apxUSD — Risk Report

Significant risk · 3.5/10

apxUSD is a $1-target stablecoin from Apyx, a young protocol (live since Feb 2026) backed primarily by Strategy’s STRC perpetual preferred shares, plus a cash sleeve and protocol-owned liquidity. The yield from the backing flows to Apyx’s sibling token apyUSD; apxUSD holders forgo yield in exchange for stablecoin functionality. Supply grew rapidly into the hundreds of millions after launch.

apxUSD broke its peg in June 2026 and ran below par on collateralization until the autumn; the attested ratio has held at or above par continuously since 2026-09-17. ⚠️ The market price has not followed — apxUSD still trades about 1.10% below $1 and has never printed at or above $1.00 in 722 readings. That event reshapes most of this report — see The June 2026 depeg directly below. Live peg, reserves, and collateralization are on the dashboard.

Peg	Yield	Exit	Age	Chains
$1 target, RWA-backed synthetic (broke peg June 2026)	None (yield routes to apyUSD)	DEX (Curve) or Apyx USDC settlement (opaque)	about 6 months	Ethereum, Base, BNB Chain

Score breakdown

Every axis now states its reason.

Dimension	Score	Notes
Stability	3.5	The June 2026 depeg is the record and the post-mortem is on the page above, but the recovery gate this report set has been met on both legs. ⚠️ It stops at 3.5 on a measurement, not a judgement: apxUSD has never traded at or above $1.00 in 722 readings (maximum $0.994716)
Backing	3.0	Attested collateral ratio is 101.5602% netted, a surplus of $2,749,710, and has held at or above par continuously since 2026-09-17 — 393 consecutive hourly readings. ⚠️ The improvement is in the divisor, not the collateral: the coverage hole closed and the “no buffer” condition cleared, while the look-through baseline fell from 5.7 to 4.71 as the collateral’s own score dropped
Liquidity & Exit	3.5	Manual mint/redeem plumbing was the thing the depeg broke, and that is unchanged. ⚠️ Half a notch, and only for a distortion removed rather than depth gained: measured depth is flat — derived mean 5.613 before 2026-09-22 against 5.570 after, with sub-5.0 readings slightly more frequent since. ⚠️ And the ladder is censored — both the 50bps and 200bps rungs report $100,000 with `is_floor: true`, `ladder_exhausted`. A floor is a lower bound and does not lift a score
Dependencies	3.0	STRC is 86.7% of what actually backs the token on the issuer’s own netted basis, against a 56.44% gross share — two of the four basket legs are the protocol’s own positions. Argued under 4 · Dependencies
Contract & Admin	4.5	Three Safes: token governance 4-of-6 behind 72h, STRCx custody 3-of-6 with no delay, bridge 4-of-7 with no delay. ⚠️ The custody leg binds — same owner set as governance, no wait, over the reserve asset. Argued under 5 · Contract & Admin
Issuer	5.0	Unchanged, and deliberately about behaviour rather than the contract layer — the June operational failures are what stepped it down
Overall	3.5	⚠️ Deliberately below the 3.75 axis mean. The mean is lifted by Contract & Admin 4.5 and Issuer 5.0 — the axes this asset is strongest on — while the binding constraint for a dollar-denominated token is the par it has not reached

1 · Stability — 3.5

The June 2026 depeg

apxUSD broke peg in the first week of June 2026. Two things happened at once:

The collateral de-anchored. Strategy’s STRC preferred — the dominant reserve asset — fell well below its $100 par, to an all-time low around $74.57 in late June. That is a direct mark-to-market hit to a reserve that is mostly STRC. (STRC has since recovered to around $95 — see the 2026-08-18 update box above for where that leaves things.) (Note: STRC’s old “drift below ~$95 forces a dividend hike to defend par” reflex was itself removed by Strategy’s 06-29 framework — par defense is now a discretionary soft floor; see the update box at the top.)
Holders headed for the exit. A meaningful share of supply was redeemed/sold over the same window, so reserves shrank faster than supply.

