When developers attempt to change the rules of a blockchain – such as the value of a new block of mined crypto-currency, or who is allowed to access data, or how a smart contract operates – the fallout can be unpredictable. Extreme cases can, and have, led to ‘forks’, and the act of splitting a blockchain in two can be disruptive.
As more financial institutions explore moving operations, recordkeeping and even trading to blockchains, forks are likely to become events to avoid. Although forks should have minimal to zero impact on a well-governed blockchain, it’s helpful to understand why they occur.
Above all, you want to be sure such events don’t touch the operations and integrity of your business.
Blockchain and other types of distributed-ledger technology (DLT) are decentralized and therefore hard to change. Bugs need fixing, systems need upgrading, conditions need amending. All require consensus, validation by a majority of ‘nodes’ (the computers of block miners, app developers and end users) on the network.
This means that even if a bank develops its own blockchain, the nature of DLT means it gives away control. The network’s consensus mechanism is in charge, and anyone can (and probably will!) propose a change.
The difficulty of consensus
This sort of thing happens all the time, and most changes are designed to avoid disruption, assuming there exist reasonable governance regarding the consensus rules (the voting process, the terms under a private blockchain’s articles of association). Nonetheless, blockchains that exist for enterprises, as opposed to those which simply mint and trade Bitcoin and other crypto-currencies, are subject to the same processes of change.
Under Bitcoin’s consensus system, a block – a history of transactions – is added to the chain every ten minutes, each one referencing the block that immediately preceded it.
New blocks must be validated by the majority of nodes on the network; if the data is confirmed, the block gets added to everyone’s database. To upgrade a network – if, for example, a miner wants the next block to be worth BTC50 instead of the current BTC12.5 – would require changing the rules so that new blocks, operating under new parameters, are accepted.
Therefore adding new players to the blockchain can change the economics, says Leonhard Weese, president of the Bitcoin Association of Hong Kong, from a talk he gave to members. “Whoever gives permission has control over the network and could force new rules.”
This is where forks come in to play. A software upgrade released by anyone involved in a blockchain – a miner, an app developer, even a user – creates a ‘software fork’. These things happen all the time and most are routine and harmless. Many are compatible and “boring”, says Weese, because it is usually in everyone’s interest to have the same apps running on their system. Ideas that are popular will get adapted; unpopular ones wither, and life goes on.
But sometimes a schism arises because the miners, developers and users of a blockchain fundamentally disagree on a proposed change. In the worst case, the system forks and two new ecosystems emerge.
“In a split, you get two competing blocks both referencing the same preceding block,” Weese said. “You get two chains running in parallel. Every bitcoin in the original chain is now spendable in both new chains. So it means that people can spend their old bitcoins twice.”
This is obviously a problem for the integrity of the system, and a fork therefore sets off civil wars among the community, as players attempt to game the system in their favor. There is no regulator to prevent front-running, spoofing and all other manner of market manipulation.
Changes that reject old formats effectively turn their back on original blocks, and demand the network shift to the new rules or risk a hard fork, in which competing and mutually exclusive chains exist.
This creates problems. Imagine a hot wallet that is holding investors’ tokens. A hard fork ensues, and those existing tokens suddenly can be spent in their original form – as well as in the new form.
This actually happened in 2016 with Ethereum, in which a dispute saw the community split between regular ether and the new Ether Classic.
Weese said: “This was the first time the question arose about custody of Ether…as a custodian, was I obligated to give [investors] back your Ether, or both your Ether and your Classic?”
Hard fork, soft fork
Forks and the turmoil they can unleash are one reason why the value of Bitcoin and other alternative currencies has been so volatile – which is a hindrance to its broader acceptance, and a brake on liquidity.
Not all forks are the same. The Ethereum split was a ‘hard’ fork, meaning the proposed change meant current nodes would no longer recognize the old blocks. This is an example of a change that rejects old formats.
