Hashprice Explained: Bitcoin Mining’s Master Metric in 2026

Every Bitcoin miner, from a hobbyist with one machine humming in a spare room to a public company sitting on a gigawatt of contracted power, wakes up to the same question: is the electricity I am about to burn worth more than the Bitcoin it will produce? Hashprice is the number that answers it. In a single quote it compresses four moving parts, the block subsidy, transaction fees, network difficulty, and the price of Bitcoin, into one figure: how much a unit of computing power earns in a day.

For most of 2026 that number told a bleak story. Hashprice opened July at $28.09 per petahash per day, the fifth-lowest daily reading ever recorded, according to Luxor’s Hashrate Index. It has since clawed back to around $32, but the recovery says as much about miners switching machines off as it does about anyone wanting to buy blockspace. To understand why, you have to understand the metric itself: what it measures, how it is built, and why three of its four inputs are quietly working against miners all the time.

This is a deep guide to hashprice as it stands in August 2026: the math, the current snapshot, the two-year slide and the summer bounce, the leap from revenue to actual profit, the derivatives market that now lets miners hedge it, and the artificial-intelligence boom that has turned every megawatt into a choice.

What Hashprice Actually Measures

Hashprice is the expected revenue a miner earns from a fixed amount of hashrate over a day. It is usually quoted in US dollars per petahash per second per day, shortened to dollars per PH per day; divide by 1,000 and you get the terahash version that shows up on machine spec sheets. Luxor, the mining-services firm behind the Hashrate Index, coined the term in 2019 to give miners a single ticker for profitability, the way a stock price gives an investor a single number instead of a spreadsheet of fundamentals.

There are two flavors, and the difference matters. USD hashprice is what most people mean: expected daily revenue per unit of hashrate, measured in dollars. Because Bitcoin is priced in dollars, this figure rises and falls with the market. BTC hashprice measures the same thing in Bitcoin terms, which strips the price out and isolates the pure network economics, difficulty and fees. As of the Hashrate Index roundup for August 3, 2026, USD hashprice sat at $32.10 per PH per day, equal to 0.00050222 BTC, down 1.3 percent on the week.

The cleanest way to think about it: hashprice is to a miner what the day rate is to a drilling rig, or revenue-per-room to a hotel. It is a revenue metric, not a profit metric. Two miners quoting the identical hashprice can be worlds apart on the bottom line, because one pays four cents per kilowatt-hour and runs the newest chips while the other pays ten cents on hardware from two cycles ago. Hold that distinction; it drives half of what follows.

The Formula, Decomposed

Luxor describes USD hashprice as a function of four inputs: the block subsidy, transaction fees, network difficulty, and the Bitcoin price. The intuition behind it is simple arithmetic. The network mines roughly 144 blocks a day, each paying the current 3.125 BTC subsidy plus whatever fees its transactions carry. Value that daily reward in dollars, then divide by the total hashrate racing to win it, and you have the revenue each unit of hashrate can expect.

Plug in current numbers. About 144 blocks per day times 3.125 BTC is 450 BTC of subsidy. At roughly $65,000 per coin, per CoinGecko, that is near $29 million a day before fees. Spread it across a network running around 932 EH per second, which is 932,000 PH per second, and each PH earns on the order of $31 to $32 a day, almost exactly where the index prints. The math is reassuringly boring, which is the point: hashprice has no hidden magic, it is total reward divided by total effort.

The published index refines two of those inputs. Rather than use estimated hashrate, which is noisy because true hashrate is never directly observed and only inferred from block times, Luxor anchors the denominator to network difficulty, the one honest, on-chain measure of how hard mining is. It also smooths fees with a 144-block lagging average and averages the Bitcoin spot price across three US exchanges. The takeaway that should stick: of the four inputs, three trend against miners over time. The subsidy halves every four years, difficulty grinds higher as machines improve, and only price and fees can push hashprice back up.

A Hashprice Snapshot, August 2026

Here is where the metric and its inputs stand as this piece goes out. Every figure is drawn from a live source; hashprice is volatile, so treat these as a moment in time rather than a fixed level.

