Hut 8 [HUT]
As of: 2026-07-29 · Framework: v1.5.1 / criteria.md 2026-07-29 · Type: MEASURED — dated events that would confirm or refute the implied path. Consumed by monitoring; blocks nothing.
Every date below is either taken from a filing or explicitly marked as an inference from filing history. Per the brief: a fabricated catalyst date is worse than an absent calendar. Nothing here is invented.
| Date | Event | Source | Why it matters |
|---|---|---|---|
| 2026-02-02 | Far North JV sold to TransAlta for $75.4m; $33.6m gain | 10-Q Q1 2026, Note 3 | already occurred — booked in Q1 2026 and must be stripped from the quarter |
| February 2026 | 2026 ABTC Bitmain Purchase Agreement: ~11,298 S21 XP miners for ~$49.4m, 80% paid by pledging 314 BTC | 10-Q Q1 2026 | already occurred — adds to the pledged-BTC balance and the miner purchase liability |
| ~early August 2026 (estimated from the 2025-08-07 prior-year filing date; NOT company-confirmed) | Q2 2026 10-Q | filing cadence | the print that matters: does the AI/HPC line inside Compute move off ~$3.9m, does revenue stop falling sequentially, and does G&A retreat below revenue? |
| Ongoing, no date disclosed | Bitmain miner redemption windows — the redemption right on the pledged BTC starts from ASIC shipment dates | 10-Q Q1 2026 | determines whether HUT recovers the 3,090 pledged BTC or settles the $360.9m liability |
| 2028 (protocol) | next bitcoin halving | Bitcoin protocol | halves the reward on the 87.5% of revenue that is self-mining |
No AI/HPC catalyst is asserted because none is disclosed. HUT's filings name no AI counterparty, no contracted AI megawatts and no remaining performance obligations. Per the brief, an absent calendar is better than a fabricated one; the two forward dates above are explicitly marked as inferences from filing history.
| Date | Event | Why it matters to HUT |
|---|---|---|
| 2028 (protocol-determined) | Bitcoin halving | halves the block reward on the mining leg. Not a forecast — it is in the protocol |
| continuous | Global network hashrate / difficulty | IREN's 10-Q quantifies the mechanism: its own hashrate rose to 36.0 EH/s and it still mined fewer bitcoin, because implied global hashrate rose faster. The denominator always grows |
| 2026-12-31 | CORZ's hosted-mining operations conclude | removes a competitor's hosted capacity from the market |
| Q3 2026 | WYFI's $865m Nscale contract reaches full revenue contribution | the cluster's cleanest read on realised contracted AI economics — a live benchmark for every claimed pipeline |