LEO Inflation Analysis · October 2026 · Mixed flows · supply roughly steady

UNUS SED LEO (LEO), the exchange token of Bitfinex and its parent company iFinex, has no new coins coming in and a buyback that takes a small amount out. In the 90 days to Oct 12 2026 the LEO buyback moved 355,203 LEO out of the supply count against zero new coins, a 0.04% drop to 919.86M LEO. Those moves have paused since Sep 1 2026 while buying goes on; if they resume, the next 90 days take out about 622,884 LEO, a 0.07% drop. The size is set by Bitfinex revenue, not by any cap.

The verdict, in one paragraph

The MrNasdog Pressure Framework reads LEO at −0.04% net over the last 90 days and −0.07% projected for the next 90. The inflation monitor reads −0.13% for its own 90-day window, a gap of 0.09 percentage points, well inside the 0.5-point line, so there is no warning chip. Both say the same thing: LEO supply edges down a little every quarter, and nothing pushes it up. Call it a quiet, slowly shrinking corporate token, where the only moving part is how much money Bitfinex makes.

Sell pressure: where new LEO comes from

Protocol inflation is 0. LEO has no block rewards and no staking pay. It lives as a token on two chains, 660M LEO on Ethereum and 307.65M LEO on Vaulta, and both totals read the same at the start and the end of the window. The keys to create more still exist: Bitfinex controls the Ethereum token, and the Vaulta version has about 692M LEO of unused room under its 1B cap. That is a capacity we watch, not something that happened.

Vesting unlocks are 0. The whole LEO supply was sold in one private sale in May 2019, with no lock-up for anyone, so no team or investor schedule is left to release coins.

Foundation and unscheduled unlocks are 0. Bitfinex-linked wallets hold most LEO, but every one of them is already counted as circulating, so moving coins out of them adds nothing new. The details are in the overhang section below.

Long-term locked or bankruptcy is 0. No estate, trustee or long lock holds LEO. The 2016 Bitfinex hack matters for LEO only on the buy side, because recovered money is promised to the buyback.

Buy pressure: where new LEO goes

The programmatic buyback took 355,203 LEO. Each month Bitfinex spends at least 27% of its gross revenue from the month before buying LEO at market prices, and has said it will keep going until no LEO is left in circulation. The coins are sent, once a day, to the token's own issuer account on Vaulta, which now holds 47.80M LEO. That account is not counted as supply: the circulating figure is exactly the two chain totals minus this one balance. All 50 moves in the window show on-chain, and they match the drop in supply to the coin. One honest caveat: these coins are parked, not destroyed, and the account is a company multisig.

The last move was on Sep 1 2026. Since then Bitfinex has kept buying, about 5,800 LEO a day, but the coins wait off-chain: 236,909 LEO by Oct 12 2026. Waiting coins still count as circulating, so they are not in the number yet. Bitfinex has paused like this twice before, from Jul 17 to Sep 15 2025 and from Jan 7 to Feb 4 2026, and both times it moved the whole waiting pile in one go. So for the next 90 days we use the last 90 days of buying, about $5.55M, at today's price of $8.91: about 622,884 LEO.

The protocol fee burn is 0. LEO pays no gas on either chain, so there is no fee to destroy; any LEO Bitfinex takes as fees can feed the same buyback, and we do not count it twice.

Foundation buy is 0 for now. Bitfinex has promised at least 80% of the net money it recovers from the 2016 hack to buy back LEO within 18 months. In early October 2026 US officials moved about 264.8 BTC to an exchange under the court-ordered return, and moved a much larger batch between their own wallets, but nothing shows Bitfinex has received any coins. With no amount and no date, the row stays at zero.

New long-term lock is 0. There is no LEO staking and no new lock contract.

Foundation and overhang

LEO is unusual because one company holds almost all of it. One Ethereum contract funded by a Bitfinex wallet holds 648M LEO, about 70% of the supply, and has not moved since 2023. A Bitfinex cold account on Vaulta holds 257.79M LEO, and the account that sends coins to the issuer holds 2.06M LEO. All three count as circulating, so a sale from them would change who holds LEO, not how much LEO exists. Outside the float sits the issuer account with 47.80M LEO of bought-back coins, which Bitfinex could in theory send back out, plus the 236,909 LEO waiting pile. We read these balances on-chain at each rebuild; if any of them falls between rebuilds, the outflow enters Sell #3 at the next refresh.

How LEO compares to other exchange tokens

Most big exchange tokens follow the same idea, buy back with company money, but the mechanics differ. BNB runs a quarterly auto-burn tied to its price and blocks, plus a gas-fee burn on its own chain, so its supply cut is larger and set by a fixed recipe. LEO has no chain of its own and no such recipe: its buyback is a fixed share of revenue, so a slow quarter at Bitfinex means a smaller cut, and the coins go to an issuer account rather than a dead address.

Compared with tokens that still have unlock schedules, LEO is simpler: there is no vesting, no emission and no reward pool, so the sell side is zero by structure. The trade-off is concentration. Where other exchange tokens spread supply across many holders, about 99% of LEO sits in a handful of Bitfinex-linked wallets. Supply is steady, but the float that actually trades is small.

LEO also has a one-off buyer that no peer has: the promise tied to the 2016 hack. If Bitfinex recovers a large share of the stolen bitcoin, 80% of the net money would buy LEO over 18 months, which could dwarf the normal revenue buyback.

What to watch in the next 90 days

First, whether the buyback moves resume and the waiting pile of 236,909 LEO (as of Oct 12 2026) reaches the issuer account; past pauses ended within two months, and this one began on Sep 1 2026.

Second, the hack bitcoin: after the US moves of Oct 6 and Oct 8 2026, any sign that Bitfinex has received coins starts its 18-month window to buy back LEO with 80% of the net money.

Third, how much Bitfinex buys each day, which follows Bitfinex revenue; it ran near 5,800 LEO a day in September 2026, lower than earlier in the year.

Fourth, any move out of the issuer account or the 648M LEO contract, or any new coins created on Vaulta or Ethereum. None happened in this window.

Summary

The MrNasdog Pressure Framework reads UNUS SED LEO at −0.04% net over the last 90 days and −0.07% projected for the next 90: no new LEO at all, against a revenue-funded buyback that moved 355,203 LEO out of supply. The buyback has paused its moves since Sep 1 2026 with 236,909 LEO bought and waiting, so the next reading depends on Bitfinex resuming them. The key risks are the parked, not destroyed, 47.80M LEO in the issuer account and the heavy concentration in Bitfinex wallets; the upside is the promised buyback from any recovered hack bitcoin. Until one of those changes, LEO supply keeps shrinking very slowly.

MrNasdog Pressure Framework analysis of LEO, Metric 1 — Inflation. Data + explanation only. Not financial advice. Checked Oct 12 2026.