Archiv für den Monat: August 2014

Blockchain 2.0 – viability and applications

Bitcoin has great potential but is in fact only one way to make use of the underlying consensus and cryptographic technology. The rather ambiguous term “Bitcoin 2.0” is regularly used to account for all those projects which apply the innovation pioneered by Bitcoin to use-cases that go beyond transferring money (i.e. bitcoins in this case). People rightfully pointed out that the term „Bitcoin 2.0″ is misguiding as it indicates a second version of the Bitcoin ptotocol. I agree with this. I therefore changed the title to „Blockchain 2.0″. What follows is a basic overview aimed at anyone with a basic understanding of Bitcoin.

Proponents of Bitcoin 2.0 projects suggest that decentralized transaction ledgers – Bitcoin essentially is a decentralized transaction ledger – offer a variety of technological advantages1:

  1. Resilience against (a) a deliberate or accidental shutdown and (b) censorship by a single party. As a result users are guaranteed the accessibility of services for current and future use.

  2. The possibility for corruption is reduced because (a) users themselves are in control of their funds and transactions are enforced without a centralized, potentially corruptible intermediary. Additionally (b) all transactionsincluding the sender’s and the receiver’s addresses, which might be tied to a real world identity/organization or not, as well as the amount of tokens (e.g. bitcoins) that were sent are publicly auditable. Consequently a decentralized and public transaction ledger can help an organization to enhance trust in the service provided and increase accountability of their bookkeeping.

  3. The absence of counterparty risk, due to the relative irreversibility of transactions, reduces costs directly (credit card fraud) and indirectly (regulatory costs).

  4. No (legal) ambiguity. The open source software run by participants of a decentralized network unambiguously defines the service any user of the network can expect. The explicit definition of the service does not leave room for ambiguity, thereby reducing the risk of costly legal proceedings.

  5. Anonymity for users (to a certain degree).

There also are some obvious downsides associated with decentralized transaction ledgers:

  1. Security and network costs. Compared to a centralized solution it is relatively costly to securely maintain a unified transaction ledger in a decentralized manner.2

  2. Dependence on the network’s security. The network’s security can fail. In the long run, a consensus network is less likely to fail due to a flaw in the underlying cryptography or software but rather because of a potentially flawed incentive structure for those securing the network collectively, which, together with economies of scale, can lead to a centralization of the control over the network. The result can be that one or a few parties control the network. It is therefore fair to argue that counterparty risk is replaced by the risk that the incentive model of a consensus network fails.

  3. Lack of flexibility. The possibilities to provide a service solely based on a decentralized transaction ledger, if not combined with other solutions (e.g. external data feeds or prediction markets), is limited to those services which can be entirely reduced to conditional rules and can therefore be encoded into software and enforced by a decentralized network of software instances.

  4. Complexity and usability for the end-user. Keeping your funds safe requires a few precautious measures. If a user loses his private keys there is no way to restore his funds.

Note that this list of applies to public blockchain networks. A private blockchain network for examples does hot have increased security and network costs (downside #1 above).

Taking account of these propositions it is possible to identify potential applications particularly suitable for utilizing decentralized transaction ledgers. Some of these are being implemented or at least theorized about by various Blockchain 2.0 projects. I will cover two basic use cases below which also seeks to illustrate the rather abstract description above.

First, the creation of synthesized assets: It is possible to issue customized tokens that are uniquely identifiable (also known as “colored coins”) without creating a separate blockchain. Those customized “assets” can be used to track ownership of anything a trusted issuer attributes to them. Examples include shares in a company, tickets that grant entrance to public transportation, bonds and other financial certificates etc. These tokens can be traded peer-to-peer like Bitcoin. All the projects below do or plan to allow for this functionality.

Second, the gambling industry. A possible, simplified scenario: Digital lottery tickets are issued that come with a certain chance of winning a growing price pool. Contrary to a centralized gambling company the random number generation which determines the winner(s) as well as the ledger that keeps track of who bought how many lottery tickets is decentralized. The advantage over a traditional lottery company lies in the system’s high resistance to corruption and a potentially low ‘house’ edge. 

