Debunking the Bitcoin Death Spiral Theory
Bitcoin critics always seem to be the loudest during bear markets, which means the most ridiculous negative statements about the digital asset are also made during these times. In addition to plenty of claims of bitcoin’s death from the likes of angel investor Jason Calacanis and economist Nouriel Roubini, there is now also the theory that bitcoin will face a “death spiral” in which bitcoin miners decide to completely stop mining the crypto asset.
Like most other types of bitcoin FUD, this is not the first time this theory has been articulated. Just last year, Bitcoin Cash (Bitcash) proponents, such as Bitcoin.com CEO Roger Ver, were giddy at the thought of causing a death spiral on the Bitcoin network through increased adoption of their new altcoin. Since then, bitcash is down roughly 81% against bitcoin.
On “mining death spirals”…
Yet another chunk of FUD we’ve seen multiple times. Specifically late 2011 and late-2014/early-2015. This has been a non-issue for a few reasons. 1/
— Dan McArdle (@robustus) December 3, 2018
The most ridiculous version of the story this time around has come from Santa Clara University finance professor Atulya Sarin in an op-ed titled Bitcoin is close to becoming worthless. Let’s take a closer look at the issues with this theory, using Sarin’s piece as the basis for a response.
The Death Spiral Scenario
A bitcoin death spiral is basically a scenario where a large number of miners decide to stop mining at nearly the same time because the activity is no longer profitable for them. Bitcoin’s protocol rules are structured in a way where a new block of transactions should be mined roughly every ten minutes. Blocks will be found more quickly if more miners throw hashpower at the network (perhaps due to a higher bitcoin price), and blocks will be found less frequently as miners start to leave the network (perhaps due to a falling bitcoin price).
The idea is that so many miners would decide to leave the network after a sharp price decline (or for some other reason) that the network would effectively grind to a halt and no new transactions could be made.
What Sarin fails to explicitly mention in his opinion piece is there is also a mining difficulty adjustment every 2016 blocks. This difficult adjustment will make finding blocks easier if there has been a drop in the network hashrate over the previous 2016 blocks. This effectively means the hypothetical death spiral Sarin predicts must happen within a specific period of 2016 blocks (roughly two weeks when everything is operating normally).
Problems with the Death Spiral Theory
So what sort of decline in hashing power would be required to cause an actual death spiral? How likely is it that such a scenario would play out in the real world?
The sharpest downward adjustment for mining difficulty in the ASIC era just happened on Monday, which made it 15% easier to find a new block. Let’s assume a scenario five times worse than this and say the network hashrate declined 75% within the same difficulty period. In this scenario, a block would be mined every 40 minutes rather than every ten minutes. If the drop happened early in the new difficulty period (let’s say a quarter of the way through it), it would then take six and a half weeks from the previous difficulty adjustment for the difficulty to be adjusted once again (rather than the normal two weeks).
Blocks being mined every 40 minutes on average instead of every ten minutes is not a big deal. Most use cases of on-chain bitcoin transactions (e.g. trading on exchanges, buying and holding, and online purchases) do not need to be completed within an hour or two, let alone ten minutes. The Lightning Network is currently being rolled out as an instant, layer-two payments solution, but it would not be affected by slower block times.
If Bitcoin’s network hashrate declined by more than 75% during a single difficulty adjustment period, it would likely be a sign that something else was seriously wrong with the system.
But let’s say there was something like a 95% drop in the network hashrate in less than two weeks. In such a scenario, transaction fees would have been skyrocketing throughout the decline due to blocks being mined at a slower rate (effectively lowering the supply of block space). This means declines in the profitability of mining could be offset by increased transaction fee-based revenue.
In the past, there have been many instances where transaction fees have accounted for a larger portion of the block reward than the block subsidy (the newly created bitcoin), which equates to at least a doubling of bitcoin-denominated miner revenue per block. Notably, the past four times transaction fees have outweighed the block subsidy in a block reward were during the popping of last year’s bubble. This is the exact scenario that proponents of the death spiral theory would say a death spiral is most likely to occur.
Additionally, there are plenty of people who are invested in the success of Bitcoin who could pay to keep the network going smoothly through ridiculously high transaction fees that effectively increase the profitability of mining.
Finally, if you’re still not convinced that the death spiral scenario wouldn’t happen, the issue could be solved through the use of a change to Bitcoin’s difficulty adjustment algorithm via a hard fork. Yes, hard forks on Bitcoin have proven to be difficult to coordinate in the past, but a situation where the network has literally stopped working is quite different from adding new features or making a change that will alter the costs associated with operating a full node.
There would be no split caused by the hard fork because everyone would move over to the new network with a different method for dealing with difficulty adjustments. It’s unlikely that another block would ever be mined on the old chain.
To be clear, the chances of the situation getting to the point of a hard fork approach zero in my view.
“But This Time It’s Different”
Sarin also claimed that the threat of a bitcoin death spiral is different this time in his recent article. The irony is extremely strong here, as “this time is different” is often what people say when they’re investing in a bubble (as noted by many bitcoin skeptics). Siran first laid out three reasons why things are different this time in a separate article back in April.
In Siran’s view, things are different this time because the magnitude of the decline in 2018 is much greater than past declines, those losing money this time around are newcomers to the space who will be quick to exit, and bitcoin futures products did not previously exist for miners.
These three points are put together by Siran to tell one story, which is that the vast majority of the hashrate is currently controlled by newcomers to the bitcoin space who are only interested in short-term profits (powered by the futures markets) and they won’t be able to continue mining at a loss because the magnitude of those losses will be much greater than losses miners operated at in the past (due to the historically high nominal price swings).
First of all, the futures market bit isn’t true. Miners have been able to hedge against bitcoin price volatility risk since at least 2013.
In terms of the magnitude of this most recent bubble, yes it was larger than past bubbles on a nominal basis; however,, it should be remembered that — as already explained — the death spiral scenario only begins to take shape if everyone decides to shut off their mining equipment at roughly the same time.
Bitcoin’s network hashrate expanded from 5 exahashes per second to 60 exahashes per second over the past year and a half, but the hashrate has already declined back down to 37 exahashes per second. Some mining operations, such as GigaWatt, have already gone bankrupt, but different mining operations have different assumptions regarding the bitcoin price built into their business plans.
It must also be remembered that as some mining operations go offline, the mining operations that remain online become more profitable after the next difficulty adjustment because there is less competition for mining blocks. In other words, a miner’s break-even point could move from $3,000 to $1,500 if half of the network hashrate goes offline.
Of course, pretty much every business plan is destroyed if the bitcoin price falls far enough fast enough, but the sort of price drop Sarin is predicting would be epic even by bitcoin’s standards. If such a price drop did occur, Bitcoin may have already failed for some other reason.
While Sarin may even be correct in saying that the current environment may make it more likely that there is a dramatic decline in hashrate, it’s sort of like pointing out that you’re more likely to get struck by lightning than win the lottery.
It should also be noted that bitcoin mining is generally a long-term investment involving the initial purchase of hardware, long-term electricity contracts, and other upfront costs. Some miners might be more trigger happy than others when it comes to turning off their equipment, but everyone isn’t going to get up and leave at the same time — and again, that is the most important point when it comes to declines in hashrate not leading to death spirals.
There are other pieces of nonsense trickled throughout Sarin’s article, such as saying the bitcoin price is mostly driven by the cost of mining, but that and other topics are out of the scope of this article.