Evernorth's XRP Treasury: A Line-by-Line Audit of a $300 Million Claim

Check the ledger, not the press release.

On October 1, 2025, Evernorth Holdings announced that shareholders had approved its merger with Armada Acquisition Corp. II. The headline was clean: approximately $300 million in gross cash proceeds, a Nasdaq listing under the ticker XRPN, and the title of the first publicly traded XRP treasury company in the United States. It read like a milestone. It was written like a milestone. Every aggregator ran the same three numbers — $300 million, XRP, Nasdaq — and moved on.

The number nobody bolded: roughly $88.5 million of that $300 million is identifiable as genuinely new, deployable, post-close gross capital.

I don't read press releases for narrative. I read them for arithmetic. Sixteen years in this market — a stretch of them spent manually auditing ERC-20 contracts before anyone cared about smart contract security — taught me that the only thing that survives contact with reality is the ledger. Announcements are marketing. Filings are evidence. And the Evernorth filing, decomposed line by line, contains a gap between the announced number and the signed number large enough to change how you position.

This is a capital structure audit, not a takedown. But the structure, traced properly, says something the current market is getting wrong.

The instrument you are actually buying

Start with what a Digital Asset Treasury actually is, because the category has been sloppily defined and the sloppiness is where retail money gets harvested.

A DAT is a public equity wrapper around a concentrated crypto position. MicroStrategy is the archetype for Bitcoin. The mechanics are simple on the surface: a listed company holds a single digital asset as its primary reserve, and the equity trades as a leveraged, occasionally premium-priced proxy for that asset. Buy the stock, get the beta. That is the pitch. It is also the only thing the pitch ever is, because a pure treasury vehicle generates no operating cash flow, no protocol revenue, no product. Its entire net asset value is a function of one price chart.

Evernorth is that structure applied to XRP. The shell was Armada Acquisition Corp. II, a special purpose acquisition company — a blank-check vehicle that raises capital in an IPO, parks it in a trust account, and then merges with a target to take it public without the friction of a traditional IPO roadshow. The merger converts the blank check into an operating entity. The ticker changes to XRPN. The trust cash gets released, minus whatever the redeeming shareholders take with them.

That last clause is the whole game. Hold it.

The structure Evernorth assembled is not a single financing. It is a stack: a private placement, a convertible note, trust proceeds, and a physical — meaning in-kind, meaning actual XRP tokens, not dollars — contribution from an entity called RippleWorks. Each layer has different conditions, different timing, and different degrees of certainty. Some of it is cash already spent. Some of it is cash that only arrives if specific boxes get ticked. Some of it is not cash at all but tokens contributed by an affiliate who retains the right to pull them back.

When you blend four different asset types with four different certainty levels into a single "approximately $300 million" figure, you are not communicating a number. You are communicating a mood. And the mood is designed to feel like buying power.

Here is the first thing worth internalizing about this trade: the market is treating Evernorth as a $300 million buyer of XRP. It is not. It is, at the close, something closer to an $88.5 million buyer with a large pre-existing position and a debt instrument attached. Those are two different assets. They deserve two different prices.

A timeline that does not close

Before the money, a warning about the text itself, because a trader who ignores inconsistent timestamps deserves what happens next.

The Evernorth disclosures contain a temporal contradiction that I could not resolve cleanly. The same document set references an October 1, 2025 announcement and a September 30, 2025 shareholder vote, a November 4, 2025 disclosure of purchases, and — separately — an Armada balance sheet dated June 30, 2026. Those dates do not coexist in a single coherent present tense. Either the material was assembled retrospectively from a 2026 vantage point, or the dates are mislabeled, or what I am reading is a scenario document rather than a filed record.

I flag this at medium confidence, not to discredit the analysis but to bracket it. Every forward judgment below inherits that uncertainty. If you are trading off this, verify the primary source timestamps yourself. I watch the blockchain, not the ticker — and I do not trade a chart whose x-axis I cannot trust.

What I can analyze is the structure, and the structure is internally consistent even when the calendar is not.

