Part 1: Explaining Bitcoin
Why Bitcoin works and why it can be trusted
Bitcoin is working and making progress and there are more people wanting to understand why this thing called Bitcoin, which doesn’t have a CEO, a board of directors or employees, can thrive like it does.
However, more often than not when someone offers an explanation as to why Bitcoin works, the explanation only includes something to the effect that “Bitcoin works because we can trust the math and software”. Sometimes the explanation also includes reference to “the Blockchain”. In my opinion, this sort of explanation is not doing Bitcoin’s growing audience any favours. Here’s why.
The breakthrough with Bitcoin was the design of a system that uses ‘economic incentives’ to constantly encourage people to perform certain acts instead of others. Building on this explanation, Bitcoin works because notwithstanding the array of different preferences and opinions in society we are all united by our desire to create rather than destroy wealth. This universal human preference allows Bitcoin to work.
Full nodes enforce the same set of rules for how their copy of the Bitcoin ledger can be updated with new transactions. These rules are referred to as the ‘consensus rules.’ By enforcing the same set of rules, full nodes stay in sync with one another and maintain consensus as to the state of the ledger, and therefore who owns which bitcoin.
Full node operators are economically incentivized to ensure Bitcoin works
Full nodes are operated and controlled by people who store wealth in bitcoin and/or transact using bitcoin (the Bitcoin ledger’s native currency). Due to the fact that these people or organisations use bitcoin, they are economically incentivized to ensure that bitcoin continues to be good money. This is because only if bitcoin stays good money will it be valued by other members of society, and therefore be able to function as a store of value. The following examples demonstrate this economic incentive.
If there were to be a departure from the system’s predetermined issuance schedule (the set times at which set amounts of new bitcoin enter circulation), then people could legitimately question bitcoin’s ability to be scarce (which is an attribute of a store of value and good money) and people would probably, therefore, be less likely to store wealth in bitcoin.
Alternatively, if people were allowed to double spend bitcoin (use the same coins to make multiple payments which would leave a supplier of goods or services out of pocket) then people would also question bitcoin’s ability to perform the functions of a money and they would therefore store less wealth in bitcoin.
If, as a result of bitcoin being bad money, less wealth is stored in bitcoin, bitcoin’s price (purchasing power) would decrease and bitcoin investors (people storing wealth in bitcoin) would be less wealthy. Of course, full node operators (who store wealth in bitcoin) do not want this to happen. In addition, there are also millions of people around the world who own bitcoin without running a full node. These people also don’t want bitcoin to be bad money and even though they don’t run a full node, their expressed preferences and economic significance influences full node operators to enforce the consensus rules.
Full node operators are therefore economically incentivized to see to it that Bitcoin works so that the system’s native currency, bitcoin, continues to be valuable and stores value. In particular, full nodes see to it that only valid (as opposed to fraudulent) transactions are cleared and settled and that bitcoin continues to have the important monetary attributes of durability, portability, fungibility, verifiability, divisibility, scarcity, a growing established history and the ability to resist censorship.
On the other side of the coin, as much as full nodes are economically incentivized to ensure that nothing goes wrong, they are also incentivized to see to it that things happen as people expect them to.
With this in mind, you can understand that provided the consensus rules are met, full nodes will without fail add new blocks of transactions to their copy of the Bitcoin ledger, thereby clearing and settling the valid transactions, updating the registry of which address owns which bitcoin and ultimately enabling Bitcoin users to enjoy their property rights.
Step 2 — Miners work to produce eligible blocks and incur costs in order to do so:
The most significant consensus rule is that the full nodes will only update their copy of the ledger with a batch of new transactions once they receive a correct answer to a specific mathematical puzzle that needs solving. This correct answer to the mathematical puzzle is referred to as the ‘proof of work’.
The enforcement of this consensus rule has a big influence on how and why Bitcoin works. Its influence is introduced in this section and Part 3 builds on these principles to describe the economic incentives and their effect in more detail.
