Stellar Votes on Bell on November 3 — but a Four-Session 0.1665 Floor Broke First, and 0.1764 Decides 0.1974 or 0.1524

Stellar validators vote on Bell, protocol 30, on 3 November at 17:00 UTC. In the 32 sessions since XLM last printed a 0.23710 high it has given back 33.5%, and today's 0.15770 low is the lowest the pair has traded since 19 August. The vote is on the calendar. The chart is not waiting for it.

What the validators are actually voting on

The Stellar Development Foundation introduced Bell on 20 October. It bundles three Core Advancement Proposals, and none of them goes live on a foundation decree: node operators enter a protocol version and a timestamp, and the network raises its version once enough voting weight sits behind the change. The testnet vote is 3 November at 17:00 UTC, the mainnet vote 23 November at 17:00 UTC, and the Foundation notes that both dates can move.

The code has been out for two weeks. Stellar Core v30.0.0 shipped as a stable release on 16 October, with the infrastructure and library versions following through 23 October. The upgrade guide is unambiguous about one thing: for anything that touches addresses, the SDK update this cycle is not optional.

CAP-84 is the one that pays, and CAP-87 is the one that got oversold

CAP-84 gives smart contracts multiplexed addresses. A contract address picks up a 64-bit identifier, SC_ADDRESS_TYPE_MUXED_CONTRACT, carried by the new Strkey prefix W — described by the Foundation as an upside-down M. The identifier lands in the payment's transfer event, so a payment provider can settle thousands of customers through one contract address instead of deploying one contract per customer. This is the change with a business case attached, because payments are where Stellar actually earns.

CAP-87 builds ML-DSA signature verification into Stellar Core rather than into contract code, so a contract can check a post-quantum signature without the compute bill that a contract-level implementation would carry. ML-DSA is one of the schemes NIST standardised in 2024. The limit matters: transaction signatures on the network stay Ed25519, and a native quantum-safe signer for classic accounts is named as work for a later protocol version. Bell does not make an ordinary Stellar account quantum-safe.

CAP-88 moves the ledger to a millisecond clock. Ledger close times do not get shorter — the five-second target stays — but the one-ledger-per-second floor that blocked any later shortening disappears. It also changes the ledger header format, so anyone decoding raw ledger data, whether a custom indexer, an explorer or an analysis pipeline, has to lift the XDR definitions to protocol 30. Decoding otherwise fails on unknown fields, while the closeTime value itself keeps reporting whole seconds.

The upgrade already cost one venue a deposit channel

Two days before the announcement, bitFlyer told customers on X that inbound XLM had stopped crediting to accounts at around 23:49 JST on 19 October, and named access restrictions at the Stellar Development Foundation's official Stellar interface, which the exchange relies on, as the cause. Sending XLM was unaffected, deposits and withdrawals of every other asset were unaffected, and no resumption date was given. Nothing was broken on the chain. The bookkeeping layer in between was.

What XLM did while the vote was being scheduled

The window high is 0.23710 from 29 September. The low is today's 0.15770, printed 32 sessions later: 33.5% down, and the lowest low since 19 August's 0.15380. The August floor, the lowest of the past three months, sits at 0.15240, with the lowest close of that stretch at 0.15490.

The break had a clean shape. From 24 to 27 October the intraday low never left a 0.0005 band — 0.16630, 0.16680, 0.16660, 0.16650 — and the 28 October session held 0.16870. On 29 October the floor gave way to 0.16040 on 68.0 million XLM, 1.27 times the 53.5 million average of the twenty sessions before it, and the session closed 5.9% lower at 0.16230. Today printed 0.15770 and then closed 4.1% above it at 0.16420, on 46.5 million XLM, 13% under the average participation of that same window.

Around the chart, the market data reads $5.79 billion of market capitalisation and rank 19, with 35.25 billion XLM circulating against a 50.0 billion total supply. The token is down 3.3% on the day, 6.9% on the week, 27.8% on the month and 47.4% on the year, and it trades 81% under the January 2018 high of $0.8756. The fourteen-day average true range is 0.00755, or 4.6% of the price, so a session of that size is now routine rather than exceptional.

0.1764, 0.1665 and 0.1524 — and what is left on the calendar

0.17644 is the 0.236 retrace of the whole 0.23710-to-0.15770 leg, and it lands within four ticks of the 15 October low at 0.17610. A bounce has to clear that before anything else in the structure matters.

0.16650 is the four-session floor that broke on 29 October, and it is resistance now. The descending ceiling drawn from 0.22710 on 5 October through 0.17400 on 29 October runs at 0.17180 today, so the 0.1665 to 0.1718 band is the supply shelf a recovery has to eat. Below the market, 0.15770 is today's low and 0.15240 is the August floor.

Two scenarios carry the weight. Reclaim 0.1665 and hold it, and the path is 0.1764, then 0.18803 at the 0.382 and 0.19740 at the half of the entire downleg; a daily close above 0.18803 would put the November vote in front of a repaired chart instead of a broken one. Lose 0.1577 on a closing basis, and 0.1524 is the next stop, with the 0.17520-to-0.15770 leg repeating to 0.1402 and the 1.272 extension of the full decline waiting at 0.1361 underneath. The middle case, chop between 0.1577 and 0.1764 while the vote is pending, is the one the current volume supports.

The rest of the schedule writes itself: 3 November testnet, 23 November mainnet, both 17:00 UTC, with wallet firmware, exchange SDKs and address validators due for a W prefix in between. The Foundation's own count puts the tokenised-asset ecosystem it tracks at $4 billion, and CAP-84 is the piece that lowers the cost of settling it.

One warning belongs with every upgrade. Bell involves no token swap, no registration and no wallet migration: existing addresses stay valid and no balance moves. Any message asking for a recovery phrase because of protocol 30 is an attack.