Ripple released 1,000,000,000 XRP from escrow at 00:00:10 UTC this morning. The five US spot XRP ETFs held roughly 906 million XRP between them at 30 June 2026 — after seven and a half months
Ripple released 1,000,000,000 XRP from escrow at 00:00:10 UTC this morning. The five US spot XRP ETFs held roughly 906 million XRP between them at 30 June 2026 — after seven and a half months of buying. In a single overnight ledger operation, Ripple unlocked more XRP than the entire American ETF complex has accumulated since it launched on 19 November 2025. That arithmetic is the reason this XRP price prediction sets a 12-month band of $0.90 on the downside and $2.35 on the upside against a spot price of $1.38, rather than the rounder, friendlier numbers circulating elsewhere. Every competing page is reprinting the same nine-day inflow streak as unambiguously bullish. Almost nobody is netting those inflows against supply, and almost nobody has opened the funds' own filings with the US Securities and Exchange Commission.
When you do open them, a second number falls out that reframes the whole debate. The three XRP ETFs that file 10-Qs disclose their exact holdings and their exact cost. Add the two that do not, and the complex carries a blended cost basis of about $1.87 per XRP — a figure this piece derives below from the filings themselves rather than borrowing. Spot at $1.38 sits 26.2% underneath it. That single line explains the dispersion nobody reports: Grayscale, holding the oldest and most expensive book, halved its position in six months, while Franklin, holding the newest and cheapest, grew its position 41% in a single quarter. $1.87 is not a technical level anyone drew on a chart. It is an accounting level — the price at which sector-wide redemption pressure structurally inverts. It is also precisely why the band below is $0.90 to $2.35 and not $0.75 to $3.00.
State the arithmetic plainly, because a forecast that does not is not a forecast. Spot is $1.38 (the 1 September daily close prints at $1.3788). The bear case of $0.90 is 34.8% below spot. The bull case of $2.35 is 70.3% above it. The base case of $1.55 is 12.3% above. The horizon is twelve months, to 1 September 2027. Bear sits below spot, spot sits below bull, and the distance is deliberately asymmetric to the upside because the downside is capped by a cost basis that a large, identifiable holder base cannot cheaply abandon.
Now the harder question: why not $3.00 and $0.75, the band this desk itself ran in August? Because both ends had stopped describing anything real. $3.00 would be a +117.6% round trip to within 17.8% of the all-time high of $3.65 set on 17 July 2025 — while the entire US spot ETF complex sits about 26% underwater on its own book. XRP has not traded at $3.00 since July 2025. A target requiring the largest institutional holders to climb out of a $746m hole and then keep buying through it is not a forecast; it is a wish. $2.35, by contrast, is a level that actually printed this year — the close on 6 January 2026 was $2.3538 — and it sits above the $1.87 accounting line where the flywheel reverses.
The bear end had the opposite problem. $0.75 is −45.6%, and it prices a failure already tested and rejected. XRP closed below $1.00 on 15 and 17 August 2026, hitting a 52-week closing low of $0.9927, then was bought back 51.9% in five sessions — from a $1.0011 close on 19 August to $1.5205 on 24 August. The market has told us what happens beneath a dollar: it does not stay there. $0.90 is the honest bear — 9.3% below the 52-week closing low, a genuine break of the cycle floor rather than a capitulation fantasy. Spot sits 38.9% above that low and 62.1% below the all-time high.
XRP daily closing prices, 12 months to 1 September 2026, with the $2.35 bull and $0.90 bear targets drawn as dashed rules. Both fall inside the year's range — the bull target is a level XRP closed at in January, the bear target sits just below August's low. Source: CoinGecko daily closes; chart by FinanceFeeds. Closing prices only. Here is the supply arithmetic that the inflow headlines omit. Ripple's programmed escrow releases 1bn XRP on the first of each month, then re-locks most of it. Reconstructing every live escrow's creation transaction on the XRP Ledger shows the re-lock has been exactly 700,000,000 XRP in each of the last eleven consecutive months, from October 2025 through August 2026. The net drip into Ripple's hands is therefore a constant 300,000,000 XRP a month — not a vague range.
Set that against ETF demand. The five funds accumulated roughly 906m XRP in the 7.4 months between launch and 30 June — an average of about 123m XRP a month. Against a 300m net monthly drip, the entire American spot ETF complex has been absorbing around 41% of Ripple's net new supply, and about 12% of the gross monthly unlock. The nine-day inflow streak is real, and it is the strongest institutional demand XRP has ever seen. It is also, structurally, not yet large enough to clear the supply arriving on the first of every month without help from somewhere else. Our earlier reporting on the XRP ETF complex passing $1.5bn tracked the demand side of this ledger; this is the other side of it.