The money, decomposed

The headline number is "approximately $300 million in gross cash." The precise sum in the proxy materials is $303 million. That three-million-dollar gap is not the story. The story is what the $303 million is made of.

Layer one: a prepayment of approximately $214.05 million, part of a broader $224.55 million affiliate private placement. This is the largest single number in the stack, and it is also the least useful to a forward-looking buyer, because it has already been spent. The filing states that $214 million was deployed in November 2025 to purchase XRP. That cash is gone. It converted into tokens at an average price of $2.54. It will not buy a second token. A converted dollar is a consumed dollar.

Layer two: a deferred subscription of approximately $10.5 million — money committed but not yet funded, typically contingent on closing conditions.

Layer three: a convertible note of $30 million, structured as payment-in-kind, maturing in 2031, carrying a 4% rate paid in additional principal rather than cash.

Layer four: trust proceeds of roughly $48 million, the residual of the SPAC's IPO trust after redemptions are settled. This number is the most fragile in the entire stack, because redemptions are not known until they happen, and SPAC shareholders redeem precisely when they dislike a deal.

Add the genuinely new, post-close, pre-fee layers — the deferred $10.5 million, the $30 million note, and the $48 million of trust proceeds — and you get approximately $88.5 million. That is the new deployable gross capital. Everything else in the "$300 million" framing is either already converted, in-kind, or conditional.

I have done this exercise before. In 2017 I audited three ICO contracts manually and found a reentrancy flaw in one — a flaw that killed the sale before it opened. The lesson from that year was not about Solidity. It was about the difference between what a project says it raised and what its contract can actually spend. Same discipline here. Same gap. Different decade.

The $214 million that is already gone

The single most important mechanical fact about Evernorth is this: the largest block of its capital has already been converted into XRP and cannot be reused.

This sounds obvious. It is not obvious to the market, because the market is aggregating the $214 million prepayment into the same "buying power" bucket as the fresh $88.5 million. That is a category error with real consequences.

Think about what a treasury vehicle is supposed to do. It raises capital, then deploys it to accumulate the asset. The accumulation is the value-add. The press narrative for Evernorth implies a company sitting on roughly $300 million of dry powder, ready to absorb XRP supply. If that were true, it would be a meaningful bid — $300 million against XRP's daily volume is not trivial, though it is not enormous either.

But the dry powder is not $300 million. It is $88.5 million, and even that shrinks once you net out fees and expenses. The $214 million was dry powder in November 2025. In the present tense of the disclosure, it is inventory, not ammunition. It was bought at $2.54. The reference price in the same materials is $1.53.

Run the arithmetic on that position. Eighty-four million three hundred seventy thousand tokens at $2.54 is an outlay of roughly $214 million. The same position marked at $1.53 is worth roughly $129 million. That is an unrealized loss in the vicinity of $85 million — on the single largest line item in the company's balance sheet.

I am not going to dress that up. A treasury vehicle whose primary asset was accumulated at an average cost roughly 40% above the current mark is a vehicle trading under water on its core position from the moment it lists. That does not make it a fraud. It makes it a leveraged long that entered at the wrong price, which is a common and survivable condition — but only if the vehicle has fresh capital and time. Evernorth has $88.5 million of fresh capital and a 2031 maturity. It has time. What it may not have is a market willing to fund the next leg at a premium.

The 473 million number and its water content

The second headline figure is the holding: roughly 473 million XRP expected at close.

Treat this number the way you would treat a token's "total supply" that includes locked, vested, and treasury allocations. It is a gross figure that blends categories with radically different economic meaning. Break it apart.

Purchased with cash: approximately 84.37 million XRP, the November 2025 buy, already converted, already underwater.

Contributed in kind by RippleWorks: approximately 211.3 million XRP. This is not a cash purchase. It is a physical token contribution by an affiliate, and — critically — it is retractable. RippleWorks retains the right to withdraw the contribution if the merger fails to complete. That is not the same asset as tokens the company bought on the open market and controls outright.

Affiliate subscription: approximately 50 million XRP, another in-kind contractual allocation with its own conditions.

