Expected Revenue, Risk, and Grid Impact of Bitcoin Mining: A Decision-Theoretic Perspective

Bitcoin mining is often imagined as a static, relentless machine, a factory that hums day and night, consuming vast amounts of electricity regardless of the cost or the needs of the power grid. In reality, these operations are far more like a high-stakes lottery where the ticket price fluctuates by the second. The miners are not passive consumers; they are active participants who constantly weigh the odds of winning a digital prize against the cost of the electricity required to play. This delicate balance between risk, reward, and energy use has become a critical issue for power grid operators, who must plan for massive loads that can vanish or surge in minutes depending on market conditions. To manage this, planners need to understand not just how much power a mine uses on average, but how likely it is to shut down or ramp up when prices change.

A team of researchers at Texas A&M University has developed a new way to look at this problem, moving away from simple historical snapshots to a forward-looking statistical model. Instead of guessing future profits based on what happened yesterday, they treat the entire mining process as a series of tiny, independent chances to win. Imagine a miner's computer trying to solve a puzzle; every single attempt is a roll of the dice. Most rolls fail, but occasionally one succeeds, earning a reward. The researchers built a unified framework that calculates the expected income, the risk of losing money, and the potential for huge profits based on these fundamental probabilities. Their work reveals that the size of a mining operation and its decision to join a shared group, known as a mining pool, fundamentally changes its risk profile and its behavior as a power consumer.

The study begins by breaking down the mining process to its simplest form. Each time a computer performs a calculation, it is essentially buying a lottery ticket. The chance of winning is incredibly small, determined by the difficulty of the network, but the reward is substantial. Because the outcome of any single calculation is uncertain, the total income over a day or a year is not a fixed number but a range of possibilities. The researchers showed that for a miner to feel confident about their income, they need a massive number of these attempts. If a facility is too small, the luck of the draw can cause wild swings in revenue, making it impossible to predict whether they will make a profit or lose money. To smooth out these swings, miners often join a pool, where many small operations combine their computing power to win more frequently, sharing the rewards and fees. While this reduces the risk of a bad month, it also lowers the average potential profit because of the fees paid to the pool.

Using real data from 2023 and 2024, the team calibrated their model to match actual industry performance. They looked at major mining companies like Riot Platforms and Marathon Digital, comparing what their models predicted miners should earn against what those companies actually reported. The results showed a consistent gap: the actual earnings were often lower than the theoretical maximum, a difference the researchers attribute to the inherent volatility of the lottery and the costs of joining pools. This gap represents an "opportunity cost," a real financial trade-off miners make to stabilize their cash flow. The model allows a miner to calculate exactly how many machines they need to run to keep their risk below a certain level, or how much of their fleet they should send to a pool to guarantee a minimum income while keeping the rest free to chase higher rewards.

One of the most practical outcomes of this research is a clearer understanding of when a miner will turn off their machines. The researchers calculated a specific break-even electricity price for a typical mining machine. If the cost of power rises above this threshold, it becomes cheaper to stop mining and wait for prices to drop. However, this threshold is not fixed; it shifts with the price of Bitcoin, the reward for finding a block, and the efficiency of the hardware. For example, if the price of Bitcoin drops by 20%, the break-even point for electricity drops significantly, meaning miners will shut down at much lower power prices than before. The study also found that miners who participate in pools have a slightly lower break-even point, making them slightly more willing to keep running during expensive periods, but also less flexible in their response to market shocks.

For the people who manage the electrical grid, this new perspective is vital. Current planning methods often treat mining as a fixed load or a simple switch that can be turned off. This research suggests that mining is a state-dependent load, meaning its behavior changes dynamically based on a complex mix of financial and technical factors. A grid operator cannot simply assume a mine will cut power when the price hits a certain number; they must understand the specific risk tolerance and pool arrangements of that mine to predict its reaction accurately. By using this statistical framework, grid planners can better forecast how much power will be available or how much can be cut during emergencies, leading to more reliable electricity systems.

The authors conclude that their approach provides a solid foundation for analyzing the economic and energy impacts of Bitcoin mining. It moves the conversation beyond simple estimates of total energy use to a nuanced understanding of how miners make decisions under uncertainty. This clarity helps both miners design better strategies for their operations and grid operators create policies that can effectively harness the flexibility of these massive loads. While the model simplifies some complex details like transaction fees and specific local constraints, it offers a rigorous, unified way to view the interaction between digital currency and physical power systems, ensuring that future planning is based on the true probabilistic nature of the mining game.