BNB: supply down by design.

Two metrics, two ledgers. 65.22M BNB out of 200M genesis are already gone (32.6%), the protocol keeps burning more every quarter, and there is nothing scheduled to replace them.

The setup

BNB launched in the 2017 ICO as an ERC-20 on Ethereum, migrated to its own chain in 2019, and now lives natively on BNB Smart Chain (BSC). Genesis supply was 200M with a publicly stated target of reducing it to 100M through systematic burns. Eight years in, the burn is already a third of the way past that target.

Live numbers, origin-first from the BSC mainnet RPC (bsc-dataseed.binance.org):

• Total supply: 134.78M BNB (down from 200M genesis → 65.22M cumulative burned, 32.6%)
• BEP-95 real-time burn address (0x489a…): 1.02M BNB held
• Universal dead address (0x000…dEaD): 14.87M BNB held (legacy quarterly burns through 2021)
• Auto-Burn direct supply destruction (post-2021): ~49.33M BNB (the remainder)
• Active BSC validators: 45 (expanded from 21 via BEP-294)
• Price: $655.75 → market cap ~$88.4B

Two burn mechanisms split the work. Auto-Burn (since BEP-95, April 2021): a quarterly destruction of BNB calculated by formula (price + BSC block count). Burns ~1–2M BNB per quarter directly from a pre-allocated team-held pool. This is a supply reduction, not a market buyback — it replaced the pre-2021 mechanic that used 20% of Binance's profits to buy BNB on the open market. BEP-95 real-time burn: roughly 10% of validator gas rewards burned every block. Continuous, on-chain, modest in volume.

The sell ledger

What the design predictably puts on the market.

Inflation: zero. BNB has no protocol inflation. New BNB is never minted on a schedule.

Vesting: zero. The 2017 allocations (10M founders, 40M angels, 50M public ICO, 80M team) all finished vesting between 2021 and 2022. Nothing scheduled to release today.

Tag B is the interesting line. Under the framework's updated source-#3 rule, only identified coordinated entities count, and exchange custodial wallets are explicitly excluded because those coins belong to depositors, not the exchange's discretion. Applied to BNB: the biggest "Binance-labeled" wallets on BscScan — e.g., 0xf977…aceC with ~6.3M BNB — are Binance Peg Token reserves backing wrapped BNB on other chains. Custodial. Excluded. CZ's personal BNB holdings are not transparently disclosed at a single public address. The honest read: BNB's Tag B is small and not enumerable to the same precision as Ondo's Foundation Safe. That's not the framework failing — it's the framework correctly identifying that Binance the company does not separate corporate treasury from custodial holdings publicly.

Bankruptcy estate: zero. No FTX-style estate distributing BNB.

The buy ledger

What the design predictably takes off the market.

Burn is the headline. Cumulative burn has retired 32.6% of original supply, and the protocol is structurally on track toward the stated 100M target — roughly another 35M BNB to go. At the current ~5–8M / yr rate, with BNB at $655.75, that's ~$3–5B / year of supply destruction — economically equivalent to that much annual buy pressure in net price impact.

A subtlety the framework cares about: Auto-Burn ≠ market buyback. The original 2017–2021 mechanic was an open-market buyback (Binance used 20% of profits to buy BNB on exchanges, then burn). Since BEP-95, Auto-Burn destroys BNB from a pre-allocated pool — no open-market purchase happens. From the framework's view this scores as Metric 2 source #2 (Burn), not source #1 (Revenue-backed buyback). Price impact similar, mechanic different, headline line item different.

Net position

Combine the ledgers:

• Sell, Tag A: ~0 (no inflation, no vesting)
• Sell, Tag B: small, hard to enumerate (Binance custodial excluded per rule)
• Buy, Tag A: ~5–8M BNB / yr from burns + on-chain gas demand
• Net structural change: supply DOWN ~3–6% / year

The structural conditions on Metric 1 + Metric 2 are favorable. This is the opposite shape of Ondo (scheduled supply up, structural buy ~0) and NEAR (inflation up, tiny burn offset). BNB's supply side shrinks by design with no scheduled offset.

The only structural risk

There is no inflation, no vesting, no estate. The only structural risk is opacity: Binance the company holds significant BNB outside identified custodial wallets, and the framework cannot fully read this. If a future leak or transparency disclosure reveals a separately-held corporate treasury, that becomes a new Tag B line. Until then, the visible sell ledger is empty.

What to watch

1) Quarterly Auto-Burn announcements at binance.com/en/bnb-burn-schedule — the amount and the formula's inputs (BNB price, BSC block count). 2) BEP-95 burn rate on 0x489a… — track quarterly to see chain-activity-driven burn. 3) BSC validator count + self-bonds. 4) Any Binance corporate transparency disclosure— would add a new Tag B line.

Data note. All numbers are origin-first from the BSC mainnet RPC. Cumulative burn = 200M genesis − current total supply (134.78M). Burn-address balances via eth_getBalance. Validator count via the BSC ValidatorSet system contract (0x0000…1000). Price from CoinGecko.