When you strip the in-kind and conditional layers out of the 473 million, the number the company unambiguously owns outright is a fraction of the headline. The rest is committed, contingent, or withdrawable. The disclosure itself concedes this: it notes that in November 2025 the figure of "over 473 million" already included tokens described as purchased and committed. Which means the 473 million presented at close is not a new number. It is the same number, re-presented, with the composition unchanged.

A treasury figure that does not grow between the pre-close disclosure and the post-close announcement is not an accumulation story. It is a re-labeling story.

I have seen this pattern at the protocol level too many times to ignore it. In 2025 I reverse-engineered an AI trading bot protocol that advertised 40% annual returns. The returns were real on paper. The slippage costs hidden in the execution logic erased them in practice. The headline metric was accurate in isolation and misleading in context. That is exactly what a blended 473 million XRP holding is: accurate in isolation, misleading in context. The number is not false. It is constructed to be read as more than it is.

The PIK note and the compounding you are not modeling

The $30 million convertible note is the smallest layer and the one most likely to be ignored. It should not be.

It is structured as payment-in-kind, maturing 2031, at 4%. PIK means the interest is not paid in cash. It is added to the principal. The debt accretes. A 4% PIK note that sits for the better part of a decade does not stay a $30 million note. It becomes a larger obligation that must eventually be satisfied — either with cash the company does not generate, or with equity that dilutes existing holders, or through conversion that hands the noteholder a claim on the upside at a fixed strike.

The 2031 maturity is a feature, not a bug, from the issuer's perspective: it means no near-term principal pressure. But it also means the obligation compounds quietly in the background while the company's only real asset — the XRP position — swings with the market. If XRP recovers, the conversion is cheap and the dilution is manageable. If XRP stays depressed, the company faces a debt it cannot service from operations, because it has no operations. It has a token.

This is the structural asymmetry of every DAT: the debt is denominated in dollars, the asset is denominated in a token, and the two do not have to move together. When they diverge against you, you are running a levered basis trade with no hedge and no cash flow. That is not a treasury. That is a bet with a financing wrapper.

The RippleWorks asymmetry

Now the part that should make any serious reader uncomfortable, and that the retail narrative glosses over entirely.

RippleWorks is contributing roughly 211.3 million XRP — the largest single block of the treasury — and it can take that contribution back if the merger fails. Read that clause slowly. It means the entity contributing the majority of the treasury's non-cash assets has retained an exit option. Evernorth bears the downside of the arrangement; RippleWorks retains the optionality.

In contract terms, this is an embedded put held by the counterparty. In plain terms, it means the relationship between Evernorth and its largest contributor is asymmetric: Evernorth depends on RippleWorks for the bulk of its holdings, and RippleWorks does not depend on Evernorth for anything it cannot walk away from.

This matters for two reasons. First, it means the 473 million figure has a hidden conditional floor — under adverse conditions, a large chunk of it can evaporate. Second, it tells you where the power sits in this structure. Evernorth is not an independent treasury vehicle that happens to hold XRP. It is, functionally, the capital markets arm of the Ripple ecosystem, and its largest asset is held at the pleasure of an affiliate.

I have written before that code is law, but human greed is the bug. Here the law is a contract, and the contract was written to protect the contributor, not the public holder. That is not illegal. It is not even unusual. It is simply the truth of the structure, and the truth is that a holder of XRPN is buying exposure to a position whose largest component is revocable by a related party.

The price gap is the whole thesis

Everything above converges on a single observable fact: the disclosed average purchase price of $2.54 against a reference price of $1.53. That is a spread of roughly 40%.

If you accept the timeline as presented, this is a vehicle that accumulated its core position at a price 40% above where the market is currently clearing, and it is about to list into that gap. The equity has to price that. A treasury vehicle with an underwater core position and no operating income trades at a discount to net asset value, or it trades at a premium on pure narrative and eventually corrects. There is no third option, because there is no cash flow to bridge the difference.