In order to provide the poof of work that the full nodes require, participants in the Bitcoin network, known as ‘miners’, use computer hardware to solve the simple mathematical puzzle. A correct answer (the ‘proof of work’) has the effect of converting a batch of new transactions into a block of data that can be added to the Bitcoin ledger.
A correct answer (the ‘proof of work’) is difficult and expensive for the miner to compute but once the answer is found it is very easy and inexpensive for the full nodes to verify that it is in fact a correct answer.
Full nodes then use the eligible block of data to update their copy of the ledger (add a new block of data to the Bitcoin blockchain). On average, every 10 minutes at least 1 miner will solve the mathematical puzzle that produces an eligible block.
Miners are economically incentivized to produce eligible blocks
Importantly, the computer hardware that miners use to produce the proof of work is application specific, i.e., it can’t be used for any other economic purpose, and these computers burn electricity in the process of producing the proof of work. Therefore, producing eligible blocks involves capital investment and operational expenditure.
Like anything else in the world, people won’t incur these economic costs for no reason at all. The same reality applies here.
Miners only make the investment and incur the ongoing costs because they can expect to receive the bitcoin that the full nodes award to the miner (or mining pool) that produces a valid block of data that can add the new transactions to the ledger.
As an incentive for producing an eligible block, full nodes allow the successful miner to create and include in the new block one additional transaction that credits newly ‘minted’ bitcoin to an address that the miner controls. This reward is currently 12.5 new bitcoin for every new block. This is how new bitcoin enters circulation in the Bitcoin system.
In addition, if any of the transactions included in the new block pay a transaction fee then the miner can also ascribe that amount of bitcoin to their chosen address. Transaction fees are a discretionary payment made by the spender/sender of bitcoin and are used to incentivize the miners to prioritize your particular transaction over other user’s transactions that also need to be cleared and settled.
This computational work of producing eligible blocks to receive newly created bitcoin is called ‘bitcoin mining’ because in a similar way to how traditional mining involves working and incurring expenses to gather precious metals from the earth’s crust, receiving newly created bitcoin involves costly computational work to receive newly created bitcoin and transaction fees paid in bitcoin.
Step 3 — Full nodes add the eligible blocks of new transactions to their copy of the ledger:
The step by step process that I’ve been describing was initiated by a user generating a transaction (step 1 above). Miners then work to include this transaction along with other transactions in a block that complies with the consensus rules (step 2 above). As soon as a miner solves the mathematical puzzle that produces the proof of work, it distributes the eligible block to the full nodes that it is connected to (step 3 described in this section).
Upon receiving the block from the miner, the full nodes set about verifying the proof of work and they will also check that the block, and the transactions in the block, comply with the other consensus rules. Some of the other rules are:
- Does the bitcoin being spent exist in a prior block of the ledger?
- Is it the owner that is spending the bitcoin or has someone else tried to fabricate the transaction?
- Has the bitcoin not already been spent?
- Does the spender own at least as much bitcoin as they are now trying to spend?
- Has the miner rewarded themselves with the correct amount of new bitcoin?
If the block and the transactions in the block comply with the consensus rules then full nodes will add it to their copy of the ledger and they’ll also pass it onto their peers in the network so that these other full nodes can also verify the block and update their copy of the ledger accordingly.
If, however, the block isn’t a valid one then the full nodes will discard it and they will wait for another block to be submitted to them by another miner. As described above, all miners are economically incentivized to produce valid blocks so it will only be a matter of time until a valid block is submitted by another miner.
Cryptography and a blockchain data structure make the verification possible
The independent verification by full nodes is made possible by Bitcoin’s use of cryptographic hashing, asymmetric-key cryptography, digital signatures and a blockchain data structure. When an explanation of why Bitcoin works includes reference to the ‘math’ and/or the ‘software’ and/or ‘the Blockchain’, this is what is being referred to.
Cryptographic Hashing –
Cryptographic Hashes are used in many places throughout Bitcoin. Some of these areas are: digital signatures; creating transactions and adding blocks of data to the ledger and ‘chaining’ those blocks together.