The result: apxUSD traded at a material discount and Apyx’s own attestation feed went to a persistent, issuer-attested sub-100% collateral ratio, where it stayed for months. ✅ That condition has now cleared: the ratio has held at or above par continuously since 2026-09-17, 393 consecutive hourly readings, attested 101.5602%. The first-week trough was into the low-90s on the dollar and a mid-90s% collateral ratio; through late June the dislocation deepened to about $0.82 (roughly −18%) with a collateral ratio near 84%. That is not a transient mark-to-market wobble; it is a real shortfall, and it has now run for months rather than weeks. It has substantially repaired — see the 2026-08-18 update box above for the current figures — but it has not closed. Because apxUSD has no atomic on-chain redemption — no contract you can call to swap one apxUSD for a dollar of collateral — there is no built-in arbitrage to force the price back to $1. Recovery depends on STRC re-rating (now on the soft-floor bid) and on Apyx restocking reserves through its off-chain pipeline.

Treat the specific discount and collateral ratio as a moving event, not a fixed number — check the live dashboard for the current values. (One caveat: Apyx pulls its own secondary-market depth off-hours by design, so weekend snapshots overstate the steady-state dislocation.)

Peg & yield dynamics

apxUSD is designed to trade at $1 — and in June 2026 it didn’t. Peg integrity rests on three things, all of which the depeg tested:

STRC + cash backing holds its value. STRC fell well below par, directly cutting reserve value. As of 06-29, Strategy defends STRC value with a discretionary soft floor (buyback / reserve) rather than a reflexive sub-$95 dividend hike — a bid under the price, not a peg. STRC is a ≈1-year-old instrument, not yet tested through a prolonged BTC drawdown or an MSTR equity-raise pause.
The Alpaca brokerage and the STRCx Safe hold what’s reported. Wolf’s monthly CPA examinations independently verify the securities sleeve at two snapshot dates per month, and the STRCx Safe at 0x37b0779a… is verifiable on-chain in real time. Wolf does not opine on the cash sleeve after the April scope narrowing.
Secondary depth absorbs flows. It did not absorb the June redemption wave at par.

What broke the peg, and what restores it: STRC de-anchoring on a concentrated reserve plus a redemption wave pushed collateralization below par, and with no atomic redemption there was no automatic arbitrage to close the gap. ⚠️ The collateral gap closed through the reserve recovering rather than through any redemption mechanism, and it closed for good on 2026-09-17 — which is the point: the same absence of arbitrage that let it open means nothing holds it shut. The path back to $1 requires STRC to re-rate up — now on Strategy’s discretionary soft-floor bid (the $1.0B STRC-priority buyback plus a cash reserve that has since built to a record $4.80B), not the old forced sub-$95 dividend hike — and Apyx to restore the collateral ratio to ≥100% on the attested feed for a sustained window. As of August 2026 the first half has largely happened and the second has not, which is why the scores are unchanged.

“Apyx 2.0” — an announced redemption redesign (June 15, 2026). In a follow-up to the post-mortem, Apyx outlined a reworked mint/redeem model intended to fix a flaw the June event exposed. Everyone — in calm and in stress — would mint and redeem at a single Redemption Value, a floor price carrying a small spread, with the dashboard’s headline NAV relabeled Total Collateral Value so the gap between the two reads directly as the overcollateralization buffer (their worked example: $1.02 of collateral behind a $1.00 redemption floor). Approved counterparties would quote against the reserve through a structured RFQ. Apyx frames the redesign as closing the prior “free put option” — under the old NAV-redemption logic, the first redeemers in a drawdown could arbitrage the buffer at the expense of everyone who stayed, exactly the dynamic that played out in June; under Redemption Value the buffer instead accrues to long-term holders. The intent is a genuine improvement in the redemption logic. The important caveat: this is blog-only as of this revision. Apyx’s own docs at docs.apyx.fi still describe the old mechanism (apxUSD “settled in USDC / not directly redeemable”), no contract, PSM, or cooldown change has been disclosed, and redemption stays off-chain and is now explicitly gated to “approved counterparties.” Redemption Value is an announced pricing policy, not an enforceable on-chain mechanism — so it doesn’t change the no-atomic-redemption reality behind the Peg and Backing scores, and the scores are unchanged. Read it as a credible fix that is pending adoption into the docs and on-chain confirmation, not as a live guarantee that Apyx now redeems at a floor.

