XRP Ledger Weekly Data Report — Week of September 21–27, 2026

How to read this report

Every metric below carries its source and its snapshot date, and we do not publish figures we cannot trace to a public source. Where two methodologies disagree — circulating supply is the standing case, and RLUSD's total supply joined it this week — we show both and say why. Where a metric's freshest verified reading is older than this week, we date it honestly rather than extrapolate. This week that policy earned its keep in a specific way: the supply-and-custody section has been carried on a July 26 snapshot for two months, and this edition replaces most of it with a September 26 on-chain read while naming exactly which lines are estimates rather than counts.

Three mechanical rules follow from the policy, and they explain most of the oddities a careful reader will notice. First, we do not average across methodologies: where a tracker and an issuer print different values for the same quantity, both are published with attribution rather than blended into a false precision. Second, we do not refresh a metric for appearance's sake — a reading that could not be re-verified stays dated. Third, we label estimates as estimates: the lost-XRP line below is an estimate with a stated basis, and the derived lines are marked calculated with their inputs shown.

So that the reader can see exactly how much of this report is this week's and how much is carried, here is the verification log for the edition.

1. Market metrics

The tape story of the week is a short squeeze with a supply shock folded into it. Monday's 8.7% session began with roughly $300 million of short positions closed across the crypto market inside an hour, and XRP — among the most heavily shorted large caps after the prior week's slide — captured more of it than any peer (24/7 Wall St.; Yahoo Finance). Tuesday took the token above $1.60 for the first time since February 4 and Wednesday printed the high at $1.6536, an eight-month peak and +17.3% on the week at the extreme (calculated); the same session reversed to $1.4821 and closed at $1.5006. The year's frame has moved with it: at $1.510 the token is up roughly 46% over three months while still down approximately 17% year-to-date and about 44% over twelve months (24/7 Wall St., Sep 27; Yahoo Finance, Sep 24).

Two source-cut caveats belong next to the price, and both matter more this week than usual. The first is methodological: this report quotes a Sunday-to-Sunday close pair of $1.4099 → $1.5098 for +7.1%, while several trackers printed +8.5% to +9.8% for the same seven days (Bitget's 7-day reading of +8.52%; coinlib's +9.78%; ForkLog's roughly +11% measured Monday-open to Friday-close). All of those are true at their own cut times; none of them is the other. The second is the distinction between the week's peak and its close: the market traded $1.6536 on Wednesday and settled $1.5098 on Sunday, so a reader who quotes "XRP gained 7%" and a reader who quotes "XRP gained 17%" are describing the same week with different endpoints. We name ours in every line for exactly this reason.

The market-wide frame is what makes the move legible. Bitcoin began the week near $81,000, reached a local high of $87,378 on Monday — the day spot BTC ETFs took in nearly $1 billion — then retreated to about $84,000 and finished the week up roughly 4.2–5% with dominance at 58.3–58.8% (ForkLog; CryptoTakeProfit; The CryptoBase). Ether added about 2.5%, Solana led the majors at roughly 11%, and Zcash gained about 14% to a new all-time high above $1,600 — up roughly 2,500% for the year and now the seventh-largest asset. Total crypto capitalization rose from $2.75 trillion to about $2.90 trillion, the Fear & Greed index held between 70 and 74 ("Greed") all week, and Glassnode's Altcoin Cycle Signal flipped to "altseason" (ForkLog; CryptoTakeProfit; The CryptoBase; Crypto Trends, Sep 26–27). XRP's +7.1% therefore beat Bitcoin and Ether and trailed Solana and Zcash — a reversal of the prior week's pattern, when it lagged every major it trades against.

2. Supply, custody & escrow

Why two circulating-supply numbers: market-data aggregators apply their own methodology — chiefly Ripple's distribution API estimate, which nets out certain classified holdings — to the ledger's raw activated supply. The 62.88B figure is what the market cap is computed against; the ledger basis, now derivable at ~67.5B from the September 26 map, is what the chain itself accounts for. Both are real and they measure different things, and this edition can show the arithmetic for the second one for the first time.

The substantive change this week is that the float became legible. A refreshed on-chain read dated September 26 distributes the full 100 billion across holder categories — and the distribution is the point, because it converts a series of separate anecdotes into one picture.

