentcontract, i.e. an agreement whereby one party invests moneyin a common enterprise with the expectation of receiving areturn on investment. This assessment is done by applyingthe so-called Howey test. If an investment agreement exists,the digital asset is classified as a security and therefore U.S.federal securities laws apply and must be considered by issuersand other parties involved in, for example, the marketing,offering, sale, resale or distribution of the respective asset [6].Other jurisdictions, e.g., Ireland, follow a similar approach ofclassifying cryptographic assets based on their qualification asa security [7]. However, the Howey Test is to be understoodless as a classification framework but more as a decision aidas to whether a cryptographic asset represents a security or not.An academically based classification framework forcryptographic assets, which goes beyond the legal perspectiveand also takes technological and economic aspects, amongothers, into account was carried out by Oliveira et al. [8]. Byapplying a design science research approach, including 16interviews with representatives of projects with blockchain-based token systems, the paper derives a token classificationframework for cryptographic assets that can be used asa tool for better informed decision making when usingtokens in blockchain applications. Their final classificationframework consists of the 13 attributes class, function,role, representation, supply, incentive system, transactions,ownership, burnability, expirability, fungibility, layer, andchain, each of which include a set of defined characteristics.A similar framework was developed by Ballandies et al. [9]. The authors established a classification frameworkfor distributed ledger systems consisting of a total of 19descriptive and quantitative attributes with four dimensions(distributed ledger, token, action, and type). The attributescomprise the distributed ledger type, origin, addresstraceability, Turing completeness, and storage in thedistributed ledger dimension, underlying, unconditionalcreation, conditional creation, transferability, burn, and supplyin the token dimension, action fee, read permission, andactor permission in the action dimension, and fee, validatepermission, write permission, proof, and type in the consensusdimension. The framework was derived from feedback fromthe blockchain community.Three further classification frameworks for cryptographicassets that were strongly driven by the industry are thoseproposed by the consulting firm MME, the InternationalToken Standardization Association (ITSA), and the EthereumEnterprise Alliance (EEA).The framework by MME was published in May 2018 andfocuses on the legal properties and risk assessment ofcryptographic assets. The paper’s resulting classification isbased on a token’s function or main use, alongside othercriteria such as the existence of a counterparty, as well as itstype and/or the underlying asset or value. The final archetypesof cryptographic assets are native utility tokens, counterpartytokens, and ownership tokens, which are each subject toadditional subcategories of token types [10].The International Token Classification (ITC) framework bythe ITSA comprises an economic, technological, legal, andregulatory vertical each containing a set of subdimensionswith different attributes. The economic and technologicalverticals include three subdimensions each, which refer toa token’s economic purpose, its target industry, and theway of distribution, and the technological setup, consensusmechanism, and technological functionality, respectively. Thelegal vertical includes the two subdimensions legal claimand issuer type, whereas the regulatory vertical focuseson assessing a tokens regulatory status in the US, China,Germany, and Switzerland. Over all verticals, a total oftwelve subdimensions are defined, though ITSA plans todefine further subdimensions in the future. Concerningthe evaluation of these individual subdimensions, as ofSeptember 2019, the ITC framework already provideddetailed information on four of the twelve subdimensions,namely for the economic purpose, industry, technologicalsetup, and legal claim. The classification into these foursubdimensions was compiled in a database covering morethan 800 cryptographic tokens. Besides the classificationframework and the corresponding database, the ITSA alsointroduced a nine digit unambiguous identifier for each token,the so-called International Token Identification Number, shortITIN [11].2he third industry-driven framework for classifying crypto-graphic tokens was published by the EEA in November 2019.Their proposed Token Taxonomy Initiative (TTI) distinguishesbetween five characteristics a token can possess. The firstcharacteristic is the token type and refers to whether a tokenis fungible or non-fungible. The second characteristic, thetoken unit, distinguishes between the attribute of being eitherfractional, whole or singleton and indicates whether a token issubdivisible or not. The value type, as the third characteristic,can assume the attribute of being either of an intrinsic value,i.e., the token itself is of value (e.g., bitcoin), or a referencevalue, i.e., the token value is referenced elsewhere (e.g.,tokenised real estate). Characteristic four, the representationtype, comprises the attribute of being common or unique.Common tokens, on the one hand, share a single set ofproperties, are not distinct from one another and are recordedin a central place. Unique tokens, on the other hand, haveunique properties and their own identity, and can be tracedindividually. The fifth and last characteristic is the templatetype and classifies tokens as either single or hybrid and refersto any parent/child relationship or dependencies between to-kens. Unlike single tokens, hybrid tokens combine parent andchild tokens in order to model different use cases. In addition,the TTI provides measures in order to promote interoperabilitystandards between different blockchain implementations [12].III. T HE (C RYPTO ) A
SSET T AXONOMY
Building on the literature review in Chapter II, this chapterproposes a holistic framework for the classification ofassets. Unlike existing classification frameworks, our assettaxonomy aims to classify all existing types of assets, i.e.,assets from both traditional finance as well as the cryptoeconomy, based on their formal characteristics. Furthermore,the taxonomy introduces a terminology that is suitable forboth traditional and the crypto assets. A morphologicalbox is chosen as the methodological approach in order tobe able to take the multi-dimensionality of the matter intoaccount. The taxonomy is illustrated in Appendix A. Intotal, we identify 14 different attributes based on whichall types of assets can be classified. They include claimstructure, technology, underlying, consensus/validationmechanism, legal status, governance, information complexity,legal structure, information interface, total supply, issuance,redemption, transferability, and fungibility, with each attributecomprising a set of at least two characteristics. Note thatcertain attributes in the frameworks discussed in ChapterII subsume some of the attributes presented here. Hence,our 14 attributes factorize these superordinate attributes tomake them universally applicable. Table I breaks down the14 attributes in terms of their inclusion in the publicationsdiscussed in Chapter II. The first column shows the attributelabels of the taxonomy we propose. Column two to ten referto the publications discussed, where an “x” indicates thatthe corresponding attribute is either explicitly or implicitlyconsidered in the classification framework given in row one.Note that the terminology regarding a particular attribute differs across these publications, for example, because theyfocus on different types of assets. The terminology we proposegeneralises these terms to ensure compatibility across all typesof assets, thus creating a common linguistic understanding.Also note that due to the extension of the taxonomy totraditional assets, some DLT-specific attributes/characteristicsin the publications discussed are summarised or generalised,while new attributes/characteristics were added in order toenable the mapping of traditional asset types. Overall, TableI shows that each of the existing frameworks covers certainattributes determined by the specific focus or objective ofthe publication. The framework of FINMA [4], for example,focuses on regulatory aspects, and thus predominantlyincludes corresponding attributes, i.e., claim structure, legalstatus, and legal structure. Other frameworks, for example theone published by the EEA [12], focus more on technologicalaspects or the design of token features. Overall, none of theframeworks discussed in Chapter II covers the full range offormal attributes identified in our taxonomy. However, ourtaxonomy is generally confirmed by the existing literature,as each attribute is considered in at least one of the existingclassification frameworks. The degree of agreement withthe classification framework we propose varies, however.While the publication of ISO [1] covers four attributes of ourtaxonomy, the publications of Oliveira et al. [8] and Ballandieset al. [9] cover ten. There are also differences in coveragefrom an attribute perspective. While the underlying of anasset is of relevance in all frameworks analysed, the attributesinformation interface and fungibility are only covered bytwo. The taxonomy we propose therefore goes further thanthe existing classification frameworks, firstly because it isindependent of the type of assets to be classified and secondlybecause it contains additional attributes and characteristics.Since some of these attributes and characteristics are notintuitively clear, they are explained in more detail in thefollowing:
Claim structure:
Does the asset represent a claim, i.e., ademand for something due or believed to be due [13]?– No claim(s): The asset does not represent any kind ofclaim.– Flexible claim(s): The asset represents certain claims, thepossession or exercise of which can depend on certainconditions (e.g., catastrophe bonds).– Fixed claim(s): The asset represents claims which canneither be restricted nor restrained under any condition(e.g., fixed income).