Two things jump out. First, the fee share is tiny; the subsidy is still doing almost all the work, an issue we return to below. Second, network difficulty is set to tick up again at the next retarget on roughly August 8, a reminder that even in a weak market, competition keeps nudging the denominator higher whenever price gives miners a reason to switch machines back on. You can watch the live figure move on The Block’s hashprice index as well.

How Hashprice Collapsed: Anatomy of a Two-Year Slide

To appreciate why $32 counts as relief, look at where hashprice came from. The April 2024 halving cut the block subsidy from 6.25 to 3.125 BTC, instantly halving the largest term in the formula. A halving is a step-change to revenue that miners cannot vote away; the only offsets are a higher price or more fees, and neither arrived fast enough.

Then hashrate exploded. Fresh, efficient ASICs shipped through 2025, and Bitcoin’s autumn record high made expansion look cheap. Per the CoinShares Q1 2026 mining report, the network first crossed 1 zettahash per second in August 2025 and peaked near 1,160 EH per second that October before easing to around 1,045 EH by late December. Every machine that plugged in diluted everyone else’s slice of a fixed daily reward.

The squeeze bit hard. CoinShares put the weighted-average cash cost to produce a coin across public miners near $79,995 in the fourth quarter of 2025, with an estimated 15 to 20 percent of the global fleet running at a loss. JPMorgan analysts led by managing director Nikolaos Panigirtzoglou pegged the industry’s all-in production cost near $78,000 and noted that Bitcoin had traded roughly 19 percent below that level for five straight months, per coverage at TFTC. When revenue per machine falls below the cost of the power feeding it, something has to give. Miners realize that revenue through pools, which convert a brutal solo lottery into steady payouts; the mechanics of who actually assembles the blocks are covered in our guide to Bitcoin mining pools in 2026.

The 2026 Recovery: When Falling Difficulty Puts a Floor Under Revenue

Here is the counterintuitive engine of the summer bounce: hashprice recovered not mainly because Bitcoin soared, but because difficulty fell. When unprofitable miners power down, fewer machines chase the same fixed daily reward, and at the next retarget difficulty drops to match. The survivors then command a larger share of that reward, so their revenue per unit of hashrate rises even if the price barely moves. It is Bitcoin’s built-in thermostat.

July 2026 was a clean demonstration. Difficulty fell across two consecutive downward adjustments, 5.00 percent on July 11 (the fourth-largest decrease of the year) and 0.74 percent on July 25, a net drop of 5.71 percent from 133.87T to 126.23T, per the Luxor Hashrate Lookback. Back-to-back negative adjustments are rare; CoinShares noted the first such streak since 2022. As difficulty gave way, hashprice climbed from that $28.09 open to a peak of $32.92 on July 21 and closed the month at $31.95, up 13.7 percent, helped by Bitcoin rising from $59,304 to $63,577 over the same weeks.

The JPMorgan desk described the mechanism precisely. As their note put it, and as reported by TFTC: “When bitcoin trades below its production cost, higher-cost miners power down, the hashrate declines, and difficulty adjusts lower.” That is the floor under hashprice. It does not guarantee profit, but it stops the bleeding, and it is why a five-year low is usually a signal that the weakest capacity is about to leave. By early August, hashprice held near $32.10 while the seven-day hashrate rebounded 6.2 percent to about 932 EH as some machines came back online.

Hashprice vs Price: Why a Rising Bitcoin Does Not Always Help

A common misconception is that a higher Bitcoin price automatically makes miners rich. It does not, at least not reliably. If price rises but difficulty rises faster, USD hashprice can still fall, because the denominator is growing quicker than the numerator. This is exactly why the BTC-denominated hashprice is worth watching: it removes the price and shows the raw network economics underneath.

That divergence was visible in the summer. The Lookback noted the BTC hashprice reached its highest monthly average since August 2025 even while the USD figure remained historically low, meaning the per-coin economics improved (difficulty fell) faster than the dollar value recovered. For miners, the denominator is the real adversary. Every competitor that plugs in dilutes the pool. Fred Thiel, chief executive of MARA Holdings, framed it starkly to CoinGeek: mining, he said, “is a zero-sum game. As more people add capacity, it gets harder for everybody else. Margins compress, and the floor is your energy cost.”

