Bitcoin: From Success to Self-Destruction?

Throughout 2013 Bitcoin was highly volatile (as figure 3 below illustrates); at the start of last year its capitalisation amounted to no more than US$200m. Due to the digital currency’s flexibility and openness, a rich ecosystem surrounding Bitcoin is thriving. Users can also embed scripts into Bitcoin transactions, opening up many potential future applications.

Background

In 2008 an individual using the pseudonym Satoshi Nakamoto designed the original Bitcoin protocol, launching a network the following year. Reports suggest that he still has around one million Bitcoins.

Bitcoin is an electronic payment system based on peer-to-peer (P2P) network of anonymous strangers without a trusted third party. How does one organise a payment system on this basis? Most known payment systems have a trusted party at their centre – most obviously a bank -which, when instructed, will credit one account and debit the other account to facilitate payment.

The trusted third party can also facilitate account-free payments with crypto currency, see for example www.MysmartEmoney.com. Bitcoin does not rely on a trusted central authority, but instead uses a network of unrelated strangers. Yet provided the majority of these strangers are honest, payments should be correctly accounted for.

In traditional payment systems all transactions are potentially secret and outside the public domain – assuming hackers do not comprise the bank’s computer. In the Bitcoin system transactions are open to the public, but instead of names or account numbers cryptographic keys are used. One may visualise the Bitcoin as a chain of digital signatures; each owner transfers the coin to the next by digitally signing the coin using asymmetric cryptography.

Asymmetric cryptography refers to a cryptographic algorithm requiring two separate keys, one secret (or private) and the other public. These two keys are mathematically linked. Bitcoin uses the public key to verify digital signatures and the address or name of the owner, as long as it stands as last in the chain; while the private key is used to create the digital signature. It is computationally easy for users to generate their own public and private key-pairs.

The strength lies in the fact that it is computationally infeasible for a properly-generated private key to be determined from its corresponding public key. Public keys are used also as the recipients’ addresses in the Bitcoin system. As there are so many potential public keys, users need not concern themselves as to whether someone else has created the same public key. It’s possible to generate as many public private key pairs as one wishes, as the Bitcoin system permits almost innumerable variations.

How Bitcoin System Works

You own a Bitcoin if your public key stands as last in a chain and you know the corresponding private key. As an example the public key could look as follows: 1EgiEMCAM7PsEnEpdsZKUeQgtmaDmDjSaw*.

This public key has only one private key. When the owner knows this private key he/she can then use it to transfer the Bitcoin to somebody else’s public key or back to his own public key. As mentioned, Bitcoin operates as a P2P system of many computers called nodes, or ‘miners’. So when an individual initiates a transfer of his/her Bitcoin the new transaction is broadcast to all nodes. Each node collects the transaction into a block and must then verify whether the transaction is a valid one.

An individual signing a transaction uses his private key. The nodes check that the correct private key was used; once satisfied the transaction is then verified (having checked that the input has not already been used) and added to the block. The node also checks previous transactions related to that specific Bitcoin. Through the mathematics behind the digital signature the nodes can verify an individual’s signature, without actually knowing his/her private key, as they use the public key. Importantly, because the signature depends on the message, it will be totally different for any change – no matter how small – and therefore cannot be used by anyone else for a different transaction. Nor can anyone else modify the transaction while passing it along the network.

Ownership of Bitcoins passes along in a kind of chain, with the validity of each transaction dependant on the validity of previous transactions. Bitcoin nodes keep track of a giant list of transactions. Owning Bitcoins means that there are transactions in the list which point to the owner’s name or, more accurately, his/her public key. To keep track of their Bitcoin the owners install the wallet software on their Smartphone or PC, which iterates through every transaction and shows their Bitcoins, which they now can spend with the help of their wallet.

There is no bank or credit card company to keep track of the Bitcoin owner’s personal record and to which (s)he can appeal. Lose the private key and any funds associated with the corresponding public key will be lost forever. To avoid that outcome it is possible to have the various transactions linked together as they are all publicly stored on various computers. The owner can generate a new public key for every incoming transaction and provide this information to the payer, as illustrated below in Figure 1.