I want to be precise about what this does and does not imply. It does not imply the vehicle is fraudulent. Buying an asset and watching it fall is the oldest experience in trading. It does not imply the structure is illegal. SPACs with conditional financing are standard. What it implies is that the narrative — a well-capitalized XRP treasury with $300 million of buying power — and the reality — an $88.5 million fresh-capital vehicle holding a 40%-underwater position with a revocable in-kind contribution — are far apart. The trade is not the XRP exposure. The trade is the correction of the narrative to the reality.

What the smart money is watching that retail is not

The retail read on Evernorth is simple and wrong: "XRP treasury company, Nasdaq listed, $300 million, bullish for XRP." That read treats the announcement as the event. It is not. The announcement is the marketing. The event is the redemption rate.

Here is the mechanism, stated plainly. When a SPAC merges, existing trust shareholders get a choice: keep their shares in the merged company, or redeem — take their pro-rata slice of the trust and walk. Redemptions are the single most revealing signal in any SPAC deal, because they are a vote by the most informed, most liquid holders on whether the deal is worth holding. High redemption means the people who can do the math chose cash. Low redemption means they chose the equity.

For Evernorth, the redemption rate directly determines the trust proceeds — the $48 million layer, which is itself part of the $88.5 million fresh capital. If redemptions run high, that layer shrinks, and the already-modest fresh capital shrinks with it. The $88.5 million is not a floor. It is a starting estimate that only gets revised downward.

This is what smart money watches. Smart money watches, dumb money chases. The dumb money is chasing the $300 million headline and the XRP beta. The smart money is waiting for the redemption disclosure and the post-close cash statement, because those two documents — not the press release — contain the actual deployable number. Until they print, every "buying power" figure in circulation is an estimate dressed as a fact.

There is a second signal the careful reader catches. The framing of the disclosure itself is tilted. The public announcement rounds to "approximately $300 million." The proxy materials carry the precise $303 million and the precise subscription amounts down to the decimal. When a company rounds up in its public communication and gets precise in its regulatory filing, the rounding is not an accident. It is a narrative choice. The public number is chosen to read as strong. The filed number is the truth. Always trade the filed number.

The category error that is going to hurt people

Step back from Evernorth specifically and look at the pattern, because Evernorth is one instance of a broader reflexive loop in the DAT sector.

MicroStrategy established the template: take a public equity wrapper, load it with a concentrated asset, and let the equity trade at a premium to net asset value. The premium exists because the equity offers something spot ownership does not — leverage, tax treatment, access for accounts that cannot hold the token directly. As long as the premium persists, the company can issue more equity, buy more of the asset, and the flywheel spins. The premium funds the accumulation, and the accumulation validates the premium.

This works until it does not. The flywheel requires two things to keep spinning: a rising or stable asset price, and a market willing to pay a premium for the equity. When the asset falls, the net asset value falls, and the premium compresses or inverts. When the premium inverts, new issuance destroys value instead of creating it, and the flywheel becomes a death spiral. Every leveraged treasury vehicle faces this. The question is only timing.

Evernorth enters the flywheel at an awkward moment. Its core position is 40% underwater. Its fresh capital is a fraction of what the narrative implies. Its largest in-kind contribution is revocable. And it is entering a sector — XRP treasury vehicles — that has no proven premium model, because it is the first. Being first is a genuine advantage in narrative terms. It is not an advantage in structural terms, because there is no comparable to anchor the valuation and no track record to sustain the premium.

The broader DAT sector is watching Evernorth as a test case. If it holds a premium, it invites a wave of imitators — treasury vehicles for every token with a large enough holder base. If it trades at a discount or falls flat, it cools the entire category, because the first mover sets the tone. Evernorth is not just an XRP trade. It is a referendum on whether the DAT model generalizes beyond Bitcoin.

The disclosure soft spots

Two disclosure issues deserve flagging, both at low confidence but both worth watching.

The first is the gap between the public "approximately $300 million" and the filed $303 million. Rounding is normal. Rounding up in a way that flatters the buying-power narrative is a choice, and if the Securities and Exchange Commission ever reads that choice as materially misleading, it becomes a disclosure liability rather than a rounding convention. I rate the probability low. I rate the consequence high enough to monitor.

