Updated: Mar 17, 2020
Authors: Brunini Gabriele & Montagnani Daniele
Following the initial release of Bitcoin in the 2009, the world of crypto assets has evolved in many directions, with many possible applications of Blockchain technology being found, with many under development, this constant evolution makes the field complex and at times confusing. In this paper we try to propose a classification of the assets that at the time of writing have a market capitalization above $100 million.We take two different approaches in doing so, one from an economic point of view and the other from a technical one.
We are the Bocconi Blockchain and Cryptocurrency Association. We foster a community-driven approach for the adoption of Blockchain and Cryptocurrencies. Our goal, as an association is to research the technology, to help other people learn about its mechanisms, potential application and challenges, and provide opportunities for discussion with experts and professionals in this broad new exciting industry.This paper is written by members of the BBCA Research Team, whose aim is to help and provide guidance and advice for companies interested in the foundations and further implementation of the technology in their operations and services.
1. Economic Classification
By addressing two main economical/logical questions, we classified the crypto assets into their natural economic grouping. The main questions we will address are:
1) What would you use this crypto asset for?
2) What is the rationale for possessing such a crypto asset? Based on the famous book written by “Chris Burniske” and “Jack Tatar”, “Crypto assets: The Innovative Investor's Guide to Bitcoin and Beyond”, from an economic point of view we can broadly divide crypto assets in three classes: cryptocurrencies, crypto commodities and crypto tokens.
1.1 Crypto assets categories
A crypto currency is designed to work as a means of exchange, a store of value and a unit of account. The usability of a crypto currency depends on how well it serves to replicate a traditional currency’s main uses, and it heavily depends on how much the blockchain network is used (e.g. number of daily transactions, which is a good approximation of how many users are using it), and its recognition as a good investment, independently on its length. Examples of crypto currencies are Bitcoin or Ripple’s XRP. “Currency” is the most common asset class among cryptos, and that is because the blockchain is an emerging technology and it is not entirely clear what the possible future uses for the technology could be.
When referring to commodities in terms of crypto assets, we are mainly considering the ones used in computation, storage and bandwidth. A user of the blockchain network can complete actions only by paying with commodities. Take for instance Neo, Cardano and Qtum, which are all crypto commodities that are used to pay the network fees of their respective Blockchains.
1.1.3 Crypto Tokens
Beyond currencies and commodities, on a blockchain we can also find crypto tokens, which broadly define finished products/services. Much like in the real world where currency and commodities support the result of the equation, generally a product or a service. The term token can be a bit ambiguous when referring to blockchain; for example, bitcoins can be called “crypto tokens” even by being simple currency, so the jargon here can be a bit confusing. To make things clear, here we will simply refer to crypto tokens as means for accessing D-Apps (Decentralized Apps) on the blockchain they are built for, and which are generally needed to access goods or services that can be in or out of the specific blockchain (e.g. online financial/investment assistance, facilitated exchange amongst blockchains).
Through an economic analysis based on the two main questions previously defined and conducted on the first 62 crypto assets (ordered by market capitalization), we categorized them in order to have a clearer understanding of the current industry situation.
The figure below (Fig1) clearly explains the situation: out of the 62 crypto assets analysed, almost 50% of them are classified as cryptocurrencies.
That shows that, as the technology is still relatively new and new applications are still to be found. There are several potential uses found every day, and for this reason it is very important for everybody to be able to explain what a blockchain is and how it works in order to not to be left behind. Unfortunately, we are still far away from unlocking the true potential of the blockchain and most of the blockchains available today are only capable of keeping track of who owns what amount of cryptocurrency in his/her wallet.
1.2 Economic rationale
It is easy to see whether there’s a strong connection between mainstream crypto asset use and the economic reasoning for choosing to invest in it (the rationale to possess). The value of a crypto asset can be in fact divided in two parts: the speculative value is assigned to it when there’s a reason to believe that the market price is going to increase in the future, and consequently from buying the asset, the owner may be able to sell it and realise profits.
The other side of the crypto asset value is given by its utility value, which is assigned to it for its particular use, whether it gives some utility to the owner, such as paying for a product or a service. Indirectly, it doesn’t mean that the token is not generating profits for the owner, but it shouldn’t be something that people invest in based purely on speculation.
These two asset values determine the type of investments that are made by the owners, whether they are long-term investments or short-term ones. Other than that, the economic value of a crypto asset is indirectly affected by the technology that is behind its blockchain network, that is to say, the consensus mechanism used to verify if a certain transaction is considered valid or not. Today, there exist models that can assign direct economic valuation of a certain crypto asset by taking under consideration its consensus mechanism.
