The 63 Billion Illusion: What XRP's Circulating Supply Never Told You

Code over hype. But numbers decay more quietly than code.

On October 8, I opened four screens to answer one question a reader had sent me at 2 a.m.: how much XRP is actually for sale right now? CoinGecko said 63.09 billion circulating. XRP Insights, using a broader definition, said 68.59 billion. Ripple's own disclosure listed 37.656 billion total holdings, with 32.6 billion still locked in protocol escrow. And buried in the same snapshot was a detail that made me put down my coffee: 140.4 million cbXRP — a wrapped, XRP-backed token — appeared simultaneously as an exchange balance and as a DeFi reserve.

The same coins. Two ledgers. One number that no longer reconciles. For an asset this large, that should be impossible. It is not, and the reason is structural.

I have spent twenty-two years in this industry watching it teach itself the same wrong lesson, and this is the one that keeps returning: we mistake a size metric for a liquidity metric. The distance between those two ideas is where retail investors bleed in a bear market. Hold the line, because this is not a price story and it is not a project announcement. It is a data literacy story, and it matters more than any upgrade.

Context: The ledger that hides behind its own transparency

XRP Ledger has run for over a decade. This is not experimental software. Its escrow mechanism is native protocol code — 31.4 billion XRP locked under conditional release, immovable until those conditions are met. I want to be precise about why that single fact carries more weight than everything else in this discussion: protocol escrow is trustless. "The team promised not to sell" is not. One is code. The other is a sentence, and sentences change with sentiment.

Around that one verifiable fact sits a fog of definitions, and the fog is the story. CoinGecko excludes escrowed supply from its circulating count. XRP Insights uses a wider "non-custodial supply" figure that lands 5.5 billion higher. Exchange wallets hold 2.197 billion — but that number bundles cold storage and customer assets, so it is not a sellable figure. Founder holdings of roughly 4.6 billion come from XRP Radar, a third-party estimate rather than a confirmation. Roughly 531 million is presumed lost or dormant. The disagreement is not a scandal; it is a definitional gap. But a definitional gap in a top-five asset is a structural vulnerability, because most buyers never learn that the definitions exist.

None of these numbers can be added together. None can be cleanly subtracted from the others. And yet that is exactly what a buyer does — mentally, in their head, every time they size a position. The aggregate feels like knowledge. It is closer to noise.

During the 2020 DeFi summer I helped build a set of ethical lending guides with the MakerDAO community, and the hardest thing to teach 2,000 users was not collateral mechanics. It was that radical transparency is not the same as radical accuracy. You can disclose everything and still communicate nothing useful. XRP's supply disclosures sit in exactly that trap. That lesson is why I stopped reading supply dashboards as answers and started reading them as questions.

Core: Where the supply math breaks

Let me walk through the actual structure, layer by layer, because the layers are the whole point — and because each layer answers to a different authority.

Layer one: protocol escrow. The cleanest number on the asset is also the least agreed upon. XRP Insights puts escrow at 31.4 billion; Ripple's own disclosure rounds it to 32.6 billion. Even the most verifiable figure disagrees between sources, which tells you how soft the rest of the stack is. The mechanism itself is elegant: monthly releases, with unused portions rolling back into escrow. Historically, most released XRP returns to escrow, which means the net supply pressure from this channel is routinely overstated by people who read only the release headlines and never the rollback. I have seen this misread drive entire bearish theses.

Layer two: exchange attribution. 2.197 billion XRP sits in exchange-affiliated wallets. Investors look at this figure and see sell pressure. They should not. Cold storage lives inside that number. Customer assets live inside that number. The portion actually resting on order books is a small fraction of the total. When exchange balances fell 44 million XRP over seven days, that could be net outflow — a holding signal — or it could be customers withdrawing to self-custody, which is not bullish at all. The direction is genuinely ambiguous, and the original data critique wisely refused to spin it. I respect that refusal more than any bullish headline, because in a bear market the honest answer is often "we do not know yet."

Layer three: the wrapped-token trap. This is the technical heart of the problem, and it is where the accounting truly breaks. 477.4 million XRP sits in wrapped-token reserves, and 140.4 million of that backs cbXRP. Here is what makes it dangerous: the same XRP that backs a wrapped token on a secondary chain is also counted as an exchange asset on the primary ledger. One coin, two ledgers, counted twice. When a native asset gets re-financialized across bridges, the single-ledger verifiability that makes blockchain accounting trustworthy simply dissolves. The moment a native asset is wrapped, its supply stops being a fact and becomes an interpretation. This is not unique to XRP, but XRP is a clean case study because the double-count is visible in public dashboards. What nobody has measured is the shadow supply this creates — the same coins pledging themselves on multiple chains at once, each chain convinced it holds the real thing.