Figure 1: Illustration of a Bitcoin Wallet:
TMS-Bitcoin-fig-1_Edit.gif
The Bitcoin system puts transactions in blocks and links the block together into something called the block chain. Each block carries a reference to the previous block. Transactions in the same block are considered to have happened at the same time and transactions not yet in the block are known as unconfirmed transactions. Any node can collect, verify and set into block the unconfirmed transaction and broadcast it to the rest of the network.

With a network as huge as Bitcoin, potentially many nodes could finish this work at the same time. Bitcoin’s solution to this challenge is that each valid block must contain the answer to a very specific mathematical problem, which means solving the proof-of-work function. Once solved, the hash is like a fingerprint that uniquely identifies that block. The node, which solves first the proof-of-work protocol, broadcasts the block to the network and is awarded 25 Bitcoins (at the time of writing around US$20,000) for its endeavour.

As there are so many nodes solving proof-of-work protocol, solving it is akin to winning the National Lottery. You could buy thousands of lottery tickets but your odds of becoming the winner are considerably less than it being somebody else. Likewise, you would need to control half of the total computing power in the entire network to have a 50% chance of solving the proof-of-work function before someone else.

Each block in the chain is protected by a hash function, whose value depends on that of the previous block in the chain. Sometimes, two blocks are published nearly simultaneously and a fork in the chain can occur. Nodes are programmed to follow the longest block chain – this means the block chain whose total proof-of-work difficulty is the largest – and to abandon other blocks. Transactions from the abandoned block go into a pool of unconfirmed transactions and will, eventually, be collected by the prevailing branch.

Mining and Pools

Bitcoins are transferred through digital signatures and transactions chains. The order of transactions is protected in a block chain. To send Bitcoins, the individual must reference previous transactions (called inputs), where he/she was the recipient of Bitcoins. As a means to slowly (every 10 minutes) generate and distribute coins, a reward is given to whoever solves first the proof-of-work protocol.

This is where the term ‘mining’ comes from. Every four years the reward is reduced by half, so eventually no more coins will be released once the total reaches about 21m. To date 12.3m Bitcoins have been created. A figure of 21m might not sound like very much but as it is possible to send down to 1 100 millionth of a Bitcoin (0.00000001) there will be enough Bitcoins to facilitate various payment transactions. Once mining rewards cease, what incentive will nodes have to process the transactions? In addition to the assignment of new Bitcoins, nodes can charge transaction fees. Currently nodes process transactions without fees, as the main incentive is mining reward.

Figure 2: Bitcoin Currency Statistics:
TMS-Bitcoin-fig-2_Edit2
As figure 2 shows, as at 23 January 2014 miners created 4600 Bitcoins and received an additional 17.01 Bitcoins in transaction fees. Total miners revenue (new Bitcoins plus transaction fees) amounted to US$4,373,157.39. Miners processed a total of 64,014, giving a cost per transaction of US$68.32. The average transaction amounted to US$1,286 and transaction costs, as a percentage of the average transaction, amounted to 5.3%. By contrast, a credit card transaction where transaction costs exceed 5% is rare. However, as far as Bitcoin is concerned, the costs do not end here. At some point it will become necessary to convert Bitcoin into dollars or another currency. Bitcoins have very high volatility, as shown below in figure 3.

Figure 3: Bitcoin priced in US dollars (USD), February 2013 to February 2014
TMS Bitcoin fig 3

It would be impractical to offer prices in Bitcoins where your costs are in dollars. When Bitcoins are used just to execute payments, then volatility is less important, as Bitcoin is converted immediately into a base currency. However, doing so incurs additional conversion costs. To convert Bitcoins to US dollars (USD) or other currency, owners can use exchanges such as Kraken (www.kraken.com) or Mt.Gox (www.mtgox.com). As per figure 4 below, on 24 January this year, Kraken showed the following rates for Bitcoin, shown by the abbreviation XBT, against USD, the euro (EUR) and the Korean won (KRW), as well as alternative currencies Litecoin (LTC); Ripple (XRP); Ven (XVN) and Namecoin (NMC).