One caveat matters more than any other here, and it is worth flagging rather than papering over: as of 11:53 UTC today, Ripple had not yet re-escrowed anything. There is not a single EscrowCreate transaction anywhere in Ripple's escrow system dated 1 September 2026. All 1bn XRP released this morning sits untouched in two Ripple accounts, each showing a single transaction since 31 August. The re-lock has historically landed between 15:00 and 19:00 UTC on the first — 17:47 UTC in August, 15:25 in July, 18:19 in June — so the window has simply not opened yet. If the eleven-month pattern holds, expect 700m back and 31.7bn locked. But that is a projection, not evidence. Until those transactions validate, the honest statement is that 1bn XRP is liquid and none of it has been returned.
A note on the escrow total itself, because the widely syndicated figure is wrong. Several outlets, including crypto.news, report 31.28bn XRP remaining, attributed to a third-party tracker. A direct audit of the ledger — paginating every escrow object across all Ripple-labelled accounts — returns 31,000,000,000 XRP exactly, in 98 objects across eight accounts. Ripple's own distribution data independently reported 32.0bn escrowed in mid-August; subtract today's 1bn and you get 31.0bn, agreeing to the token. The tracker's figure is 280m too high, and its own framing gives the game away: 31.28bn is exactly 31.28% of the original 100bn supply, which is circular.
The most durable myth in XRP commentary is that transaction fee burns make the asset meaningfully deflationary, offsetting escrow releases. This is measurable, and it is not close. Comparing the ledger's total_coins field across a clean 24-hour window this session — ledger 106,666,309 at 11:53:51 UTC on 31 August to ledger 106,688,634 at 11:55:30 UTC on 1 September, spanning 22,325 ledgers at a 3.875-second average close interval — the total supply fell by 331.02 XRP. That is a burn rate of about 330.65 XRP per day, or roughly 9,900 XRP a month.
Against a 1bn monthly gross release, the burn offsets one part in 100,813. Against the 300m net drip, one part in 30,244. Cumulatively, 14,378,208.77 XRP has been destroyed since genesis — 0.0144% of supply in fourteen years. At the current rate, burning the 1bn released this morning alone would take roughly 8,300 years. Anyone can reproduce this in two ledger calls to a public node. The deflation argument is not merely weak; it is off by five orders of magnitude, and any price model leaning on it is mis-specified. That does not make XRPL uninteresting — the ledger's actual case is throughput and settlement, which we examined when Mastercard added XRPL as one of eight supported chains, and which the validator base has been steadily extending through its recent amendment votes. But the case is utility, not scarcity.
Now the derivation, because this is the number the band is built on. Three XRP ETFs file quarterly reports with the SEC, and each discloses holdings and cost to the token and the dollar at 30 June 2026:
Those three hold 567,242,994.66 XRP at a combined cost of $996.43m — a blended basis of $1.7566 and an unrealised loss of $403.31m. Their combined fair value of $593.12m implies an XRP price of $1.0456 at the measurement date, which is the key that unlocks the other two funds. CryptoSlate reports the five-fund complex at $947.3m fair value on $1.693bn of cost; divide $947.3m by that implied $1.0456 and the complex held about 906 million XRP. Divide $1.693bn of cost by 906m and the blended basis is $1.8687 — call it $1.87.
That derivation also validates the secondary source. The three filings read directly here sum to −$403.31m. CryptoSlate's figures for the two funds that do not file 10-Qs — Canary at −$229.2m and 21Shares at −$113.5m — add −$342.70m. Together: −$746.01m, against CryptoSlate's stated −$746.1m, reconciling to within $0.1m. (For the record, the SEC shows five XRP funds with 2026 filings, not the seven that circulate in some coverage.)
The behavioural consequence is visible in the filings. Grayscale, carrying the highest cost basis and the oldest holder base, cut its position from 122,230,386 XRP at 31 December to 55,035,728 — a 55% reduction, with net asset value per share falling from $45.46 to $20.22. Franklin, the cheapest and newest book, went the other way: from 159,659,740 XRP at 31 March to 225,368,820, up 41% in one quarter. Same asset, same quarter, opposite direction — separated by entry price. That is what a cost-basis threshold looks like in practice. Below it, legacy holders redeem and new money accumulates; above it, the redemption bid disappears and the complex becomes a one-way buyer. Reclaiming $1.87 is the single most important precondition for the bull case.
The bull case does not rest on the token. It rests on whether Ripple's institutional business grows fast enough to make XRP infrastructure rather than inventory. Ripple Prime now clears more than $3 trillion annually for over 300 institutional customers, and Brad Garlinghouse said in August that he expects the company to more than double annual revenue during 2026 despite market weakness. Speaking at the SALT Wyoming Blockchain Symposium on 20 August, Ripple's chief executive framed the strategy directly:
"More and more people are realizing that the infrastructure side, the institutional side is where it's at." — Brad Garlinghouse, CEO, Ripple, via crypto.news
Ripple's stablecoin RLUSD now carries a market capitalisation of $2.26bn on 2,264,392,875 tokens, with $147.3m of 24-hour volume (CoinGecko, 11:53 UTC today) — we covered its move through $2bn last week, alongside Ripple Prime's launch of a delta one service for US equity derivatives. The uncomfortable truth for holders is that none of it requires the token to appreciate. A stablecoin business and a prime brokerage sit beside an asset whose price is set by a different mechanism entirely — escrow supply meeting ETF demand.
