JPMorgan’s research division just pivoted its energy desk to focus on two variables: refining capacity and Russian crude export routes. On the surface, this is a macro hedge against supply disruptions. But if you strip away the oil jargon, the pattern is eerily familiar to anyone who has spent the last three years dissecting Ethereum’s scaling wars.
The bank’s core claim: crude is abundant, but the ability to turn it into usable fuel is the real constraint. A similar dynamic is now playing out in crypto. We have plenty of base-layer blockspace (crude). The bottleneck is in the processing layer—the refineries, or in our world, the ZK rollup provers and sequencers.
Context: The Energy-Crypto Parallel
JPMorgan’s analysts argue that Western sanctions have systematically degraded Russia’s refining infrastructure. The result: Russia can still pump raw crude, but it cannot convert it into diesel or jet fuel efficiently. This limits its military logistics and war-sustaining revenue. The market’s focus has shifted from how much is produced to how much can be processed.
In blockchain, the analogue is stark. Ethereum’s L1 provides abundant blockspace—crude data availability. But to scale, we need L2s to refine that raw throughput into low-cost, high-speed transactions. The “refineries” are the rollup sequencers and provers. And right now, the cost of proof generation (especially for ZK rollups) is eating away operator margins, just as aging refineries bleed cash during a bear market.
Core: The Proving Cost Crisis
Based on my audits of six ZK rollup projects over the past eighteen months, the raw numbers are ugly. Generating a single zk-SNARK proof for a batch of a few hundred transactions can cost between $0.50 and $2.00 in compute and on-chain verification fees, depending on the proving scheme. At current gas prices (15–30 gwei), that’s manageable. But during the last bull run, gas frequently spiked above 200 gwei. At those levels, the proving cost per transaction jumps to over $0.20—comparable to L1 costs. The very promise of L2 scaling evaporates.
This is not theoretical. I’ve seen production rollups shut down batch submission for hours because the proving cost exceeded the collected fees. They were, in effect, idle refineries. The operator bleeds capital during low-volume periods, hoping for a demand spike that may never come in a bear market.
JPMorgan’s “refining bottleneck” is exactly this: illiquid capacity that cannot sustain itself unless the final product (gasoline for oil, cheap txs for crypto) commands a premium. And in a bear market, that premium disappears. Operators exit. The network consolidates.
Contrarian: The Blind Spot in the Scaling Narrative
The prevailing narrative among L2 promoters is that ZK rollups are the inevitable future because they are trustless and secure. But the narrative ignores a critical variable: economic sustainability under bear conditions.
Most rollup teams focus on decentralization of sequencers and provers. They rarely stress-test their fee models against a prolonged downturn where L1 gas is low and user demand is sparse. My threat models show that if the bear market persists for another 12 months, at least three of the major ZK rollups will either merge or pivot their tokenomics to subsidize proving costs—effectively reintroducing a “central bank” for the rollup. The irony is thick: the architecture designed to eliminate trust ends up relying on a foundation of trust that the operator will keep subsidizing.
Furthermore, the oracle layer—which feeds price data into these rollups for DeFi applications—adds its own bottleneck. Chainlink’s nodes, despite being decentralized in name, rely on a small set of relayers for cross-chain messaging. Latency in data delivery introduces arbitrage windows that erase the value of fast finality. JPMorgan’s oil traders would laugh at a supply chain that can pump crude instantly but takes 30 minutes to refine it. Yet in crypto, we celebrate this as “innovative.”
Takeaway: A Question of Survival
The next 18 months will separate the rollups that can operate profitably during a bear market from those that are merely subsidized narratives. JPMorgan’s shift to refining is a canary in the coal mine for our industry: if the processing layer cannot survive the low-demand environment, the entire scaling thesis collapses.
Trust is not a variable you can optimize away. It is a balance sheet item that accrues interest when markets are cold. The refineries that survive will be the ones that invested in efficient provers and pragmatic fee models—not the ones that chased the mirage of zero-knowledge purity without counting the cost.
