Author: Cryptocurrency Researcher at Coin Metrics; Translation: Xiaozou from Golden Finance

Key points:

  • In the third quarter of 2024, Bitcoin mining revenue fell to $2.5 billion, down from $3.7 billion in the previous quarter, mainly due to the impact of the halving event in April 2024.

  • Transaction fees remained low throughout the third quarter, with only a brief spike around the halving date, forcing miners to consider other sources of revenue.

  • To stay competitive in a changing environment, miners are exploring a variety of business models, including renting rack space to generative AI and even considering tokenizing whiskey barrels.

  • The empty block problem is still common. Compared with other mining pools, the proportion of empty blocks generated by mining pools such as SpiderPool is abnormally high.

1. Introduction

As we head into Q3 2024, we revisited the Bitcoin mining space, continuing our quarterly update on the state of proof-of-work stakeholders. The impact of the Bitcoin halving event in April continues to reverberate across the industry, with mining yields facing continued pressure due to relatively stable BTC prices and a depressed fee market. This article will explore these factors in depth, and in addition to focusing on BTC price performance, will also comprehensively study and assess the health of the mining industry.

2. Bitcoin mining: a mixed bag

The third quarter of 2024 has proven to be a particularly difficult period for Bitcoin miners, as the mining industry as a whole grapples with the lingering effects of the block reward halving scheduled for April 2024. The reduction in block issuance revenue has put significant pressure on mining profit margins, exacerbating an already depressed fee market.

Our analysis shows that fee income, which previously provided a large supplemental revenue stream for miners, has remained the weakest over the past few quarters. This is very clear in the chart below, where we can see a brief peak around the halving date, caused by a flurry of activity as people tried to mine the valuable halving sats and create inscriptions on the block, but otherwise an almost complete downturn throughout July to September.

This trend is forcing miners to increasingly rely on alternative revenue streams, such as transaction acceleration services and diversification into adjacent infrastructure opportunities — including renting out rack space for generative AI or tokenizing whiskey barrels (a particularly interesting case) — while maintaining lean and efficient operations.

Despite these headwinds, it’s not all bleak when looking at the mining revenue picture. Block rewards, while down from pre-halving levels, remain relatively strong, contributing around $2.5 billion in revenue, though down from $3.7 billion in the previous quarter.

In response to these challenges, many miners are adjusting their business models to diversify beyond pure mining. We are seeing a trend of miners reinventing themselves as general infrastructure providers, actively seeking to provide hosting contracts for power-hungry AI applications. This strategic shift reflects how the mining industry is adapting to changing market dynamics.

At the same time, technological advances in chip efficiency continue to advance at an unstoppable pace. This progress forces miners to face a critical decision: whether to continue using old-fashioned ASIC hardware or to upgrade their mining machines. The final choice may have a significant impact on the competitiveness and profitability of mining companies in the coming quarters.

3. The mystery of the empty block

The empty block phenomenon is an interesting aspect of Bitcoin mining operations. One could argue that empty blocks are somewhat counterintuitive, as the miner who mines an empty block will not receive any income, which would otherwise come from fees paid by users who wanted their transactions to be included in the mined block. So, why is this the case?

First, we have to understand that Bitcoin mining is a very competitive industry. No Bitcoin miner worth their salt is going to forgo revenue unless they have a very good reason to do so. What's worth considering here is the use of block templates in the block mining process. Block templates are used to optimize the transactions in a block to maximize revenue. One of the problems is that creating block templates takes time, so one of the arguments about these empty blocks is that these blocks are mined quickly one after another - because Bitcoin mining is a random process, sometimes miners are destined to be "very lucky" and happen to create a new block shortly after the previous block was found and distributed.

However, if we look at the distribution of Bitcoin block times in 2024, we notice that it is very much like an exponential distribution - as we would expect, the time between blocks can be viewed as a Poisson process, with an average time between blocks of about 10 minutes - and even as high as 47 minutes for empty blocks. Here, we use the consensus_time database from Coin Metrics to avoid the problem of inaccurate self-reporting of block times by miners.

This makes us start to suspect that the cause of these empty blocks is actually that miners do not have time to replace the block template as we assume. In fact, if we dig deeper and see which pools are generating these empty blocks, we see a pattern where some pools (such as Foundry and Binance Pool) show zero empty blocks in a specific time period, while other pools (most notably SpiderPool) contain a relatively large number of empty blocks.

So how can we explain this? There are many arguments to explain this, but one of them supports the fact that mining pools seek to maximize their profits, and if the mining pool has just mined a block, it is in their interest to start working on the block without a template, so as to start working as soon as possible after broadcasting the previous solution. This is because it takes about 15 seconds for a block to propagate throughout the Bitcoin network, which would have given the last mining pool to publish a block a head start in finding the next block.

To test this argument, we can look at the time interval between consecutive blocks (i.e. blocks mined consecutively by the same pool) and compare it to the theoretical distribution of block times to see if a larger proportion of consecutive empty blocks are broadcast earlier than expected. This time we will expand the dataset and look back from 2022.

The above chart shows that the chart content is not what we would expect if miners ignored transactions in the first few seconds, as the distribution is very similar to the expected distribution of a concentrated trend around 10 minutes, and shows quite a lot of such blocks after 10 minutes. This seems to contradict the hypothesis that miners are more likely to produce empty blocks in the first few seconds after mining a block. It is worth noting that the above chart is based on a large dataset that goes back to 2022 and includes 44 blocks, as only 8 consecutive empty blocks were mined in 2024. However, further research is needed to make a clear statement about the continuous empty block mining pattern.

4. Conclusion

The third quarter of 2024 shows an overall transformation trend in the Bitcoin mining industry, struggling to cope with the impact of the halving event in April. Miners are facing huge challenges, including reduced block rewards and continued low transaction fees. In response, many mining companies are diversifying their businesses, exploring other sources of income, and rethinking their hardware strategies.

The empty block phenomenon continues to trouble analysts, with significant differences between mining pools. This trend, along with the ongoing debate over optimal mining strategies, highlights the complex dynamics of the Bitcoin ecosystem.

As the industry has evolved, miners have also demonstrated resilience and adaptability, transforming into general infrastructure providers and exploring new business models. This shows that the industry is actively reinventing itself in the face of changing market conditions. We are likely to see more innovation in the coming quarters as miners strive to maintain profitability and relevance in an increasingly competitive environment.