Another way to put it: a hard fork broadens the rules, with proposals that let you do new things – such as increase the value of newly minted blocks, or change permissions, or redefine smart contracts. To enable the new rules means turning your back on how things were done previously.
The alternative is a soft fork, which has a softer ring to it. But soft forks are troublesome too.
Soft forks tight or restrict the existing rules. Current nodes will accept new blocks, and new nodes (new network players) may recognize both old and new blocks (depending on how the rule change is framed).
This sounds very amicable. But, Weese says, it’s actually “coercive”. Once a change wins consensus under such changes, the minority has no leverage to oppose it. Under a hard fork, the minority can carry on the old rules by sacrificing membership in the broader network.
Under a soft fork, Weese says, the consensus becomes a tyranny, and new blocks created under the old rules are likely to lose some or all of their market value, because no one will validate them: exchanges won’t trade them and investors won’t buy them. Yet the miners are expending a lot in electricity costs to mine new blocks, so real money is at stake.
Good citizens vs. malign meddlers
Proposals that lead to forks are often benign, intended by a developer to improve the system, or by a miner to increase the value of what they produce. In theory, users (exchanges, investors, enterprises and anyone with a node) can also propose changes. Thousands of such changes have been made without controversy.
But sometimes a proposal hits a nerve, threatens someone’s economic interest, or challenges political goals. And then the fights, though geeky, get nasty. And then the risk of a fork arises, with civil wars breaking out based on raw financial incentives versus the damage from a split.
The community is learning how to avoid the messiest or least responsible ways to initiate changes to Bitcoin or other alt-currencies. A little bit of care can ensure changes don’t lead to confusion that can cause tokens to disappear from people’s accounts – which has happened in the past, but at a time when the community was small and a bitcoin wasn’t worth a dollar. Today, however, alt-coins' market cap is $50 billion, with bitcoins trading at over $1,900.
Changes can also be proposed with high thresholds of consensus and conditional new rules, so that soft forks are rendered harmless.
Some people in the community also would prefer new rule changes not only require new blocks be invalid for old nodes (i.e., for holdouts to the change), but also insist that new nodes cannot recognize old blocks. This ensures a safe hard fork: you’re either with the old currency, or you’re with the new one, and your coins cannot be spent twice.
It also protects against the ambiguity of two rival systems that can be accepted by both old and new nodes – in which case, inexorably, the blocks with the longer string of code continue to attract patronage and those blocks with the shorter string of code die off and the coins lose all value.
Does a Bitcoin fork matter?
These esoteric rules in the computer-science world could have real-world spillover: there is a proposed rule change to Bitcoin doing the rounds that would double the size of newly minted blocks. This is meant to increase supply, to accommodate the expanding community of users, and therefore increase liquidity and improve stability. But some developers hate the idea, and a hard fork in Bitcoin looms.
What does that mean for investors – and for blockchain itself?
Forks disrupt blockchain’s decentralized consensus. They create speculative events, which means there are winners and losers – and the result can mean tokens that lose some or all of their value, or which are paralyzed because of double-spending issues that can take time to resolve.
For investors, Weese advises against keeping coins on an exchange. They surrender the ability to deploy their coins during these bouts of turbulence. “A fork can happen any time, and I predict they will happen more,” he said, because a lot of people try to generate crises to game the market. The best defense is to run your own node, so at least you can control your buy and sell decisions, as well as have a say in the consensus process.
Don’t panic, either. Weese said, “It’s easy to design wallet software or nodes that can ignore these events…most of the drama is for the people who like to speculate.” But he does predict more volatility for alternative coins, as some crash and others become established tools.
As for enterprises using blockchain for their own purposes, forks are either a threat or a non-event depending on their proof of work and the algorithms they deploy for consensus, Weese says. For example, what permission rules exist in a financial institution’s blockchain? Who holds the power to propose new rules? What happens to old transactions? What’s the voting process with regards to a blockchain’s articles of incorporation?
“The idea of a Bitcoin fork is scary,” Weese said, “but forks happen, and they are becoming easier to mitigate.”