Technology:
Which technology is the asset based on?– Physical: The asset exists in a physical form (e.g., goldbullion).– Digital: The asset exits in a digital form, but is not basedon the distributed ledger technology (e.g., electronicshare).– Distributed ledger technology: The asset is based on thedistributed ledger technology, structured either as a native3able I: Coverage of the 14 attributes in existing classification frameworksAttribute ISO[1] B.&M.[2] FINMA[4] O. et al.[8] B. et al.[9] MME[10] ITSA[11] EEA[12]Claim structure x x x x x xTechnology x x x x x xUnderlying x x x x x x x xConsensus/Validation mechanism x x xLegal status x x x xGovernance x x xInformation complexity x x xLegal structure x x x xInformation interface x xTotal supply x x x xIssuance x x x xRedemption x x x xTransferability x x x x xFungibility x xtoken, i.e., a token that is native to a specific blockchain,or as a protocol token, i.e., a token issued on an existingblockchain protocol [8] such as, for example, ERC-20 orERC-721 tokens for the Ethereum blockchain.
Underlying:
Which underlying or collateral is the asset’svalue based on?– No underlying: The asset’s value is not a derivative of anunderlying asset (e.g., bitcoin).– Company: The asset’s value represents a stake in acompany (e.g., equity).– Bankable asset: The asset’s value represents a bankableasset, i.e., an asset that can be deposited in a bank orcustody account (e.g., fiat currencies).– Cryptographic asset: The asset’s value represents a cryp-tographic asset, i.e., an asset based on the distributedledger technology (e.g., derivative of a cryptographicasset).– Tangible asset: The asset is in a physical form [14] (e.g.,real estate).– Contract: The asset’s value represents a contract (e.g.,license agreement).
Consensus-/Validation-mechanism:
How is the agreement onthe finality (e.g., property rights or ownership transfer) of theasset reached?– Instant finality: Consensus is final. Mechanisms thattypically, but not necessarily, belong to the deterministictype are, for example, notary services or qualified writtenform.– Probabilistic finality: Consensus is not final, but reachedwith a certain level of confidence. Mechanisms thattypically, but not necessarily, belong to the probabilistictype are, for example, proof-of-work or proof-of-stake.
Legal status:
What is the regulatory framework governing theasset?– Regulated: There are regulatory requirements for theissuance, redemption and governance of the asset.– Unregulated: There is no specific regulatory frameworkfor the issuance, redemption and governance of the asset.
Governance:
In which way is the asset governed?– Centralised: The asset is governed by an authoritativeparty or consortium.– Decentralised: The asset is governed without centralisedcontrol (e.g., certain types of cryptographic assets suchas bitcoin).
Information complexity: What type of information complex-ity is associated with the asset?– Value: The asset represents a specific value (e.g., curren-cies).– Contract: The asset encompasses conditional informationin addition to its value (e.g., coupon bonds or DLT-basedsmart contracts ).– Turing completeness: The asset is based on a Turing-complete («universally programmable») computationalmodel (e.g., Ethereum). Legal structure:
What is the legal form of the asset?– No legal structure: There is no legal structure governingthe asset.– Foundation: The asset is governed by a foundation/ truststructure. Note that the characteristics of this attribute build on each other, i.e., eachcharacteristic contains additional information compared to the previous one. Note that such (smart) contracts, as in the case of bitcoin, are notnecessarily based on a Turing-complete system. : The asset has an alternative legal structure (e.g.,central bank money). Information interface:
How does the asset receive and/orsend relevant information?– No interface: The asset has no kind of informationinterface.– Qualitative: The asset manages relevant information in-directly through an authorised instance (e.g., generalassembly).– Quantitative: The asset manages relevant informationfrom authorised sources automatically (e.g., IoT sourcesor oracle interfaces in the case of DLT-based smartcontracts).
Total supply:
To which limit can the asset be generated?– Fixed: The total supply of the asset is fixed.– Conditional: The total supply of the asset is dependenton predefined conditions.– Flexible: The total supply of the asset is managed flexiblyby authorised parties.
Issuance:
How is the asset generated?– Once: After an initial issuance, no additional units of theasset are issued.– Conditional: Additional units of the asset are issuedonce predefined conditions are met (e.g., newly issuedcryptographic assets through mining).– Flexible: Additional units of the asset can be issuedflexibly by authorised parties (e.g., increase in sharecapital).