The second is the disclosure of the RippleWorks retraction right. Related-party transactions with asymmetric optionality are exactly the kind of arrangement that draws regulatory questions about fairness and independence. If the retraction clause was disclosed in full, the company is protected. If it was buried, the company has a problem. The proxy materials are detailed — precise to the decimal on the subscription amounts — which suggests the disclosure is at least nominally complete. But completeness of disclosure is not the same as adequacy of disclosure, and the difference is where liability lives.

Neither of these is a reason to avoid the analysis. Both are reasons to read the primary documents rather than the summaries.

What the counterparty structure tells you about the deal

There is a governance question underneath all of this that the numbers only hint at.

Who is actually in control of Evernorth after the merger? The public shareholders hold a minority, fragmented position. The sponsor — the entity that created the SPAC — typically retains a meaningful stake and an incentive structure that rewards deal completion over long-term performance. The affiliate contributors hold in-kind positions with retraction rights. The convertible noteholder holds a claim on the upside at a fixed strike.

The public holder is the residual claimant in a structure where everyone else has a preferred position or an exit option. This is not unique to Evernorth. It is the defining feature of SPAC economics, and it is why SPAC-merged entities so often underperform their sponsors' expectations. The incentives are misaligned by design: the sponsor wins when the deal closes, regardless of how the equity performs afterward.

I have said before that the promise of "code is law" in governance is largely fiction, because upgrade rights always sit with a few multisig admins. The same fiction operates here in corporate form. The governance documents look like shareholder democracy. The economics are sponsor-advantaged. Reading the economics, not the governance theater, is how you avoid being the residual claimant who funds everyone else's exit.

What would change my mind

A good analysis names its own falsifiers, so here are mine.

First, a high redemption rate is not automatically bearish if the deal structure compensates. If redemptions come in low and the trust proceeds land near $48 million, the fresh capital holds near $88.5 million and the story improves marginally. Watch the redemption disclosure. It is the single most important post-close data point.

Second, if XRP recovers toward or above $2.54, the underwater position becomes a break-even or a gain, and the entire "40% below cost" thesis inverts. This is a macro bet I cannot make for you. The treasury thesis is a leveraged XRP long, and if XRP runs, the equity runs harder. That is the bull case, and it is real. I am not dismissing it. I am pricing it against the structural drag.

Third, if the company deploys the $88.5 million efficiently and the market rewards it with a premium, the flywheel spins and the revocable contribution matters less, because the company's own capital base grows. Execution could rescue the structure. It usually does not, but it can.

What would not change my mind is another press release restating "approximately $300 million." I have the filed number. I am not trading the rounded one.

Where this leaves the reader

Strip it to the essentials.

Evernorth is a Nasdaq-listed XRP treasury vehicle, the first of its kind, built through a SPAC merger with Armada Acquisition Corp. II. The public narrative is $300 million of buying power. The filed reality is roughly $88.5 million of new deployable gross capital, an $84 million purchased position that is approximately 40% underwater, a 473 million token holding that blends purchased, in-kind, and conditional allocations, a $30 million PIK note that compounds toward 2031, and a largest-asset contribution that can be withdrawn by an affiliate.

The trade is not "XRP is going up." The trade is the gap between the announced buying power and the signed buying power, and that gap is roughly a factor of three. Markets correct that kind of gap. Sometimes slowly, sometimes violently, but they correct it.

Watch the redemption rate. Watch the post-close cash statement. Watch the XRP price against the $2.54 cost basis. Those three numbers tell you everything the press release does not. Everything else is rounding.

I will be watching the ledger. I always watch the ledger.

A note on the timestamps: the material underlying this analysis contains inconsistent dates — 2025 and 2026 references that do not reconcile in a single present tense. I have analyzed the structure, which is internally consistent, while flagging the calendar as unreliable. If you are acting on this, verify the primary source dates yourself. A trade is only as sound as the timeline it sits on.

A note on the price: the disclosed average purchase price of $2.54 against a reference price of $1.53 is the single most load-bearing fact in this analysis. If the reference price is stale or mislabeled, the underwater thesis weakens proportionally. Check it. Everything downstream of it depends on it being right.