If you haven't had an opportunity to review the previous article, it is at http://forums.prohashing.com/viewtopic.php?f=11&t=782. At the time, it was estimated that the Bitcoin halving would occur on Sunday, not tomorrow. There, I presented a diagram of intersecting possibilities, and attempted to gauge the probability of each occurring. I concluded that the future was grim for Bitcoin and positive for Ethereum. I think that the original predictions are mostly on target, but as they did not account for the DAO theft and we now have two forks to consider, the discussion should be reframed into "fork/no fork" possibilities for each coin.
What has changed in the past three months?
There were two major changes that I didn't foresee when I posted the price predictions in March: the low knowledge level of bitcoin investors, and the DAO theft.
The Bitcoin bubble rose higher than I predicted because I overestimated how much knowledge investors in Bitcoin have. My incorrect opinion was that the majority of people investing in something are very knowledgeable about the fundamentals of that thing. With Bitcoin at least, that doesn't appear to be the case. The lack of fundamental awareness of Bitcoin's flaw is demonstrated by the large number of articles that are just now starting to appear that discuss the blocksize issue and how it will affect Bitcoin in the future. For example, there was an article published on the NASDAQ website - of all places - where the author seemed to have just discovered the problem and thought to mention it to investors. A simple Internet search yields lots of these articles discussing the blocksize problem as if nobody had known about it years ago.
While I could be wrong in placing significance in these articles, their increasing frequency now suggest to me that many journalists, investors, and people with money may have been aware of the discussion, but had misunderstood the significance of the issue. Some of these people undoubtedly bought bitcoins and are now surprised to find that they are largely useless for any reason other than large money transmissions or speculative investment. A 50 cent fee for a transaction that often takes hours to confirm is unacceptable for a purchase of $50 when credit cards have no fee, include consumer protection, and provide instant confirmation. Even people who spend their days reading Bitcoin news websites seem to be disconnected from the reality that it is no longer economical to use the currency for everyday payments, and that businesses (like ours) have already switched to other coins to do that.
The lower level of knowledge among Bitcoin investors than what I predicted caused the price of Bitcoin during the bubble to peak higher than I thought it would, and to last longer than I thought it would before the crash comes.
The DAO also took me by surprise. At the time of the predictions, the DAO had not yet launched. Even when the DAO was on the public radar, however, few predicted the size it became. I chose not to purchase tokens on the grounds that the threshold for action was too high and there were no defined goals for the organization. It turns out that was the right decision for the wrong reason.
The wild success of the DAO, and its subsequent theft introduced uncertainty into the Ethereum markets and pushed the Ethereum price prediction to the right. The price of Ethereum will remain suppressed until a hard fork for Ethereum succeeds.
On the other hand, I was correct with other predictions. Segregated Witness, as predicted, had no effect on transaction throughput. The Chinese said it is dead on arrival because the Core did not include the promised blocksize increase. Similarly, I stated that the Core was wasting its time on features that nobody needed while ignoring the important issue. Despite the Core's work on many new features, only a minority has adopted the latest Core versions months after they have been released. If an alternative implementation becomes popular, it is probable that all of the work on these features will be completely discarded since none of the other implementations has implemented them, nor have users expressed interest in them.
It all comes down to forks
The fundamental assumption in these predictions, and one that has been borne out so far, is that there is a demand for two services - a way to send money, and a way to implement smart contracts. We are past the stage where people should be questioning whether there is demand for either of these things. I think that most also agree that there will eventually be much more demand than there is now for both uses. The primary unknown that will impact the price of Ethereum, Bitcoin, Litecoin, Monero, Lisk, and other systems is which coin people will use to accomplish these tasks, not whether there is a market for the services.
I also believe that none of the systems has a particular feature that allows it to supercede the others in its class. The only major difference in the systems is the people involved. Litecoin does not provide any special features that would convince people to switch to it other than that it is there. Similarly, Lisk's Javascript support is not so far superior to Ethereum that people would choose to use it for that reason alone, since someone can write a compiler from Javascript to Solidity for Ethereum.
Therefore, the communities of the two usage models will decide their fate. If they can come together to fix their problems, they will succeed; otherwise, users will try to shoehorn the other system to accomplish the purpose of the failed one. Coincidentally, both Ethereum and Bitcoin are facing moments that will define their futures for years to come - Bitcoin on July 9, and Ethereum on July 21. Whether the systems hard fork or not will be the most critical decisions community members have ever made. The decisions are so important that I will present this update as a simple four-part table, with the relevant factor that will decide the future direction (and value) of either network being whether it forks or not.
Before getting to the possibilities, I should review why I believe that both networks must fork to remain relevant, and why failure to fork would be disastrous.
Why Bitcoin must fork
Bitcoin's major limitation is that its blocksize is too small and the network is overloaded. There are many articles that provide arguments for and against the Bitcoin hard fork, and they are good reading. That said, the arguments being presented for the Bitcoin fork often focus on the technical advantages or disadvantages of different paths forward for Bitcoin. I stopped being interested in these long ago, because I don't believe that there are any relevant technical issues in the blocksize debate.