2 · Backing — 3.0

The two collateral ratios in circulation are the same book measured two ways, and both are the issuer’s:

basis	ratio	what it excludes
netted — the headline attestation	101.5602% (2026-10-04)	protocol-owned liquidity and minted-but-unsold inventory, removed from both sides
netted, earlier reading	98.0177% (2026-08-23)	same basis — both ends dated, so the move is the comparison
gross	98.5901% (2026-08-23)	nothing — reserves over supply. ⚠️ Not re-read on this pass

⚠️ This is not a disagreement, and not a gap between the issuer and us. Both figures are published by Accountable, and the netted one reconciles exactly: (reserves − POL − inventory) / (supply − POL − inventory) reproduces the attested headline from the raw bucket split, to within 2.6e-05 pp across 723 of 723 readings. ⚠️ That tests the arithmetic, not the inputs — it is an internal-consistency check on a single feed, not a second source, which is why the trust banner above still stands.

Our independent lower bound is 98.5289%, computed gross after stripping the roughly $190,970 premium at which on-chain STRCx trades over the underlying STRC NAV. Against the issuer’s gross figure that is 0.06pp stricter — exactly what a conservative bound should be, and exactly the size of the premium removed. ✅ Compared like for like — gross against gross — our bound is stricter than the issuer’s netted headline, which is how it is designed to behave.

✅ The recovery gate set out below is now met on both legs, and it was a hard gate rather than a hopeful one. It demanded STRC recovering toward about $95 and Apyx restoring collateralization to par or better on the attested feed for a sustained window. STRC marks $99.41, −59bps to par (⚠️ a last close — market_session: closed). The attested ratio reads 101.5602%, a surplus of $2,749,710, continuously at or above par since 2026-09-17T19:49:54Z across 393 hourly readings, with none of the 293 readings since 09-22 below par.

⚠️ The word “sustained” earned its keep, and this is the part to take from it: between 2026-09-04 and 09-17 the ratio crossed par 21 times, with 49% of readings in that window still below 100% and a mean of 99.998. Two weeks before the clean break, a reader checking the headline would have seen par and been wrong. A gate that had asked only for par would have opened on noise.

⚠️⚠️ And the discount did not close with the backing, which is the structural finding rather than a leftover. apxUSD trades about −1.10% below par. It has never traded at or above $1.00 in 722 readings — maximum $0.994716. Backing closing on the balance sheet while the discount stays open in the market is only jointly consistent with there being no atomic redemption, so the recovery confirms the thesis this report has argued rather than refuting it.

Read that as recovery, and read it accurately. Three things keep it from being a bigger story than it is:

The move since our last stated figure is about +6 percentage points on the attested basis, not +14 — and only +4.5 points on the conservative one. The June trough (about 84% collateral ratio, apxUSD near $0.82) is the number in the older sections below, but it was superseded on 2026-08-01, when the ratio had already recovered to roughly 92% headline / 94% on the lower bound. Measure from there, ⚠️ and measure like against like: attested 92% → 98.02% is +6.0pp, lower bound 94% → 98.53% is +4.5pp. Comparing the 92% headline anchor against our 98.59% computed figure would give +6.6pp, and that is a basis mismatch rather than a bigger recovery. Most of the repair happened in July, not in the weeks since.

The recovery gate is met on both legs. STRC marks $99.41 and the attested ratio reads 101.5602%, continuously at or above par since 2026-09-17T19:49:54Z across 393 consecutive hourly readings, none of the 293 since 09-22 below par. ⚠️ “Sustained” is what made the gate worth having: the ratio crossed par 21 times between 09-04 and 09-17, 49% of readings in that window still below 100%, mean 99.998. A gate that asked only for par would have opened a week early on noise. ✅ The netting is not taken on trust — recomputing (reserves − POL − inventory) / (supply − POL − inventory) from the raw bucket split returns the attested figure to within 2.6e-05 pp across 723 of 723 readings.