The regulatory picture, meanwhile, has moved from existential threat to structural plumbing, and that shift has a date attached. Evernorth Holdings — a digital asset treasury vehicle merging with SPAC Armada Acquisition Corp. II (Nasdaq: XRPN) — had its registration statement declared effective on 27 August 2026, and filed its final prospectus the same day under Registration No. 333-294417. The document states Evernorth expects to launch with corporate holdings of at least 473,276,430 XRP at closing, and sets the shareholder vote for 12:00 ET on 30 September 2026 (SEC Form 424B3). Note the scale: one vehicle would hold roughly 52% as much XRP as all five US spot ETFs combined — the most datable catalyst in the next thirty days, and materially underpriced in the commentary.
Three predictions, with the causal chain stated so they can be falsified.
First: the September re-escrow lands at 700m, and if it does not, that is the story. Eleven consecutive months at exactly 700m is a strong prior. A re-lock materially below it — or none at all — would mean Ripple intends to deploy more than its usual 300m net, the single most bearish supply signal available. Watch the 15:00–19:00 UTC window today.
Second: the Evernorth vote on 30 September is the near-term swing factor, not ETF flows. A vehicle that must source 473m XRP is roughly four months of the ETF complex's average accumulation arriving at once. Approval pressures the float upward into the $1.55 base case; a failed vote removes the only demand catalyst large enough to matter against a 300m monthly drip.
Third: $1.87 gates the bull case. XRP cannot reach $2.35 without first spending time above the ETF complex's blended cost, because below it every rally meets legacy redemption supply of the kind Grayscale's halving demonstrates. Above it, that supply is gone. The path to $2.35 runs through $1.87, and the crossing will be contested rather than clean.
Prediction markets, for what little they are worth here, lean the same way. Polymarket's "What price will XRP hit in September?" event prices a move to $1.40 at 93.5¢, $1.80 at 18¢, $2.00 at 9¢ and $3.00 at 1.5¢, against a dip to $1.20 at 42.5¢, $1.00 at 11¢ and $0.80 at 4.5¢ (Polymarket, re-queried 11:52 UTC). Treat these as indicative only: total event volume is $130.30 — a rounding error, not a market. At least one leg in the same event is quoted off a book holding barely $100 of liquidity; we excluded it rather than repeat a number that would not survive a single fill. The distribution is nonetheless coherent with the band here. For how contested the near-term range has become, see the debate over XRP's road to $2.
The twelve-month picture: a token 62.1% below its all-time high and down 50.5% on the year, with the strongest institutional demand in its history sitting 26.2% underwater, meeting a supply schedule that delivers 300m net tokens a month with metronomic regularity. Not a collapse, and not a moonshot. A $0.90-to-$2.35 range with a $1.55 centre of gravity, and a single accounting number at $1.87 deciding which half of it matters.
This analysis sets a twelve-month band to 1 September 2027 of $0.90 in the bear case, $2.35 in the bull case and $1.55 as the base case, against a spot price of $1.38. That is 34.8% downside, 70.3% upside and 12.3% upside respectively. The band is anchored on the US spot ETF complex's blended cost basis of about $1.87, not on chart patterns.
Because $2.35 is a level XRP actually closed at this year — $2.3538 on 6 January 2026 — and it sits above the $1.87 accounting level where ETF redemption pressure inverts. A $3.00 target implies a 117.6% rally to within 17.8% of the all-time high while the ETF complex is roughly 26% underwater on its own book. XRP has not traded at $3.00 since July 2025.
A direct audit of the XRP Ledger at validated ledger 106,688,634 on 1 September 2026 found exactly 31,000,000,000 XRP locked across 98 escrow objects on eight Ripple-labelled accounts, following this morning's 1bn release. Widely syndicated figures of 31.28bn come from a third-party tracker and appear to be roughly 280m too high.
Not in any economically meaningful sense. Measured on-chain over a 24-hour window on 1 September 2026, XRP's total supply fell by 331 tokens, a rate of about 331 XRP per day. Total destruction since genesis is 14,378,208.77 XRP, or 0.0144% of supply in fourteen years. The burn offsets roughly one part in 100,000 of the monthly escrow release.
Not as of 11:53 UTC on 1 September. No EscrowCreate transaction dated 1 September 2026 exists in Ripple's escrow system, and all 1bn released tokens remained untouched in two Ripple accounts. Ripple has re-locked exactly 700m XRP in each of the previous eleven months, typically between 15:00 and 19:00 UTC on the first, so the window had not yet opened at the time of writing.
On the upside, a sustained reclaim of $1.87 alongside continued ETF creations would argue for the top of the band or above. On the downside, a September re-escrow materially below 700m, a failed Evernorth vote on 30 September, or sustained redemptions from the cheaper ETF books would open the path toward $0.90.
Analysis, not investment advice. Digital assets are volatile and you may lose your capital. All price and on-chain data verified 1 September 2026.