A Cryptographic Hash Function converts any form of digital data (file, photo, text etc.) into a unique string of text know as a Cryptographic Hash.
The same digital data will generate the same Cryptographic Hash but if you change the data even slightly the Hash will change completely in a way that’s impossible to predict. In addition, the chance of finding two different inputs (pieces of digital data) that provide the same output Hash is equal to, in percentage terms, the amount of space that 1 atom occupies in relation to all other atoms in the universe.
Asymmetric-Keys and Digital Signatures –
Every bitcoin or part of a bitcoin in circulation is cryptographically linked to a Public Key. A Public Key is itself generated by a Cryptographic Hash Function using the Private Key as the input data. The cryptographically linked Private Key is kept secret and is used to encrypt transactions spending or sending the bitcoin linked to the Public Key. The encryption of the transaction data is known as the Digital Signature for the transaction.
Using the known Public Key, full nodes decrypt the Digital Signature and if the decryption produces the same transaction details that accompanied the Digital Signature then the full nodes can be sure that:
- the transaction was generated by the person/entity that owns the bitcoin; and
- the transaction details were not tampered with in the period of time between: (1) when the transaction was created by the owner of the relevant bitcoin and; (2) the transaction details reached the full node.
A Private Key therefore acts as the tool to generate something akin to the password (the Digital Signature) that full nodes require before they will allow the particular transaction to be cleared and settled (included in a block that is added to their copy of the Bitcoin ledger).
Blockchain data structure –
In order to be able to always verify whether or not the Digital Signature for a given transaction is valid, full nodes need a record of the Public Keys that own (are cryptographically linked to) each bitcoin (or part of a bitcoin) in circulation. This is where the blockchain data structure comes in.
As part of enforcing the consensus rules, full nodes require miners to group new valid transactions in a block of data and cryptographically link the block (using a Cryptographic Hash Function) to the previous valid block in the chain. That previous block is also itself cryptographically linked to the block before it, and so on. In this way, a ‘chain’ of blocks of data (transaction data) is created — hence the name ‘blockchain’.
A blockchain data structure’s primary benefit is that it immediately highlights if anything in the previously recorded information (historical transaction data) has been changed / tampered with.
Due to the fact that full nodes can rely on the authenticity of the data in previous blocks (because any change to the previously agreed state will immediately announce itself), they are able to independently enforce the consensus rules listed in the bullet points under the heading “Step 3: Full nodes add the eligible blocks of new transactions to their copy of the ledger” above.
You should please note, however, that the blockchain data structure alone does not make the data immutable (unchangeable) and therefore trustworthy. The immutability of the data in Bitcoin’s blockchain is as a result of Bitcoin’s decentralization and the strong economic incentives.
Don’t get overwhelmed by the technology –
As stated in Part 1, all of these technologies, the ‘math and software’ and/or ‘the blockchain’, are not the focus of this article because alone they wouldn’t have enabled Bitcoin to thrive as it has. If someone hammers on about these technologies while they are explaining why Bitcoin works, then tell them to stop because the technologies have actually been around for ages. Linked lists go back to the 1950’s, public-key cryptography came about in the 1970’s and proof of work is from the 1990’s. This alone should help one understand that these are not the things that make Bitcoin special.
But I digress… Back to the economic incentives, the interesting information and the stories that help us engage with these principles.
Difficulty adjustment consensus rule protects the issuance schedule and scarcity
The repetitive cycle of the new transactions going into a block and a new block being added to the ledger occurs on average every 10 minutes.
This timing does not change all that much in response to how much computational power (mining equipment) is working to solve the equation (produce the proof of work) because as part of enforcing the consensus rules, after every 2016 blocks, the full nodes in effect adjust the difficulty of the equation to make it harder or easier so that if the amount of computational power mining bitcoin stays constant for the next 2016 blocks (about 2 weeks) then proof of work will be produced by one of the miners on average every 10 minutes. This adjustment every 2016 blocks is referred to as the ‘difficulty adjustment’.
