In the last week of a debt-restructuring negotiation that almost nobody in crypto was watching, a sentence began circulating in the anglophone press: Senegal is "exploring" Bitcoin mining.
No wattage. No hashrate. No named partner. No ministry, no budget line, no procurement document. Just a verb in the conditional tense, wrapped in the word "pilot." Tracing the ghost in the machine, I went looking for the parts that normally exist โ an energy source, an ASIC order, a pool agreement, a signatory โ and found only the outline of a project. Not a mining operation. A narrative.
That distinction matters more in a bear market than it would have in 2021, because the cost of being wrong is no longer a drawdown. It is the last of your liquidity. So let me be precise about what can and cannot be verified here, because that precision is exactly what the headline skips.
Context: why Dakar, and why the timing is not accidental
Senegal's economy is roughly $30 billion. It runs on the CFA franc โ the XOF โ pegged to the euro at a fixed rate and managed by the BCEAO, whose headquarters sit in Dakar. There is an irony worth holding onto: the regional central bank is in the same city as the government now rumored to be weighing a dollar-denominated, non-sovereign reserve asset.
The political context is not incidental. The administration elected in March 2024 campaigned on renegotiating energy and fishing contracts, auditing extractive deals, and a broadly pan-Africanist posture. By late 2024 it had surfaced roughly $7 billion in previously undisclosed debt, which forced a renegotiation of the country's IMF program. When a finance ministry is staring at a restructured repayment schedule and a revenue gap, the appeal of an asset that appreciates in dollar terms while liabilities are denominated in a euro-pegged currency is not irrational. It is simply not sufficient.
It is also worth pausing on the word "pilot," because in state planning vocabulary it is load-bearing. A pilot is what a government calls something when it wants the political credit of announcing without the fiscal commitment of executing. It implies a small budget, a short horizon, a review clause, and an easy exit. Nine out of ten pilots in infrastructure never scale; they terminate quietly at the end of the review window, and the only trace is a line item that stops appearing.
Sovereign mining itself is not new. Bhutan ran it for years through Druk Holding before anyone noticed โ we reconstructed it from on-chain flows, not press releases. Ethiopia opened hydro capacity to Chinese miners after the April 2024 halving and the Kazakh crackdown, and now hosts a meaningful share of global hashrate. El Salvador announced constantly and mines a rounding error. The Central African Republic made Bitcoin legal tender in 2022 and quietly reversed it. The pattern is consistent: states that actually mine tend not to talk, and states that talk tend not to mine much.
Core: the arithmetic a pilot cannot escape
Strip the narrative away and sovereign mining is an energy arbitrage. Three numbers decide whether it works.
The first is issuance. Since April 2024 the block subsidy is 3.125 BTC. Network-wide that is 450 BTC per day. Against a hashrate somewhere near 900 EH/s and a bear-market BTC price around $60,000, the entire network grosses roughly $27 million daily โ about $30,000 per EH/s per day, before a single watt of power is paid for.
The second is power density. One exahash per second of modern ASIC capacity draws roughly 17.5 megawatts continuously โ call it 5,000 S21-class machines at 200 TH/s and 3.5 kW each, plus cooling overhead. That is 420 megawatt-hours a day. At six cents a kilowatt-hour, power alone costs $25,200 against $30,000 of revenue. At ten cents, the operation is underwater. There is no configuration in which a miner survives on expensive marginal electricity. This is not a matter of strategy or hashrate management; it is a floor set by physics and the difficulty adjustment.
The third is Senegal's grid. It is gas and heavy fuel oil, with the first volumes from the Greater Tortue Ahmeyim field only now arriving. Industrial marginal cost in that generation mix is not six cents. If the pilot draws from the grid at unsubsidized rates, it loses money at any realistic BTC price; if it draws at subsidized rates, the state is transferring fiscal capacity into a volatile foreign asset. Both branches end badly, and neither is disclosed.
That is the whole economic argument, and it fits in a paragraph. Which is precisely why the interesting version of this story โ the one I would assign maybe 30% probability โ is not grid mining at all. It is flare-gas monetization. Associated gas that would otherwise be burned or stranded carries a near-zero opportunity cost, and converting it into hashrate is a genuinely coherent use of a wasted resource. If a Senegal-linked entity eventually discloses an offtake agreement with a gas producer, the model changes completely. If it discloses a grid connection, it does not.
Then there is the supply chain, which the coverage has ignored entirely. ASIC manufacturing is concentrated in a handful of Chinese vendors. Pools are predominantly offshore. A state can call its mining sovereign while purchasing every input from a foreign supplier and selling output into a foreign market โ that is sovereignty over the balance sheet, not over the stack. I saw a milder version of this in 2017 while auditing Uniswap's V1 contracts in Buenos Aires: the constant-product formula looked neutral and wasn't, quietly transferring risk from traders to liquidity providers. Mechanism design always hides a directional bet. Here the direction points away from Dakar.
The currency layer compounds it. The XOF is pegged to the euro, and reserve management runs through a framework built for euro-denominated assets. A sovereign miner converting BTC into XOF touches capital-account rules that were not written with this in mind. The friction is not a statute that says no โ it is a set of institutions that say not like this, and institutions move slower than narratives.
And beneath all of it sits the mismatch I keep returning to. I spent three months in Patagonia after Terra, and what I carried out of that silence was not a lesson about math. It was a lesson about the assumption that math is stable. A debt schedule needs predictable cash flows. Bitcoin mining revenue is denominated in an asset with 50โ80% annualized volatility, against a subsidy that halves every four years. Pledging hashrate revenue against a restructuring schedule is a duration mismatch wearing a technological costume. The code remembers what the market forgets: reward per unit of work only ever declines.
Contrarian: the beneficiary is not Senegal
Here is the angle nobody is taking. The likely beneficiaries of this story are not Senegalese taxpayers. They are the ASIC vendors, the pool operators taking a cut of every block, and above all the narrative packagers who need a fresh sovereign-adoption datapoint to sell into a market that has run out of new ones.
Consider the asymmetry. A mining pilot requires hard-currency capital expenditure on hardware, paid to foreign suppliers. Revenue arrives in BTC. The debt is in euros. The state absorbs the FX and volatility risk; the supply chain gets paid regardless of outcome. That is a transfer, not an investment.
Second, the market impact is approximately zero, and anyone telling you otherwise is selling something. Senegal's GDP is $30 billion. A full exahash would be roughly a tenth of one percent of network hashrate. There is no listed equity, no token, no instrument through which this headline can be expressed. Any BTC price move attributed to Dakar is misattribution, and in a bear market misattributed headlines are how retail gets liquidated. When the herd wakes, the signal has already faded.
Third, and least comfortable: announcement and execution are inversely correlated in this sector. A pilot with no named partner, no stated megawatt figure, and no disclosed counterparty sits at the extreme end of the announcement-only tail.
Takeaway
The only number that will settle this is wattage. Watch for three specific disclosures: a named energy source, a named ASIC supplier or pool, and a stated megawatt figure. Any one of them converts a rumor into something with a runnable model; none of them means it stays a sentiment artifact.
Reading the silence between the blocks is the only honest way to price a story like this. The question was never whether Senegal will mine Bitcoin. It is whether anyone will be able to tell if it does.