The collateral leg has re-rated, and the bid funding it has been re-authorised rather than exhausted. STRC — about 61% of Apyx’s reserves, or roughly 69% once the net-zero inventory line is stripped out — now marks at about $97.30, corroborated three ways within 0.3%: live secondary at $97.30, the 8-K implied repurchase price at $97.48, and Apyx’s own on-chain mark at $97.24. ⚠️ This does NOT improve the collateralisation figures on this page. Apyx already marks at market, so the attested ratio and deficit on the 2026-08-31 read — 98.7953% and −$2.59M, against the 98.0177% / $4.40M of the 2026-08-23 table above — were computed on a market-consistent price all along. The mark moved; the collateralisation did not, because it was never computed off the old one.

⚠️ And the “spending more each week for less price response” reading is withdrawn — it reversed. $136.4M moved the average execution price $0.11; the next $151.8M moved it $2.18. An average repurchase price is not a market print and cannot separate a secondary re-rating from more aggressive bidding, so neither direction was ever strong evidence. ⚠️ What is not in doubt is the runway: about $364.8M of the $1.0B authorization remains as at 2026-08-31, roughly 2.4 weeks at the current pace, putting the end around mid-September. Coverage is stable; the bid underneath the collateral is close to exhausted. See the STRCx report for the detail.

⚠️ One composition fact worth carrying, because it bears on what can be checked at all: of the $183.0M STRC bucket, only $67.6M — 36.9% — is on-chain STRCx. The remaining $115.5M, 63.1%, is issuer-stated. The largest single reserve line is majority off-chain, which is why the independently verifiable portion of this reserve is the smaller one.

Reserve composition has also drifted since the sections below were written: net of the inventory line the reserve is now roughly 69% STRC family, 19% protocol-owned liquidity, and 12% cash — the STRC concentration is a little lower than the 74% described below, but the shift went into POL, which is reflexive backing rather than cash, not into the cash sleeve.

Backing is verifiable through two layers. (a) A continuous TEE-attested proof-of-solvency feed at accountable.apyx.fi signs each snapshot from a secure enclave with a key registered on-chain. (b) Monthly third-party CPA-firm attestations by Wolf & Company, P.C. (Boston; AICPA examination standards) published at docs.apyx.fi — March 2026 and April 2026 reports both signed. The live dashboard shows reserves, supply, and collateralization; as of this revision the attested ratio is below 100% (see The June 2026 depeg above).

Reserve composition — STRC-heavy on a net basis (corrected June 2026). Apyx’s June post-mortem clarified that the Accountable feed breaks reserves into five buckets: STRC family, Cash & Equivalents, Protocol Owned Liquidity (POL), Inventory, and SATA. Apyx’s public dashboard had been folding POL and Inventory into “Cash,” which made the basket look more cash-heavy than it is. Reading it correctly:

Inventory nets to zero and is not backing. It’s apxUSD that Apyx has minted but not yet sold — an asset offset by an equal burnable liability. Stripping it out is the honest way to count reserves.
Net of that Inventory line, the reserve is STRC-dominant — about 69% STRC family, plus roughly 19% POL and 12% cash on the current feed (it was closer to 74% / 13% / 13% when this correction was first made). Either way the point holds: net of inventory the basket is more STRC-concentrated than the gross figure suggests, and the drift since has moved weight into POL rather than into cash.
POL is reflexive. It’s protocol-owned DEX liquidity plus USDC lent into Apyx-collateralized Morpho markets, capped at 15% of reserves. It’s liquid, but it is deployed against Apyx’s own assets, so it is lower-quality backing than plain cash.

That correction — more STRC concentration than previously scored, and a thinner true-cash cushion — combined with a collateral ratio that ran below par continuously from June until September is why the Backing axis sat at 2.0. ✅ That continuity has ended — par or better since 2026-09-17, 101.5602% attested — and the axis is re-derived to 3.0 — the axis also rests on the concentration and cash-cushion findings above, which are unchanged, and the re-rate was done on its own pass rather than as a reaction to a few days of readings. ✅ apxUSD has held at or above par continuously since 2026-09-17 — above 100% on the netted collateral ratio, on the gross basis, and on an independent conservative re-mark. ⚠️ Read the level carefully: the attested surplus is 1.56%, $2,749,710. That is a recovery and now also a thin buffer, on a book that was in deficit through the summer. ⚠️ The margin was far thinner than it looks for most of September: between 09-04 and 09-17 the ratio crossed par 21 times, with 49% of readings in that window still below 100%. ✅ What it removes is the realised undercollateralisation; it does not remove the backing concern, which rests on concentration and on what the reserve is made of rather than on the ratio alone. A 50% STRC writedown would still leave the cash and POL portion, but that cushion is thinner than the gross numbers implied. Strategy’s buyback program has since lifted the STRC mark substantially, which carried the attested ratio back up to roughly 97% and then first above par in mid-September — real repair, and the reason the trajectory reads as recovery. ⚠️ It has not lifted this axis, and the reason is no longer the level. The ratio is above par but by about 25 basis points and for three days, and the improvement came from an issuer bid on the collateral rather than from Apyx restocking its own reserve. The recovery here is borrowed from Strategy’s balance sheet, and Strategy’s buyback authorization is finite.

