Most capital destroyed in Bitcoin mining infrastructure wasn’t lost to bear markets or hashprice compression. It was lost before the first machine went online. The common thread is not bad timing. It is diligence that was either absent entirely or structured around the wrong questions. The deal type determines the risk profile, and the risk profile determines the checklist.
The Cost of Getting It Wrong
Hosting: The Trust Verification Problem
Greenfield: Where Feasibility Studies Lie
Acquisition: Buying the Liability Too
A First-Pass Scoring Framework
Where Each Deal Type Actually Breaks 🔒– Premium
FBOX’s 10MW modular data center solution is designed for rapid, turnkey deployment with prefabricated cooling, power, and IT modules. With major subsystems pre-tested before shipment, FBOX helps operators reduce on-site complexity, accelerate infrastructure rollout, and achieve faster time-to-compute in high-density AI environments.
The Cost of Getting It Wrong
The mining industry moves fast. Term sheets get signed before site visits happen. Verbal commitments from operators get treated as contractual ones. Power figures in pitch decks don’t always reflect what’s in the PPA, or sometimes there is not a signed PPA at all. None of this is unique to any one geography or market cycle; it’s structural to an industry where many deals are done informally and where physical assets are often thousands of kilometres away from the capital deploying into them.
The three deal types — hosting, greenfield, acquisition — each carry fundamentally different risk profiles. Hosting is an operator trust problem: you’re deploying assets you don’t control into a facility you can’t always inspect. Greenfield is a feasibility problem: building from scratch means nothing is confirmed until it is. Acquisition is a liability discovery problem: the asset has a track record, but that track record is precisely what needs to be examined hardest.
A structured diligence process doesn’t slow down good deals. It kills bad ones faster. The deals worth pursuing survive scrutiny. The ones that don’t were always going to fail and the only question was when, and whose capital was at risk when they did.
Hosting: The Trust Verification Problem
Hosting deals are fundamentally contracts for services you can’t directly supervise. You’re deploying hardware into a facility operated by a third party, secured on terms negotiated at signing, enforced at a distance. The due diligence question isn’t whether the pitch deck or website looks credible but whether the entity behind it is real, capitalized, and contractually bound to deliver what it’s promising.
Entity verification is the first checkpoint. The legal entity on the hosting agreement must match the actual counterparty not a holding company, not a related entity, not an entity registered in a jurisdiction where enforcement is unworkable. Operators sometimes present a US entity in pitch materials while the actual contracting counterparty is a shell in a jurisdiction with no effective legal recourse.
Data center infrastructure and power deserve the most scrutiny of any hosting criterion. Claimed rackspace capacity, deployment speed, grid stability, and uptime guarantees must be cross-referenced against the actual footage, power supply agreement and pool watcher links, not just taken from the facility’s promotional materials. What is a realistic deployment timeline? What is the curtailment risk? What happens when the local grid goes down?
The SLA uptime guarantee needs to be not just contractually stated but enforceable with defined penalty or compensation mechanisms, incident notification procedures, and remedies that don’t require litigation in a foreign jurisdiction to trigger. Monitoring access must be guaranteed in the contract, not just offered informally. You should see your machines running in real time.
On the commercial side: verified references from existing clients matter more than claimed portfolio size. How long has the operator been running at this facility? What’s the machine count under active contracts? Can those clients be contacted independently?
Key hosting red flag: billing on nominal consumption rather than metered consumption. You pay for contracted MW capacity whether your machines are running or not. This clause is usually buried in the billing model section, not on the headline term sheet.
Greenfield: Where Feasibility Studies Lie
Greenfield is the highest-risk deal type because the asset doesn’t exist yet. Everything is contingent: the power, the timeline, the cost, the regulatory clearances. A persuasive feasibility study and a working facility are separated by a gap that regularly consumes more capital and more months than the original plan.
Power grid access is the binding constraint. A signed Letter of Intent with a utility is not a grid connection agreement. Interconnect studies regularly surface surprises like a substation upgrades. In competitive power markets the interconnect queue in some areas has extended to multiple years. Greenfield projects that haven’t locked a firm grid connection agreement before raising development capital are just speculative.
CapEx estimate quality is the second most critical factor. Third-party or detailed bottom-up estimates are the baseline standard. A contingency budget below 15% signals either optimism or inexperience. Geotechnical surveys, civil works assessments, and site drainage evaluations should be completed before any CapEx figure is considered reliable. Desktop studies often miss the conditions that add 10-15% to civil costs.
Greenfield checklist priority: (1) signed grid connection agreement, (2) third-party validated CapEx with 15%+ contingency, (3) EPC contractor committed (not just named), (4) permits and zoning confirmed not ‘in process’. Any other order is backwards.
Acquisition: Buying the Liability Too
An operating asset looks efficient on paper: existing infrastructure, an established track record, a client base already generating revenue. But the track record is precisely the thing to examine hardest in acquisition due diligence, because it contains everything the seller has chosen not to disclose in the information memorandum.
Start with the ASIC fleet. Ask for the average fleet age, the J/TH distribution, and the sellers’ depreciation methodology when self-mining. Then build your own replacement timeline.
Trailing 12-month financials should be audited or independently verified. EBITDA inflated by above-market hosting rates charged to related entities under common control is a pattern that surfaces in acquisition diligence more often than it should. The delta between the hosting rate on managed contracts and the market hosting rate is the number to watch.
Key person risk is frequently underweighted. If the founder or primary operator is leaving at close and the facility relies on their relationships with the utility, the local permitting authority, or the client base, you’re acquiring an asset that may not function as modelled once they’re gone. A retention plan for critical staff should be a closing condition, not an afterthought.
PPA assignability is non-negotiable. If the power contract cannot be transferred to the acquiring entity, the most valuable infrastructure asset in the deal doesn’t transfer either. Confirm assignability with consent from the utility if required, before any other valuation work proceeds.
Do a valuation benchmark. The $/MW or EV/EBITDA should be cross-referenced against comparable transactions. Entry price above greenfield replacement cost requires a specific justification. This is usually a locked long-term PPA, an established client base with contracted revenue, or expansion optionality that isn’t available on the open market.
A First-Pass Scoring Framework
The DMS Bitcoin Mining DD Scorecard is a 100-point first-pass filter built across three deal modes: hosting, greenfield, and acquisition. Each mode maps the deal’s specific risk pillars to weighted criteria, with scores reflecting the relative materiality of each factor.
The scorecard produces one of three verdicts. A score of 70 or above means no structural blockers, the deal moves to full due diligence. Between 45 and 69, the deal can proceed but specific gaps must be resolved before capital commits, or it should be reflected in a discount. Below 45, the deal requires either major renegotiation or an exit, no financial model should open until the structural problems are addressed.
The tool includes an N/A toggle for criteria that don’t apply to a specific deal so that partial scope deals still score fairly rather than being penalized for irrelevant criteria. Each criterion has a notes field for follow-up items. The full summary exports in one click for use in investment memos or deal files.
ESG is tracked as a bonus outside the 100-point core: up to +8 points for hosting, +6 for greenfield and acquisition. It reflects energy source accuracy, emissions profile, and community risk, without distorting the core investment signal.
The scorecard is a first-pass filter, not a substitute for full Due Dilligence. Legal entity verification, power contract analysis, SLA stress-testing, and a complete investment memo are built separately. The scorecard tells you whether a deal deserves that work or whether you should be looking at the next one.
🔒 Premium subscribers only — what follows goes deeper: the specific failure modes sorted by deal type, and the exact criteria that disqualify a deal before any financial model opens.
