Given the governance exercise happening in Montreal on September 12 and 13, see here, I thought it timely to discuss how bitcoin is governed.
As a pure p2p network, the bitcoin blockchain is based on these two fundamental tenets – amongst others:
– There is no centralized agent to rule over the governance of the blockchain.
– A significant majority of peers always need to use the same protocol and technology for the blockchain to remain one.
The recent debate over block size shows it is not easy for a network to be purely decentralized and to prosecute change in an orderly fashion. The implications surrounding block size and transaction size are fascinating:
– Who is best placed to decide what needs to be changed when?
– How are decisions being made?
– BitcoinXT or not?
– Fork or no fork?
– How to propagate a change?
– What will happen to the overall performance of the blockchain if block size increases are adopted – speed of transaction validation, cost to remain a miner?
– Will there be more miner concentration or less as a result? Is more miner concentration good or not?
– Will transaction fees rise or not?
From a pure bitcoin-centric point of view the current process may be convoluted and may lead to forks, but the externalities are somewhat manageable. A blockchain that is inwards focused and used by crypto enthusiasts, for crypto enthusiasts, mainly interacting between themselves can survive pretty much anything, even balkanization.
From the point of view of a platform that has aspirations to be an integral part of the financial services industry, the process as it is laid out and managed will eventually cripple any shred of credibility and make it a non starter for financial services companies to interact and work with it.
Why do I make such a stark prediction? For two reasons.
Macro reason: Any startup, company, network, marketplace needs the right type of governance which will deliver order instead of chaos, orderly direction instead of a multitude of view points and rules instead of anarchy. Too much or too little governance kills the golden goose – too much leads to bureaucracy, too little leads to destructive behaviors, too much leads to the Soviet Union, too little leads to Enron. Every successful platform/network/marketplace/entreprise architected the right governance that supervised the right business model. Think of eBay or Facebook to name but a few recent and disparate examples and at the root you will find the “right” governance model. This is not a choice, it is a requirement. This is not a bug, it is a feature. Success depends on good governance.
Micro Reason: Basel I, II & III as well as national banking regulatory initiatives have made sure that banks – similar reasoning and regulatory initiatives apply to asset managers, broker dealers, insurance companies – view their risks holistically. One of these risks is that of vendor management. Banks go to extreme pain in choosing, vetting, approving and then reviewing and managing vendors. Vendors need to show maturity, sustainability, resilience, foresight. They need to show they are credible, have the right processes around software development, product management, financial health. Erratic and unprofessional behaviors as well as out of control management is a killer. Without the right governance one will find it extremely difficult to become a vendor to banks.
Few questions come to mind which I do not have an answer to:
1) Can a purely decentralized network devise and abide by a proper governance framework?
2) Can a p2p architected technology propagate technology change?
If the answer to both of the above questions is no, then the bitcoin blockchain in its entirety is faced with a choice:
Remain as it currently is, face future potential balkanization and have limited links to the financial services industry.
Fundamentally reinvent itself, shed its “decentralized” nature, turn into a proper platform with the right amount of centralization – probably killing a few sacred cows in the process – in the hope of becoming a widely used protocol.
Which is more likely?