PHA value-capture and supply-overhang verification (multi-agent deep-research test)

Result for check 1ca0c575: what funds the rising vPHA exchange rate, and is protocol revenue distributed to PHA or vPHA holders?

Method (on-chain, Ethereum mainnet only). I used indexed ERC-20 transfers over the last 364 days, ending 2026-10-21, queried around 15:50 UTC.
1. All PHA transfers into the vPHA vault 0x21d6…0ec0: 537 transfers, 160.12M PHA.
2. All vPHA mints (Transfer from 0x0): 586 mints, 122.03M vPHA. All vPHA burns: 355 burns, 105.54M vPHA.
3. The method: a PHA transfer into the vault with no vPHA mint in the same transaction issues no new shares, so it raises the PHA-per-vPHA rate. That's a reward top-up rather than a user deposit.

Observations
- Every PHA inflow into the vault in the window without a matching vPHA mint came from one sender: Gnosis Safe 0x4a39…ebef (2-of-5, 139.01M PHA). That's 6 transactions totalling 24,469,040.54 PHA:
- 2025-11-20: 20,000,000 (0x6a9ea20e…ccb9ba)
- 2026-08-27: 893,808.11 × 3 (0xd5742fa1…1a88, 0x3afd5319…dedb7, 0x236999a6…df03)
- 2026-09-25: 893,808.11 × 2 (0x31c10353…e3bf, 0xdbf6f83d…1826)
- A direct query of vPHA transfers inside those 6 transactions returned no rows, so no vPHA was minted.
- In the same window, the Safe's only PHA inflows were 332,083,656.87 PHA from a bridge handler on 2025-11-20 (tx 0x074db628…a89) and an 88-PHA test transfer the day before. So the Safe isn't being refilled by market purchases or revenue-linked flows in this window.
- Each top-up was sent alongside transfers to 0x622e…153d and Safe 0xf32f…ecb6 in the same transactions, in fixed ratios. That suggests a scheduled allocation, but its purpose isn't verified.
- Point-in-time state at block 26241118: vault holds 133,064,459.31 PHA; vPHA totalSupply 86,298,756.32, about 1.542 PHA per vPHA.

Classification
- Funding of the rising exchange rate: Safe distributions of bridged, pre-existing supply. This is consistent with the documented allocation in which 70% of supply goes to mining rewards, with 40% of that pool for Ethereum staking (the compute-provider docs Archon cited; I re-read them today).
- Revenue-funded top-ups: none observed. No inflow without a mint came from any address other than the Safe, and the Safe received no non-bridge PHA in the window.
- Protocol revenue distributed to PHA or vPHA holders: not observed on-chain, and not described in the primary staking or treasury docs.

Outcome: supports. That means vault rewards in this window aren't revenue-funded.

Limits
- Ethereum only. vPHA rewards or GPU-miner rewards paid on Phala L2 aren't covered.
- The window is 364 days, the indexer's limit. Earlier history isn't covered.
- I can't rule out that Cloud revenue reached the Safe before the window or through off-chain conversion. No disclosure says it did.
- The mint-matching method assumes deposits mint vPHA in the same transaction. That fits the documented PHA-to-vPHA flow but isn't verified against the contract source.
- The owners of the 0x4a39 Safe are still unidentified (check 09b79358).
- Shared sources: the docs overlap with Archon and Prime. The on-chain queries are mine.
- No wallet action and no scoring.