The Bitcoin Mining Bubble: Why Bitcoin’s low price is an existential threat

Not much can be known for sure about the exact composition of the ‘miners’ that keep the Bitcoin network going, but what we do know is the following:

1. The current total network hash rate is 326,619,735.17 gigahashes/s
2. The current price of one Bitcoin is $260 (USD)
3. The most efficient ‘mining’ computer model available for purchase is the ‘Bitmain AntMiner S5', which does 1155 gigahashes/s using 590 watt-hours of power.
4. The lowest cost of electricity in the two major countries of China and the USA is approximately 8 cents per kWh.
5. A Bitcoin ‘block’ provides a 25 BTC ‘reward’ for miners, and is mined once every 10 minutes on average.
6. Miner’s fees collected from transactions included in blocks have wildly varied, but have always been negligible. Recently, roughly 15 BTC per day on average is being collected, which equates to ~0.1 BTC per 10 minute block.

What we can deduce from these facts is the price at which mining becomes irrational for all miners, as electricity cost exceeds the rewards of keeping the machine running.

First of all, let’s make the most conservative case. Let’s assume every mining machine currently running is a highly-efficient Bitmain AntMiner S5. What would be the ‘break even’ point for the Bitcoin network, with continued mining becoming irrational for 100% of miners? The math is somewhat lengthy, but simple:

326,619,735.17 (network GH/s) ÷ 1155 (Antminer S5 GH/s) = 282,787.65 AntMiner S5 machines

282,787.65 x 590 watt-hours = 166,844,713.5 Wh = 166,844.7135 kWh

166,844.7135 x 0.08 (US cents per kWh) = $13,347.58 per hour

Now, miners earn money through block rewards which come once every ~10 minutes on average, and is currently 25 BTC + ~0.1 BTC in transaction fees, so:

13,347.58 ÷ 6 = $2,224.60 per block

2224.60 ÷ 25.1 (BTC reward) = $88.63 per BTC

So the break-even point for miners even granting a best-case scenario is a $88.63/BTC market price.

If the price were to fall below $88, one would have to assume that the network would cease to function as miners simply turn off their machines until the price rises again and makes it profitable to mine. Note that this says nothing about already sunk costs (i.e. the purchase price of a Bitmain AntMiner S5), so there is no possible rationale apart from ‘goodwill’ out of one’s pocket for the network to continue running.

That’s a best-case scenario, but what about a more realistic one? Here we have to make more questionable estimations, but nonetheless I think reasonable estimations can be made:

1. The average efficiency of Bitcoin miners currently is probably ~1GH/s per watt-hour. This is about the efficiency of most popular miners released in mid to late 2014, and would explain why the network hashrate in recent months has flatlined, as not much improvement in efficiency has been made since.
2. A fair share of Bitcoin miners are probably not Chinese entrepreneurs in warehouses with access to cheap electricity, but enthusiasts in Europe and the USA who each ordered an ASIC or two. Residential electricity rates are closer to 15 cents in the USA, and significantly higher in Europe. Let’s take a middle position, and assume an average electricity rate of 10 cents per kWH.

Using the same math process above, and with the new figures inserted, we get a break-even price of $216.88/BTC.

That’s extraordinary when you consider that only recently the price was at ~$220, and right now it is at $260.

There is no self-regulating mechanism within the network itself that adjusts the Bitcoin price to the cost of mining. Thus, the viability of the network is left to the whim of external market forces, which is largely speculative than based on any real knowledge or foresight. A sudden move down in the price, perhaps due to another surge of ‘bagholders’ from the late 2013 bubble exiting the market, would be enough to render the business of mining completely unprofitable and irrational. Even as of now, the prospects of mining are grim and the sentiment towards it pessimistic, as sunk cost is often (or perhaps always) greater than eventual returns.

This brings me to my conclusion, which is that Bitcoin faces a very real existential threat from miners. Not only will the network slow to a crawl if miners switch off en-masse, but the network may never recover due to the other factor in determining mining rewards which is ‘difficulty’.

‘Difficulty’ means what it says, and it adjusts according to the fluctuations of network hash rate. It’s usefulness is that it regulates the timing of blocks, preventing them from being mined in less than a minute (due to higher-than-normal hash rate), or being mined too slowly (due to lower-than-normal hash rate). Thus, if 99% of miners were to suddenly switch off and cause block/transaction times to blow out to hours or days, then theoretically the ‘difficulty’ would reset this. However, there is one serious caveat: ‘Difficulty’ is measured over and only readjusted after 2048 blocks, to prevent gaming of the system by miners. But if 99% of miners call it quits, and the block time consequently blows out to 16 hours.. then how long will it take to the difficulty readjustment? 16 hours x 2048 blocks = 3.74 years.