.png){kind=logo}
Bitcoin cleared $80,000 on August 25, 2026, its first time above that level since May 15, reached roughly $81,000 and settled back below $80,000 the next day. For Bitcoin miners who spent this year watching hashprice slide toward $27, that week brought the first real good news since March.
Three things improved at once, and none of them are about sentiment. Hashprice is back above $39 per PH/s per day. Bitcoin mining difficulty has retargeted downward ten times this year against seven increases, easing the cost side for every miner still hashing. And spot ETFs took in roughly $1.9 billion in the week to August 21, which is money buying bitcoin rather than traders shuffling positions.
The qualification is narrower than the headlines suggest. Bitcoin mining equities barely moved on a 21% week: Riot down 1.1%, CleanSpark down 3.4%, only Marathon keeping pace at +15.9% as of the August 24 close. The hashrate forward market held above $35 into January, still below the current level. Two markets that live on mining revenue refused to price the improvement as permanent.
EMCD reads it from the operator's side: what hashprice is, what actually pushed it to $39, what has to hold for it to stay there, and where the threshold sits for hardware at a given J/TH and power price.
Hashprice at $39.06 per PH/s per day is more than 43% above the June low near $27, per Luxor's Bitcoin Hashprice Index. That is the first sustained rise in mining revenue since the first-quarter squeeze.
How the move was built matters more than its size. Futures open interest measured in BTC fell about 11% between mid-July and August 23, dropping to a five-month low of 587,584 BTC per Glassnode. Funding rates stayed low. Bitcoin gained more than 20% while borrowed money was leaving the market, which is the reverse of the rallies that failed earlier this year.
Third, part of the buying was for cash rather than on leverage. U.S. spot funds took in roughly $1.9 billion in the week to August 21, then $337.6 million on August 24 and $314.3 million on August 25.
A mining stock is a bet on one number: expected revenue per unit of hashrate. When investors leave those stocks flat through a 21% move in Bitcoin, they are saying it will not last.
The balance sheets explain the caution. All four major listed Bitcoin miners were unprofitable and burning cash as of August 24, with breakeven Bitcoin prices estimated between $45,000 and $65,000 depending on power costs and hardware efficiency. That 20,000-dollar spread is the difference between the fleets that came through the post-halving squeeze and the ones that did not.
Hashprice cleared $39 because two inputs moved in the same direction inside two weeks. The Bitcoin price rose, and network difficulty fell. Both are genuine, but they have different half-lives.
The August 22 retarget cut difficulty 1.31% at block 963,648, from 127.48 trillion to 125.81 trillion. That puts mining difficulty 0.7% above the June low of 124.93 trillion, effectively on the floor for the year.
The full-year picture is starker. Difficulty entered 2026 near 148.25 trillion and now sits 15.1% lower. Difficulty data implies roughly 150 EH/s of effective hashrate has exited, against a 7-day average network hashrate near 920 EH/s in mid-August. Live estimators disagree by around a tenth on any given day, so the arithmetic below runs against the 7-day average rather than a spot reading. Miners still running are taking a share of blocks that a fuller Bitcoin network would not leave on the table.
CoinWarz estimates the next difficulty adjustment around September 6 at roughly minus 1.86%, based on observed block times.
If that holds, sidelined hardware has not come back yet, and the difficulty discount widens slightly before it narrows.
The first part of the move came from liquidations. Roughly $3 billion in leveraged short positions were closed out on August 19 and 20, with shorts accounting for about $2.77 billion, or 92% of the total, and $1.29 billion of that closing inside a single hour.
That buying is real, but there is a fixed amount of it. A short that gets closed cannot be closed again, and with open interest at a five-month low most of them already have been. ETF inflows are what has to take over, which makes them the weekly number worth more attention than the Bitcoin price itself.
The second part is arithmetic, which makes it easier to check. Hashprice moves inversely with difficulty. Return the full 150 EH/s to a 920 EH/s Bitcoin network, more miners join, network hashrate rises about 16%, difficulty follows inside one or two retarget epochs of 2,016 blocks, and hashprice gives back roughly 14% on that input alone, landing near $33.5.
A partial return is the more realistic case. Half of that capacity, roughly 75 EH/s, lifts network hashrate about 8% and takes hashprice to around $36. Still above the June level, and no longer a comfortable margin for anyone above 25 J/TH.
Holding $39 through a difficulty increase of the full size would require the Bitcoin price to cover it, roughly 16% above where it sat when hashprice reached $39, near $93,000. It is the rate at which difficulty and price trade off against each other, and every miner is exposed to it whether they track it or not.
Reactivation is unlikely to be instant. Hardware came offline against power contracts, hosting agreements and seasonal energy availability, and none of that reverses because of one good week. That lag is the window miners actually have.
Hashrate forwards above $35 into January sit below the current $39 and above the $27 of June. Read plainly, that says the floor has moved up and the spike has not been confirmed.
It is a more useful signal than any commentary, because it is a live price on the same revenue a farm produces. A miner who believes $39 understates future mining revenue can sell forward against it and find out.
Mining revenue has recovered. Production costs have not changed, and part of the current hashprice level is on loan from miners who have not yet come back.
That points to one approach: use the window. Curtailment contracts and forward sales are worth more now than they were in June. Buying hardware against a hashprice that depends on competitors staying offline is buying at the wrong point in the cycle, since ASIC prices track hashprice and move fastest once everyone reaches the same conclusion.
