Inside Multichain: A Build-Your-Own Blockchain Service for Banks

"Blockchain technology" may be the buzzword of the moment in the FinTech world, but like any blanket term, actual definitions vary.

Rather than any one solution, it may be beneficial to think of blockchain technology as a loosely defined set of goals financial institutions are hoping to unlock through distributed ledgers. The transition finds these entities seeking to define the attributes of the bitcoin blockchain they prefer – its secure ledger and payments rail – and the ones they don't – its controversy-grabbing native asset and largely anonymous miners, while determining solutions that allow them to further define these needs.

Entrepreneurs are now stepping in to facilitate this education process, attempting to serve a clientele that is unique from that of the industry surrounding the bitcoin blockchain. Among them are private blockchain firms such as Blockstack, Eris Ltd and Coin Sciences.

Led by CEO and founder Gideon Greenspan, Coin Sciences recently launched a new solution this summer called the Multichain Private Blockchain. The product takes what could be described as a build-your-own blockchain approach that aims to free banks from the more rigid options of competing offerings.

Greenspan believes there's potential profit in allowing banks to experiment with the technology while gleaning insights into their observations, telling CoinDesk:

"It became clear that what [banks] were doing is grabbing their best developers, pointing them at the source code for Ethereum and Bitcoin Core and telling them to start tinkering with it... We saw the opportunity to create an off-the-shelf platform where they don't need to spend a year breaking their heads on this stuff."

Multichain is the latest offering from Coin Sciences, following CoinSpark, a software tool for web assets and legal contracts on the bitcoin blockchain. The problem, Greenspan alleges, wasn't that financial institutions weren't interested in this solution, or removing intermediaries from financial processes, but rather they didn't believe the bitcoin blockchain was tested or stable enough for their needs.

"It became clear people are looking for things that are similar to bitcoin but different, rather than cryptocurrencies, traditional entities are more interested in the ledger," Greenspan said.

In this light, Multichain emphasizes end-user choice, allowing customers to control whether the chain is private or public; the target time for blocks; who can connect to the network; how these entities interact; and the maximum block size and metadata that can be included in transactions, among other features.

Moreover, Greenspan argues the offering is proving popular, having been downloaded nearly 850 times from mid-June to mid-Septemberr. He declined to mention any large financial clients that may be using the software.

The alpha version of the platform is currently available, with a beta to be released in late 2015.

'Bitcoin's shortcomings'

As in the Multichain white paper, Greenspan is keen in conversation to elaborate on why his firm believes bitcoin is ill-suited for the needs of financial conglomerates, while steering clear of suggesting the bitcoin blockchain is not a needed innovation.

The paper, for example, seeks to note that bitcoin currently supports only 300,000 transactions per day due to the current limit on how many megabytes worth of data can be processed on the network every 10 minutes. Further, it argues that bitcoin's fees – roughly 2.5 cents for a transaction at the time of publication – are set to rise as the number of new bitcoins produced to incentivize miners for transaction processing decreases, both factors that simply don't appeal to institutional users.

Complicating matters, Greenspan argues, is that the blockchain is designed to move bitcoins, meaning that it can only effectively exchange a record of the asset, not the asset itself.

"That conceptual idea ... doesn't carry over to other assets. Unless another asset has been issued, like dollars onto a blockchain, you're never going to be directly an asset. You're always going to be owning a promise that that assets exists," Greenspan said.

Notably, startups such as Symbiont are developing technology to represent asset ownership on the blockchain, though financial giants such as Nasdaq have offered a differing view. In a recent interview, Fredrik Voss, Nasdaq's head of blockchain strategy, for instance, argued that bitcoin representations of assets are still beneficial due to the tracking benefits shared ledgers provide.

Multichain, however, believes its workaround will find a market. As described in the white paper, the technology aims to make tracking blockchain assets easier using bitcoin's scripting language to encode metadata in transaction outputs, which it argues allows miners to verify the quantity of an asset that differs from the network's native token in any transactions.

Customizable permissions

Like other private blockchain solutions, Multichain aims to remove perceived problems associated with bitcoin by limiting the visibility of the ledger to certain participants, allowing institutions to set controls on transactions permitted and by forgoing distributed mining.

On the subject of privacy, Multichain allows users to set a list of permitted users that can act as nodes that refer information on the network and 'miners' that verify transactions, including a method by which nodes can verify whether other nodes have been approved.

Privileges are granted using transactions with special metadata, and the miner of the Genesis block, the first block on the blockchain, is given all established rights on the network and acts an 'administrator'. This administrator can then appoint other administrators on the network, with any changes to consensus needed to be validated by a set number of participants.

Permissions can also be grated on a limited basis, according to the white paper, or to a fixed range of block numbers.

This differs from the current consensus strategy in bitcoin, in which a distributed community of developers must introduce changes that then require a majority or all of those processing transactions on the network to upgrade to new versions of the software.

Though often not difficult technically, the challenge inherent in this approach was perhaps best exemplified by the ongoing debate on the size of blocks on the bitcoin network. Differing proposals have been introduced as far back as this summer, though no consensus on solutions has so far been reached.

Proof of work alternative

As miners on the multichain network do not need to conduct proof of work, Multichain has also introduced a novel way by which these entities can trust decisionmaking.

Called mining diversity, the process enables miners processing transactions to approve transactions in a random rotation. Greenspan explained a situation in which a Multichain user set a blockchain network with 10 mining entities, and set a "relatively high" mining difficulty by which the approval eight of these entities would be needed for a block to be recorded.

"That means they have to wait eight more blocks before they sign another one. Having that, what it means is there's no possibility for a small group of miners to mine a chain in private, unless eight of them do it," he said.

Greenspan argues that this structure allows more miners to participate in the approval of transactions, while ensuring there is no fixed order of verification that could be corrupted.

Transaction fees and block rewards are set to zero by default in Multichain, though this is also customizable, as the white paper suggests participants could be charged a yearly service fee.

Still, Greenspan believes that version of a private blockchain offers benefits over a centralized database, a claim often espoused by critics, contending that the benefits are that each participant retains full control over assets; control of the database is distributed across entities; and that, due to the mining diversity concept, the loss of any one server would not compromise the network.

"This is a new type of database that enables multiple parties to share a database and to be able to modify that in a safe and secure way even if they don't trust each other. I think that's a very valid direction and new types of databases can be an extremely powerful thing."

Missing link

Multichain users also aren't limited to the number of blockchains they create, and further, they have the ability to connect their blockchains to the bitcoin network or to a bitcoin testnet, a version of the blockchain on which no real bitcoins move.

Interoperability, however, is so far limited to bitcoin, as Greenspan said Multichain cannot connect to Ripple or Ethereum, which he said were "in a separate world from Multichain".

"Bitcoin's model is hugely more scalable," Greenspan continued. "You can run many transactions in parallel as long as you can see they're not modifying the same data."

In contrast to these alternatives, Greenspan still sees bitcoin as a preferred technology, basing Multichain off of a fork of Bitcoin Core, using the protocol and architecture as well as its command-line and API interface.

In this way, the Multichain white paper positions private blockchains as an early version of a more robust, global bitcoin blockchain, one that can allow private institutions to begin working with the technology in the way that Intranets exposed entities to the Internet.

"Private blockchains will likely be a more attractive solution for financial institutions wishing to deploy this technology during the next decade," the paper reads, adding:

"Twenty years from now, if bitcoin or another blockchain is processing billions of transactions monthly at very low cost, with mining controlled by large identifiable corporations, bitcoin may start looking like an attractive platform for institutional financial transactions."

Assembly image via Shutterstock