Layer four: ETF custody. Bitwise holds 420 million XRP. Franklin holds 296 million. That is 716 million coins locked inside two funds, and the structure matters as much as the size. Retail cannot redeem shares directly. Only authorized participants can create or destroy baskets by delivering XRP or cash. That pulls coins out of free float and inserts an institutional middle layer into the holder graph — a layer that did not exist in 2020. The lock is real, but it is a lock on float, not a reduction of supply. It is worth noting, too, that XRP Insights excludes unlabeled US spot ETF custody wallets from its non-custodial figure, so the data providers themselves cannot fully agree on where institutional holdings live.

Layer five: the reconciliation problem. Now put it together, the way a real buyer does. Start with 63.09 billion. Subtract escrow. Subtract founder holdings. Subtract ETF custody. Subtract dormant coins. The result feels precise and is almost certainly wrong, because several of those categories overlap. CoinGecko's circulating figure already excludes escrow, so subtracting escrow again is a double subtraction. The cbXRP backing already appears in both exchange balances and DeFi reserves, so adding those two categories is a double count. You cannot subtract overlapping categories and get a meaningful remainder. You can only get a confident-looking number that means nothing. This is the single most important discipline in reading XRP supply data, and it is the discipline almost nobody practices.

What the effective float actually looks like. Strip away the layers that cannot sell at market and a different picture emerges. Escrow is locked by code. ETF holdings are locked by fund structure. Cold storage is not on the book. Dormant coins are gone. What remains — the coins genuinely available to absorb a large order — is a number no dashboard publishes, and it is materially smaller than 63.09 billion. I am not arguing XRP is scarce. I am arguing that the number investors quote and the liquidity they can actually access are two different quantities wearing the same units. When a size metric is mistaken for a liquidity metric, position sizing fails before price does. In a bear market, that mistake is the difference between a managed drawdown and a forced exit.

The depth reality. Here is the number that should unsettle anyone sizing a trade. As a top-five market-cap asset, XRP shows a +2% order-book depth of roughly $2.4 million on Binance's XRP/USDT pair and $4 million on Coinbase's XRP/USD pair. Combined, about $6.4 million.

Six million dollars of depth behind a 63-billion-token float.

Nominal market cap is not the ability to absorb size. A large market order here does not move through a deep pool. It falls through a thin one. I have watched readers assume a multi-billion-dollar asset trades like a multi-billion-dollar asset. It does not. Depth is local, it is pair-specific, and it disappears the moment you need it most. There is a further subtlety worth preserving: order-book depth records orders within a price band, and those orders move in real time. Every snapshot is instantaneous and perishable. A depth figure is not a property of the asset. It is a photograph of a moment that has already passed. And note that the $6.4 million figure covers only two centralized venues — it ignores over-the-counter desks, which is where large trades actually happen. The real market is more liquid than the dashboards show, and less liquid than the market cap implies.

A data quality flag. One more thing, because I am constitutionally unable to ignore it. The snapshot is dated October 8, yet it cites disclosure documents timestamped January and June 2026. Either the article was written later than its snapshot date, or the data was misassembled. I flag this not to dismiss the analysis — the core insight holds regardless — but because a piece about data literacy must itself survive a data-literacy check. Truth decays slowly, but timestamps decay faster.

Contrarian: The numbers people trust are the ones nobody audited

Here is the counter-intuitive part, and I want to state it plainly.

The XRP number with the highest trustworthiness is the one people discuss least: the protocol escrow. It is native code, conditionally locked, publicly verifiable, and tamper-resistant. The numbers people build entire theses around — circulating supply, founder holdings, exchange balances — are estimates, aggregations, or definitions still in flux.

The industry has this exactly backwards. We treat the CoinGecko figure as gospel because it sits on a clean dashboard, and we treat the escrow schedule as background noise because reading it requires effort. Truth hides in the least glamorous places, and the escrow ledger is one of them.

There is a deeper asymmetry the original analysis never fully resolved. Regulatory transparency for XRP has actually improved. Spot ETFs exist, which is a de facto acknowledgment that the asset is not an illegal security, and Ripple now discloses quarterly. But market-data transparency has not improved alongside it. You can have a legally legible asset sitting on top of a statistically illegible float. That mismatch is its own risk category, and no compliance framework is designed to catch it. Regulation asks whether an asset is legal. It never asks whether its supply number is coherent.

I will admit something I rarely put in writing. During the FTX collapse in 2022, I retreated for six months and audited decentralized identity code because I had lost faith that any intermediary's numbers meant what they said. I came back believing that sovereignty is technical, not rhetorical. The XRP supply question is the same lesson wearing a different face: if you cannot verify the number yourself, you are not holding information. You are holding someone else's summary of information, and summaries drift.

Takeaway

The 63 billion figure is not a lie. It is a size metric doing a liquidity metric's job, and it is failing quietly, every day, in every position size.

Build anyway — but build your own reconciliation. Track escrow releases and rollbacks. Separate cold storage from order books. Remember that wrapped supply lives on more than one ledger at once, and that ETF custody is a lock, not a float. In a bear market, the reward does not go to the person holding the biggest number. It goes to the person who knows which numbers can be added together and which ones never could.

So ask the only question that survives a snapshot: what is actually for sale, and who told you so?