Bitcoin will succeed if the Thunder network is rolled out into production, if block sizes are scheduled to increase on a regular schedule, or whatever permanent solution is adopted. The Bitcoin community largely ignores that no matter how bad a decision is made, it can't be worse than what they are doing now, which is nothing. Making some sort of decision allows them to move forward and fix problems that may come up as a result of the choice, which isn't happening now.
The reason that Bitcoin has not wildly exploded is because businesses have no certainty that the network will be capable of supporting their transactions. It seems intuitive to suggest that it wasn't necessary to fix the problem until the network became overloaded, but the reality is that many businesses would have long ago started development on new products and services if there were a solution in place earlier. Instead, the opposite is happening: businesses still plan for the future, but they need to make plans based upon the facts on the ground. We did that two weeks ago; even though transaction fees had only reached $2.00, we announced the change of the default payout coin to Ethereum to lower our costs now before the fees become completely unacceptable. Coinbase, long a Bitcoin-only shop, hedged its bets with the GDAX.
The critical reason to hard fork bitcoin is because every day that passes sees more businesses plan around the limitations by using other coins or private blockchains instead.
I place the odds of a bitcoin fork in the near-term at 5%, based on the history of little action on this topic. The 5% is almost entirely accounted for by the fact that I think the probability that the Chinese miners will actually deliver on what they said June 30 is around 20%, and that the odds of them actually coming to a more permanent solution other than a one-time 2MB increase are less than 1 in 4.
I still think that it is better to force an immediate decision on the Bitcoin fork instead of leaving Bitcoin for dead, because I believe that most people would agree to a fork if a crisis can be created that presents an unlimited blocksize as the clear alternative. In my next post, I will present a proposal to create a crisis in bitcoin that would force an immediate decision on whether to upgrade or not, and which would cause people who attempt to block whichever decision is supported by the community to lose a lot of money.
Why Ethereum must fork
Ethereum's major problem right now is the catastrophic failure of The DAO, which caused about 15% of the entire currency to be stolen. If nothing is done by July 21, the attacker, whose tainted coins cannot be used to buy anything or sold at an exchange, is likely to begin using them for the only thing they are good for: to destroy the Ethereum network.
The attacker can destroy the Ethereum network by sending the stolen coins to random (or targeted) addresses. This action will require the Ethereum developers or someone else to create and release a blacklisting client that burns stolen coins, something that would cause immense controversy throughout the community. The client would be necessary to protect businesses from the legal liability incurred by receiving and spending stolen money.
While a simple blacklist would cause enough problems with services like ours that rely on precise wallet balances, the attacker could launder his ether through a tumbler into so many addresses that it would be impossible to determine what money is stolen and what money is not. One out of seven ethers would then be stolen and the entire network would be poisoned with these coins, and many businesses would be forced to stop using Ethereum altogether. Failure to hard fork Ethereum would allow the attacker's coins to be spent, and would probably result in the destruction of the Ethereum network and loss of all its value.
Another possible attack is for the attacker to simply write a bot that uses up all the gas in every single block for years. Nobody would be able to get the network to do anything else and Ethereum nodes would be stuck at 100% CPU doing wasted work, resulting in the node count dwindling and nobody being able to afford their contracts doing anything useful.
I place the odds of Ethereum hard-forking in the near-term at 80%, because there is widespread consensus among Ethereum miners that a fork is desired. Most of the remaining 20% of probability is accounted for by the scenario where nobody can agree on exactly which hard fork plan to choose before time runs out on July 21, rather than on whether the idea of a hard fork is a good one.
The four scenarios
Here is a chart of the four scenarios, followed by some ideas of how they could play out. I'll review the possibilities in order from most to least likely.
One currency (Bitcoin forks while Ethereum does not) - 1% probability
In the "one currency" scenario, the Bitcoin miners agree on a fork which permanently resolves the blocksize problem, while Ethereum fails to do so. This scenario is so unlikely that I have difficulty imagining the circumstances that would cause the Bitcoin miners, whose incentives are not aligned with those of users and businesses, to adopt a solution like BIP101, the "adaptive block size" limit, or a removal of the limit altogether. While crises are often effective in forcing immediate change, it's difficult to see how even the worst crisis accomplishes anything more than a temporary solution, like a one-time 2MB increase.
While a 2MB increase would increase reliability in the short term, it does not count as a "fork" in this article because it doesn't resolve the critical issues facing Bitcoin - the wasted time, the censorship, and the inability of businesses to invest due to uncertainty.
If, however, this situation does come to pass, I think that we would see Ethereum lose almost all value and solutions for Bitcoin that allow Turing-complete code to run would start appearing. A Bitcoin bubble would occur, and it would dwarf the previous bubbles in size and duration. Litecoin and other coins that have also been unable to agree on a blocksize increase and which provide few features that Bitcoin does not would see a dramatic decline.