The other half of the equation, the price itself, is set far from the mining rig. It moves with spot demand, exchange-traded fund flows, and monetary policy, forces a miner cannot influence with a single watt of hashrate. When the Federal Reserve held rates in the summer of 2026, crypto barely flinched, as we covered in our FOMC reaction; that macro backdrop, not anything happening on-chain, is often the swing factor in USD hashprice from one month to the next.

From Hashprice to Margin: Turning Revenue Into Profit

Hashprice is gross revenue per unit of hashrate. To get profit, subtract power, hosting, pool fees, overhead, and depreciation. The gap between the first two of those and the last is where miners live or die, and it explains why the industry talks about cash cost versus all-in cost.

Riot Platforms offers a clean worked example. In its Q1 2026 results, the company reported a cost to mine one Bitcoin, excluding depreciation, of $44,629 against a production value of $75,964 per coin. On a cash basis, comfortably profitable. But include depreciation on its enormous fleet and the cost jumps to $96,283, or about 127 percent of production value, which is a GAAP loss in the same quarter. Same hashprice, two very different answers, depending on whether you count the wearing-out of the machines.

The most useful lens for an individual operator is the energy-only break-even: the power price at which a given machine exactly covers its electricity. A rough rule works well. Break-even dollars per kilowatt-hour equals hashprice divided by 1,000, then divided by the machine’s efficiency in joules per terahash times 0.024. Run it at the current $32 hashprice for a few common models.

These are energy-only numbers; they ignore hosting, pool fees, overhead, and depreciation, so the true break-even power price is lower still. The message is blunt: at $32, the newest chips tolerate near-ten-cent power, while a legacy S19 XP needs electricity around six cents or cheaper just to break even on energy. The Lookback pegged July’s marginal break-even efficiency near 27 J/TH, with the 25-to-38 J/TH tier unprofitable for a second straight month. On a per-megawatt basis, Hashrate Index estimates efficient sub-14 J/TH fleets earned roughly $109 per MWh mining, while inefficient 25-to-38 J/TH machines earned about $41; keep those two figures in mind, because they set up the AI comparison later.

Fees and the Security-Budget Question

Look again at the snapshot table and the smallest number on it: fees were about 0.77 percent of the block reward in early August, per Hashrate Index. The subsidy still supplies roughly 99 percent of miner revenue. That is fine today. It is the long-run problem hanging over the whole model.

The subsidy halves on a fixed schedule: 3.125 BTC now, 1.5625 in the spring of 2028, and onward toward zero around 2140. Each halving knocks the biggest term in the hashprice formula in half overnight. For hashprice to hold through the next one, either Bitcoin’s price has to keep roughly doubling every cycle, which no asset does forever, or transaction fees have to grow into the gap. Many researchers argue fees eventually need to reach north of 20 percent of revenue to fund a credible long-term security budget. From 0.77 percent, that is a long way off.

What makes fees the great unknown is that they depend on demand for blockspace, and that demand is fickle. Inscriptions, token protocols, and layer-two settlement have all produced fee spikes, then faded. The ongoing fight over what belongs in a Bitcoin block is a live one; see our coverage of Bitcoin’s blockspace war over Taproot and BIP-110. And the direction of travel on rollups matters too: if layer-two networks keep pulling activity onto their own rails, the base-layer settlement demand that ultimately drives fee revenue becomes the swing factor for how much miners can ever earn from fees rather than subsidy.

The Hashprice Curve: Derivatives and Hedging

For most of Bitcoin’s history, hashprice was something you endured. A miner with fixed power costs and volatile revenue simply rode the swings. That is changing. A market has grown up around the metric, and miners can now hedge it the way an airline hedges jet fuel.

Two venues anchor the space. Luxor runs over-the-counter hashrate forwards out to twelve months plus daily-settlement contracts, and Bitnomial lists exchange-traded Hashrate Futures under the ticker HUP, in one-petahash monthly contracts. The mechanics are straightforward: a miner worried about a soft market sells hashrate forward, locking in revenue today. A lender financing new machines can price the loan against a real forward curve instead of a guess. The implied yield embedded in these contracts is effectively the miner’s cost of capital, and in mid-2026 it ran roughly 6 to 13 percent annualized, per the Lookback.