Figure 4: Recent Exchange Rates:
TMS-Bitcoin-fig-4_Edit.gif

As shown by figure 4 the spread between bid and offer for XBT/USD amounts to US$15.13 or 1.9%. The half of this spread (middle rate to bid or offer) can be seen as conversion costs.
In addition to the spread, Kraken typically charges a percentage fee of 0.3% for amounts below US$1,000 and a fee of 0.07% or lower for amounts over $ 500,000. So conversion costs can easily amount to 1.25%.

Conversion costs and Bitcoin transaction costs amount together to 6.5%, meaning the Bitcoin system has the highest payment fees when the reward for mining is included than any system since Croesus first manufactured coins from gold around 650 BC. At the moment users do not directly pay most of these costs, because miners are paid with new Bitcoins. On average every 10 minutes 25 new Bitcoins are created out of the thin air, so to speak.

Problem of Double Spending

Nodes do not accept an invalid transaction. Hashes protect every transaction in the block, with the hash of each block depending on that of the previous block. If an attacker attempted to make a change – for example adding, removing or changing a transaction in the block – he/she would need to recalculate and provide new hashes for every following one. Each hash includes the proof-of-work protocol, so it would take considerable time, even for the fastest computers.

An attacker can try only to change one of his own transactions and attempt to take money he has already spent, because he does not know the private keys applying to any other transactions. In other words a double spending attack is an attempt to convince the payee that the transaction has been confirmed, and then convince the entire network to accept some other transaction.

The more central processing unit (CPU) the attacker has in relation to that of the entire network, the higher the potential for such an attack to succeed. Normally the CPU power of each individual computer is miniscule against that of the network and solving the proof of work protocol can be compared to winning the lottery. It would take, on average, several years for a node to solve the proof-of-work protocol whereas the network needs 10 minutes, so the odds of ever solving one before the rest of the network is very low.

To receive a steadier stream of income, many people join groups called mining pools and work collectively to solve proof-of-work protocol. One such mining pool, BTC Guild, has created six blocks in a row. As of 9 January 2014 a mining collective known as GHash.IO accounted for 45% of the-then present computing power. Normally double spending should not be an issue and waiting for six confirmation – the creation of six blocks after the respective transaction has been announced to the network – might represent overkill for a casual attacker but potentially is not sufficient if any group accumulates more computing power. With more than 50% computing power the group has the ability to confirm all Bitcoin transactions on their own.

Recent problems experienced by the Mt.Gox exchange can be seen as resulting from the double spending issue. Mt.Gox suspended transactions from wallets that it holds to external Bitcoin addresses, but continues to allow conversion of Bitcoins. The exchange claimed that the issue was due to a bug in the Bitcoin software, which makes it possible to alter the transaction details of a previously-executed transaction. However, the Bitcoin Fondation rejected the claim, insisting that Mt.Gox was experiencing its own technical problems, which had been known about for some time. It appears that Mt.Gox wallet software was not handling the double spending issue in accordance with the original Bitcoin protocol. As a result, the Mt.Gox wallet believed that some coins were available for spending which, in fact, had already been spent and allowed them to be spent again. The episode underlines the fact that Bitcoin is still a work in progress and the technology in use is partly immature.

According to the current technological standard the attacker can only change his own transactions to which he knows the private key. The possibility cannot be ruled out that sometime in the future a powerful supercomputer will be able to deduce the private key from the available public key. So possibly Bitcoins whose users lost their private keys are not lost for ever after all.

Future of Bitcoin

One should distinguish between current Bitcoin system and Bitcoin protocol. The ‘Achilles heel’ of Bitcoin is its incentive system and technical progress of computer power. Bitcoin specifies an exponentially decreasing rate of money creation that ultimately sets a cap on how many Bitcoins will be created. As ‘free’ payment in the form of new Bitcoins is successively phased out, a transaction fee will be required for motivating honest nodes to verify and account for new transactions.

The incentive to obtain ‘free’ Bitcoins has encouraged miners to invest in better, faster, stronger and more expensive computers. Initially it was relatively easy to solve the proof-of-work protocol as an average desktop or laptop could handle the task. Today, anyone with just a regular PC has little chance of mining Bitcoins – special computers created solely for the purpose and carrying US$15,000+ are required. It provides a bonanza for companies such as HashFast, KnCMiner and Batterfly Labs – somewhat like the gold rush of 1849, which transformed California while providing little benefit to the miners.

