This interpretation seems to depend on thinking the bitcoin just is some random asset people are speculating about. But it’s not. It’s blockchain technology, which is unquestionably important and will unquestionably change the way we do a lot of things in the technology and financial space. Bitcoin itself might lose out to another cryptocurrency, though network effects and first move advantage play a large role here. But the underlying tech is important and game changing and that pushes against your view of it as nothing but a speculative bubble. Put another way, that the dotcom bubble was a bubble and crashed doesn’t mean the internet was nothing but a speculative bubble, because the underlying tech was sound. Bitcoin, being a protocol, is more like the early internet than it is like a tulip.
It’s anonymous because ownership is tied to a private key, not an identity. Whoever has the private key owns the currency. It’s very easy to prove that you have the private key if you do have it. You can do this without revealing the key, just the fact that you have the key. It’s secure because it’s almost impossible for someone to figure out the private key.
You do lose privacy if someone manages to associate your key with your identity. But that doesn’t happen in the blockchain itself, it happens at the edges, when you need to exchange real currency or goods.
Bitcoin is produced by mining. Miners solve cryptographic puzzles, the purpose of which is to update the blockchain to reflect any new transactions. As a reward for assuming this transaction cost, miners receive payment in new bitcoins. This will continue until the middle of the next century, after which all bitcoins will have been mined. Then miners will have to charge a fee in exchange for their services. Meanwhile, the actual transaction cost is the electricity used to power the many thousands of computers that work on mining bitcoin. This is substantial, and today bitcoin is mined primarily in places with cheap electricity.
But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?
You could start with a bank issuing crypto-currency. If the bank wants to cater to a certain breed of paranoia, it can back its crypto-currency with gold. If it just wants to cater to normal people, or cater to someone like me, it can back its crypto-currency with dollars.
I don’t understand blockchain, but let me try another analogy. Think of a blockchain real-estate title service that is perfectly robust (allowing no disagreement over who owns the property). When I want to put an addition onto my house, the contractor needs to know that I am the one with title to the house. My private key allows me to prove that. If I had some anonymous way of communicating with the contractor, then the contractor might not know who the addition is for, other than it is approved by the legitimate owner of the property.
Similarly, as the bank’s crypto-currency circulates, the folks who maintain the blockchain record system get their fees from people making the transactions. Nobody can take someone else’s currency without permission. When they obtain currency, they only know who they are taking it from if identity disclosure is an element of the transaction.
Getting back to the bank, it is a potential point of failure. It could sell its currency for dollars, then abscond with the dollars, and then never redeem its currency. The bank is not decentralized. It constitutes a single point of attack should the government in the jurisdiction where the bank is located want to shut it down or demand to see customer lists (although the bank could destroy the latter after every transaction).
So, there are risks with a crypto-currency started by a bank. But I would take those risks any day over the risks of trying to carry wealth in the form of Bitcoin.
Keep in mind that if Bitcoin does end up being like a chain letter, then it is a mathematical certainty that most of the people speculating in Bitcoin will end up losers. The only winners will be those who actually sell in time to take their paper profits, and it is mathematically impossible for most people caught in a pyramid scheme to walk off with a profit. So we can be certain that the majority of Bitcoin speculators will not sell in time.
Hyman Minsky said, “Anyone can issue currency. The trick is getting it accepted.” If a consortium of banks around the world were to issue a crypto-currency backed by dollars, then that would be a lot easier for me to accept than Bitcoin. But I am so far from comprehending the technology that I am probably just showing off my ignorance.
But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?
Because the designer found it much easier to pay early adopters in Bitcoin than in money. This attracted enough early adopters that eventually the money came. It’s the same reason startups pay in stock options (with a minimal life-support wage in real money).
Note that the blockchain uses over 200 kWh per transaction (currently, and it’s getting worse). In places with cheap electricity, that’s about $8 per transaction. That’s going to seriously limit its appeal once transaction costs stop being subsidized by the mining process (about 8-11 years).
I think you may have been misled about the important characteristics of blockchain, which is emphatically NOT that it is anonymous (it’s pseudonymous, not anonymous –
like an IP address), but rather that it provides a way to implement the characteristic you assume away in your example – robustness. Extreme robustness.
A blockchain is simply a ledger, maintained as a large number of copies on servers around the world. Each copy records every single transaction since the chain began. To the extent that the identities of the parties have to be included in the transaction record, those identities are included in every copy, forever. Bitcoin appears “anonymous” because the bitcoin identity and the common identity of the participant cannot easily be connected – but that’s a property Swiss banks once offered (no more), and that, say, online commenters have today, and we all know better than to assume our online comments could not be tracked back to us with enough effort. It’s simply the lack of infrastructure and effort to connect the one to the other – but they are connectable, and once a common identity is matched to a bitcoin identity, every single transaction, ever by that bitcoin identity is traceable, as are all the counterparties’ bitcoin identities.
