Friday, December 07, 2018

A "failed" BSV surpassed the "winner" BCH

I no longer find it too likely that Craig Wright is Satoshi Nakomoto – Nick Szabo could be the right man (which would be somewhat similar, anyway) – but Craig Wright has surely become the most impressive "unexpected late winner" in the cryptocurrency capitalization wars.

If you look at the now, you will see that the total value of the world's cryptocoins is some $108 billion, about 1/8 of the peak value in January 2018. The regular Bitcoin, BTC near $3333, keeps below $60 billion out of the sum. It is followed by the Ripple and the Ethereum. Ripple did better during the recent stage of the bursting cryptobubble than Ethereum because the former is seen as slightly institutionally backed which was an advantage while the latter is not.

But what's even more intriguing are the cryptocurrencies at the 4th-7th places – their capitalizations are almost the same, around $2 billion each, within a few percent. The fourth is still Stellar, waiting to be surpassed, however.

The fifth is BCH, the Bitcoin Cash, the most important fork of the Bitcoin after the Bitcoin without adjectives. Note that the peak price of Bitcoin Cash was over $4,000. Now it's close to $100 and about 39/40 of the wealth has disappeared if you didn't preserve the fork. The sixth is Tether pegged to the U.S. dollar – I remember it was 35th or so, and knew it was going towards the #1 status as the cryptobubble was going to burst while USDT has the intrinsic value of the dollar.

And the seventh is BSV, Bitcoin Satoshi's Vision, a recent fork of the Bitcoin Cash masterminded by Craig Wright.

However, just hours ago, BSV was already above BCH, by 10% or so according to the price. (Update: 30 minutes after I completed this blog post, BSV jumped to the fifth place again, ahead of Tether and BCH.) This factoid was cool and unexpected for almost everybody who followed that enterprise. Why? Because BSV and BCH (during the contest, the latter was called BCHABC) were competing in the "hash war" just a few weeks ago. And BCHABC "won", people agree – a larger amount of computation resources went into the mining of that – which is why it kept the "prestigious" brand BCH. Cryptoexchanges and others renamed the loser to something else, BSV: by losing the war, it has lost the copyright.

Now, it's expected that the winning fork is more expensive, too. People vaguely assume that there is one kind of a victory, one type of attention that focuses on the winner. Bitcoin Cash itself was the most successful but still losing fork of the Bitcoin – and correspondingly, its price never went approximately above 1/4 of the price of the Bitcoin. However, Craig Wright has falsified that wisdom. The "loser" may do better according to the price than the winner.

This is an outcome that was "softly" considered very unlikely in Satoshi Nakamoto's paper. In fact, he didn't discuss "peacefully co-existing forks" at all. When a cryptocurrency forks, it's because someone tries to attack it. And Nakamoto explains that the miners' search for profitability will make it very likely that except for one, all other forks die away. This outcome should result from the miners' realization that they're wasting money if they focus on mining the "doomed, minority version of the chain". The survivor is the fork with the greatest computational resources dedicated to it. It's labeled the "consensus truth" and all the minor branches are labeled "fraudulent chains".

We just saw that this wisdom was extremely oversimplified. Because Nakamoto has "almost proven theorems" about the disappearance of the "fraudulent chains", how is it possible that the outcomes of the theorems are violated in the real world? Well, it's because he had to make some assumptions which aren't obeyed in the real world.

In particular, what isn't obeyed is the assumption that miners always try to maximize the immediate profit – the expectation value of the cryptocoins that have a chance to be found in the next block i.e. within 10 minutes or so. Or at most a few blocks. Clearly, we saw that this assumption was violated. Craig Wright and miners on his side were clearly computing BSV hashes even though a different kind of computational work was predicted to produce a higher immediate profit. Why did they do so? Well, they had a grander plan in the more distant future and they (or he) had an ego to protect.

(Lots of miners keep on mining at a loss, anyway, so the assumption that the miners are generally profit seekers – while this assumption is important in Nakamoto's paper – is violated in the real world, and rather completely so.)

It's plausible that BSV became comparably expensive as BCH because BCH was seen as too centralized ( did too big a fraction of the mining, perhaps a majority). And the software changes done by BSV were seen as "better" or "more decentralized" or "more smoothly scaling" than those in BCHABC. But it's also possible that Craig Wright is a master trader and he was selling and buying BSV and other currencies at the right moments to "pump" the BSV price. And maybe he has included some glitch that allowed him to double-spend or change the number of coins temporarily or something highly irregular like that.

