Bitcoin is an example of a cryptocurrency for which special computer chips are being used. All transactions with Bitcoin are recorded in something called the blockchain. To add new transactions to the blockchain (also known as ‘mining’), Bitcoin requires you to first solve very complicated puzzles, using powerful computers. This is to make sure that not just anyone can simply add or erase transactions. As a reward for the work, parties who manage transactions receive new Bitcoins and a transaction fee.
Now imagine you own some Bitcoins and want to spend them on a luxury yacht. How nice would it be if you could then erase that transaction from the log, allowing you to spend the same Bitcoins again on something else? In order to do so, you would need to solve puzzles faster than all others within the network. You could then start an alternative version of the blockchain from the moment just before you pay for the luxury yacht. In that alternate blockchain, you omit your previous purchase. The moment your alternative blockchain is longer than the original sequence, your alternate blockchain is the only valid blockchain.
Van Oordt: "To solve those puzzles quickly, you can use ASICS, specialised computer chips designed and optimised to solve Bitcoin's mathematical puzzles. Those chips are expensive, and you can't really use them for anything else. That is exactly why it is expensive to launch an attack on Bitcoin."
The attack, wiping out your transaction or even a complete series of transactions, will initially make you a lot of money. But the expensive chips you bought to carry out the attack could also have been used to continue to mine new Bitcoins. And that, of course, would also make you a lot of money, especially if the value of Bitcoin increases. Van Oordt: “If you carry out an attack, it is likely that the exchange rate of Bitcoin drops. After all, you have proven that using the currency is not secure. With those expensive chips, you can still mine Bitcoins, but the returns will be lower than before the attack. This makes it much harder to earn back your investment in the specialized computer chips.
In previous studies into the safety of cryptocurrencies, researchers did not take into account the high fixed costs of specialised hardware. “In our new study, we show that if a cryptocurrency requires specialised hardware, the costs of carrying out an attack can increase by a factor of 1,000. That means it is only financially attractive to carry out an attack if you can make a lot of money with the initial transaction you erase from the blockchain,” van Oordt said.
Attacks on other cryptocurrencies have already happened. Van Oordt: ‘An attack is only successful if one or more parties who control 51 percent of the computing power within a network are willing to use it for an attack. This has happened in the past for other cryptocurrencies, for example with Bitcoin Gold. But the Bitcoin network is so big and the hardware required for mining is so expensive, that it is very unlikely that a successful attack on Bitcoin would be profitable.”
Van Oordt adds that the puzzles that make Bitcoin safe do have a big downside: the computers that solve the puzzles consume a lot of energy. In other words, security comes at a high price.
More information
The research has recently been published in Management Science.
There is also a publicly available working paper.
Would you like to know more about research at the VU School of Business and Economics? Contact science editor Yrla van de Ven, y.f.van.de.ven@vu.nl or 06-26512492.