Crypto asset valuation has been a topic of interest for many crypto-users and lately many models have been invented in order to value crypto assets as precisely as possible. In the following chapter we will try first to give a broad definition of what a consensus mechanism is, and we will then explore the most used consensus mechanisms by using some of the 62 crypto assets we mentioned above.
2. Technological Classification
In order to extract a classification based on the technology behind crypto assets, some things should be clarified. Crypto asset ownership is registered on a distributed database, that from now on we will call “blockchain”. A blockchain is a specific kind of distributed database. Many assets live on different structures, for example: tangles, trees and distributed hash tables. We are going to make a simplification and refer to all such data structures as “blockchains”. The platform on which the asset grants services might be related to the blockchain or might perform completely different tasks and issue tokens registering their possession on a third party blockchain. One obvious example being ERC-20 tokens built upon Ethereum.
The major trade-off that a blockchain must face is that between efficiency and decentralization. Indeed, in order to perform better in terms of speed, scalability and efficiency, it could leverage some trust and thus sacrifice real decentralization.The architectural aspect where this trade-off takes place is the consensus mechanism, the algorithm that essentially extends honesty from the majority of the nodes to the entire system. This mechanism is of fundamental importance as it influences, through incentives, how nodes will behave. Eventually, nodes behaviour will decide some economic aspects of the network (for example centralization, velocity).
A classification based on the consensus algorithm is best suited to assets deeply related to their blockchain, since making changes would be much more complex than just using a different general-purpose platform as an entity issuing an ERC-20 token could do.We divided consensus mechanisms into 4 macro-classes containing different versions all taking their main design ideas from the class representative. These 4 classes are: proof of work (POW), proof of stake (POS), federated byzantine agreement (FBA) and byzantine fault tolerance (BFT).
2.1 Proof of work (Bitcoin)
Proof of work is the original consensus mechanism. It was first invented in order to avoid spam and was later used and modified by the mysterious Satoshi Nakamoto in order to sustain the Bitcoin Blockchain. What is the main idea behind such mechanism? The participants must find a hash value under a certain difficulty threshold, and because the hash function gives values in a non-predictable way, a suitable hash can be found only with brute force (i.e. guessing the nonce [“number only used once”] until the correct value is found). This is enough to prove that behind the last chain of block the greatest amount of computing power has been used. Hence, if the honest nodes in the network are able to maintain most of the CPU power, their blockchain will always be longer than the malicious ones and will be adopted by the network.
Proof of work comes with two main disadvantages:
In order to compute the correct hash value, miners need to use a lot of computing power which translates into tons of electricity being used. Hence, the entire architecture is inefficient.
As the certainty of a transaction can be obtained only after a certain number of blocks, transaction validation procedure requires some time which highly affects the speed of the network.
2.1.1 Tangle (Iota)
Iota functions on a structure called tangle. This structure is not technically a blockchain, but we consider it to be working on top of an algorithm that resemble a proof of work algorithm. The most important parameter is called cumulative weight and is defined as “the weight of a transaction plus all the weights of the transactions approving it”. The weight of a transaction is said to be “proportional to the amount of work that the issuing node invested into it”, and is assumed that “no entity can generate an abundance of transactions with ‘acceptable’ weights in a short period of time”.
2.2 Proof of stake (Stratis, Ardor)
Proof of stake aims at diminishing the energetical expense needed to maintain a blockchain based on the proof of work consensus mechanism. It assures nodes’ honest behaviour by assigning to each node an amount of probability of forging the next block proportional to their token holdings. Metaphorically speaking tokens act as small mining rigs, but it must be noted that this remains a metaphor, as no actual mining happens. This mechanism however incentivizes token saving and may undermine the network’s vitality. Some other assets have tried to improve the proof of stake mechanism while maintaining the underlying idea:
2.2.1 Leased Proof of stake (Waves)
In a simple proof of stake mechanism every node must participate in the “forging game” on its own, or at least, in order to build unified forging pool, tokens must be moved into unified accounts. This problem has been solved by “waves”, which uses a leased proof of stake consensus mechanism. In this network minor nodes can lease their forging rights to full nodes and as soon as the full node adds a block to the chain, the reward is split according to predefined rules.
2.2.2 Delegated proof of stake (EOS)
In the delegated proof of stake mechanism, token holders vote for block producers with their votes being weighted on the number of tokens they hold. The algorithm then proceeds to choose the most voted candidates (let’s call it “n”, which must be an odd number). These “n candidates” are now block producers. Each of them will forge one block in a predetermined order decided by the algorithm, and after n blocks being produced, the production order is randomly shuffled, and production can start again. At this point, in order to choose between eventual forks, the longest chain rule, similar to the one of bitcoin, is adopted. A block can only be on the chain, which will eventually turn out to be the fastest growing one, if it has been confirmed by more than 2/3 of the producers.
