Bitcoin Miners in West Texas Made More Switching Off Than Hashing
Luxor's West Texas model shows power prices can beat hashrate margins outright, and Riot's quarterly filing explains why curtailment is now a real revenue line. The catch: this trade only exists on grids built for it, which is why the capital keeps landing in Texas and not in Tokyo.
Shutting a bitcoin mine off shouldn't be the profitable part of the business. In West Texas, it's.
Luxor ran the numbers on a West Texas mining operation and the model produced an answer that should make every operator in the industry uncomfortable. Past a certain power price, the site earned more from selling its electricity back than from producing hashes. Not marginally more. Enough to change the economics of the entire facility.
Riot's quarterly filing makes the same point from the other side of the ledger. Curtailment isn't a survival tactic for a bad month anymore. It's a business line, and it's carrying real weight in the results.
The 4CP arithmetic
Here's the mechanism, stripped down. Riot runs the Corsicana site in Navarro County, Texas, about 700 megawatts of contracted load sitting inside ERCOT, the most aggressively deregulated electricity market in North America.
ERCOT sets a big chunk of its transmission charges using four coincident peak intervals, known as 4CP. Four fifteen-minute windows per year, and whatever your load is during those windows shapes your transmission bill for the next twelve months. So a large power consumer has an enormous incentive to drop load at exactly the moment the grid is straining. The grid pays. The miner saves. Both sides walk away happy.
For bitcoin miners, this is close to a perfect hedge. Hash price has been brutal, hovering in the low tens of dollars per petahash per day for stretches. Mining at a loss to keep machines warm is just burning capital. Curtailing, by contrast, converts a bad power day into a credit on the books.
Riot has booked tens of millions of dollars in power credits across recent quarters, including a stretch north of $13 million in a single three month window. That's not a rounding error for a company that also has to justify its hashrate guidance every earnings call.
Is that a bitcoin miner, or an electricity trader with extra steps? At this point the honest answer is both.
Who this kills
And here's the part Western coverage tends to skate past. This trade isn't available to everyone, and that's precisely why it's valuable.
A miner on a fixed-price industrial contract in a vertically integrated market can't do this. There's no 4CP, no real-time price signal to respond to, no counterparty paying for demand flexibility at scale. That operator just eats the margin compression and hopes the next halving doesn't finish the job.
So the gap between a Texas miner and a miner anywhere else isn't operational skill. It's jurisdiction. It's whether the grid you plugged into was designed to pay large loads for being interruptible.
My take: half the mining companies publicly traded today are structurally uncompetitive and haven't figured it out yet. They'll keep reporting hashrate growth while quietly bleeding, and the market will keep rewarding the Texas operators with premium multiples. The ones with flexible power contracts and demand response revenue are the only ones whose cash flow survives a sub-$60,000 bitcoin price for a full year.
Second take, and this one won't be popular. Curtailment credits are a subsidy, dressed up as a grid service. Sometimes they're a genuinely efficient one, because the grid really does need that load to vanish at 4:30pm in August. But let's not pretend this is pure free-market alpha. It's a policy construct, and policy constructs can be rewritten. ERCOT has already been tinkering with how large loads are treated. A rule change is the single biggest risk to the whole model, and it won't come with a press release.
What's a mining company worth when the grid stops needing it to shut off? That's the question nobody wants to model out loud.
The view from Singapore
Tokyo and Seoul are writing different playbooks, and neither of them involves paying miners to switch off.
Japan has been cautiously opening its electricity market, but the kind of granular, financially settled demand response that makes curtailment lucrative in Texas doesn't exist at the scale a 100 MW site needs. Korea is worse. KEPCO's tariff structure is heavily regulated, industrial rates are administered, and there's no four-peak trick to arbitrage. Singapore is a genuinely open market, but it's small, expensive, and the government has been openly cool toward energy-intensive mining.
So where does the capital go? It goes to the grid that pays it. Asian family offices and funds that want bitcoin mining exposure have spent two years figuring out they can't get it domestically at competitive cost, so they buy Riot and its peers instead. Texas becomes the exposure, and the operators who learned to monetize curtailment become the benchmark.
The capital isn't leaving crypto. It's leaving your jurisdiction.
Here's the concrete take. Stop valuing miners on hashrate alone. Hashrate without a flexible power contract is a liability that depreciates on a three-year schedule. The number that should be on your screen is the power credit line, and how big a share of gross profit it represents. For Riot, that share is now meaningful enough that a weak bitcoin quarter doesn't automatically mean a weak earnings report. For most of the sector, it's zeros across the board.
Asia moves first on a lot of things. On this one, we're watching from the cheap seats.
Explore More
Key Terms Explained
Valuable, non-public information or insights that give you a trading edge.
Profiting from price differences of the same asset across different markets.
The first cryptocurrency, created in 2009 by the pseudonymous Satoshi Nakamoto.
A company's profits, typically reported quarterly.