What I was wrong about, Part 1
Fork Arbitrage and Fork Futures
1) Market maturity and timing
I assumed the market could come to the right answer in a short period of time. This was a rookie mistake. The market is always right...eventually. It can stay irrational longer than you can stay solvent (Keynes); in the short term it’s a voting machine and in the long term a weighing machine (Graham). The newer the field and the more complex the question, the longer the market will take to come to the right answer. Years is clearly the minimum we can expect to wait, no matter which camp you're in.
What about the legendary accuracy of prediction markets I often invoked when pushing this idea? When the market is mature and the question is one that is frequently asked and answered, with well worn research pathways, it can indeed be produced reliably and quickly. This explains how Intrade was able to call 50 out of 50 state elections correctly. However, there is nothing remotely close to a mature market for evaluating splits on the cutting edge of cryptoledgers.
Verdict: The fork futures do serve to get a read on the speculation market result in advance, which can alert backers of each side of the split early as to the outcome if they care about immediate price, but it isn’t any kind of ultimate guide to the truth or to the most prudent path. (We've seen how hashpower minorities can in fact survive on the same algorithm; the reasons are something to detail another time as this post is overlong already.)
2) Protocols
I fell for the idea that innovation should happen on the base protocol layer itself, partly because Core led us to believe Bitcoin was very limited in what it can do and Satoshi never made much effort to argue for the rightness of his choices so it looked like maybe he had chosen Bitcoin's parameters out of a hat. The narrative was that Bitcoin was just an evolution, stepping just high enough out of the mud to be useful but with no reason to expect it was a fully formed product.
After extensive research and thought, mostly prompted by he who must not be named, it is now clear to me that Bitcoin’s design was incredibly well thought out by someone with a multi-decade perspective, far more than ample academic expertise, and plenty of real-world experience in business from various angles, who had spent two decades hellbent on building the necessary chops to solve all the problems of the previous attempts at a workable money P2P system.
Bitcoin was delivered as a well-formed product, protocol-wise (not codewise). It cannot be perfect, as nothing is, but it remains in my honest estimation 20 years ahead of anything else out there. By the time anyone catches up it will be far too late. Well before that, and arguably from the very beginning as this is money we’re talking about, the protocol needs to have been ossified so that layers can be built on top by miners and all manner of businesses for their own needs, with complete confidence that nothing - no matter how seemingly innocuous or reasonable - will shift in the ground beneath their feet. Not to mention that simple things like time-delayed transactions and deals made since Bitcoin's founding, made on the assumption of eternal stability, would get screwed over by any protocol change (note: blocksize is not part of the protocol; it's a miner acceptance Schelling point). I wrote on the utter necessity for a stable base protocol here.
Verdict: Fork arbitrage assumes the base protocol should be shifting sand to make way for "innovation." Maybe 80 years from now, as that’s the kind of timeline financial institutions operate on, miners will draw on long-range prediction market dynamics to help in determining the new protocol.
3) Law
I assumed this was cypherpunk world so law could be ignored, or that there were no applicable laws currently and they’d have to be written when the legislators catch up, by which time Bitcoin would be “too big to fail.” Turns out, essentially nothing in Bitcoin is new from a legal perspective. Digital signatures, smart contracts, common carriers (miners), copyrights, and of course design documents and partnerships (mining as rule enforcement). It's only a matter of lawyers and judges getting up to speed on how existing laws apply. An immutable evidence trail means laws can be enforced in a delayed fashion; as they always did apply, there's no special allowance needed for retroactive application.
Miners have gigantic cost-sunk operations that use tremendous power and cannot hide from governments, at least while having a remote chance at remaining profitable. They are subject to the laws of their jurisdiction for the foreseeable future. We aren’t anywhere near a cryptoanarchy nor are we even moving in that direction; but even if we were, laws still apply in any anarcho-capitalist society, and I think any here who aren’t classical liberals or minarchists are ancaps. I know of no famous ancap who wasn’t strongly pro-law and law enforcement. Do you? Any society that supports an infrastructure that spawns unstoppable criminals is a failure at being a society in the first place. Somehow ancaps forgot that law is paramount and fell into the cypherpunk dreamworld.
Time to wake up.
Miners are simply businesses trying to make a profit, with legal and compliance departments like any other business. I'm no legal expert, but I now strongly suspect it’s not as simple as miners voting to add in “any needed rules and incentives” (last sentence of the whitepaper I often used to invoke) to change the protocol any way they like; they must defend these changes in court, otherwise be - quite rightly - accused of choosing winners and losers, as that’s what all protocol changes do. You can’t mess with the foundations of a money system without rewarding some at the expense of others, and any who know of the changes in advance or have a hand in shaping them - especially the developers themselves but also rent-seekers close to them - stand to gain tremendously on this knowledge and influence.
This should have been obvious in retrospect, regardless of law. Law here just reinforces common sense, and (2) above. The protocol really is set in stone -- practically, economically, legally, and via Schelling point dynamics / Keynesian beauty contest convergence amplified by all three.
Verdict: “The nature of Bitcoin is that its protocol was set in stone at version 0.1” (Satoshi) and there can be no hard or soft “forks” of the protocol to arbitrage for at least a few generations of our descendants if we want a system that gets used by business.
