Bitdeer Technologies Group [BTDR]
As of: 2026-07-29 · Spot: $8.92 · EV (ex digital assets): $3,207m · Framework: v1.5.1 / criteria.md 2026-07-29
Cluster context: CLUSTER_Analysis.md
Two outputs are required on every name: a 12-month target and the implied-path test. Reporting only one is a defect. Both are addressed below, and where an input does not exist it is declared UNIDENTIFIED rather than substituted with a peer median.
BTDR grew total revenue +77.3% in FY2025 (from $349.782m to $620.253m). No comparator in the AI-compute set
brackets 77% — CoreWeave grew +111.6% YoY in Q1 2026 and +168% in FY2025; the mature contracted-datacenter set
grows 5–15%. Per valuation.md, the multiple is declared UNIDENTIFIED rather than defaulted to a peer median.
And a second, more fundamental reason the multiple is unidentified for BTDR specifically: the company is three businesses and the analysis has a comparator for none of them.
| Business | FY2025 revenue | % | Appropriate comparator set | Status here |
|---|---|---|---|---|
| Bitcoin self-mining | $396.0m | 63.9% | mining peers | UNIDENTIFIED, floor at zero — see below |
| ASIC design and sale | $108.3m | 17.5% | semiconductors | NOT BUILT — the largest gap in this file |
| Hosting (membership + general + subs) | $98.4m | 15.9% | declining service business | UNIDENTIFIED |
| AI cloud services | $6.8m | 1.1% | CRWV 6.6x, WYFI 10.9x | identified but immaterial |
The mining leg is scored UNIDENTIFIED with a floor at zero, on the cluster-wide evidence: CORZ discloses a −56% self-mining gross margin and HIVE's gross mining margin fell 49% → 24% in two quarters. A stream with a negative gross margin at one operator and a halving margin at another has no defensible positive revenue multiple.
Holding the pure-play anchor of 6.6x EV / annualised AI revenue and today's EV ex-crypto of $3,207m fixed:
This is a category error, not a valuation gap — the same conclusion as HUT (89x required) and for the same reason. BTDR is not an AI-compute holding at any multiple, and pricing it as one is precisely the mistake this cluster exists to prevent.
BTDR is the cheapest miner in the cluster on blended revenue at 5.2x EV/Sales (HIVE 2.3x is cheaper but is a $678m company). The honest variant view is not the AI pivot but the ASIC hardware business: $108.3m of revenue from zero, $153.9m of R&D, own-designed SEAL silicon, and a position as one of very few credible alternatives to Bitmain — whose pricing power every other name in this cluster is exposed to (see HUT's $360.9m Bitmain miner-purchase liability).
But that view cannot be scored here, for three stated reasons: 1. No semiconductor comparator set was built. Applying a datacenter multiple to a chip designer is the same category error in reverse. 2. Gross margin fell from 21.1% to 9.8% while the hardware line scaled — so it is not yet demonstrated that the hardware business is accretive to group margin. 3. The data is seven months old and there is no interim XBRL to check whether the hardware line persisted into 2026.
| Exit multiple on annualised AI revenue | AI revenue required for $3,207m EV | vs $6.8m demonstrated |
|---|---|---|
| 6.6x (CRWV today) | $486m | 72x |
| 10.9x (WYFI today) | $294m | 43x |
| 25.0x | $128m | 19x |
| 118.4x (BTDR today) | $6.8m | current |
Every row is implausible before the last. As with HUT, the exit-multiple sensitivity is the wrong instrument for BTDR — the sensitivities that matter are the bitcoin price (64% of revenue), the durability of ASIC third-party demand (17.5%), and the refinancing of $753.8m of current borrowings. None of the three is an exit multiple and none should be treated as one.
Both required inputs fail, and a third input is missing entirely:
WeightedAverageShares 12.66bn vs 108.7m for the same periods across two 20-Fs), the appearance of a hardware
business from zero, and a swing from a $599.2m loss to a $65.6m profit driven entirely by a derivative mark.Declared UNIDENTIFIED. No target is manufactured.
This is the methodological core of the cluster and it applies directly to BTDR:
| Bitcoin mining | AI/HPC hosting or cloud | |
|---|---|---|
| Revenue driver | commodity price × share of a rising global hashrate | contracted $/MW or $/GPU-hour |
| Counterparty | none | named, with a term |
| Contract | none | 5–10 years, often with prepayments and RPO |
| Terminal value | halves every ~4 years by protocol; global hashrate grows continuously | a datacenter with a signed tenant |
| Observed gross margin, 2026 | HIVE 24% (from 49%); CORZ −56% | CORZ colocation 59%; WYFI 61% |
A single blended exit multiple applied across these two streams is the exact defect that inflated ADMA by 51pp and INOD by 32pp. This analysis therefore ranks on EV / annualised AI-HPC revenue and EV / RPO — a price-for-what-exists and a price-for-what-is-signed — and scores the mining leg at UNIDENTIFIED with a floor at zero, because a stream with a negative gross margin at one operator and a halving margin at another has no defensible positive revenue multiple.
valuation.md| Requirement | This memo |
|---|---|
| The parameter solved for, and every parameter held fixed, named | Stated above: AI-HPC revenue required, holding the exit multiple, share count and net debt fixed |
| Margin: demonstrated − required, in percentage points | Stated above — this is the number the strategy ranks on |
| The exit multiple used, and the implied compression from today's trading multiple, as a number | Stated above |
| Sensitivity over the exit multiple (never over scenario probabilities) | Table above |
| Reverse DCF mandatory where terminal value > 60% of EV | Terminal value does exceed 60% of EV. reverse_dcf.py was not solvable because the exit multiple is UNIDENTIFIED; the required-parameter analysis above is the substitute and is disclosed as such rather than presented as a DCF output |
| Growth-matched anchoring, or declare UNIDENTIFIED | Declared above |
| Do not set the base exit multiple below every stated anchor | Not applicable — no base multiple is asserted |
| TTM revenue, never last fiscal year | TTM used throughout; see the model notes for the derivation |