Weekly Thoughts #21: bitcoin miner breakeven theory & miner efficiency


NOVEMBER 1, 2018

For the entire existence of bitcoin, bitcoin mining difficulty has been consistently increasing with very few periods of negative growth. But recently, the bitcoin protocol has lowered the difficulty target by a small amount – an event that has only occurred 13 percent of the time throughout the protocol’s history. The few instances of negative bitcoin difficulty adjustment have primarily occurred around block reward halvings, bear market cycle bottoms, and periods when the protocol itself has been under attack. Negative difficulty growth is impactful because it signals that bitcoin price declines have reached the point where miner profitability is at the threshold for unprofitable miners to exit the industry.

Miners secure the bitcoin network by validating transactions. They do this by collecting transactions into a candidate block and appending the block to the bitcoin blockchain that represents the public ledger of past transactions. A miner can only add a block to the blockchain by solving a computationally-difficult proof-of-work problem and adding this proof-of-work in the candidate block. The problem that miners attempt to solve is to find a hash of the block’s header (combined with a nonce) that is below a given difficulty target defined by the protocol. The probability of finding a hash that is below the difficulty target is low, so miners make many attempts and compete with each other to be the first to solve the problem. As more miners join the bitcoin network, the protocol adjusts the difficulty target every 2,016 blocks such that, on average, a new block is found every 10 minutes.

Join our mailing list to be notified about future reports, research publications and other ideas.

The following plot shows the difficulty level since bitcoin’s inception in January 2009. Difficulty is an unit-less value and was set to an arbitrary value of 1 at inception. The current difficulty level is 7.18 x 10^12 times harder than it was at inception. As more miners have joined the network and as mining technology has improved, the protocol has regularly readjusted the difficulty upward to compensate.

Periods of declining difficulty are rare but occur when either price declines or block reward halvings cause mining to be unprofitable for inefficient miners. During these times, miners exit the industry and the decline in mining hash power causes blocks to be found slightly longer than every 10 minutes. When the next difficulty re-targeting occurs, the protocol adjusts the difficulty downward such that, on average, the next 2,016 blocks will be found every 10 minutes. The following plot shows the two-week growth rate of the difficulty which illustrates how infrequently this occurs. A two-week change is chosen because, on average, the protocol adjusts the difficulty every two weeks.

Historical cycles illustrate that periods of negative difficulty growth occur immediately after the block reward halvings that occurred in November 2012 and July 2016. This makes sense because the halving of the block reward translates into a halving of aggregate bitcoin miner revenue, assuming no change in the price. Why is this significant? It shows that miners are rational actors that respond in predictable ways to mining profitability and do not necessarily mine altruistically to support the network. If mining profitability declines, we should expect miners to exit the industry. More importantly, block reward halvings serve as important points at which inefficient miners are culled from the network. The remaining miners possess more efficient miners, access to cheaper electricity, and have greater economies of scale.

The culling of inefficient miners is an important factor for why bitcoin price has rapidly increased following the block reward halvings in bitcoin’s history. The rapid price increase is driven by three factors:

  1. An increase in perceived scarcity of bitcoin by investors brings capital into the space.
  2. Miners have less bitcoin available to them to sell following the block reward reduction which reduces the constant selling pressure from miners driven by the need to liquidate bitcoin to cover their fiat-denominated expenses. In other words, the maximum amount of bitcoin that miners can sell is reduced because of the decline in the block reward.
  3. Remaining miners are more efficient in that their fiat expenses are lower as a percentage of their bitcoin-denominated revenue (relative to miners that exited the network) which means that they can afford to sell less of their bitcoin to cover their fiat-denominated expenses, thereby further reducing the selling pressure. In other words, the amount of bitcoin that miners need to sell is reduced because of the increase in aggregate efficiency.

The following plot shows bitcoin annualized inflation and the impact of block reward halvings can be clearly seen. Compelled selling by miners to cover their fiat-denominated expenses are a large and consistent source of selling pressure. At the time of this writing, bitcoin annualized inflation is running at about 3.8 percent, and miners have the ability to sell around $12 million worth of bitcoin every day. Therefore, the market needs to be able to absorb this selling pressure in order for the price to remain stable.