The qualifiers worth knowing:

The “Cash & Equivalents” line is not itemized in the Accountable feed — could be bank deposits, T-bills, USDC, or some mix. Not disclosed publicly.
Wolf is mid-tier, not Big-4. Solid AICPA examination engagement (named auditor, professional liability) but a tier below Deloitte / PwC / EY / KPMG.
The April Wolf engagement narrowed scope to securities only — cash, stablecoin, and dividends-in-motion (covered in March) all dropped out. The narrowing has held. Wolf’s June 2026 examination, signed 2026-07-22, covers $193.3M of STRC/STRCx securities at 06-30 and again excludes cash. So no CPA firm has attested the cash sleeve for any date after 2026-03-31, and this now reads as the standing arrangement rather than a one-month transition.
apxUSD is issuer-push, not user-pull. Mints do not atomically pull USDC from a depositor — Apyx mints apxUSD to itself and sells it on Curve / CoW Protocol for USDC. The proceeds are swept from Apyx’s operational multisig to a Kraken deposit address (Kraken is a US-licensed exchange — the team uses it as a USDC→USD off-ramp), then routed off-chain to Apyx’s brokerage at Alpaca for STRC / SATA purchases. The pipeline is institutionally normal; the disclosure gap is that Apyx hasn’t published the off-ramp arrangement, and the June depeg showed this manual leg is too slow to defend the peg under a real run.
STRC + SATA are held at Alpaca brokerage in the name of Preference Foundation (or a subsidiary). STRCx (the on-chain tokenized form of STRC) sits in 0x37b0779a…323a555, a 3-of-6 Gnosis Safe with the same six owners as Apyx’s other admin Safes (identified on-chain; the 4/30 balance matches the Wolf attestation within 1%). ⚠️ That sameness was re-measured on 2026-08-25 and is exact, across three Safes rather than two: this custody Safe, the 4-of-6 token-governance Safe 0xABdd8c8e…65e96, and the bridge governor 0xf9862EfC…3cE2. ⚠️ Re-read on-chain 2026-09-09, that sameness is no longer exact and the bridge is the reason: the token-governance and custody Safes still hold the identical six owners, but the bridge Safe moved to 4-of-7 at block 25,877,589 on 2026-08-31 — one owner added, threshold raised — so it now holds those six plus one exclusive signer.

Six people still hold three hats. Four of them satisfy token governance, three of them satisfy this custody Safe, and four of them satisfy the bridge, so the six reach all three doors and the seventh signer opens none they cannot. ⚠️ The 3-of-6 custody hat is now the only one with both a low quorum and no delay: any three of those six can move this collateral without touching the 72-hour delay that governs the token. The bridge closed its quorum gap on 08-31 and keeps its delay gap. The owner-set identity and thresholds are measured on-chain; the role attributions are read from this coverage’s records. See the apyUSD report for the full write-up and the correction history behind it. The Safe has no time-delay on transactions — same admin gap as the cross-chain bridge layer.

The Accountable TEE feed proves what Apyx feeds the enclave; it does not audit the custodian or the wallet keys themselves. Wolf’s monthly examinations close part of that gap on the securities balances, but no examination opines on the on-chain wallet’s key-management posture.

A note on “solvent.” In its post-mortem Apyx says it “remained solvent throughout — reserves exceeded the market value of supply.” That is a weaker claim than reserves covering supply at par ($1). By the stricter, standard measure — collateral ratio versus par — apxUSD has been below 100%. Both framings are true; the par-based one is the conservative one, and it’s the one that matters when you’re holding a token that’s supposed to be worth a dollar.

