HOOK
A Steak 'n Shake cheeseburger costs ten to fifteen dollars. That is the number the entire story has to reconcile against.
On September 10, 2025, the Indianapolis-based chain announced double-digit same-store sales growth across its franchise network, quoting a 19% figure. Management tied part of that momentum to its decision to accept Bitcoin at the point of sale, a rollout that began in May 2025. It also recycled a claim first floated in the spring: Bitcoin settlement cuts payment processing costs by roughly half versus card rails.
Two numbers. One is auditable in principle and was never audited in practice. The other cannot be audited at all, because ordinary Lightning Network payments leave no public footprint. I count the cracks before the dam breaks. Both figures have structural fractures running through them.
This is not a story about whether a restaurant took Bitcoin. It is a story about what a payment ledger looks like when the mechanics sit inside a black box, the marketing runs loud, and the arithmetic of a twelve-dollar ticket does not naturally produce the outcome being sold.
CONTEXT
Steak 'n Shake runs roughly 500 to 550 locations across the United States. The brand is controlled by Biglari Holdings (NYSE: BH), a holding company led by Sardar Biglari that also owns Western Sizzlin and the Maxim apparel brand. The parent is publicly listed. That matters. Any material claim about franchise economics eventually has to survive a filing, not just a tweet.
In May 2025, Steak 'n Shake began accepting Bitcoin directly over the Lightning Network, Bitcoin's second-layer scaling stack built for low-cost, high-throughput microtransactions. The chain paired the rollout with a treasury decision: a $10 million allocation to a corporate Bitcoin reserve, joining the "corporate Bitcoin treasury" wave that MicroStrategy industrialised between 2020 and 2025 and that has since spread from software firms into consumer-facing businesses.
The September disclosure added the financial layer. Franchise same-store sales rose 19% year over year. Company-operated locations improved sequentially, with quarter-over-quarter growth of 11% in Q2 accelerating to 15% in Q3. Management placed the chain ahead of McDonald's, Taco Bell, and Domino's on the same metric. An executive named Michael Boes was quoted discussing the results, and the brand's social accounts amplified a "Bitcoin is working" narrative.
One procedural flag before the analysis. The source material places a Boes appearance at "Bitcoin 2026." That conference runs in late May each year. In September 2025, a Bitcoin 2026 keynote had not happened. This is not a clerical slip. It is a tell. When promotional timelines drift, the numbers attached to them deserve harder scrutiny. I learned that over eight years of auditing token launches. Trust code, not calendars.
CORE — The Mechanics of a Lightning Merchant
To test whether the fee claim holds, trace what actually happens when a customer pays for a milkshake in satoshis.
Three layers are involved. The customer holds BTC, either in a self-custody wallet or a custodial app like Strike or Cash App. The merchant needs a payment endpoint. Between them sits a processor. In the 2025 Bitcoin payments stack, the realistic options are OpenNode, Strike's merchant suite, or a self-hosted BTCPay Server instance.
The flow that most large chains run looks like this: consumer wallet → Lightning channel or custodial transfer → merchant processor → real-time or batched conversion to USD → bank deposit.
The processor layer is where the economics live, and it is the layer the announcement never names. That omission is the first crack.
A Bitcoin-only processor typically charges 0.5% to 1% to accept Lightning payments and convert to fiat. A standard card processor charges roughly 2.5% to 3.5% once you blend interchange, assessments, and acquirer margin. On the surface, that supports a "half off" claim. But only if the merchant fully offloads custody to the processor.
Now run the alternative. If Steak 'n Shake wants to hold some settlement in BTC, which the $10 million reserve strongly implies, it is not fully offloading custody. It is carrying balance-sheet exposure and managing hot and cold wallets. That costs money: custody infrastructure, key management, reconciliation tooling, and the working capital tied up in the reserve. None of that appears in the 50% headline.
During the 2020 DeFi Summer, I ran high-frequency arbitrage across Uniswap and Sushiswap and captured over $45,000 in spreads during the UNI airdrop window. The lesson was not about returns. It was that fee estimates almost always omit second-order costs — gas, slippage, rebalancing — until the P&L forces them into view. Merchant payment economics work the same way. The headline rate is not the cost. The full settlement lifecycle is the cost.
So the 50% claim holds only under a specific structure: zero-custody, processor-managed conversion, near-zero-fee rails. The moment the merchant holds BTC, the arithmetic changes. The moment the merchant runs self-hosted infrastructure, the labour cost appears. The ledger bleeds faster than the logic holds.
Lightning's Privacy Is an Audit Wall
There is a harder reason to distrust the attribution. Lightning payments are not publicly auditable.
Unlike on-chain Bitcoin transactions, which broadcast to a public ledger and can be counted by anyone running a node, Lightning payments route through private channels. An outside analyst cannot reconstruct how many people paid with Bitcoin at a specific restaurant, on a specific day, for a specific amount. The privacy that makes Lightning usable as a payment rail is the same property that makes the "Bitcoin drove our sales" claim unfalsifiable from the outside.
This cuts both ways. It protects the merchant from competitors reading their transaction flow. It also lets the merchant attach any narrative to the metric without external contradiction.
I have audited this blind spot before. In 2017, I manually reviewed the ERC-20 implementations of three mid-tier ICOs. One of them, CoinDash, had an integer overflow in its fundraising logic that the team had missed entirely. I sent the finding to the developers through GitHub, not Telegram. The takeaway was not that the project was fraudulent. It was that teams routinely report numbers they have never stress-tested. That pattern has not changed. It has migrated from token contracts to point-of-sale systems.