At any rate, BSV could catch up with BCH – and it has briefly surpassed BCH. BSV may surpass BCH again, when it comes to the capitalization. Don't forget that BSV is a "failed fork of a failed fork". But this double failure didn't prevent BSV from becoming the fifth cryptocurrency by capitalization hours ago.

Software of the Bitcoin kind can be written – and has been written – to power the nodes (miners) and traders. But when the cryptocurrency such as the Bitcoin isn't backed by anything, the lore and theorems presented to defend the view that it is a good idea are fishy. Some of the theorems – about the disappearance of the failed forks – have to assume that the miners are profit-seeking in some old-fashioned sense.

However, what is the correct way to estimate the profit if the Bitcoin price is a completely emotional, social construct? In the real world, the miners aren't really independent from the holders or traders. And this is a problem – in the fiat money world, that would be like the absence of the independence of the central banks. Instead, the miners are people who have some beliefs about the future evolution of the prices of all the relevant cryptocurrencies – and they have some exposure combined with some form of profit-seeking through non-mining channels, too. Because of these beliefs and preferences, their mining may also violate some rules of "short-term profit seeking". And when the assumptions of the theorems are violated, the proven propositions may fail to be true, too.

BCH price in recent 7 days, the financial world's best approximation to a dropping parabola ever.

In particular, there is really no guarantee that the blockchain paradigm produces "one and only consensus truth" after several blocks. Even if it seems to be the case after a few hours or dozens of blocks, the identity of the truth may very well change after a few weeks or any other period of time because the miners are connected with people who have some long-term plans. They may convince the system that their software, algorithms, and version of the blockchain is the "more valuable one" despite the smaller resources dedicated to it in the past. Why? Because no one really cares about how much energy has been wasted. The Marxist theory of value is gibberish and because the Bitcoin "somewhat" relies on that flawed theory, it's too bad.

Now, the Bitcoin's or blockchain's ability to isolate the "only and unique truth according to the consensus" is really the one and only important alleged "achievement" of that software trick. It's the reason why the miners are wasting as much electricity (watts) as half of Czechia at every moment. But as we just said, even that "achievement" doesn't really work. There may be "several truths" about the transactions that have taken place and a truth (a version of the blockchain) that looks far more compatible with the "consensus" may become the less relevant one a few weeks later.

So I think that it's correct to say that even the basic "achievement" of the Bitcoin – a method to prevent double spending without a trusted third party – doesn't really work and the price success of the BSV has proved this assertion of mine. When you are allied with a group of miners, you may rewrite the whole history by turning your fork of the Bitcoin into the more expensive one and therefore the more relevant one. And it may be unclear in the initial hours after the fork which of them will be more expensive and therefore more relevant in a few weeks, months, or years.

In other words, the blockchain paradigm has one big contradiction. On one hand, it only works assuming the users' and miners' financial rationality; on the other hand, its very introduction basically requires the bulk of these people to behave irrationally, as self-sacrificing activists etc.

Would this problem be resolved if the cryptocurrency were backed by something? I don't think so. First of all, the backing requires some "human enforcement". If someone promises to pay at least $0.99 for one Tether, you are never quite guaranteed that the promise is correct and the law enforcement may always fail to force the Bitfinex to do so. But even if that enforcement were a part of the cryptocurrency, and Ethereum comes close to this within the smart contracts, there could still be long-term projects that involve forks that temporarily look hopeless but they become the most important ones, anyway.

So I think that the whole blockchain idea has no merit at all – and this problem exists on top of the pyramid-game problems with the intrinsically worthless currencies. Blockchain is just a very awkward way to give the responsibility for the "integrity of the transaction records" to some rather randomly and dynamically picked "trusted parties" where you don't know who the trusted party is. But the party that actually decides which ledger is the "true history" does exist and may have lots of murky goals that may be harming you, among other things. I think that knowing the identity of the trusted party is better than not knowing (e.g. because you may choose a bank from many competing choices), and therefore the blockchain is strictly worse for maintaining the information about account balances of payers' and recipients' than the regular banking.

No comments:

Post a Comment