Skin in the Game of Centralised Mining

Last week a research firm Bernstein published a report stating that the Chinese bitcoin mining company Bitmain made between $3 billion and $4 billion operating profit in 2017. The estimated profit even exceeds the profit of leading graphic manufacturer Nvidia, which saw high growth of sold GPUs in last year due to mining demand led by retail.

When it comes to Bitmain and its mining operations, the crypto community often reacts negatively towards such news as one of Bitcoin’s main attributes is also decentralised mining. Namely, Bitmain controls about 70% of all mining rigs for Bitcoin through its control of special mining devices ASICs. Further, it controls the largest mining pools and even mines for its own account. What makes the centralisation of the Bitcoin mining industry even more severe is Bitmain’s ownership, which is effectively in the hands of its founder Jihan Wu.

The concept of decentralised mining is important to prevent a 51% network attack, where the attacker would be able to make Bitcoin transactions in its own favour. But let us think about this further what would actually happen in reality if such an attack were to occur? I think this video from Andreas Antonopoulos explains it really well. In the position of mature network such as Bitcoin, vast network resources are needed to make the attack. Assuming this calculation, around $7.6 billion would need to be spent for mining capacities to perform a successful attack.

Let’s assume that Bitmain would be able to take control over the network with some coordination and would transact bitcoins in its favour. By the time attacker tried to convert them all into fiat money, the network and community would already recognise the attack, make a hard-forked Bitcoin and adjust the protocol in their favour — (make ASIC mining irrelevant) and Bitmain would be thrown off the network.

Now, there are two important things to this.

First of all, we need to know that all Bitcoins’s stakeholders (miners, developers, users, wallets, exchanges, etc.) are in this game because of Bitcoin’s value which is determined by certain attributes — one of them being censorship-resistant and limited supply to enable store of value function. What gives Bitcoin its value is mostly its users who will determine the price of an asset because of its attributes. And the crowd can change its mind in a second and cause price and liquidity to crash both to zero and migrate to other hard-forked blockchain. Even the miners who mine in a pool controled by Bitmain might leave because they felt threatend to loose access to community.

Second of all, let’s think about attackers in the case where they are thrown of the network, leaving them with a large amount of non-useful Bitcoins and having no easy access to the new readjusted network mining setting. They end up with non-yielding mining equipment and lose all cash flows associated with it in the future, measured in billions. In such a scenario, the attacker would actually perform the single most stupid business decision ever made.

This is why I believe centralised mining in Bitcoin’s case might actually be very safe for the Bitcoin itself because the centralised miner has huge skin in the game. But of course, it is not as simple as that.

Some would probably argue that centralised mining leads to unfair distribution of mining and transaction rewards and is not aligned with the whole idea of each entity having equal rights to the network and to contribute to it. But then again, one needs to ask oneself what is the most valued thing in Bitcoin, equal rights of mining participation or having a secure network by way of centralised mining and the skin-in-the-game model. Of course, a combination of both would be the ideal security model (decentralised skin-in-the-game) and actually reminds us of Ethereum’s planned proof-of-stake model.

The other argument against centralised mining could also be that the person in charge of it could simply become evil or completely non-rational and would be able to cause damage to the network for other non-monetary reasons. Yet again, at most he would harm the network only for a limited amount of time and the network would not take his efforts seriously (Bitcoin Cash scenario). To prevent such a scenario, some decentralisation of Bitmain would be actually beneficial to have a bit more robust setting. As far as we know, ownership of Bitmain is not very well diversified. There are perhaps a few people involved in the ownership of Bitmain as leaks from the Paradise Papers revealed.

Note also that centralised mining might only be effective if the network is mature and has enough users who will defend their currency by coordinating the development of a “defensive protocol”. In the early stage projects, decentralised mining is much more preferable to centralised mining — skin in the game of a single actor is too low and one could easily overtake the network which could not easily defend itself.

Considering all the facts above I think that in the early stages of crypto network development, users would best be protected from coordinated attacks by having a decentralised mining model, whereas once the network grew to levels where high stakes were needed for an attack, centralised mining might practically be looking just as beneficial for network security as decentralised. Apart from the “reward unfairness” fact, which of course is not something to overlook.

To be clear, personally I would always prefer decentralised mining over centralised, not just because it is superior but because it also goes hand in hand with Bitcoin’s vision of decentralisation and giving all parties equal participation to the network, either making transactions or mining blocks. This article just tries to highlight the skin-in-the-game aspect of Bitcoin’s centralised mining that makes the network more secure than one would normally expect. Importantly, mining is a function of Bitcoin’s price and Bitcoin’s price is a function of expectations of its wide array of users.