3 · Liquidity & Exit — 3.5

Entry: Mint at Apyx (manual, EIP-712 signed order workflow) or buy on the Curve apxUSD/USDC pool on Ethereum. The Curve pool is the realistic retail entry and exit venue.

Exit:

Retail size: Sell on Curve apxUSD/USDC (live depth on the dashboard’s Secondary Liquidity panel). During the June depeg, observed exit cost rose sharply and the pool traded at a discount to par — the “institutional-grade, $100K clears under 25 bps” depth described in earlier revisions did not survive the stress event. Apyx pulls its own secondary depth off-hours by design, so depth and slippage swing with the time of day and the stress state; check the live panel before sizing.
Larger size: Apyx offers redemption in USDC, but the mechanism isn’t publicly documented — no published timeline, no PSM, no on-chain guarantee. Effectively trusts Apyx to liquidate STRC into USDC and pay out when requested.

The durable finding on this axis is that secondary depth is variable and issuer-discretionary, not the dependable book earlier scores implied — which, together with a peg that has been dislocated continuously since June (currently around $0.94, having been as low as about $0.82, with KyberSwap aggregator depth rather than weekend pool depth the canonical exit mark), is why the Liquidity axis sits at 3.0. The narrowing discount improves the level of the exit mark without changing the finding: depth still varies with the issuer’s own participation and the time of day. Two related constraints live on other axes: coordinated-run capacity (secondary depth is a small fraction of supply, a backing/buffer concern) and redemption-mechanism opacity (the undocumented off-chain settlement path, a peg-mechanism concern).

4 · Dependencies — 3.0

Gross basket (2026-09-07): STRC 56.44%, Inventory 17.55%, protocol-owned liquidity 17.33%, cash 8.67%. ⚠️ The gross share understates the exposure, and the issuer’s own basis says so — the attested ratio is netted on both sides, and on that basis STRC is 175.5M / 202.5M = 86.7% of what actually backs the token.

⚠️ Two of the four legs are circular. Protocol-owned liquidity and Inventory are the protocol’s own positions sitting inside its own backing, and netting them out is exactly what produces 86.7%. For a dependency read the netted figure is the honest one, because a basket that counts your own positions as diversification is not diversified.

5 · Contract & Admin — 4.5

Three Safes, not one. Token governance is 4-of-6 behind a 72-hour timelock; STRCx custody is 3-of-6 with no delay; the bridge is 4-of-7 with no delay.

⚠️ The binding constraint is the custody leg: three keys and no wait, over the reserve asset, on an owner set identical to token governance. A delay is only a delay if the people it binds have no faster door, and here they do.

✅ Held at 4.5 rather than lower on a genuine audit stack — Quantstamp, Certora formal verification and Zellic — and a guardian role that can cancel scheduled operations.

6 · Issuer — 5.0

Backers + audits + contract layer are solid; the June depeg exposed the operations:

Backed by DeFi Development Corp (DFDV), a Nasdaq-listed company. Joseph Onorati (DFDV CEO) is publicly named. Investors include ParaFi, Pantera, Kraken Ventures, Wintermute Ventures, GSR.
Three audits: Quantstamp (Feb 2026), Certora (Mar 2026 — formal verification, 1 high-severity finding fixed), Zellic (Mar 2026). Solid stack.
Admin posture (verified on-chain): Token governance is a 4-of-6 Safe behind a 72-hour timelock, with a distributed guardian role that can cancel scheduled operations. ⚠️ Read that as one of three doors rather than as the admin posture. The same six people also hold a 3-of-6 custody Safe over the STRCx collateral with no delay at all, and a 4-of-7 bridge Safe, also undelayed. The binding constraint is not four keys plus 72 hours — it is three keys and no wait, over the reserve asset. See the topology above; this is what Contract & Admin scores.

