Blockchain Technology and Institutional Finance

This is the first in a series of non-technical blog posts we will publish in the coming months, which will help users understand how Symbiont's technology works and why we made design choices that make Symbiont different than competing platforms and a better choice for many uses in institutional financial markets. We're built from the ground up for institutional finance. Smart Securities®, Intelligent Markets!

With interest in blockchain technology gripping the institutional financial sector today, there is a great deal of confusion about the true nature of the technology, and specifically how it is different from what has come before. The promise to dramatically improve financial infrastructure is well established, but it’s often not clear why the benefits associated with blockchain technology cannot be had with traditional technology.

The problem that the financial sector hopes to solve by using blockchain technology is that of duplication and reconciliation of administrative processes and data. To illustrate, standard practice in the syndicated loan market involves a dozen banks receiving the same fax, each entering the same data contained on that fax into their internal ledgers, and then reconciling any discrepancies which arise. Blockchain technology solves the duplication-and-reconciliation problem by creating a distributed ledger: a single record of financial transactions that may be shared among multiple institutions. The adoption of a distributed ledger can solve very real, very great inefficiencies that financial institutions face in the coordination of all of their back-office systems and associated operational costs.

The naïve solution to this problem is to appoint a single entity to maintain a master set of records for the entire market to use.  However, this third party must be trusted to keep all records accurately, securely and (if applicable) confidentially. A blockchain, by contrast, allows multiple parties to share records in a manner that all can trust as valid, simply by the design of the protocol.

But if the problem is solved by allowing parties to share a master copy of data, why not use just a distributed database instead of a distributed ledger? The majority of the most popular SQL and NoSQL databases can function as distributed databases, in fact... Problem solved? No.

A distributed database is a type of database that shares its data among multiple computers. We want a database that can share information among multiple institutions. That’s something else entirely. In fact, no existing database was designed to do this. Even the most scalable, globally distributed databases are all designed to be hosted by a single party, and this assumption is baked into their design at a very low level. You can’t just take an Oracle database, give each bank a node and call it a day. The security model is totally different. Everyone would be able to see and change everyone else’s data, as well as bring down the whole network on a whim.

For a distributed ledger to work what you really need is not only a distributed system but also a decentralized system: one that multiple, independent parties can use to share information. There are plenty of decentralized systems out there, though you may not think of them as such. Faxes are a great example of a decentralized system. Email, the Web, BitTorrent, and Git are others. The problem with any of these systems, though, is that everyone’s data is inconsistent, based on which messages or files each particular user happened to send and receive. What you really need is something like Google Docs, but without Google. And for finance!

There are two crucial pieces to building a distributed ledger: decentralization and a single source of truth.1 Blockchain technology meets both of these requirements, and no other kind of software does.

What blockchain technology allows for, beyond a distributed ledger, is nothing less than a fundamentally new class of computer systems: decentralized like email, but with a single source of truth like a traditional database. Note that so-called “permissioned blockchains” (such as the one Symbiont is building) are networks in which the identities of all of the participants are known to each other. There exists a more general kind of decentralized system which additionally allows for anonymous network participation (such as Bitcoin). Because of the stronger assumptions around the identity of network constituents, permissioned networks can be faster, scalable and more reliable than permissionless ones.

Blockchains aren’t magic; they aren’t a panacea. There are many problems which blockchain technology cannot solve, and many places where other technologies are more appropriate. The fact is that many companies in the “blockchain technology” space are building either 1) decentralized systems with no single source of truth, or 2) a centralized systems that rely on trusted third-parties to validate network events. In the case that either of these systems is sufficient, there exist well-established software solutions that will work better (and faster!) than any blockchain. Nevertheless, blockchain technology represents a real innovation in computer engineering. And, in the right circumstances, it can radically improve transparency, efficiency and security of the very core of financial markets infrastructure.

  1. The idea for this definition is taken shamelessly from an excellent blog post by Albert Wegner.