Bitcoin’s predetermined issuance schedule and scarcity is therefore guaranteed by the fact that (1) miners are only able to solve the mathematical problem as fast as full nodes permit them to; (2) the miner reward is the only way that full nodes permit new bitcoin to enter the system; and (3) full nodes don’t allow miners to reward themselves more bitcoin than is provided for in the predetermined issuance schedule.
We can of course expect that the rise in the value of any money will lead to more resources being dedicated to the production of that money and thus an increase in its supply. However, due to the difficulty adjustment in Bitcoin, as bitcoin’s value rises and the incentive to produce bitcoin increases, the continued enforcement of the difficulty adjustment by full nodes sees to it that no more bitcoin can be produced than is provided for in the predetermined issuance schedule. Genius!
Part 3 — Credible answers to some important questions
Question 1 — What keeps the full nodes in line?
Do full node operators have too much power in the Bitcoin system? If they are the custodians of Bitcoin’s ledger, what stops a smaller group of nodes from, for example, disregarding the difficulty adjustment rule or from ignoring the predetermined issuance schedule?
The answer is ‘nothing’! At the outset, one should know that there is actually nothing stopping anyone from doing this! Bitcoin is open-source software so anyone can at any stage change the software and release it for others to use with them. However, what’s important is that it doesn’t actually matter if anyone does this. Here’s why.
Remember the important principles
To understand what the disincentives are, it’s worth repeating certain key principles.
The first such principle is that bitcoin is money and a money only gets you something if others accept it from you in exchange for their goods, services or another money that they have. It is therefore in anyone’s best interest to converge with what others think, rather than create and use something that others don’t value.
Secondly, people value a money because it is able to be durable, portable, fungible, verifiable, divisible, scarce, has established history and can resist censorship.
In the next section I’ll use practical examples to: (i) illustrate these principles; and (ii) explain how and why Bitcoin full node operators are incentivized to stay in line.
Full node operators are economically incentivised to enforce the established consensus rules
Let’s assume for a moment that a group of nodes start developing and maintaining a new ledger that includes a miner reward in each block that is more than is provided for in Bitcoin’s predetermined issuance schedule. Better still, let’s assume the ledger has a miner reward that is less than the miner reward in Bitcoin. This lesser reward would in theory make this ledger’s native currency scarcer than bitcoin, right? At face value, yes, but actually not. Let me explain.
Despite this ledger having fewer units in circulation (a ‘scarcer’ supply), the break from the established consensus in Bitcoin requires us to ask some important questions:
- Will this new issuance schedule stay this way in the future? It has changed once so why won’t it change again? The mere fact that there has been a change (no matter how big or small) from the predetermined issuance schedule (established consensus) implies that there could be further changes to this new ledger’s rules at any stage in the future. By way of example, it’s now conceivable that there will be other decisions to radically increase the reward (which would dilute the value of the existing supply and impact scarcity).
- Is this new network going to treat all users equally, or will it favour the people who created it?
- Given the uncertainty raised in the bullets above, will this network get enough support in the form of full node operators? If not, will the network be able to resist any attempts to censor transactions in the network and what would stop anyone from seizing or confiscating the ledger’s currency from the owners/users.
These are just some of the questions that become important when people create a new ledger with its own native currency.
In summary, we now can’t with any degree of certainty predict the future for this new ledger and its own native currency and there is no guarantee that our property rights will be enforced. In these circumstances, it would be ill advised to use this ledger’s native currency to store wealth.
Don’t get me wrong, the creators are still free to use this new ledger going forward. All I wish to point out is that for the reasons explained above, very few, if any, other users will follow which means that the ledger’s native currency won’t be as valuable (have as much purchasing power) as bitcoin.
Bitcoin full node operators are therefore better served if they stay in sync and enforce the Bitcoin consensus rules along with the other nodes. They are economically incentivised to do so.
Lessons from Bitcoin Cash and Libra
Here are a few examples of how these principles have played out in real life.
