The Pattern Nobody Talks About

Bitcoin miners don't die once. They die twice.

The first death is operational. BTC price drops, hashprice collapses, revenue evaporates in USD terms — but the power contracts, equipment leases, and debt service don't move an inch. 60–70% of a typical miner's cost base is denominated in USD. Revenue is entirely in BTC. When the exchange rate moves against them, the business suffocates.

This is standard commodity distress. Everyone understands it. Lenders restructure, operators cut costs, the cycle turns, and the survivors emerge leaner.

The second death is quieter, and it's the one that actually kills the company.

After the restructuring, the lender recapitalises with more USD debt. Same capital structure. Same currency mismatch. The miner has to dig out — but they're shovelling in USD while their earnings are still denominated in BTC. The restructuring didn't fix the disease. It just restarted the clock.

Every major Bitcoin mining restructuring in this cycle tells the same story: USD DIP financing in, USD exit financing out, USD capital table at the end. NFN8 Group took a $2.75M USD DIP from Twelve Bridge Capital. Core Scientific restructured through roughly $400M in USD convertible note-to-equity conversion plus a $500M Morgan Stanley loan at approximately 7.8%. Rhodium Enterprises defaulted on a $54M USD loan and is still litigating the Whinstone dispute years later. BlockFills filed Chapter 11 in March with a USD structure.

Zero BTC-denominated DIP financing. Not in 2022. Not in 2023. Not in 2024. Not in 2025. Not in 2026.

The industry is five cycles deep and nobody has tried matching the medicine to the disease.

This Has Happened Before

In 1997, Thai companies borrowed USD at low international rates to fund operations denominated in baht. When the baht collapsed, USD debt service became unpayable overnight. Not bad management. Not excessive risk. A currency mismatch that was always there — invisible until stress arrived.

Hungarian households did the same thing in the mid-2000s, borrowing Swiss francs because the rates were lower. When the franc appreciated 60% against the forint, the payments became unpayable. They weren't reckless. The mismatch was structural.

Brady Bonds resolved Latin America's USD debt crisis — with more USD bonds. It worked because USD export earnings eventually outran USD obligations. The macro turned in their favour. Bitcoin miners don't have that luxury. Their revenue doesn't converge toward USD. It diverges from it by design.

Bitcoin miners are running the same book as every emerging-market borrower who ever blew up on a currency mismatch. The difference is that BTC isn't a weak currency trending toward USD parity. It's a separate monetary system with its own dynamics. The mismatch doesn't resolve itself. It compounds.

The Collateral Problem

Traditional asset-based lenders underwrite collateral that holds value when the business fails. In oil and gas, the reserves still have intrinsic worth in the ground. In real estate, the land provides a floor. ASICs have one job — SHA-256 hashing — and they reprice to near-zero in every bear market.

NFN8 Group had over 5,000 machines. Estimated total liquidation value: under $50,000. That's roughly $10 per machine for equipment that sold for $5,000–$10,000 each at peak. An ASIC isn't equipment in the traditional sense. It's a hashprice futures contract wearing a metal jacket, and it depreciates in both BTC and USD terms simultaneously.

The lenders who wrote loans against this collateral weren't wrong about Bitcoin. They were wrong about what the collateral actually is. You can't restructure what has no independent floor.

The AI Pivot: Right Move, Wrong Denominator

The migration to AI/HPC hosting is the latest response to this structural problem. And the operational logic is sound — the skills transfer, the sites transfer, the power infrastructure transfers. Bitfarms shareholders voted in March on renaming the company to Keel Infrastructure. CEO Ben Gagnon said plainly that they are no longer a Bitcoin company.

Among the six publicly listed miners with HPC contracts, CoinShares projects Bitcoin mining revenue falling from approximately 85% to under 20% of total revenue by late 2026. That's not a pivot. It's a full identity transplant.

But the AI pivot solves the wrong problem. These companies still have USD debt. They're now servicing it with USD AI revenues. Multi-year GPU hosting contracts look superior to BTC mining revenues — but only if you run the comparison in USD. At the experienced inflation rate of the last five years, a USD revenue stream worth 100 today is worth roughly 61 in ten years. The numeraire problem hasn't gone away. They've swapped Bitcoin volatility for a different kind of cyclicality — one that depends on hyperscaler demand, contract renewal rates, and the pace of AI infrastructure buildout.

The ones who survive aren't the ones who found the best pivot. They're the ones who never needed to pivot because they controlled their energy cost so tightly the cycle couldn't touch them.

The Fix

The structural solution is BTC-denominated debt. Loan in BTC, service in BTC, covenant in BTC. When hashprice drops, the capital structure doesn't break — because the debt denominator dropped with it. When difficulty adjusts downward (as it did with the -11.16% correction on March 15, 2026), the remaining operators see improved economics without the debt becoming relatively more expensive.

Oil and gas has reserve-based lending — debt denominated in the production currency, repayment from the reserves. Bitcoin mining has nothing like this. The closest analogues are Lombard-style lending against Bitcoin reserves, where the productive asset itself secures the obligation in the same denomination.

The difficulty adjustment is, in a sense, the most elegant restructuring mechanism in financial history. No creditors' committee, no lawyers, no fees. Every two weeks, it recalibrates the entire industry's economics. Operators whose cost structures are aligned with BTC fundamentals benefit directly from this self-correcting mechanism. Operators carrying USD debt don't — because their obligations are denominated in a system that doesn't adjust.

Mining credit works. But only if you underwrite the energy position and the hashprice sensitivity — not the machines. The operator at $0.025/kWh with flexible capacity will be running in 2028. The operator at $0.07/kWh with an ASIC-backed loan will not.

The 2028 Preview

The next halving is approximately 25 months away. Block reward drops from 3.125 to 1.5625 BTC. At current hashprice levels, roughly 20% of installed capacity is already below breakeven. That number gets worse before it gets better.

Every operator taking USD DIP financing today will emerge from Chapter 11 with the same mismatch, facing a 50% revenue cut they can see on the calendar. The protocol doesn't care who it displaces. That's the design.

The miners who make it to April 2028 will be at the energy frontier — stranded gas, flare mitigation, flexible load agreements, sub-$0.03 power. They won't have ASIC-backed USD loans. They'll have either BTC-denominated credit or no external credit at all. They'll have survived not because the price bailed them out, but because their cost structure was so far inside the breakeven line that the cycle was almost irrelevant.

That's not a prediction. That's what the surviving operators from every prior cycle look like in retrospect.

Ben Vincenzi is the founder of BTSF, a Bitcoin-native credit institution focused on structured lending to Bitcoin miners. He is the author of Beyond Digital Gold: Bitcoin and the Architecture of a New Monetary System.