Historical cycles also illustrate that periods of negative difficulty growth occur near bear market cycle bottoms. Difficulty growth turned negative in during late 2011 and into 2012 which marked the bottom of the first major bubble-crash cycle. Difficulty growth also turned negative in 2015 which marked the bottom of the previous cycle. These periods serve as important empirical examples to test the miner breakeven theory that the breakeven cost of mining bitcoin somehow serves as a support for the price of bitcoin. Several explanations have been proposed to support this theory, but intuitively it makes sense: miners must expend work to produce new bitcoin and the cost of this work should reflect how much that bitcoin is worth.

The more logical explanation has its roots in the increase in aggregate miner efficiency that occurs when inefficient miners leave the network. As inefficient miners leave the network, difficulty declines, which means that for the miners that are still mining, the revenue for the average miner for the next 2,016 blocks (denominated in BTC) should be slightly higher than the previous 2,016 blocks since the miner base has contracted. And since the remaining miners are more efficient, they are compelled to sell less bitcoin to cover their fiat-denominated expenses thereby reducing the natural selling pressure by miners. This reduction in natural mining pressure may cause the price to rise, which could lead to miners needing to sell even less bitcoin – this creates a virtuous cycle and is the seed from which a new bubble forms. The following plot shows bitcoin price but overlayed with periods where bitcoin difficulty growth is negative.

The difficulty for the current set of 2,016 blocks has been adjusted downward by 3.65 percent. This is impactful because it is only the second time this year that the difficulty has been adjusted downward and it is occurring in the absence of any block reward halving or a coordinated attack by supporters of bitcoin cash. Historical cycles suggest that we may need to experience 6 to 12 months of negative to flat difficulty growth for prices to bottom. Monitoring bitcoin difficulty growth is an important indicator of how close we are to the mining breakeven threshold and if we are going to experience a period where inefficient miners begin to leave the network.

Thanks for reading everyone. Questions or comments, just let us know.

Portfolio Management Team

Thejas Nalval  | Kevin Lu

This Commentary is for informational purposes and does not constitute investment advice, any type of recommendation or an offer to sell or a solicitation to purchase any securities from the Element Digital Funds or an entity organized, controlled, managed by or affiliated with Element Capital Group, LLC (“Element Group”).  Any offer or solicitation may only be made pursuant to a confidential private offering memorandum which will only be provided to qualified offerees for careful review prior to making an investment decision. We aim to educate, report and/or opine on certain developments relating to the digital asset market. These are our subjective views, based on information and sources we believe to be reliable as of the date we publish, but we make no representations or warranties with respect to the accuracy, correctness or completeness of our opinions or any information herein and have no undertaking to update it.  Please do not rely on it.
Element Group and/or its affiliates and personnel have made investments in some of the assets or instruments discussed here, and may in the future make additional investments (long or short) in such assets or instruments without further notice. Element Group does not, here or anywhere, provide accounting, legal or tax advice, or make investment recommendations.  Please do not rely on us for the foregoing. We strongly suggest that prospective investors seek independent advice with respect to any investment, financial, legal, tax, accounting or regulatory issues discussed herein.
Certain information contained in this Commentary constitutes “forward-looking statements,” such as statements that include the words“may,” “will,” “should,” “expect,” “anticipate,” “target,” “project,” “estimate,” “intend,” “believe”.  There are so many risks and uncertainties that actual events or results or the actual policies, procedures, and processes of Element Group and the Element Digital Funds and the performance of the Element Digital Funds may differ materially from those reflected or contemplated in such forward-looking statements.  Please do not rely on forward-looking statements. Past performance is not necessarily indicative of or a guarantee of future results.
This Commentary is confidential, is intended only for the person to whom it has been provided, and under no circumstances may be shown, copied, transmitted or otherwise given to any person other than the authorized recipient.