On 1 August 2017 at the time of the 478 559th Bitcoin block, certain full nodes adopted a different set of consensus rules that made their new blocks incompatible with the existing Bitcoin blocks. The main changes were to allow for an increased amount of transaction data to be included in each block and to remove segregated witness from the code. Without going into the detail, their thinking was that these changes would easily enable more transactions per unit of time and that this was required at that point of Bitcoin’s development as a money and peer-to-peer network.
Notwithstanding that a small number of full nodes adopted this change, the majority did not because they didn’t think that the changes were good for bitcoin’s monetary attributes. In particular, the majority of full node operators believed that having blocks with more data would make it harder for ordinary people to run a Bitcoin full node.
This was not an outcome that the majority of users wanted because their view is that Bitcoin needs a lot of people running full nodes in a lot of different places if it is going to remain decentralized. Only if Bitcoin is decentralized will the network be able to resist censorship, and will the native currency, bitcoin, have the important monetary attribute of censorship resistance.
Indeed, just recently we got some more evidence of the fact that if any money is going to take on governments, central banks and other custodians of the existing financial system then it ought to be able to resist censorship from these powerful institutions. These institutions have a lot to lose if an alternate to fiat money is fully monetized so it is at this point highly unlikely that they will hand over the reins without putting up a fight.
The recent censorship that I’m alluding to is the immense scrutiny and pressure that Libra has faced (Libra is a digital money proposed by Facebook through the Libra Association). If you haven’t been following along over the past 6 months, you can read articles here and here. Some of the more interesting quotes from the politicians around the world (also contained in the articles) are:
- “with all of these problems I have outlined, and given the company’s size and reach, it should be clear why we have serious concerns about your plans to establish a global digital currency that would challenge the U.S. dollar,” Maxine Waters, chairwoman of the US House Committee on Financial Services on 23 October 2019.
- “We believe that no private entity can claim monetary power, which is inherent to the sovereignty of Nations” A joint statement from France and Germany at the meeting of eurozone finance ministers in Helsinki on 13 September 2019.
Given the backlash, one could argue that there is no guarantee that Libra will ever launch.
What about Bitcoin Cash? Well, the majority of people simply don’t value this ledger’s native currency. At the time of writing, there is only about 4% as much wealth stored in bitcoin cash as there is wealth stored in bitcoin (as represented by the market capitalization of each currency). This is a serious decline for Bitcoin Cash, something which has an identical ledger to Bitcoin up until 1 August 2017.
Bitcoin cash’s poor performance as a store of value relative to bitcoin is clear evidence that node operators are economically incentivised to stay in sync and enforce the Bitcoin consensus rules along with the other nodes.
Question 2 — Can miners sabotage Bitcoin?
Part 2 described how miners work to convert a batch of new transactions into a block that can be added to the Bitcoin ledger and it explains that they perform this work because they receive newly created bitcoin and transaction fees paid in bitcoin as a reward for doing so.
Given that miners can clear and settle transactions themselves by adding new blocks to the ledger, what stops a miner from working on two different blockchains, one version that pays someone and another version that effectively overturns that payment? Said differently, what stops a miner from ‘double spending’ their bitcoin?
Like most things in Bitcoin, the economic incentives encourage people to do one thing instead of another.
In the Bitcoin system, a particular miner’s likelihood of producing proof of work before any of the other miners do is equal to, on average, the percentage amount of computational power that the miner has in comparison to all of the computational power that is mining bitcoin at that moment in time.
Therefore, if the miner controls 1% of the computational power mining bitcoin at that moment then he/she has on average a 1% chance of being the person that owns the computer that produces the proof of work before anyone else in that 10-minute window.
There is however a chance that the miner can get lucky and produce a few consecutive valid proofs of work in a short amount of time.
Simulation of a double-spend attack
If we simulate a double-spend attack, it would develop as follows. Miner has spent bitcoin (“transaction 1”), released that transaction to the network and has now produced another block that contains a different transaction (“transaction 2”) that in effect overturns transaction 1.
The problem for the miner trying to double spend with transaction 2 is that all of the other miners will at the same time be working to extend the ledger that includes transaction 1. Remember that due to their ability to receive newly ‘minted’ bitcoin and transaction fees paid in bitcoin, other miners are at all times incentivised to produce proof of work that has the effect of converting a batch of new transactions into a block that can be added to the Bitcoin ledger.
