Why mining rewards lag bitcoin's price

The mechanism, in four steps

What a TH earns in dollars per day —its hashprice— is a multiplication: satoshis the TH mines × bitcoin price. Price is the half you see. The other half depends on how much hashpower you are competing with:

1. Bitcoin rises and mining becomes more profitable.
2. Machines switch on: idle older ones right away; new ones when they arrive and get installed.
3. Every 2,016 blocks, about two weeks, the network resets difficulty so that a block still comes every ten minutes.
4. With the same block reward split across more hashpower, each TH mines fewer satoshis.

The useful question isn't whether this happens —it happens by design— but how much, and how late. That is what the tables and charts below measure.

Since August: almost proportional

Whole network, no pool factor. Sats per TH computed from the difficulty in force and the average reward per block (subsidy + fees) over the previous 7 days. Sources and formulas at the end.

Over these two months, for every 10 points the price rose, hashprice got almost 8. The "it doesn't rise proportionally" thesis holds, but only just: difficulty hasn't had time to react yet.

There is a sign that it will. Hashrate over the last two weeks (981 EH/s) is already ahead of what current difficulty assumes (950 EH/s): so far this period, blocks are arriving every 9.5 minutes on average, not 10. On 8 October, mempool.space estimated the next adjustment at +5.6% around 16 October. If it lands like that and price doesn't move, hashprice since August would end up around +15%, against +28% for price.

Two years: price up, each TH earns less

To see the full effect you need a longer window. The last two years have one advantage: the block subsidy has been 3.125 BTC the whole time (the halving was in April 2024), so the only things changing are price, hashpower and fees.

Bitcoin is worth a third more than two years ago, and a TH earns fewer dollars than it did then. Over two years the network's hashpower grew faster than the price, and each TH took a smaller slice of the same reward.

The lag, with an example: 2025

The clearest way to see it is to take one rally and watch what came after. 2025 is almost a textbook case, in two phases: bitcoin went from $78,247 on 8 April to $117,860 on 11 July, and from there to its peak, $124,231, on 6 October.

In phase 1 hashprice rose as much as price. In phase 2 the hashpower arrived: difficulty climbed 29% in three months with price nearly flat. The result is the pair of dates that sums it up: hashprice peaked for 2025 on 11 July; bitcoin on 6 October. Difficulty didn't top out until 29 October (156.0 T), with price already falling.

Something similar happened at the end of 2024, only faster: between 4 November and 17 December bitcoin rose 55.6% and difficulty 13.4% in the same six weeks, so hashprice got 36.7%.

How many weeks of lag? There is no clean number, and we won't make one up. We compared the 13-week change in price with the change in difficulty some weeks later, using data since June 2024. The relationship shows up almost immediately (correlation 0.48), stays around 0.5 for about 20 weeks and fades between weeks 24 and 30.

In plain terms: part of the response arrives within weeks and the rest is spread over roughly 4 to 5 months. That fits idle older machines switching back on as soon as they pay, and new hardware ordered during the rally and plugged in months later. These are moderate correlations over about 90 weeks: good for seeing the shape, not for forecasting.

And when bitcoin falls, the same thing in reverse

The lag works both ways. From 6 October 2025 to 1 July 2026 bitcoin fell 52.9%. Hashrate dropped only 8.2% and difficulty 11.3%: the machines are already paid for and stay on as long as they cover their power. Each TH earned 12.4% more satoshis, which offset almost nothing: in dollars, hashprice fell 47.0%.

That is why difficulty doesn't protect you much in a drop, and in a rally it takes the gain back later than it seems. Between the two, hashprice tends to drift down over the long run even when price goes up.

What this means for a GoMining miner

On GoMining your revenue is the hashprice above (minus the pool factor), but your costs aren't: maintenance is billed in dollars and doesn't depend on price. For 100 TH at 12 W/TH and $0.05/kWh, with no discounts, that is $2.33 a day, always.

That produces two effects at once, and it pays not to mix them up:

• Operating leverage, in your favour. Since 5 August the gross mined by those 100 TH went from $3.06 to $3.73 a day (+21.9%), and the net from $0.73 to $1.40 (+91.6%). With a fixed cost, every extra dollar goes straight to the net.
• Difficulty, against you and late. The same leverage cuts the other way. With the estimated +5.6% adjustment, the 8 October net would drop to about $1.20 without price moving. With 21.9% more difficulty, it would be back at August's $0.73 with bitcoin still above $80,000.

Over two years the balance is what the chart shows: bitcoin +31.7%, net for 100 TH from $1.98 to $1.40 a day (−29.3%). That is why the calculator projects with a monthly difficulty growth you can adjust, instead of freezing the current network.

What about discounts?

The figures above are with no discounts, the worst case. Paying maintenance in GOMINING token takes up to 20% off, and with VIP and the service button the ceiling is 29%. With a lower fixed cost the net is higher and amplifies less: it rises less with price and falls less with difficulty.

Daily net for 100 TH at 12 W/TH, same assumptions as the chart. Maintenance drops from $2.33 to $1.86 with 20% and to $1.65 with 29%.

Discounts don't change the conclusion: over two years, with bitcoin 32% more expensive, 100 TH earn less than they did. They only soften it.

What you do control: efficiency (the W/TH tiers decide how much margin is left when difficulty squeezes) and discounts (the maintenance fee, line by line). What you don't: the 2028 halving, which will cut the subsidy in half on a fixed date. For the current snapshot per TH and at five prices, see GoMining with bitcoin above $80,000.