The post-mortem is two-sided — both halves matter:

What Apyx admitted went wrong: overnight/weekend liquidity was pulled (a TradFi/DeFi off-hours mismatch), the manual mint/redeem plumbing was too slow to defend the peg at scale, communications lagged, and Apyx’s own dashboard briefly showed an inflated NAV from a STRCx pricing bug. These confirm the operational fragility the depeg surfaced — and are why the Issuer axis steps down to 5.0 (behavior, not the contract layer).
What actually held: the apyUSD unlock window prevented a bank run, redemptions were processed proportionally (so the remaining basket didn’t concentrate into the illiquid leg), the apyUSD yield ratchet held, and no Morpho lending market booked bad debt — the apyUSD/apxUSD Morpho market took zero liquidations because its oracle keys off the redemption rate, not the spot price. This demonstrated resilience is why the overall score lands in the mid-3s rather than lower despite the backing markdown.

Caveats:

Apyx as a legal entity appears separate from DFDV (US/EU/EEA users are geo-blocked). In a solvency event, holders’ claims may route through an unnamed offshore entity rather than the Nasdaq-listed DFDV. Standard offshore-RWA structure but worth understanding.
No Apyx team members are individually doxxed beyond DFDV’s CEO. The issuer entity is Preference Foundation (Director Carolyn Kelly signs the Wolf attestation). Brokerage is Alpaca; the bank(s) holding cash are not named.
The cross-chain bridge (audited Chainlink CCIP; Ethereum ↔ Base, with BNB Chain contracts added in mid-2026 — a third chain means a third bridge and admin surface) is governed by a smaller 3-of-6 multisig with no time-delay — weaker than token governance. Separately, some Morpho markets ran a stale self-managed price oracle that lagged as apxUSD left $1; Apyx is migrating those to Chainlink.
No bug bounty program.

Who it’s for · Who should avoid

For: Risk-tolerant DeFi users who want exposure to MSTR/STRC dividend yield via a stablecoin wrapper, understand they’re effectively making a leveraged bet on Strategy’s ability to keep paying STRC dividends, and are comfortable holding a token that has already broken peg once and depends on off-chain RWA custody and opaque redemption mechanics.

Avoid if:

Treating this as a “core” stablecoin allocation. apxUSD is a thin wrapper around a single off-chain security that has traded below par; the on-chain ticker reading $1 doesn’t remove concentration risk.
Needing instant guaranteed exit at $1 — the documented redemption mechanism is opaque, secondary depth is issuer-discretionary, and although the collateral ratio has held par or better since 2026-09-17, the market price has not: apxUSD has never traded at or above $1.00 in 722 readings.
Needing a Big-4 audited financial statement — what’s published is a monthly Wolf & Company AICPA examination (real CPA-firm opinion, but mid-tier and currently scoped to securities only), not a full financial-statement audit.

What to watch

Collateral ratio back to ≥100%, sustained. The single most important recovery signal — live on the dashboard’s Backing panel. A sustained return to par (not a one-snapshot blip) is what would justify re-rating.
STRC price + the soft-floor bid. STRC closed at $99.88 on October 12, while Strategy repurchased 1,932,145 shares for $192.4M at about $99.58. The direct collateral mark improved, but only $354.8M of authorization remains and USD Cash fell to $640.5M, $140.5M above the nearest trigger. Exhaustion alone is not automatic, since the board can re-authorize, but no Apyx reserve was restocked by this disclosure. The issuer release is live; EDGAR still ends at the October 5 8-K, so no October 12 accession is cited.
Wolf attestations — whether cash ever returns to scope. The June 2026 examination (signed 07-22) again covered securities only. Each further securities-only report leaves the cash sleeve without CPA coverage for longer; a return to the March full-balance scope would restore the disclosure stack to its high-water mark.
Curve apxUSD/USDC pool depth and discount (live on dashboard). Recovering depth and a narrowing discount = healing; widening = renewed stress.
MSTR / BTC drawdowns. A severe BTC crash compresses MSTR equity → threatens STRC dividends → degrades apxUSD backing further.

A note on the apyUSD sibling

If you’re considering the yield-bearing apyUSD wrapper, see the apyUSD report. apyUSD captures the dividend stream (≈13% APY ongoing) but exits through a 3-to-20-day unlock window with a declining fee (and now sits on top of the same below-par collateral). The two products are claims against the same Apyx + STRC backing — holding both doesn’t diversify.

Revision history

Update (2026-08-18) — the gap has closed most of the way, and the scores do not move. Peg 2.5 / Backing 2.0 / Liquidity 3.0 / Issuer 5.0 / Overall 3.0, all held.