So, the miner trying to double spend will have to get lucky continuously in order to be able to extend this second version of the ledger at the same speed as all of the other miners would be extending the version that includes transaction 1.
Due to the fact that there will be significantly more computing power working to extend the version of the ledger that includes transaction 1, over a sufficient period of time, much like the law of gravity, the law of averages will take effect and this version of the ledger will become longer (have more transaction history) than the version that includes transaction 2.
Full nodes will eventually completely ignore the shorter version of the ledger that includes transaction 2 because, as explained in this article, they are economically incentivised to ensure that bitcoin remains good money and that only valid transactions are cleared and settled.
For these reasons, when selling valuable goods for payment in bitcoin it is advisable to wait for 4–6 block to be built on top of the block that contains your transaction because after that time period it becomes less and less likely (and eventually almost impossible) for someone to successfully double spend (by reorganizing the blockchain).
Bitcoin needs more than 50% of miners to mine honestly
As described above, the chances of a single miner, or even a smaller group of miners, being successful with a double spend in the short term is equal to the amount of computational power that they control in relation to all of the computational power in the system at that moment in time. However, due to the amount of proof of work that would need to be redone and caught up, even with more than 50% of the computational power a miner has little to no chance of being successful with an attempt to double spend bitcoin from a transaction that was cleared and settled in a block much further back in the chain.
While it is possible in the short term, so long as more than 50% of all mining power is working to extend an honest version of the ledger then any other version will eventually fall behind (have less transaction history) and become insignificant in the system.
Due to the fact that the probabilities do not favour a successful double spend, all miners are economically incentivised to direct their efforts towards producing proof of work for valid blocks that contain valid transactions. Only if they do this is there a chance that they will be rewarded the newly created bitcoin and transaction fees paid in bitcoin in the version of the ledger that continues into the future.
This is important because only if the miner’s reward and transaction fees are written into a ledger that continues into the future will the miner have rights to bitcoin that other people value. Only then will the miner be able to cover its operational costs (including electricity) or make a profit.
Miners wouldn’t want to ‘shoot themselves in the foot’
In addition, if for argument’s sake a group of miners were to succeed with a double spend attack these miners would in effect be harming themselves more than they would be scoring. This is because a successful double spend would bring into question the Bitcoin system and the native currency, bitcoin, would probably no longer be valued by society. Less demand for bitcoin would reduce bitcoin’s purchasing power which in turn would have the effect of destroying miner wealth saved / held in bitcoin, and it would also make their application specific mining equipment worthless.
As a result of these economic incentives and the fact that people will want to create rather than destruct wealth, we can expect that the majority of miners will continue to mine correctly. This will, in turn, contribute to ensuring that the Bitcoin system continues to operate as it is meant to.
I hope that this article helps everyone think about the economic incentives that make Bitcoin work.
I for one can’t trust Bitcoin only on the basis that it uses math and software and I doubt that many others will be able to trust Bitcoin on this basis alone either.
Rather, I trust that we are all united by our desire to create rather than destruct wealth and that this universal human preference will see to it that the majority of network participants will do certain things instead of others.
This is why Bitcoin works.
If you would like a deeper dive into how Bitcoin works ‘under the hood’, the book “Grokking Bitcoin” by Kalle Rosenbaum is a brilliant resource. In particular, the book will help you learn about cryptography and how this technology is used in Bitcoin. I have not dealt with these aspects in any level of detail in this article so if you want to go further down the rabbit hole / are looking to take your understanding to the next level then this is the book for you to read and study.
Disclaimer: The comments, views, opinions and any forecasts of future events reflect the opinion of the quoted author, do not necessarily reflect the views of his employers or other professionals working for the same employers, are not guarantees of future events or results and are not intended to provide financial planning, investment or legal advice.
Why Bitcoin Works was originally published in Data Driven Investor on Medium, where people are continuing the conversation by highlighting and responding to this story.