Fork Arbitrage and Fork Futures
1) Market maturity and timing
I assumed the market could come to the right answer in a short period of time. This was a rookie mistake. The market is always right...eventually. It can stay irrational longer than you can stay solvent (Keynes); in the short term it’s a voting machine and in the long term a weighing machine (Graham). The newer the field and the more complex the question, the longer the market will take to come to the right answer. Years is clearly the minimum we can expect to wait, no matter which camp you're in.
What about the legendary accuracy of prediction markets I often invoked when pushing this idea? When the market is mature and the question is one that is frequently asked and answered, with well worn research pathways, it can indeed be produced reliably and quickly. This explains how Intrade was able to call 50 out of 50 state elections correctly. However, there is nothing remotely close to a mature market for evaluating splits on the cutting edge of cryptoledgers.
Verdict: The fork futures do serve to get a read on the speculation market result in advance, which can alert backers of each side of the split early as to the outcome if they care about immediate price, but it isn’t any kind of ultimate guide to the truth or to the most prudent path. (We've seen how hashpower minorities can in fact survive on the same algorithm; the reasons are something to detail another time as this post is overlong already.)
2) Protocols
I fell for the idea that innovation should happen on the base protocol layer itself, partly because Core led us to believe Bitcoin was very limited in what it can do and Satoshi never made much effort to argue for the rightness of his choices so it looked like maybe he had chosen Bitcoin's parameters out of a hat. The narrative was that Bitcoin was just an evolution, stepping just high enough out of the mud to be useful but with no reason to expect it was a fully formed product.
After extensive research and thought, mostly prompted by he who must not be named, it is now clear to me that Bitcoin’s design was incredibly well thought out by someone with a multi-decade perspective, far more than ample academic expertise, and plenty of real-world experience in business from various angles, who had spent two decades hellbent on building the necessary chops to solve all the problems of the previous attempts at a workable money P2P system.
Bitcoin was delivered as a well-formed product, protocol-wise (not codewise). It cannot be perfect, as nothing is, but it remains in my honest estimation 20 years ahead of anything else out there. By the time anyone catches up it will be far too late. Well before that, and arguably from the very beginning as this is money we’re talking about, the protocol needs to have been ossified so that layers can be built on top by miners and all manner of businesses for their own needs, with complete confidence that nothing - no matter how seemingly innocuous or reasonable - will shift in the ground beneath their feet. Not to mention that simple things like time-delayed transactions and deals made since Bitcoin's founding, made on the assumption of eternal stability, would get screwed over by any protocol change (note: blocksize is not part of the protocol; it's a miner acceptance Schelling point). I wrote on the utter necessity for a stable base protocol here.
Verdict: Fork arbitrage assumes the base protocol should be shifting sand to make way for "innovation." Maybe 80 years from now, as that’s the kind of timeline financial institutions operate on, miners will draw on long-range prediction market dynamics to help in determining the new protocol.
3) Law
I assumed this was cypherpunk world so law could be ignored, or that there were no applicable laws currently and they’d have to be written when the legislators catch up, by which time Bitcoin would be “too big to fail.” Turns out, essentially nothing in Bitcoin is new from a legal perspective. Digital signatures, smart contracts, common carriers (miners), copyrights, and of course design documents and partnerships (mining as rule enforcement). It's only a matter of lawyers and judges getting up to speed on how existing laws apply. An immutable evidence trail means laws can be enforced in a delayed fashion; as they always did apply, there's no special allowance needed for retroactive application.
Miners have gigantic cost-sunk operations that use tremendous power and cannot hide from governments, at least while having a remote chance at remaining profitable. They are subject to the laws of their jurisdiction for the foreseeable future. We aren’t anywhere near a cryptoanarchy nor are we even moving in that direction; but even if we were, laws still apply in any anarcho-capitalist society, and I think any here who aren’t classical liberals or minarchists are ancaps. I know of no famous ancap who wasn’t strongly pro-law and law enforcement. Do you? Any society that supports an infrastructure that spawns unstoppable criminals is a failure at being a society in the first place. Somehow ancaps forgot that law is paramount and fell into the cypherpunk dreamworld.
Time to wake up.
Miners are simply businesses trying to make a profit, with legal and compliance departments like any other business. I'm no legal expert, but I now strongly suspect it’s not as simple as miners voting to add in “any needed rules and incentives” (last sentence of the whitepaper I often used to invoke) to change the protocol any way they like; they must defend these changes in court, otherwise be - quite rightly - accused of choosing winners and losers, as that’s what all protocol changes do. You can’t mess with the foundations of a money system without rewarding some at the expense of others, and any who know of the changes in advance or have a hand in shaping them - especially the developers themselves but also rent-seekers close to them - stand to gain tremendously on this knowledge and influence.
This should have been obvious in retrospect, regardless of law. Law here just reinforces common sense, and (2) above. The protocol really is set in stone -- practically, economically, legally, and via Schelling point dynamics / Keynesian beauty contest convergence amplified by all three.
Verdict: “The nature of Bitcoin is that its protocol was set in stone at version 0.1” (Satoshi) and there can be no hard or soft “forks” of the protocol to arbitrage for at least a few generations of our descendants if we want a system that gets used by business.
Last edited:
