Bitcoin is in a bear market. That much is not in dispute.
What Jim Ferraioli, Director of Digital Currencies Research and Strategy at Charles Schwab, argued Wednesday on Bloomberg is more precise and more structural: this selloff has a measurable cost floor, and that floor is built not from sentiment or chart patterns, but from the physics of energy consumption.
The numbers frame the drawdown in context. Bitcoin peaked at $126,000 in the fall before collapsing to roughly $60,000 in February — a 50% correction that, while brutal for recent buyers, falls far short of the 75%-plus implosions that defined prior Bitcoin bear markets.
Ferraioli’s core analytical framework centers on one question: what does it cost to manufacture Bitcoin? The answer creates a natural gravitational floor that has held across multiple cycles.
For the most efficient miners — those operating at scale with next-generation ASIC hardware and access to the cheapest wholesale energy — the cost to produce one Bitcoin sits at approximately $60,000, Ferraioli said.
That figure is not arbitrary. It represents the all-in expense of powering a facility at roughly $0.07 per kilowatt-hour with the most advanced semiconductor fleets available.
The less efficient miners — those with older ASIC hardware, higher energy costs, and thinner operational margins — carry a production cost of approximately $95,000 per BTC, according to Glassnode data cited in Schwab’s May 2026 research report. That gap between $60,000 and $95,000 defines Bitcoin’s current valuation range.
Bitcoin’s energy floor: Why $60,000 may mark the bottom
Ferraioli argues that in deep bear markets, the cost of production for the best miners has historically served as the bottom. February’s low near $60,000 aligns almost precisely with that level, as well as BTC’s 200-week moving average.
The BTC selling pressure is not random. It is demographically specific. The investors driving forced liquidations are those who acquired Bitcoin during the past 18 months — buyers who rode the asset from sub-$80,000 up to $126,000 and then watched gains evaporate in full.
Schwab tracks two cost-basis metrics to quantify this pressure: the average acquisition cost for U.S. spot ETF and ETP holders, which stands near $83,000, and the active investor cost basis — excluding coins rewarded to miners — which sits near $78,000.
Both figures sit well above current spot prices, putting the majority of recent entrants into unrealized loss positions and reinforcing $83,000 as a ceiling of overhead supply rather than a floor of support.
Glassnode’s on-chain data corroborates this dynamic. Bitcoin’s latest attempted rally stalled at the aggregate ETF cost basis near $83,000, with total realized losses spiking to $1.35 billion per day and long-term holders capitulating from cycle-top positions. Hedge funds represent roughly 30% of spot ETP ownership but are operating market-neutral, executing basis trades rather than taking directional views — meaning they provide no natural bid when prices fall.
Here is where Ferraioli’s analysis turns constructive. Every major publicly traded Bitcoin miner has announced a pivot toward high-performance computing (HPC) for AI inference workloads. The economics on their face appear to favor abandoning mining: inference generates higher net revenue per megawatt-hour than Bitcoin mining during peak demand windows.
But demand for AI inference is not uniform across 24 hours. Models run hard during business hours and sit idle overnight and on weekends.
That creates a structural opportunity that does not displace BTC mining — it layers on top of it. Schwab’s analysis models Bitcoin as the optimal baseload monetization of power during off-peak hours, with inference overlaid during peak business-hour demand.
A data center operating this hybrid model maximizes utilization across the full 24-hour cycle rather than leaving capacity dark when inference demand falls away. For miners, this translates to more stable revenue, reduced forced BTC sales to cover operating costs, and lower structural risk across bear market cycles.
Bitcoin is backed by energy
The underlying thesis is one of energy economics. Bitcoin has no earnings, no free cash flow, and no CEO issuing guidance. Its value, in Ferraioli’s framework, derives from the energy cost required to produce it — a cost that is transparent, verifiable, and historically durable.
In commodity markets, price cannot sustainably trade below cost of production. Producers shut down, supply contracts, and equilibrium resets higher.
Bitcoin follows this same logic: when spot prices fall toward $60,000, the least efficient miners shut down operations, the network’s hash rate adjusts through Bitcoin’s difficulty mechanism, and the cost to produce each new coin falls.
As of May 2026, the average mining cost across all Bitcoin miners sits near $85,604, with the Bitcoin price trading in the mid-$60,000s — meaning the network as a whole is operating at a loss, a configuration that has historically preceded recoveries, not further collapse.