Redemption:
How is the number of outstanding assets re-duced?– No redemption: The number of outstanding assets cannotbe reduced.– Fixed: The reduction of the number of outstanding assetsfollows a predefined protocol.– Conditional: The reduction of the number of outstandingassets is initiated once predefined conditions are met.– Flexible: The reduction of the number of outstandingassets can be carried out flexibly by authorised parties(e.g., share buyback).
Transferability:
Can the asset’s ownership be transferred toanother party?– Transferable: The asset’s ownership can be transferred toanother party.– Non-transferable: The asset’s ownership cannot be trans-ferred to another party, for example, by sale or giveaway(e.g., some types of registered securities). The characteristic “Other” subsumes the broad range of alternative legalstructures for reasons of simplicity and practicability.
Fungibility:
Can the asset be interchanged with another assetof the same type?– Fungible: The asset is substitutable with another asset ofthe same type.– Non-fungible: The asset is not substitutable with anotherasset of the same type (e.g., artwork).IV. C
LASSIFICATION E XAMPLES
This subchapter seeks to test the above-mentioned taxonomywith selected examples. First, the taxonomy is used to comparecash to bitcoin, as both are intended means of payment .This comparison is followed by the classification of Ether,a utility token, Crowdlitoken, an asset token, CryptoKitties,and a traditional share. A. Comparison between Cash and Bitcoin
As both cash and bitcoin follow the purpose of a means of pay-ment, both assets share certain similarities (see Appendix B).Neither cash, in the case of a fiat money system, nor bitcoinhave a direct underlying asset. The value of the two assets israther based on the public’s trust in the issuer of the currencyor in the underlying technological protocol, respectively. Thereis also no oracle interface, i.e. no specific source that interacts(e.g., directly provides information) with cash or bitcoin. Sinceboth assets are designed as cash equivalents, their units aretransferable from one party to another and individual units areinterchangeable. Besides these commonalities there are somesignificant differences. While cash represents a certain valuewhich depends on the denomination, bitcoin is of contractualtype as it is transferred via smart contracts in the Bitcoin-script which is not Turing-complete. Bitcoin is furthermore notsubject to any type of legal claim and has no legal structure.In contrast, cash is regulated as legal tender under nationallaw. Since cash is of physical form, consensus on its stateis given deterministically by the owner of the asset. Bitcoin,on the contrary, is a digital representation of value based onthe distributed ledger technology. It is the native token of theBitcoin blockchain, the consensus of which is based on theproof-of-work mechanism and thus finality of the system is notguaranteed but only probabilistic. This implies a decentralisedgovernance of the asset, which is in contrast to the centralisedgovernance of cash by central banks. Both assets also differin terms of their total supply as well as in their ways tomanage the number of outstanding units. While the maximumsupply of bitcoin is fixed at 21 million units, there is no suchrestriction for cash. The issuance of additional units of bitcoinis conditional on the mining of new blocks and reducing thenumber of outstanding units is not possible . The issuanceand redemption of cash, on the contrary, is handled flexiblyby central banks. Bitcoin is often considered to be a store of value, but the original intentionis to provide an alternative means of payment. It is possible to send units of bitcoin, or other cryptographic assets, toan address without a known private key, so that these units are no longeraccessible. However, this does not reduce the number of total units in thesystem. . Ether Ether (see Appendix C), which is classed as a utility token, isthe native token of the Turing-complete Ethereum platformwhich is governed by the Ethereum Foundation located inthe Crypto Valley. The token itself is unregulated. Althoughmultiple decentralised systems which can act as a quantitativeoracle interface for the platform exist, there are no legal claimsand no underlyings associated with the token. Consensus onthe Ethereum platform is, at the time of writing, achievedbased on the proof-of-work mechanism, and therefore is ofa probabilistic nature. As a consequence, the governance ofthe token is decentralised. Like with bitcoin, the issuance ofEther tokens is conditional on the creation of new blocks, i.e.,when miners get awarded with newly mined units, and thedestruction of existing units is not possible. However, currentlythe total supply of Ether is not limited. All Ether tokens aretransferable between parties and are fungible.