The figures in the sections below describe a materially worse position than the current one, and the correction matters in both directions. Current state, re-read 2026-08-23 (feed timestamp 09:45:59Z) there is one shortfall, not two: $4,399,857.64 against supply of $312,073,514.82. That is Accountable’s own surplus_usd figure, and it is simply supply minus reserves of $307,673,657.18.

Update (2026-06-29/30) — scores cut to current: Peg 3.5 → 2.5, Backing 2.5 → 2.0, Liquidity 3.5 → 3.0, Overall 3.6 → 3.0 (Issuer held 5.0). Plus: the STRC collateral now carries a discretionary soft floor. Two things this catch-up reflects. (1) The depeg deepened through late June. What earlier sections describe as a “low-90s on the dollar / mid-90s% collateral ratio” was the first-week trough, not the current state — apxUSD has since traded as low as ~$0.82 (≈−18%) and Apyx’s attestation feed fell to a collateral ratio of ~84% (a ~16% buffer deficit vs par), sustained for weeks. That drives the score cuts to the levels above (live values on the dashboard). (2) Strategy’s 06-29 8-K put a soft floor under STRC. The “Digital Credit Capital Framework” changed how Strategy defends STRC — from a reflexive “hike the dividend if STRC trades sub-$95” ratchet to a discretionary soft floor (Strategy may buy STRC back via a $1.0B STRC-priority program, or draw a near-doubled $2.55B reserve, but makes no commitment to return STRC to par). For apxUSD’s ~74%-net STRC basket this is modestly stabilizing, not score-improving: a discretionary issuer bid now cushions STRC’s mark, but it does not restock Apyx’s reserve or lift the ~84% collateral ratio. Mark the STRC sleeve to its live secondary price with that floor beneath it, not to a par-defense recovery. The old “watch for STRC back above ~$95” recovery trigger is superseded — recovery now tracks STRC re-rating on the soft-floor bid plus Apyx restoring collateral ≥100%.

2026-10-04 — the recovery gate is met; Stability 2.5 → 3.5, Backing 2.0 → 3.0, Liquidity & Exit 3.0 → 3.5, Overall 3.0 → 3.5 (Dependencies 3.0, Contract & Admin 4.5, Issuer 5.0 held). The attested collateral ratio holds 101.5602%, a $2,749,710 surplus, at or above par continuously since 2026-09-17 across 393 hourly readings, and STRC marks $99.41. ⚠️ Three things hold the re-rate down and are not incidental. apxUSD has never traded at or above $1.00 in 722 readings (max $0.994716), so the discount outlived the backing recovery — which is what no atomic redemption looks like. Backing’s gain comes from the divisor: the coverage hole and the “no buffer” condition both dropped out, while the look-through baseline fell 5.7 → 4.71 as the collateral’s own score weakened. And measured depth is flat — derived mean 5.613 before 09-22 against 5.570 after — with both ladders censored at $100,000 (is_floor: true), so Liquidity moves half a notch for a removed distortion rather than for depth gained.
2026-08-23 — collateral figures re-read; scores held (Peg 2.5 / Backing 2.0 / Liquidity 3.0 / Issuer 5.0). The attested collateral ratio reads 98.017743% — a netted figure excluding $50.20M of protocol-owned liquidity and $39.92M of minted-but-unsold inventory from both sides, published beside gross totals. The shortfall on gross supply is $4,399,857.64; an independent lower bound reads 98.57%. The recovery gate is par or better on the attested feed, sustained. ⚠️ That gate was not met at this date, and the early readings that looked like it were noise: the ratio crossed par 21 times between 09-04 and 09-17. The clean break is 2026-09-17 — see the 2026-10-04 entry. What can be said is that the condition has started being met rather than remaining unmet.
2026-08-18 — collateral and peg figures refreshed; scores held.
2026-07-13 — STRC weekly 8-K anchor refresh; scores held.
2026-06-29 — STRC soft-floor reframe following Strategy’s 8-K.
2026-06-25 — depeg deepened.

Live backing dashboard

Live data Open in new tab ↗

External dashboard

Open Live backing dashboard ↗

Open live dashboard