Source: XRP Radar on-chain read, Sep 26, 2026, compiled from issuer addresses, exchange wallet attribution and DeFi protocol contracts. Categories are not exhaustive of all 100B — lost coins, unlabelled wallets and the burn are itemised separately in metrics 9–10.

Three things that table says that no single headline does. First, exchanges hold thirty times what ETFs do (20.32B against 1.18B) — so the marginal source of price discovery in XRP remains venue order books rather than vehicle creation, which is the opposite of the arrangement in Bitcoin and Ethereum. Second, the ETF custody line is now roughly 1.7–1.8% of supply, and it is the only category in the table with mechanical creation and redemption: ETFs shrink when investors redeem, while exchange balances and Ripple's escrow do not. Third, corporate treasury holdings are small in absolute terms but real and verifiable — 473.3M XRP, single-company concentration, publicly attributed. The 20.32B exchange figure also gives the week's security incident its proper denominator, and we use it in section 5.

Escrow is the line that moved. 31.98B XRP in escrow on September 26, against 32.445B on July 26 — a decline of roughly 465 million across two monthly release windows, which is exactly what the mechanism predicts: up to 1B unlocks on the first of each month and the large majority is re-locked inside the same cycle, so the balance drifts down slowly while the monthly headlines swing by a billion. The next window opens October 1, with up to 1B XRP (~$1.5B at current prices) released and most of it expected back into new contracts. The honest reading of an unlock story is that roughly 200–400M XRP per month reaches the open market, and most of that sits in Ripple's operating inventory rather than in a sell book — an estimate we flag as an estimate, since the split is not publicly itemised.

3. Network activity

No fresh weekly network series surfaced this week, so the transaction line carries its September compilation date alongside the single dated XRPScan reading, and nothing in this report leans on a refreshed count. The structural read is unchanged: throughput above 2M transactions a day across 40% fewer active accounts than a year ago means activity per account has roughly doubled — recurring, settlement-shaped flow rather than retail churn. The September 18 print adds texture in the same direction: payments of about 757,400 moved roughly 1.16 billion XRP, more than double the 30-day average, while active accounts rose 21.5% to about 18,700 and total active users fell 17.3% to about 155,300. More participants transacting with fewer distinct users, on much larger tickets, is the settlement-rail signature rather than the retail-exchange one.

The DEX line is the week's best example of why methodology disclosure matters. DefiLlama puts XRP-paired DEX volume at $55.1 million over seven days. A separate provider, XRPL.to, reported 2.82 billion units of seven-day AMM volume — a number forty times larger — of which 97.24% came from two token-token pools, XPM/TIX and RLUSD/TIX, that between them logged 185 trades, were both created on September 21, share the same issuer and pool creator, and contain no native XRP at all (CryptoSlate via CryptoVideos, Sep 25). The two figures are not in conflict; they measure different things, and the second one measures almost nothing. A volume series that a single pair of freshly created pools can move by 97% is a volume series that has been gamed, and the only defence is to keep quoting the pools that actually hold XRP.

One more dated ledger fact is worth recording because it illustrates the same measurement problem at the transaction layer. On September 13, ledger 106,965,249 closed with 3,254 transactions — a single-ledger record — but 2,000 of them were identical 1-drop payments fired by just 20 accounts, moving 0.002 XRP in total while burning about 0.04 XRP in fees; at least 890 further transactions in the window failed with error codes and still consumed fees (Bitcoin.com News via CryptoVideos). Across 7,600 consecutive ledgers, the average was about 84 transactions per ledger. A record can be a record and still be noise, and the ledger's own dynamic ledger-size mechanism kept fees at the 10-drop minimum through it.

4. RLUSD stablecoin

The chain split remains the most consequential number in this section, and this week it widened in the wrong direction. Ethereum holds $290 million more RLUSD than the XRP Ledger does, and the ledger's share of the token's supply slipped from roughly 43% to 44% only because the total grew — the absolute gap is what matters and it is not closing. The rest of the ledger's stablecoin book continues to thin around RLUSD: USDC fell 21.86% over the month to $4.62 million, leaving Circle's token with 0.4% of the market, while Valtorum USD and EUR CoinVertible also declined. RLUSD's 92.1% share is best read as a market with one functioning participant rather than as a competitive win.