The Economist magazine estimates the total current computational capacity of various Bitcoin nodes amounts to 50,000 petaflops, equivalent to 100 times the performance of the world’s top 500 computers combined. According to Bitcoin wallet and block explorer Blockchain there are approximately 75,000 Bitcoin transactions daily.

The computer infrastructure of 50,000 petaflops is estimated as costing around US$8bn as at the end of 2013. Assuming that over next three years 75,000 transactions will occur daily, while depreciating the current computer infrastructure over three years and without taking used electricity into account produces a computing cost per transaction of US$97.41. Yet as figure 2 shows, the current costs of Bitcoin transaction amount to US$68.32, with less than 25 cents – or only 0.37% – generated by fees. The huge majority of transaction costs are paid for by ‘free’ Bitcoins.

Nevertheless these extraordinary high transaction fees are still not enough to cover the costs of current computer infrastructure. Bitcoin nodes as a community are losing money with little prospect of change on the horizon. The self-destructive feature of Bitcoin lies in the necessity to have more available computer power in the network than any malicious super computer can assemble. This thwarts any attempt to implement a vicious Sybil computer security attack that could gain disproportionally large influence on the network and defraud people by stealing back the payments or using it to generate new coins.

According to Moore’s law, computer power doubles every 18 months. Therefore nodes have to invest again and again in the infrastructure to stay ahead of the curve. This scenario makes it inconceivable that fees required to sustain the Bitcoin infrastructure will ever come down. However the mining process will contribute less and less. In 2017 the reward will be reduced to 12.5 Bitcoins and in 2021 to 6.75 Bitcoins, etc. On the other hand it is also not plausible that Bitcoins user will be willing to pay transactions fees amounting to 10% or more of the average transaction. Therefore the Bitcoin system as we know it currently is ultimately doomed and will collapse.

Future of Bitcoin Protocol

Bitcoin protocol has great potential to revolutionise the payments world. The currency has already produced many offshoots, such as Litecoin, PPCoin, Mastercoin and Ripple. Bitcoin protocol is open source protocol.

According to MIT License: “Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (‘the Software’), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so..”

The rich ecosystem surrounding Bitcoin is thriving; opening many opportunities for anyone who wants promote or develop digital currencies. He/she need only adopt and amend Bitcoin according to their needs. The important objective of Bitcoin’s creator was establishing a system without central authority. As the current Bitcoin system shows this is achievable but also comes with huge disadvantages.

Bitcoin software has to run on many thousands of computers and each transaction must be populated to every computer as well. The network of strangers can run a monetary system but the costs are huge. Any bank could implement many features of Bitcoin protocol without its P2P network and implement the system on the central computer. The issue of double spending would disappear, any transaction could be confirmed immediately, and significant electricity and computer power would be saved. The cost per transaction would fall sharply, potentially to nothing. It’s not only a bank that could implement digital currency – a nation state could declare its own version of Bitcoin as legal tender.

Everyone has the right to privacy and transactions that can be executed as per current Bitcoin system. However the state also has an interest in ensuring that the system is not undermined by underworld activities. Were a central authority to run a system derived from the current Bitcoin system it could ensure that transactions over a certain amount can only be executed to payees whose public keys are known or can be made known to tax and other governmental authorities – thereby avoiding money laundering and tax avoidance. Ultimately, it seems likely that while digital cash will continue to evolve it will be differ from the current Bitcoin system, although part of Bitcoin protocol will have a huge impact on future payment systems.

Conclusion

Bitcoin is not the first digital currency, but to date it has proved the most successful while opening a door for other crypto currencies to thrive. Bitcoin in its current version has a built-in self-destruct mechanism. The costs per transaction will need to rise to sustain a necessary computer power infrastructure but the rewards from mining – issuing free Bitcoins – will progressively go down. Even the current cost per transaction of US$68.32 is not enough to cover necessary investments. Miners as a group are losing money from their current process, yet it is unrealistic to assume users are willing to pay full transaction costs. The logical conclusion from this is that the system has to collapse.

*It is possible to Google this address and see transactions related to this public key.

62 views

Related reading