2.3 Federated Byzantine Agreement (Ripple, Stellar)
Some other platforms have tried to improve scalability, network speed and energy efficiency by simply relaxing the decentralization requirements. Indeed, the founders of this protocols aimed at leveraging some of the trust which already exist among business entities, to build a consensus protocol which extends to the whole network. The various algorithms differ both in assumptions and require different levels of trust overlap. The two leading platforms which adopt this kind of consensus protocol are Ripple and Stellar.
2.4 Delegated Byzantine Fault Tolerance (NEO)
In a blockchain which uses a delegated byzantine fault tolerance algorithm, not all nodes in the network participate in the validation of blocks. Instead, some “bookkeepers” called consensus nodes are elected by NEO holders and participate in the validation of blocks.Consensus nodes assume two roles in an alternating fashion: Speaker: a node which proposes blocks to the system, Delegate: a node that, together with other “delegates”, must reach consensus on the validity of the block proposed by the Speaker. If 2/3 of the Delegates are honest, we can extend honesty to the whole network by electing a new speaker if the one under consideration is found to be dishonest. The honesty of the speaker is asserted by having the nodes exchange the message they receive with each other.
In this large class we can put together some minor consensus mechanism we have found, that were only being adopted by a single asset, some of them are:
2.5.1 Proof of Storage (Siacoin)
Siacoin is a decentralized cloud storage platform which operates on the blockchain and allows nodes to be both hosts and clients. A host is a node which stores files on his hard drive, receiving compensation in the form of siacoins, while a client then borrows storage from a host and pays for it. The agreement between the parties takes the form of a self- enforcing contract on the blockchain.Each host must prove to the network (and to his clients) that it is storing the file. This is proven through the proof of storage mechanism, which asks the host to provide a randomly chosen segment of the file together with a list of the files’ Merkle tree hashes. If the host fails to provide such pieces of information, it incurs in a penalty, which must be agreed upon in advance.Proof of storage allows nodes to assess the existence of the file, but the contracts must be mined someway in a blockchain, and this activity requires the proof of work consensus mechanism.
2.5.2 Proof of Believability (IOStoken)
In the proof of believability mechanism, transactions are validated only by a subset of nodes called ‘believable league’. These nodes are chosen randomly with the probability distributed according to factors like balance and reputation. Transactions then are randomly assigned to believable blocks to be validated, and each transaction is verified by one block only. Because transactions are validated only once by a block that could potentially be malicious, transactions are to be checked by nodes in the normal league (those nodes who didn’t get elected as believable).The system specifies punishment for those believable nodes found misbehaving, consisting in the loss of their whole balance and reputation and this is judged to be a sufficient deterrent.
2.5.3 Proof of Authority (VeChain)
The company “VeChain” built its blockchain ecosystem, which is named “VechainThor” and uses another type of consensus mechanism, called proof of authority. Once again, this mechanism sacrifices real decentralization in order to achieve better scalability and diminish energy consumption. It allows only a predetermined number (101 to be precise) of nodes (called “Authority Masternodes”) to participate in the validation of new blocks. In order to become an Authority Masternode, an entity must make its identity and its reputation available and must go through a KYC (‘know your costumer’) procedure. This of course requires some trust in believing that the foundation will be able to appropriately select AMs so that they won’t collude and perform an attack.
We applied our technological framework to the 62 crypto assets that had at least 100 million USD market cap, and obtained the following findings:
In conclusion, we observed that Proof of Work (POW) remains the most widely adopted consensus mechanism at the moment, while proof of stake (with Ethereum planning to implement it in the years to come) is most likely going to have a major role in shaping the future of the industry.With respect to applications, currencies are the largest asset class, with most of them being built on top of protocols very similar to Bitcoin’s one, but with minor changes. Crypto commodities could become an efficient and fair means of allocating digital resources which will be of vital importance in the technology-dependant world of the future.
Chris Burniske, Jack Tatar. Cryptoassets: The Innovative Investor's Guide to Bitcoin and Beyond 2017.
Satoshi Nakamoto. Bitcoin: A Peer-to-Peer Electronic Cash System, 2008.
Serguei Popov. The Tangle, 2018.
Brad Chase Ethan, MacBrough. Analysis of the XRP Ledger ConsensusProtocol, 2018.
David Vorick Nebulous, Luke Champine. Sia: Simple DecentralizedStorage, 2014.
Ve Chain development plan and whitepaper, 2018
Internet of Services: The Next-generation, Secure, Highly ScalableEcosystem for Online Services, 2017
EOS.IO Technical White Paper v2, 2018
WAVES whitepaper, 2016
NEO White Paper