The 30-day growth rate is the counterweight to the split. RLUSD's ledger-side value grew 11.55% in 30 days, faster than the ledger stablecoin market's 9.8%, and the token's total supply is up about 6% in nine days. Ripple's positioning language has moved with the number: the company now describes RLUSD as a settlement rail for payments, tokenized finance and lending rather than as a trading stablecoin, consistent with the corporate-treasury tooling pitch and with the FXRP-collateralised borrowing market that went live on Morpho in the prior week. Two new distribution channels opened this week — Binance began paying $800,000 in XRP to RLUSD holders across a four-week campaign running to October 23, and Ripple's Middle East and Africa managing director Reece Merrick used a conference panel on September 24 to describe stablecoins as an always-on layer and RLUSD as a product built "not to replace bank money, but to let it travel" (Pluang; crypto.news). Neither is a flow; both are attempts to create one.

The number that would confirm the settlement-rail claim is the XRPL share, and it is $1.10B against Ethereum's $1.39B. The mechanism most likely to invert that ratio is collateral demand from XRP holders rather than liquidity demand from traders — and that is precisely the mechanism whose usage data has still not been published, which is why the coverage note below carries it for a second consecutive edition.

5. ETF flows & derivatives

The ETF week reversed the prior one almost exactly. Last week produced +$9.57M on a one-issuer baseline; this week produced +$75.6M across four consecutive green sessions after a zero-flow Monday, and the streak now stands at eleven consecutive positive weeks (SoSoValue via Finobird, BeInCrypto, U.Today, Blockonomi). Bitwise led every session and reached about $677 million cumulative; Franklin Templeton's fund added $4.26M on Friday to reach roughly $501 million and passed Canary as the second-largest complex. The honest caveat is composition: Bitwise took about 81% of Friday's $22.65M, so the complex is still effectively one issuer with a credible second — better than last week's one issuer, not yet a market.

The gap between cumulative inflows and net assets is the most under-reported number here. At $1.79B contributed against $1.77B held, the complex's average dollar is slightly below water even though XRP is up 46% over three months — because most of the capital arrived at higher prices during the August record month. This converts every future flow print into an ambiguous signal: in a complex where the average holder is down, outflows are redemption pressure and inflows are averaging-down, and neither behaves like the allocation bid the coverage assumes. For scale and pace, this week's total also has to be set against September's month-to-date figure of roughly $117.5M ($41.9M through Sep 18 plus $75.6M through Sep 25, calculated from sourced weekly totals), still below August's record ≈$159.2M. And the hesitation was not XRP-specific: spot Bitcoin ETFs took in about $2.39 billion for the week, their biggest weekly haul since October 2025, with $108.4B in total assets, while Ether funds added $689.9M and returned to about $18B in assets (ForkLog; CryptoTakeProfit; The CryptoBase). XRP's complex re-engaged in the same week the whole market did.

Derivatives supplied the mechanics. Open interest climbed to $4.1 billion on September 23 — a nine-month high — and moved XRP above both Hyperliquid and Zcash in futures depth, with CME open interest up more than 16% in a day to $756 million and Binance futures OI up 10% to almost $700M; futures volume hit $9.4B against $2.2B of spot, and the session's short closures totalled $9.50 million, including a single $5.09 million XRPUSDT liquidation on Binance (CoinGlass via Blockonomi). The rebuild is what turned a $300 million market-wide short unwind into an 11.6% intraday band on Wednesday and a 6.9% daily reversal off the high. Two readings frame the risk: open interest at a nine-month record means a leverage stack now exists on both sides, and Coinbase Markets' options data put XRP's implied one-standard-deviation move for the week at ±8.9% — the widest among the four majors, and about 1.79 times its own historical median. A market pricing the largest expected move in its peer group is a market that has told you the next surprise will be large.