There are four Blockchain 2.0 projects which are suitable to introduce different types of approaches, each having a focus on different use cases and/or a different technological approach towards delivery of “decentralized services”. This selection is primarily based on the educational purpose of this paper and will introduce three categories of approaches: Utilizing the Bitcoin blockchain for specific applications (Mastercoin and Couterparty), providing one dedicated blockchain for various applications (Ethereum) and relying on one dedicated blockchain per application (Bitshares). There are other innovative projects like NXT3, Ripple4 (privately secured transaction ledger mostly utilized to transfer IOUs between trusted gateways) and Truthcoin5 (a proposal for a mechanism to get reliable real-world data into a blockchain which can be used as a basis for prediction markets; not deployed yet) which are not described here in detail.

Counterparty and Mastercoin both allow for the issuance of colored coins (see above) and both run “on top” of the Bitcoin Blockchain, thereby making use of the security6 provided by the Bitcoin hashing network. Both projects allow for synthesized assets, financial derivatives such as contracts for difference and distributed betting between peers – the last two feature are based on price feeds. 

The other two projects, Bitshares and Ethereum, have a wider scope, as they both try to establish a whole ecosystem of applications running on their platforms. The most apparent difference between the two lies in the number of blockchains and the security model.

Ethereum, in its current implementation, is developing a proof-of-work based blockchain, which is intended to represent a base layer on top of which applications and contracts are to be written with relative ease using one of multiple built-in scripting languages. In contrast, the Bitshares project provides a software toolkit allowing for the creation of applications, with each application based on a separate proof of stake7 blockchain aiming at maximum scalability.

In conclusion, financial and commercial applications of peer-to-peer technology go far beyond Bitcoin. They hold the potential to redefine the relations between individuals and companies and challenge our assumptions about what constitutes an economic entity.

The article above should be seen as a basic introduction. A more detailed discussion is necessary to sufficiently address questions as: which types of applications are most suitable to make use of this new technology, a more detailed comparison and valuation of the different approaches (the described projects above), security issues of consensus networks etc. 

1Please note that the advantages 1 (a) and (b), 2 (a) and 3 hinge on the security model of decentralized transaction ledgers and the decentralization of ‘voting power’. The two most wide spread solutions to secure a decentralized transaction ledger are “proof of work” (POW) and “proof of stake” (POS). In each case the goal is to decide on one unified version of the transaction ledger without a central authority. As there is no central party to decide, because everyone can participate in the decision process over which ledger is the valid one and because one vote per participant would result in an arm’s race to create as many “virtual participants” as possible, there is a need to define a valuable resource which determines the voting power of each participant. With POW voting power equals computational power, while with proof-of-stake voting power is proportional to the amount of stake (share of tokens that are native to the network) a voter controls. In both systems one voter is selected every [x] minutes/seconds to add a new block to the blockchain, that means this voter can extend the transaction ledger by adding transactions that were newly broadcasted by users. The chance of becoming that ‚one voter‘ depends on the share of the valuable resource (computational power or stake) one has compared to other „voters“. The one ‚winning‘ voter is rewarded with transaction fees and/or newly created tokens. Any of the two alternative models above fail to provide security if voting power gets too centralized. In the extreme case where one party gets more than 50% of the voting power for a prolonged period of time this party can change the parts of the transaction ledger that have been added since the voting power majority was reached. This is because the longest blockchain is always considered to be the valid one by all network participants and anyone with 50 + x % of the voting power has a higher chance of producing a longer chain over time than the rest of the voters. An attacker could therefore build a secret chain which does not include the attacker’s transaction to a merchant and not broadcast it to the network, while the same transaction is publicly included in the chain that is considered to be official at a time where the attacker did not yet publish the secret (longer) chain. When the attacker publishes the, until then, hidden blockchain the network will accept it (if no manual/social consensus based intervention takes place) as the longest blockchain and the merchant has been defrauded of the Bitcoins he/she though to have received. In conclusion: In order to maintain a unified transaction ledger without a trusted third party it is necessary to maintain decentralized control over the ledger.

2 In addition to proof of work, other more ‘efficient‘ (when efficiency is defined as the cost of security) but yet to be (further) tested proposals have been made such as POS (Peercoin), POS (NXT), DPOS (Bitshares) and CPOS (Stephen Reed).

3 http://wiki.nxtcrypto.org/wiki/Whitepaper:Nxt.

4 https://ripple.com/ripple-gateways.pdf.

5 https://github.com/psztorc/Truthcoin/tree/master/docs.

6 See footnote 1.

7 See footnote 1.