World takeover (Both networks hard-fork with a permanent solution) - 4% probability
In the "world takeover" scenario, both networks successfully execute hard forks that permanently resolve the issues that they face. In Bitcoin, miners recognize that they need to spend a little more money on computing power and equipment so that they can maintain the network's competitiveness in the long-term, and do so permanently. In Ethereum, the community bands together to prevent a catastrophic distribution of attacker coins across the blockchain.
The barrier to investment that has been preventing Bitcoin from growing is removed, so companies flock to Bitcoin and start developing applications to take advantage of the guaranteed growth of Bitcoin's transaction capacity. Bitcoin's network becomes usuable again, and people can rely on it for prompt payment. Ethereum's developers continue their research into the proof-of-stake transition and a new, more secure, DAO is formed.
This outcome ushers in a new golden age of cryptocurrency, where the brightest minds focus on developing applications for the two currencies that rapidly increase adoption, instead of wasting more effort on arguments over the correct solutions to these problems. But again, it's difficult to see how powerful interests unconcerned with ethics, like theymos, will allow such an agreement to succeed in Bitcoin.
Dark ages (Both networks fail to come to consensus) - 19% probability
In the "dark ages" scenario, both Ethereum and Bitcoin fail to come to consensus on a solution and no forks are implemented. Ethereum will be in danger of complete collapse if the attacker begins using his money to destroy the network. Bitcoin, having failed to double in value from the previous mining equilibrium of $440, is likely to see hashrate drops, even more unreliability, absurd transaction fees, and continued price decline.
When the smoke clears, the winners may be Monero and Lisk. Monero is a blockchain focused on money that is far more advanced than Bitcoin. Lisk is a world computer similar to Ethereum where contracts are written in Javascript.
This outcome has implications for the future of distributed networks, and is the second most likely result. It will have proven that it is almost certainly impossible for a network to successfully hard fork after it becomes large. It will mean that future coins must get it right the first time, or, if that turns out not to be possible, then we can expect a future where cryptocurrencies come and go every few years. So far, we haven't had enough data on whether hard forks are even possible, and it may come to be accepted that networks cannot function without some level of central planning.
It's possible that instead of features being added to existing coins, the future may well be that a coin is developed under the radar until it becomes better than the existing ones, the existing coin reaches a point where it is unable to come to consensus, and the coin with better features supersedes it.
Slow shoehorning (Bitcoin fails to fork while Ethereum prospers) - 76% probability
In the most likely "shoehorning" scenario, Bitcoin fails to agree, takes too long to agree, or agrees on a temporary fix that resurrects the issue a few months later. While the Bitcoin developers continue to waste time and money at worthless conferences or developing Core features that are not being requested or adopted, they ignore the key issue at hand. At the same time, Ethereum successfully executes a permanent hard fork that refunds DAO contractholders. By August, the DAO is over, and the Ethereum developers are spending their time on making the network more stable and secure.
There will probably be another DAO created, and the attention given to it will drive interest and adoption towards Ethereum. People who need to conduct business will discover that it is far cheaper to do so with Ethereum. At the same time, Bitcoin would be grappling with a string of negative stories about the fallout from the hashrate drop at the halving, the inability for yet another conference to resolve the blocksize issue, "Bitcoin is dead" articles with charts of the crash from this current bubble, and so on.
We saw this outcome start to happen before the DAO uncertainty unfolded in how all the exchanges that were previously Bitcoin exclusive rapidly added Ethereum wallet and exchange support. While Ethereum has shortcomings when used for sending money, it is possible to write smart contracts that simplify and even further cheapen the process. Additionally, while some people discount Ethereum as an investment due to its inflation rate, Buterin's plan sets inflation far lower than that of Bitcoin for decades to come. People who criticize Ethereum for inflation are worried about something that will happen long after current life expectancies expire.
The DAO theft set back this outcome by a few months, but I still think that this result is the most likely outcome.
Conclusion
In conclusion, despite two significant events having occurred since March, I think that the predictions made then will be largely accurate. When the halving occurs tomorrow, Bitcoin hashrate will drop, block times will increase, and fees will rise. People will panic sell even if that doesn't happen because they think others will sell. The crash failed to happen a few weeks ago because I overestimated the level of knowledge of Bitcoin investors, who I thought would use game theory and be more forward looking. The DAO theft also temporarily dampened interested in Ethereum, keeping some money in Bitcoin.
Ethereum has been hit hard by the DAO theft, but unlike with the censored and fractured Bitcoin community, Ethereum's developers are able to work together and most people are in agreement to execute a hard fork. The odds of a hard fork occurring are better than not, and the fork's success will give businesses confidence that Ethereum is governed by people who are interested in supporting rapid growth.
There is a small but significant probability that the cryptocurrency industry in a year will have seen the failure of two coins, and the leaders will be Monero, Lisk, NXT, or DASH.
However, the most likely outcome right now is that, after this initial setback is resolved, Ethereum will continue to spread and slowly take over Bitcoin's functions across the cryptocurrency community. Cryptocurrency has reached the point where there is enough demand that it is a necessary part of many people's lives. While smart contracts will be a focus, it is businesses like ours switching to a low-cost send-money alternative that will drive the industry's software focus towards Ethereum.