Scale check, for calibration. The 102,926,478 XRP taken in the September 24 Bitget breach is worth $157.5 million, the largest single asset in a haul provisionally valued at $387.5 million and about 44.1% of the itemised total (Lookonchain via Crypto News, COINOTAG, Coinpedia). Set against the custody map above, that is 0.10% of the 100 billion supply and 0.51% of the 20.32 billion held on tracked exchanges — which is why the token closed higher on the day of disclosure. The exposure is not arithmetical; it is that an unknown holder now controls a hoard whose disposition is unknown.

6. The week in one table

A squeeze week is best read as a before-and-after, not as a snapshot. The table below puts the prior-week reading beside this week's for the six metrics that carried the most information.

The interesting row is the open interest one, and it is interesting because it moved against the pattern of the prior week in both direction and magnitude. Two weeks ago the book thinned into the events and stayed thin — the site documented $3.06B falling to $2.86B — and the market's response to a bearish resolution was muted because there was no leverage left to cascade. This week the book rebuilt 43% in six days, and the market's response to bullish positioning was an 11.6% single-session band and a 6.9% reversal. The same asset, the same price level, two completely different tapes, and the variable that changed was leverage rather than news. That is the most transferable finding in this report.

Sentiment & the week ahead

The Fear & Greed Index held in "Greed" all week, printing 74 on September 25, 70 on September 26 and 70 on September 27 against readings of 70–71 at the prior week's close (Alternative.me via Crypto Trends; The CryptoBase). Unlike the prior edition's full round trip from greed to neutral and back, this week's sentiment was range-bound even through Wednesday's 6.9% reversal — mildly surprising, and consistent with a market that reads the week as a recovery rather than a new high.

The policy calendar replaced the legislative one. The SEC's Division of Corporation Finance published FAQs on September 25 holding that services to secure, maintain, improve or enhance a functional crypto network — and funding development that builds network effects — do not by themselves constitute "essential managerial efforts" under the Howey test; that an announced buyback of a non-security token on a functional system generally is not a promise of managerial efforts, though the analysis changes where the system is not functional and the buyback is marketed as yield; and that a staking receipt token evidencing ownership of a deposited digital commodity is a digital tool, or a digital commodity when issued by a protocol-based liquid-staking provider, provided the issuer cannot transfer, lend, pledge, rehypothecate or otherwise use the deposited assets (Crypto Daily; CoinAlertNews; Blockonomi; CryptoRank). The document builds on the March 17 interpretive release and the August 18 Regulation Crypto Assets proposal, and staff were explicit that it carries "no legal force or effect."

Read with the prior week's three rules, the pattern is now a sequence rather than an event: the SEC and CFTC are filling the framework Congress declined to build, in instruments that each expire. For XRP the practical position is unchanged and worth restating precisely — the commodity treatment rests on the March 2026 joint interpretation, not on a statute, and therefore carries revision risk rather than near-term reclassification risk. The new FAQs do not classify XRP; what they retire is the specific argument that ongoing development keeps a functional network's token inside the investment-contract definition.

The week ahead runs on three protocol dates. September 29: BatchV1_1's projected activation, conditional on validator support holding through its countdown; the amendment bundles up to eight linked transactions into one atomic operation with delivery-versus-payment as the headline commercial case. October 1: the monthly escrow release, up to 1B XRP (~$1.5B), most of it expected to be re-escrowed. October 5 at 11:18 UTC: PermissionDelegationV1_1's earliest activation, after entering its 14-day window on September 21 with 29 of 35 validator votes — and conditional on support staying above 80% for the entire period, with any dip resetting the clock to zero (CoinAlertNews; U.Today; Banx Network; Finance Review Daily). Beyond those, Ripple's Swell conference runs October 27–29.

The macro calendar is quiet by comparison, which is itself a change: no Fed meeting until late October, a Bitcoin complex that just posted its best ETF week since October 2025, and a market-wide capitalization that added about $150 billion on the week. The structural question the data poses is the one the custody map made explicit — exchanges hold thirty times what ETFs do, so XRP's price will keep being set by venue order books rather than by mechanical vehicle flows, and the 20.32 billion on exchanges is both the market's liquidity and its largest single point of concentration risk.

FAQ

Related reading

• XRP & Ripple Weekly News Review — Week of September 21–27, 2026
• XRP Market Intelligence — Week of September 21–27, 2026
• The Largest Line: Reading the Bitget Breach Through Its 102.9 Million XRP