The forward curve also carries information for everyone else. In July, USD-denominated forward sellers came out ahead at two of five settlement horizons, and the August-to-December curve rose about 8 percent over the month even as the market marked down its implied December hashrate from 1,054 EH to 985 EH, per the Lookback. Translation: traders are pricing in a smaller network by year-end, which, all else equal, means a firmer hashprice for the machines that remain. A liquid hashprice curve turns a raw physical resource into something investors can underwrite, hedge, and forecast, a maturation crypto has repeated in one sector after another.

The AI Opportunity Cost: What a Megawatt Earns Mining vs Leasing

The defining tension of 2026 is that a hashprice of $32 is no longer judged against zero. It is judged against what the same megawatt could earn doing something else. The power contracts, substations, and cooling that feed an ASIC farm are much of what an artificial-intelligence data center needs too, and AI compute can pay far more per megawatt, often on fixed multi-year terms rather than the whims of a volatile hashprice.

Recall those mining figures: roughly $109 per MWh for the most efficient fleets, about $41 for the least. Now weigh them against the deals miners have signed to host AI instead. Core Scientific committed around $10.2 billion over twelve years to CoreWeave; Hut 8 struck a roughly $7 billion, fifteen-year lease; TeraWulf has contracted some $12.8 billion of high-performance-compute revenue; and IREN signed a $9.7 billion, five-year cloud deal with Microsoft. CoinShares tallies more than $70 billion of announced AI and HPC contracts across the public mining sector.

John Todaro, an analyst at Needham, put the capital-markets logic plainly to CoinGeek: “The revenue per megawatt and EBITDA margins are far higher for HPC and AI colocation than for mining.” CoinShares research head James Butterfill told The Block that miners could draw “as much as 70% of their revenues from AI by the end of this year, up from roughly 30% today.” The point for hashprice is subtle but important: even a healthy hashprice now competes for capital with an alternative use of the same watt, and when that alternative pays more with less volatility, machines get unplugged and shells get repurposed. The economics of paying for compute with a token, and whether such flywheels actually spin, is a theme we examine in our look at Gensyn’s buy-and-burn machine. Thiel’s forecast to CoinGeek captures where this leads: “By 2028, you’ll either be a power generator, be owned by one, or be partnered with one. The days of being a miner plugged into the grid are numbered.”

Where Hashprice Can Mislead You

Hashprice is powerful precisely because it collapses a complicated business into one number, and that is also its weakness. A few caveats keep it honest.

• It is an expected value, not a guarantee. Block discovery is a lottery, and a small miner can go a long stretch above or below the average through pure variance. The index describes the network mean, not any one operator’s luck.
• Your payout scheme changes what you actually receive. Under a full-pay-per-share model the pool pays you the estimated fees whether or not your blocks carried them; under a pay-per-last-N-shares model you share the pool’s real luck. Realized revenue can drift from the quoted hashprice for weeks.
• Fee spikes are smoothed. Because the index uses a 144-block moving average for fees, a single frantic blockspace day barely nudges it, even though the actual hashprice that day may jump.
• Two miners at the same hashprice can post very different margins. Demand-response credits, hedges, and power contracts sit outside the metric entirely.
• One hashprice hides a wide fleet. Real machines span roughly 13 to 38 J/TH, and power prices vary three or four times over across geographies. Hashprice is uniform; margin never is.

None of this makes the metric less useful. It just means hashprice is the headline, not the whole story, and the smart reader always asks what a given miner’s costs, hardware, and payout terms do to it.

Regulation and Disclosure: The SEC’s Posture

Hashprice is not just a trading metric; it underlies what US-listed miners tell regulators. Companies like Riot and MARA disclose cost-to-mine, hashrate, and fleet efficiency in their quarterly filings, and those numbers only make sense against the prevailing hashprice. Anyone evaluating a mining stock is, in effect, reading a leveraged bet on where hashprice goes.

On the legal footing of the underlying activity, the Securities and Exchange Commission gave the domestic industry a clear answer in 2025. The SEC’s Division of Corporation Finance stated on March 20, 2025 that certain proof-of-work mining does not involve the offer or sale of securities, reasoning that miners earn rewards through their own computational effort, and that combining resources in a pool does not change that analysis, as laid out in the Division’s statement. In plain terms, running or joining a mining operation is not, by itself, a securities transaction under US law.