One piece of arithmetic does most of the work. Daily energy draw per TH/s equals efficiency in J/TH times 24, divided by 1,000. Divide hashprice per TH/s by that figure and the result is the power price at which mining revenue covers electricity, and nothing else.
Hashrate Index puts compute revenue near $95 per MWh for fleets under 19 J/TH against roughly $50 per MWh for hardware in the 25 to 38 J/TH band. The same hashprice pays those two fleets almost twice as differently, because the second converts far more energy into the same number of hashes.
Which is why hardware composition outranks site scale. A 5 PH/s farm on current-generation ASICs earns more than a 30 PH/s farm on older hardware at the same power price, and no payout model closes that gap for the miners running it.
One input would invalidate it: fees. Miners collected an average of 0.0210 BTC per block per day in transaction fees in the week to August 17, down 7% week over week, under one percent of the block reward. Fees are the only component that rises without pulling mining difficulty behind it, because demand for space inside blocks does not care how many miners join.
Nothing in the current data points that way. A lasting change in the fee share of block rewards would change everything above, and it deserves the same attention as the ETF numbers.
Three dated items decide whether the window stays open. All three are observable, and none of them require a view on where the Bitcoin price goes.
The September 6 difficulty estimate is the first. A print near minus 1.86% confirms hardware is still offline. A flat or positive print is the early signal that reactivation has started, and it arrives about two weeks before the effect shows in hashprice.
Weekly ETF flows are the second. They have to keep replacing forced covering, and the daily figures through August 25 suggest they still are.
The August 28 options expiry is the third, at roughly $6.44 billion with the largest open interest sitting at the $75,000 and $80,000 strikes. Concentrated expiries move price around a level rather than through it, and hashprice inherits whatever the price does.
For a farm that never switched off, a price that stalls here is worth more than a price that runs. Stalling keeps the sidelined 150 EH/s offline, holds difficulty near the floor, and leaves the current hashprice with the miners who stayed on.
Hashprice is a benchmark index for the expected daily mining revenue of one unit of hashrate. It is gross revenue. Whether a farm is profitable depends on hashcost, a miner's own production cost per unit of hashrate.
Two conventions circulate, with a factor of 1,000 between them: $39.06 per PH/s per day is roughly $0.039 per TH/s per day. The BTC-denominated series removes the currency component and reflects subsidy, fees, and difficulty, providing a clearer view of how difficult it is to mine Bitcoin blocks.
The Bitcoin network produces about 144 blocks per day, and each block pays a reward consisting of the subsidy plus fees. The ndex applies a 144-block lagging simple moving average to the fee component, so a single congested block does not distort the figure. Live hashprice, by contrast, moves with every new block added to the Bitcoin blockchain.
Hashprice is positively correlated with the Bitcoin price and negatively correlated with network difficulty, which retargets every 2,016 blocks, roughly every two weeks. The April 2024 halving cut the block subsidy from 6.25 BTC to 3.125 BTC, and the next halving around April 2028 takes it to 1.5625 BTC. Fees are the fourth input and currently the smallest.
The four inputs interact, so no single one explains a given level. Reading the Bitcoin price alone tells Bitcoin miners nothing actionable about their own hardware.
Miners run the ASICs whose production cost sits under live hashprice and curtail the ones above it, particularly where a demand-response contract pays more than the operating margin. Selling hardware is a different decision: it makes sense when a unit is permanently below the line, not when it is temporarily above it, and every buyer is reading the same difficulty data.
The threshold governs pool mining, solo mining and cloud mining identically. Solo mining changes payout variance. Cloud mining inserts a counterparty between the miner and the hashprice. Neither alters revenue per unit of hashrate.
At $39 per PH/s per day, a single percentage point of pool fee is worth about $0.39 per PH/s per day, near $1,400 a year on 10 PH/s. At current margins that is a line item worth auditing.
Payout model carries comparable weight. FPPS pays for accepted shares using the block subsidy plus estimated transaction fees, which smooths variance, while PPLNS pays only when the pool finds a block and hands the miner the pool’s luck. EMCD Mining pool runs both, with a pool fee from 1.5% and stable daily rewards for miners worldwide.
Hashprice is the expected daily gross mining revenue from one unit of Bitcoin hashrate, quoted in USD per PH/s per day, USD per TH/s per day, or in BTC. It combines block subsidy, transaction fees, network difficulty and the Bitcoin price into a single figure, and measures revenue before costs rather than profit.
Hashprice sits near $39.06 per PH/s per day in late August 2026, more than 43% above the June low around $27. Hashprice reprices with every block, so miners should rely on a live hashrate index reading rather than a weekly average.
A single hash carries no meaningful value, which is why miners quote a rate. At $39.06 per PH/s per day, one TH/s produces roughly $0.039 a day in gross revenue.
There is no absolute target. Efficiency in J/TH against a specific power price decides the outcome: 200 TH/s at 19 J/TH and 200 TH/s at 38 J/TH produce identical mining revenue and pay very different electricity bills. Compare production cost to hashprice before comparing raw TH/s.
Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp
_____________
Disclaimer: Analytics Insight does not provide financial advice or guidance on cryptocurrencies and stocks. Also note that the cryptocurrencies mentioned/listed on the website could potentially be risky, i.e. designed to induce you to invest financial resources that may be lost forever and not be recoverable once investments are made. This article is provided for informational purposes and does not constitute investment advice. You are responsible for conducting your own research (DYOR) before making any investments. Read more about the financial risks involved here.