With Lightning, the verifier's toolkit is empty. No explorer, no auditor, no analyst can confirm the payment mix. When a restaurant says Bitcoin accelerated same-store sales, the statement is structurally uncheckable. You are trusting the operator. That is not a technical proof. It is a promissory note.
The 19% Attribution Problem
Set the fee question aside and examine the sales number itself.
A 19% same-store sales increase is a real result. It is also a metric that measures revenue, not foot traffic. Through 2025, US restaurant menu prices stayed elevated, and same-store growth across the sector was frequently driven by repricing rather than additional customers. A chain that raised menu prices and ran heavy promotional spend can produce a double-digit same-store number without a single incremental transaction.
Three forces plausibly drive the 19%: menu repricing, a low-base effect from a prior weak quarter, and the marketing halo of the Bitcoin announcement itself. The Bitcoin payment option is a fourth, and its contribution is almost certainly the smallest of the four, because the population of customers who (a) hold Bitcoin, (b) want to spend it on fast food, and (c) live near a participating location is small relative to the chain's total traffic.
A ten-to-fifteen dollar ticket is also the wrong product for Bitcoin's 2025 narrative. BTC behaves as a savings asset, not a spending currency. The people holding it are, on average, the least likely cohort to part with it for a cheeseburger. The demand-side logic for "Bitcoin lifts sales" is thin before you look at a single data point.
Compare the disclosure to what other chains have done. Starbucks piloted Bitcoin and Lightning settlement through Bakkt and then quietly scaled the program back. Jersey Mike's has accepted Bitcoin and other assets since 2024 without building a marketing campaign around sales attribution. McDonald's, the sector's volume leader, has not touched it. Steak 'n Shake's distinguishing move is not that it accepted Bitcoin. It is that it fused acceptance into a sales narrative. That fusion is a media product, not a monetary one.
The Treasury, the Tax Drag, and the Compliance Layer
The $10 million Bitcoin reserve deserves its own examination, because it is the part of the story that actually shows up in accounting.
Under US GAAP, FASB's ASU 2023-08 requires companies holding crypto assets to measure them at fair value, effective for fiscal years beginning after December 15, 2024. For a public parent like Biglari Holdings, that means the reserve is marked to market each reporting period. A 30% drawdown in BTC becomes a non-cash impairment line that flows into the income statement and into reported earnings.
For a restaurant holding company whose core business is thin-margin food service, a $10 million BTC position is a meaningful volatility vector. If BTC falls hard, the reserve produces an accounting loss that has nothing to do with burger sales and everything to do with the equity story.
Then there is the tax layer. Every BTC payment is a taxable event. The customer incurs capital gains or losses on the spent coins. The merchant recognises revenue at fair market value on the date of receipt and must track the basis of any coins retained. Across hundreds of potential daily transactions over hundreds of stores, that is a reconciliation workload that did not exist before. It requires accounting system changes, staff training, and ongoing compliance. None of it appears in the 50% fee claim. It never does.
And there is the licensing question. Accepting BTC for goods is not, by itself, money transmission. But the moment a third-party processor holds customer funds, that processor needs state-level MSB registrations. A chain operating across dozens of states inherits that complexity through its vendor. The announcement discloses no processor. That silence makes it impossible to verify where the compliance burden sits.
I spent six months in 2024 cross-referencing BlackRock's IBIT and Fidelity's FBTC flow data against on-chain exchange outflows. The lesson there transfers directly. Institutional money moves on mechanics — custody, settlement, reporting — not on narrative. A restaurant accepting Bitcoin is a narrative event. A restaurant building a compliant settlement stack that holds up under audit is a mechanics event. Steak 'n Shake has announced the former and hidden the latter.
CONTRARIAN — The Halo Is the Product
Here is the counter-intuitive read. The Bitcoin acceptance is probably not a payment strategy at all. It is a customer-acquisition and public-relations strategy wearing a payment costume.
The measurable value of "first major burger chain to accept Bitcoin" is enormous in earned media. It generates crypto-native press, brand affinity among a high-income demographic, and a durable talking point. All of it drives traffic independent of whether a single satoshi changes hands at the counter. The 19% same-store figure is consistent with a well-timed marketing campaign riding a low base.
If that interpretation is correct, then the fee savings, the reserve, and the sales lift are three separate stories bundled into one press release. The fee savings depend on an undisclosed custody model. The reserve is a balance-sheet bet. The sales lift is unattributable. Build the cage, then watch the beast jump in — except here the "cage" is a settlement rail the company will not describe, and the "beast" is a metric no outsider can verify.
There is an ecosystem angle worth noting. Competitors can replicate the technical stack in a single day. OpenNode and BTCPay Server are open and low-friction. Steak 'n Shake's first-mover advantage is not technical; it is reputational. The moat is shallow. It consists of the timing of the announcement and the strength of the parent's franchise agreements, not of any proprietary infrastructure.
Where the model could scale is inside Biglari Holdings itself. If the parent ports the same settlement stack across Western Sizzlin and its other brands, the synergy becomes structural — shared processors, shared treasury policy, shared compliance. If it stays a single-brand stunt, it was always a media asset that happened to settle on a Lightning node.
TAKEAWAY — Watch the Filing, Not the Feed
The signal to track is not BTC's price reaction, which will be negligible. It is whether Steak 'n Shake names its payment processor and discloses a payment mix in the next quarterly filing. If the processor stays unnamed and the mix stays hidden, treat the 50% and the 19% as unverified claims, not as data.
If Biglari replicates the stack across its other brands, the model has legs and the corporate-Bitcoin-treasury thesis extends into food service. If it stays a single-brand publicity event, it was never a payment story. Liquidity is just borrowed time with a premium, and marketing is just borrowed credibility with the same structure. Survival is the only alpha that compounds. Watch the filing.