The boundary worth noting is that hashprice derivatives are a different animal. Futures and forwards on hashrate are commodity products; Bitnomial, which lists the HUP contract, operates as a federally regulated derivatives exchange under the Commodity Futures Trading Commission. So spot mining sits outside securities law, while the instruments miners use to hedge it fall squarely inside commodities regulation, a distinction that matters as more capital flows into the hedging market.

What Could Move Hashprice Next: The Bull and Bear Case

Start with the bull case. The single biggest lever is price. Butterfill told The Block that a recovery to $100,000 was “not an unrealistic assumption,” a move that would lift hashprice back toward $37 per PH per day. Sustained exchange-traded-fund demand is the most plausible route there; how those products cleared the regulatory gate is the subject of our explainer on crypto ETF approvals in 2026. A genuine fee renaissance, driven by durable blockspace demand, would help at the margin, and further difficulty declines as weak miners quit would keep firming the floor.

The bear case is structural. The spring 2028 halving will cut the subsidy to 1.5625 BTC, a step-down that price and fees will have to fully absorb just to keep hashprice flat. Fresh, efficient hashrate, much of it funded on the back of AI-driven capital, can pour back in the moment economics improve, pushing difficulty up and diluting revenue again. Thiel’s warning is the one to sit with: if Bitcoin does not grow at 50 percent or more a year, he told CoinGeek, the math “gets very tough after 2028, and even tougher in 2032.”

The honest synthesis is that even a return to $37 leaves the oldest fleets stressed. The durable fix is not a single price rally but a combination of efficient machines, genuinely cheap or flexible power, and revenue diversification into compute. Hashprice will keep telling you, in one number, whether Bitcoin’s security is being paid for. It is worth watching like a vital sign, because when it sits below the cost of the electricity behind it for too long, the network quietly reshapes itself until it does not.

Frequently Asked Questions

What is a good hashprice for Bitcoin miners?

There is no single threshold, because break-even depends on a miner’s power price and machine efficiency. As a rough guide, in 2026 a hashprice around $32 per petahash per day covers energy costs for efficient sub-15 J/TH fleets paying cheap power, but leaves older 20-plus J/TH machines underwater unless electricity is very cheap. Large public miners have reported all-in costs that imply they need a materially higher hashprice, or a higher Bitcoin price, to profit after depreciation.

Why is Bitcoin hashprice so low in 2026?

Three forces pushed it down. The April 2024 halving cut the block subsidy in half, network difficulty and hashrate climbed to records into late 2025, and transaction fees stayed near 1 percent of the block reward. More machines splitting a smaller, mostly price-dependent reward means each unit of hashrate earns less. Hashprice reached five-year lows near $28 in mid-2026 before recovering toward $32 as unprofitable miners switched off and difficulty fell.

How is hashprice calculated?

Hashprice is a function of four inputs: the block subsidy, transaction fees, network difficulty, and the price of Bitcoin. Intuitively, it is the total daily network reward (about 144 blocks times the subsidy plus fees, valued in dollars) divided by the total hashrate. Luxor’s Hashrate Index, which coined the term in 2019, uses difficulty as the denominator and a 144-block moving average for fees to publish a daily USD figure per petahash per day.

What is the difference between USD hashprice and BTC hashprice?

USD hashprice is expected daily revenue per unit of hashrate measured in dollars, so it moves with Bitcoin’s price. BTC hashprice measures the same thing in Bitcoin terms, stripping the price out, so it isolates pure network economics like difficulty and fees. In 2026 the BTC hashprice reached its highest levels since 2025 even while the USD figure stayed historically low, showing the network economics improved on a coin basis while the dollar value lagged.

Can you hedge or trade hashprice?

Yes. Miners can sell hashrate forwards and futures to lock in revenue, and traders can take positions on where hashprice is heading. Luxor offers over-the-counter forwards out to twelve months, and Bitnomial lists exchange-traded Hashrate Futures. The implied yield on these contracts reflects miners’ cost of capital, which ran roughly 6 to 13 percent annualized in mid-2026.

Yuki Tanaka is a senior markets writer at HOGE Wire covering Bitcoin mining, market structure, and the economics of proof-of-work.