C. Crowdlitoken
Crowdlitokens (see Appendix D) are classed as asset tokensand are tokenised real estate bonds, regulated under theexisting law. They are issued on the Ethereum Blockchainunder the ERC-20 standard and represent a contract includingfixed claims (e.g., voting and interest payment). The tokenvalue is derived from the fundamental value of the issuingcompany, and only indirectly by its real estate portfolio. Due tothe underlying distributed ledger technology, consensus on thestate of the tokens is not final but only probabilistic. Crowdl-itokens are structured as notes/bonds. They are governed in acentralised manner through a qualitative oracle interface sincetoken holders are allowed to vote on changes proposed by themanagement. They can be issued and burnt (e.g., through tokenbuybacks) flexibly by the corresponding company, implying aflexible token supply. The Crowdlitoken is both transferableand fungible, whereby only persons who have successfullycompleted the KYC/AML audits can subscribe to the bondsand exercise all rights relating to them.
D. CryptoKitties
CryptoKitties (see Appendix E), as the last example fromthe crypto space, are collectible digital representations of catscreated on the Ethereum blockchain. The corresponding smartcontracts can generate over four billion variations of phe-notypes and genotypes (CryptoKitties, 2019). CryptoKittiesneither represent claims against a counterparty, nor a specificunderlying. They are non-fungible - every cat is unique - buttransferable ERC-721 tokens, without any regulatory or legalgovernance. Although the front-end as a traditional web appis managed by the development team, the token’s governance,e.g., ownership, is decentralised. Since consensus of the un-derlying Ethereum protocol is reached via a proof-of-workmechanism, the finality of the state of a CryptoKitties tokenis probabilistic. Also, there is no oracle interface related toCryptoKitties tokens. The creation of additional units is doneby breeding two CryptoKitties, resulting in a new unique kitty,represented by a newly issued unique token, while destroying a unit is not possible. The corresponding smart contract allowsfor a total limit of around four billion cats that can be bred,implying a fixed total supply.
E. Traditional Share
Traditional shares (see Appendix F), as the one example fromtraditional finance, are either physical or digital in nature andrepresent a contract including fixed claims (e.g., voting and/orprofit participation) against a counterparty, with its fundamen-tal value also representing the underlying of the asset. Shares,as a legal form, are governed in a centralised manner andare subject to the existing law (e.g., national corporate law),with the general assembly of shareholders being the supremeorgan of a stock corporation, i.e., acting as a qualitative oracleinterface. Consensus on the state of a share is deterministicallygiven by the share registry. The creation of new shares as wellas the reduction in share capital, for example through sharebuybacks, is left to the general assembly of the corporation.As a consequence, the total supply of traditional shares isflexible. Shares are typically transferable, with exceptions suchas restricted shares, and fungible, i.e., substitutable with othershares of the same company.V. C
ONCLUSION
Various classification frameworks for traditional and cryp-tographic assets already exist and are applied in practice.However, a universal approach linking the two worlds hasnot yet been developed. In this paper we fill this researchgap by proposing a taxonomy that extends existing classi-fication frameworks. We identify 14 different attributes thatare supported by the existing literature and by which eachtype of asset can be properly classified. These attributesinclude the claim structure, technology, underlying, consensus-/validation mechanism, legal status, governance, informationcomplexity, legal structure, information interface, total supply,issuance, redemption, transferability, and fungibility. With thehelp of a morphological box, various possible characteristicsthat an asset can have are identified and assigned to theseattributes. In this way, our taxonomy bridges the gap betweenphysical, digital, and cryptographic assets, where sometimesthe same asset can appear in all three forms, thus creatingclear terminology. Thanks to the methodical approach, theindividual attributes can be expanded or broken down at anylevel of detail without changing the overall framework. Theclassification of selected assets, such as cash and bitcoin,has also shown that the proposed taxonomy is applicable inpractice. In a next step, the robustness and practical relevanceof the taxonomy could be further tested, for example byinterviewing experts in the field.6
PPENDIX AA SSET TAXONOMY A PPENDIX BC LASSIFICATION OF CASH ( GREEN ) AND BITCOIN ( ORANGE )7 PPENDIX CC LASSIFICATION OF E THER A PPENDIX DC LASSIFICATION OF A C ROWDLITOKEN TOKEN PPENDIX EC LASSIFICATION OF A C RYPTO K ITTIES TOKEN A PPENDIX FC LASSIFICATION OF A TRADITIONAL SHARE EFERENCES
The Journal of Risk Finance
Avail-able at SSRN [cs.CY]