A castle under siege defeats the besiegers?

It sounds like an episode of Game of Thrones. But it’s not. It’s a retelling of what’s commonly referred to as the Byzantine Generals’ problem.

In its essence, the problem is one of unprovability. The four attacking captains can only communicate through a messenger. The messenger could be captured, or otherwise fail to deliver a message. Any one of the captains, including the one who initially proposes the time of the attack, could be a traitor. Any one could back out simply due to cowardice—or expediency.

And, while the messenger is carrying the message around, any captain can modify the original message and not inform the others. Also, the one who initially proposed the time of attack may back out later and not inform others.

To succeed, all four must come to a consensus based on trust: attack together successfully or demur, and watch the others fail. Which will they choose?

The Byzantine Generals problem highlights the importance of forging consensus when multiple stakeholders are collaborating, where there is no central authority anyone can trust. The stakeholders themselves have to decide to trust for each individual project to succeed.

Byzantines meet Bitcoin

Digital currency Bitcoin is a peer-to-peer system that records financial transactions between users directly, without an intermediary. Since the system works without a central repository or a single administrator, Bitcoin has been categorized as a decentralized virtual currency.

As of February 2015, the number of merchants accepting Bitcoin for products and services surpassed 100,000—signifying if nothing else an important psychological threshold for the currency’s wider adoption.

Key to this rapid adoption? Bitcoin’s users themselves create consensus on verifying the digital currency’s transactions—all the transactions that have ever happened. This verification was an important process in the system, as there was no central user or machine that could perform the job. Bitcoin’s users have enough motivation to do the job of verification themselves, since the peer-to-peer system encourages trust by rewarding with more of the virtual currency users who verify their transactions.

Banks and the benefits of consensus-based systems

Bitcoin’s decentralizationBitcoin’s decentralization eliminates concern about a single point-of-failure. If one system recording transactions fails, the others keep the system running. Furthermore, Bitcoin’s users no longer a need trust a central computer in a transaction.

The consequences are significant. Transactions can be completed faster and more cheaply. The benefits of leveraging such a model have attracted even the banking sector to start experimenting in Blockchain, the technology underlying Bitcoin.

Curiously, examples exist of standardized, consensus-based systems in the history of banking and capital markets. Today, more than ten thousand financial institutions across 200 countries use the messaging network SWIFT to send approximately 24 million messages daily. And financial markets across the world use DTCC, a post-trade firm, for clearing and settlement services.

Each system acts as an intermediary for banks to collaborate. Both carry sensitive information. Each thrives due to the associated trust factor: All banks are similarly dependent on each other for the system to work. But, with the good also comes bad: SWIFT’s software was compromised in the 2016 Bangladesh Bank heist, and the security lapse at one bank affected all.

Consensus without an intermediary

Global banks today face enormous security risks and regulatory hurdles. Add to that increasing pressures on profits, and the need to shed non-productive or non-compliant lines of business. The continuous search for cost efficiencies and the need for digital investments to streamline business processes are prompting major Wall Street banks to collaborate with each other directly—to save money while meeting regulatory needs.

Interestingly, in 2015, J.P. Morgan Chase, Goldman Sachs and Morgan Stanley contemplated forming an entity dubbed SPReD (Securities Product Reference Data) that would pull together and clean reams of reference data at a lower cost than what each institution would spend individually.

Banks are also considering Blockchain, which has the potential to replace many central systems. They’re betting that the promise of reduced cost, increased transparency, enhanced reliability, and a better customer experience outweigh the loss of their intermediary status.

For example, R3 CEV, a financial technology (FinTech) startup that works on Blockchain technology, had established the R3 consortium with nine member banks. Within seven months of operation, an additional 36 member banks came on board. R3 CEV has already conducted two trials: one involving eleven banks and another involving forty, in which banks collaborated with each other using Blockchain.

Consensus vs. Intermediaries

Consensus is the backbone of Blockchain. Algorithms like Proof-of-Work, Proof-of-Stake, Proof-of-Burn, Practical Byzantine Fault Tolerance (PBFT) etc., and alternative protocols like Ethereum, Ripple, Stellar and HyperLedger have all tried to solve the problem of forging a consensus-based peer-to-peer system, overcoming counter-party risk while allowing faster, more cost-effective transactions.

Collectively, all these efforts illustrate the importance of the concept of building such a system. But while consensus has been solved technically by Blockchain, the challenge lies in ensuring the involvement—and trust—of the businesses for whom it is being designed. The future of Blockchain in financial services will depend on whether banks come together without an intermediary.

That’s something a Byzantine General would appreciate! Is blockchain in your future? How have you built trust to forge organizational consensus? I’m interested in knowing. 

Pery Venkat

Pery Venkat

Venkat has 17 years of experience in the IT industry and is a Senior Director managing Cognizant's business strategy, digital, M&A and... Read more