What blockchain does offer is this: it’s practically impossible to change the ledger, the transaction record. You literally don’t have to trust anyone – interestingly, not even yourself. Suppose, to make up an example, you are leading a government with a track record of corruption, trying to establish solid land titles. No-one trusts you. You don’t trust your successors. You don’t even trust the people who work for you now. With blockchain record-keeping, you don’t have to: not even the government can mess with those records. Certainly, that’s only a piece of the puzzle: you have to validate the transactions being recorded there, but once they’re on there, they’re on there forever. And, of course, you can record who validated the entry, and that can’t be changed either.
The “mining charade” is fee for service, but it’s a way of implementing fee for service that prevents fraud by the service provider (or at least makes either sustained fraud, or fraud targeting a specific transaction, impractical). The one used by bitcoin – which relies on massive computing power – is generally recognized as being clever but the wrong solution to this part of the problem: it’s too expensive, and it’s too slow. But there is extensive research going on, into alternative approaches.
This article (from a blockchain startup) talks a lot about “decentralized applications” a la your real-estate registry example
https://blog.chain.com/a-letter-to-jamie-dimon-de89d417cb80
I think it largely matches what you’re saying (though swapping “robustness” for “decentralized”)
“But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?”
You answer this question yourself later with the Minsky quote
“Anyone can issue currency. The trick is getting it accepted.”
Mining increased early stage acceptance. If the inventor of bitcoin had simply said “I have 21 million bitcoins, I promise to sell them into the market at a certain rate for 50 years” then what is the incentive to adopt? The mining allowed early adopters to produce their own coins and encouraged them to evangelize. This represented a near perfect commitment of the originator to not exercise central control over the currency and providing an incentive for increasing numbers of adopters.
The answer is yes, banks can issue cryptocurrency, be perfectly safe and never use blockchain. And any trader in the world can ask the bank computer for a secret number, and microprocessor foundry can tell you if the secret number is correct. If the number is correct, then the bank computer protocols are tied back tol hardware security, at the kernel level, and any given transaction can verify the proper protocol was used.
I think Intel can do that today. This process is called ‘watermarking the digits’ (I just named it) and the bank protocol erases the watermarks from one account as it moves them to another. Oddly, it is an exact duplicate of watermarked paper cash. It is cellulose become silicon, ink becomes digital pulse, and paper watermark becomes an encrypted code with secret properties. Which bank uses thi? Well, soon, the one in your hand, this is simply an upgrade of the chip based debit card, making it dynamic and counterfeit proof. Intl and a company called Trezor are doing exactly that.
“But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?”
What I don’t understand is why they don’t just keep the mining charade going, increase the supply by 2% per year or so. That way we don’t have to mess with transaction fees and everyone who owns bitcoin pays for the miners rather than just those that use it for transactions. The cost of transactions has become a huge flaw. Steam recently dropped Bitcoin because of it.
Who is “they” that has the authority to increase supply by 2% a year?
Nobody. The number of possible bitcoins is built into the math. If somebody figures out tomorrow how to compute all of them, mining stops forever.
I guess there is no they per se, creators of digital currency maybe, what I was advocating was building the 2% increase of coins in to the code that controls the coin, rather than having it stop at a hard limit. I don’t know if I’ve seen any coins that do this yet.
“What I don’t understand is why they don’t just keep the mining charade going, increase the supply by 2% per year or so.”
As an Internet 2.0, Ethereum ostensibly does this (mild inflation) and much more (moving to proof of stake, sharding).
That said, the market doesn’t seem to be pricing ETH (vs BTC) to reflect this so I’m probably missing something. Disclaimer: long ETH.
I’ll try to say what others have said in a different way.
Forget bitcoin. Blockchain is a technology for a database that is permanent, indelible, and immune to chicanery from anything less than a majority of the computers involved in the network. If no participant has an outsizes number of computers, that makes this a record that you can trust to keep its promises.
“You could start with a bank issuing crypto-currency.” That would require me to trust the bank in the future. Doable, but its a lot of work. Look at our regulatory state. Is that really less work than, say, the $8 per transaction quoted above?
Which would you rather have? An $8 tax per transaction, our current tax and regulate regime, or would you like to expend the energy to learn about every bank you do business with and prepare to follow through with your own lawsuit if the bank 10 years from now doesn’t follow through on the bank’s present day promise?
Which system do you expect to be the most responsive to advances in human understanding? That $8 per transaction, surely, followed by the DIY option.
It depends on what kind of transaction. For real estate, blockchains could be great. For fast food, the $8 transaction fee is a showstopper. And keep in mind that the transaction fee increases slightly every time a transaction is recorded (because the blockchain itself increased).
This is exactly correct. If transaction costs stay where they are, I don’t expect them to be use for small, regular transactions.
But I don’t expect transaction costs on small transaction to stay where they are. I do expect the market to come up with a way to make that cheap, and that may not involve placing transactions on the blockchain. Think of say a prepaid debit card. Buy it once for $108, spend it like a Visa or Mastercard. Walmart is currently selling $100 face-value prepaid Visa gift cards for $104.44, so this seems like it should be something the the market could figure out.
Doesn’t a Chinese mining consortium actually control enough of the mining that it could take over 50% of the mining? And if that’s the case, isn’t the whole chain susceptible to takeover by one entity, the Chinese?
a couple of thoughts…
If you are considering typical, familiar use cases like point-of-sale consumer transactions, or, SWIFT-like interbank payments, or, Paypal, or, etc. you are missing the incentive for Bitcoin adoption. Bitcoin has too many technical problems for those use cases; transaction costs are too high, transaction throughput is too low, latency is waaaaay too long.
Bitcoin use cases are specifically for avoiding banks, avoiding government regulations, avoiding other costs and risks that are even higher.
Wikileaks took donations in Bitcoin very early for obvious reasons. Assange is a gazillionaire now. Cross border money laundering that used to be done via trade finance transactions through banks is much more easily done via Bitcoin. Care to avoid Zimbabwean inflation and avoid holding US dollars which are illegal and subject to confiscation, use bitcoin. Avoid Russian government confiscation of deposits, take payment and hold balances in bitcoin.
‘Currency’ is a collective delusion. People faced with these use cases have adopted the delusion of bitcoin. The market for ‘gray’ payments is measured in $US trillions.
I tend to agree with Powell in terms of bubbles and investments. For all the complaints about the US bubble, we have to remember we had almost 60 year (mostly) Bull market on housing and a good house is valuable to families. (My retired parents a sitting on $700K resale when they downsize.)
Additionally, what is the size of the global investment market and remember China is adding trillions every year. 300T? Bitcoin total value is .3T so .1% of the investment market so it is still a small portion. And I still the aspect you are underestimating is the Get Your Money Outta Dodge reality here. So a lot of ‘Bitcoin’ investors from China and Russia get a benefit of moving their money away their country. (For all the complaints about our government, it still nothing compared to Russia or China or other governments.)
So it is reasonable to doubt Bitcoin much like you doubt Amazon or I doubt Uber. And it is your choice to avoid purchasing Bitcoin and Amazon stock. If Bitcoin is sort of legal ‘money laundering, for foreign nations that means:
1) Bitcoin has a value in global investment market. And if 50% of investors don’t like it, that means 50% investors may choose to invest.
2) Probably not the safest investment as it is controlled by some weird techno mining charade and foreign governments cracking down on Bitcoin investments. (Don’t buy if China is nearing a Financial Crisis.)
For the record, I don’t get it either, Arnold.
There are already banks that issue cryptocurrency, Ripple, backed by currency. Yes, you should be wary of ICOs as some likely are taking the money and running.
Bitcoin should be thought of as digital gold and not a currency. Bitcoin Cash and Dash do a better job emulating an actual currency.
“But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?”
Unlike your description subsequent to the quote, there is no central record of transactions. The record (blockchain) is dispersed in lots of places, each of which can be updated independently. The problem bitcoin has is to make sure that most copies of the blockchain agree with each other after a transaction. If they don’t, then the transaction is rejected as invalid.
In order to establish a transaction, a cryptographic puzzle has to be solved for which a solution exists only if the public and private keys match, while keeping the private key secret. That is, it’s an application of trap door cryptography. If the puzzle was too easy, then there would not be enough miners to guarantee enough copies of the blockchain to ensure security. If the puzzle is too difficult, then it becomes impossible for any miner to reach an answer in a reasonable amount of time. One genius of the bitcoin algorithm is that it finds the happy medium between the two extremes.
The miners all try to solve the puzzle by trial and error. The miner which by luck gets the answer is the one who is paid in bitcoin as a reward. (Eventually that miner will take a fee from the transactors.) Once the answer is discovered it is broadcast around the world, and it is easy for all other miners to verify that the answer is correct. Each miner updates his version of the blockchain so that they all agree.
Once a minimum number of miners have verified the transaction, it is assumed to be valid.
The difficulty of mining is what guarantees the security of the blockchain. If you reduce the mining cost to zero then you have a database, which depends on the honesty of the database manager. So bitcoin is secure precisely because mining takes 10 minutes (or several hours) to solve.
It’s not a charade.
“But what is the point of the whole “mining” charade? Why not just go to the fee-for-service model now?
Arnold,
“Centralized Bitcoin” sans mining would look like a website (app?) where you could use your password as an identity to send tokens to other users of the site. For instance, a bank could put 21 million dollar bills in a vault and then allocate 21 million tokens on its website. I could send you 1 token by identifying myself to the bank with my password and then ordering that 1 token be taken from my account and placed in yours (assuming we both had accounts of course).
Obviously you don’t actually need to “back” your tokens with actual dollar bills as long as you can convince enough people to trade them. (I’m taking “back with” to mean “redeemable for”) Someone will probably trade tokens for dollars, although probably not at a fixed rate.
Bitcoin is like that bank, except there’s no bank. Instead of asking a web site, you inspect the public bulletin of transactions to determine for yourself all account balances. The reason you can trust the public account of transaction is because the miners have put significant wealth into making the problem of forging a fake transaction list impossible to counterfeit. They get paid for this service by printing new Bitcoin today, and in the future will get to collect a portion of transactions.
Bitcoin is a Schelling point around which people can coordinate long-term wealth storage. That’s something that national fiat currencies fail to provide.