yzPP — Risk Report

High risk · 2.5/10

What this actually is

yzPP — the Yuzu Protection Pool — is a first-loss junior tranche. An ERC-4626 vault on Plasma, denominated in USDT0, whose function is to absorb losses before the collateral pool behind Yuzu’s stablecoin does.

⚠️ It is designed to lose money before anyone else does. That is not a flaw in the product; it is the product. Losses hit yzPP first, and that is what allows yzUSD to hold a collateralization ratio above 100%.

So the figures below do not read the way the same figures would read on a stablecoin. ⚠️ A drawdown here is evidence the structure functioned — the senior was protected, which is precisely what the junior is paid 29% to do. If you take one thing from this page before the numbers start, take that.

Do not confuse yzPP with yzPRIME. They are different books with different ratios, and Yuzu runs three products with adjacent tickers.

Yield	Denomination	Primary exit	Secondary exit	Chain
29% headline, discretionary	USDT0 (not yzUSD)	30-day window, 5,000 yzPP minimum, accreditation required	Open to anyone	Plasma

1 · Stability

Reference: NAV. yzPP accrues; it has no peg. Live NAV is 1.184000, read as totalAssets/totalSupply on chain against the feed’s 1.183954 — a gap explained by its 15-minute cycle.

The loss absorption has been observed once. Per Yuzu’s published history, drawn from Accountable’s timeline:

2026-04-23   NAV 1.137380   TVL 4,029,415      <- peak
2026-04-30                                     <- deepest drawdown, -3.74% from peak
2026-05-01   NAV 1.095482   TVL 3,275,693      -3.68% vs 04-23
syzUSD APY   7.0% on 04-25 · 7.0% on 05-01 · 7.0% on 05-07   UNCHANGED
recovery above the prior peak: early July 2026, roughly two months

The junior absorbed and the senior’s yield did not move. That is the waterfall doing what it says.

⚠️ Sourcing, stated plainly: the current NAV is chain-confirmed and the April event is not. The drawdown comes from the issuer’s published history, and verifying it independently needs historical vault reads that Plasma’s 10,000-block eth_getLogs cap makes expensive. We have not measured it, and this report does not claim to have.

⚠️ And −3.74% is not a bound. It is one observation of a mechanism with no documented floor, on a claim against a book that is roughly 70% levered with Ethena at 34.7%. A holder who reads it as a worst case has read a sample as a limit.

2 · Backing

The claim is on the same Yuzu strategy book that backs yzUSD — the composition, the coverage and the attestation channel are identical, and they are treated in full there. The reserve is not different at this layer; the position in it is.

⚠️ The subordination is deliberately not scored here. Being first in the waterfall is a fact about position rather than about what the reserve holds, and it is scored under axis 4 where the circularity already sits. Scoring it in both places would count it twice.

⚠️ One figure from that book governs this page: the surplus is $4,049,785, and yzPP is $4,729,857 — 116.8% of it.

3 · Liquidity & Exit

Both exit paths, and the axis takes the worse one. Here both are bad.

Primary — accredited-only, and the term is the sharpest thing on this page. A 30-day redemption window with a 5,000 yzPP minimum and accreditation required. ⚠️ First-loss exposure continues for the entire 30 days. An exit request is not an exit: you carry the risk you are trying to leave, for a month.

Secondary — open to anyone, and shallow. ⚠️ 133 holders on Plasmascan. That is the route a retail buyer would actually take, and it carries no gating and no disclosure at all — the instrument’s own eligibility rules do not reach the venue where it is most likely to be bought.

⚠️ The two legs fail in opposite directions and that is why this axis is 2.0. The gated leg is slow; the open leg is thin. Neither is a way out at size, and the gate that makes the primary leg respectable is exactly the one the secondary route removes.

4 · Dependencies

⚠️ This is the worst axis in the Yuzu family, at 2.0 against syzUSD’s 2.5 and yzUSD’s 3.0, and the ordering is the finding. All three depend on one strategy book. syzUSD passes through to it. yzUSD is it. yzPP absorbs its losses first.

So every exposure below arrives here before it arrives anywhere else:

Exposure in the underlying book	Share	Reaches yzPP
Levered loop positions	~70%	First
Ethena (USDe + sUSDe)	34.7%	First
`yzPRIME` (Yuzu’s own product)	4.8%	First

⚠️ And the position is circular. yzPP’s value is a claim on the very book it cushions, so it stops being a cushion precisely when the book takes losses — the moment the protection is needed is the moment the protection is impaired.

⚠️ Read the other way, this is a risk to the SENIOR asset, and that is the more important direction. yzUSD’s 6.92% overcollateralization is not retained earnings — it is redeemable third-party capital. yzPP is 116.8% of the entire surplus, so the cushion a yzUSD holder relies on is money belonging to yzPP holders, subject to their redemption rights rather than locked in place. Ex-junior, the senior’s ratio is 100.87%, not 108.96%.

⚠️ No cap, attachment point or detachment point is documented anywhere. The protocol describes the waterfall qualitatively and states no threshold at which absorption stops. So 8.09% is an observation about today’s balances, not a disclosed parameter — nothing commits Yuzu to maintaining it, and nothing tells a holder how thin it may become.

yzPP                $4,729,857    = 8.09% of yzUSD supply   (first-loss thickness)
yzUSD surplus       $4,049,785    -> yzPP is 116.8% OF THE ENTIRE SURPLUS
+ Reserve Fund        $503,429    -> junior + reserve = 8.95% of yzUSD supply

5 · Contract & Admin

⚠️ This is where yzPP is the strongest of the Yuzu products, and a reader comparing the three pages would otherwise assume the reverse.

Both owner() and the ProxyAdmin sit behind the 48-hour timelock, and the bare EOA that controls the syzUSD bridge holds no role here. That is why this axis scores 4.0 against syzUSD’s 2.0.

Both assets land at 2.5 overall, for opposite reasons. ⚠️ syzUSD’s is a bare-key problem; yzPP’s is subordination plus a gated exit. The contracts here are better run and the position is worse.

6 · Issuer

The issuer is a named entity — YUZU FINTECH LLC, a registered Virtual Asset Service Provider in the Republic of Georgia — which makes an accreditation gate coherent rather than anomalous.

⚠️ But registration is not prudential supervision, and the operators are not named. No disclosed CEO or founders, so the accreditation gate is enforced by people a holder cannot identify.

⚠️ The attestation channel is related-party: Accountable’s founder is a Yuzu investor. The April loss-absorption event above rests entirely on that source.

⚠️ The headline yield is not underwritable. 29% is syzUSD’s base rate plus a discretionary supplement set daily at 04:00 UTC under an unpublished risk-premiums policy, part-funded by a Reserve Fund holding $503,429 — 0.86% of yzUSD supply. A rate set under a policy nobody outside can read is not a rate a holder can model.

Who should avoid this

Anyone who wants yield without being the first to absorb losses. That is the entire function of this position (axis 4).
Anyone who might need the money inside a month. The window is 30 days and the exposure runs through all of it (axis 3).
Anyone buying on secondary because primary was gated. The gate exists for a reason the secondary route does not remove (axes 3 and 6).
Anyone relying on the −3.74% drawdown as a worst case. There is no documented floor (axis 1).
Holders of yzUSD who have not read this page. The senior’s cushion is this tranche’s redeemable capital (axis 4).

What to watch

The first-loss thickness as a share of senior supply. 8.09% today, undisclosed as a target, and nothing commits the protocol to it (axis 4).
Whether an attachment or detachment point is ever published. Its absence is the single largest disclosure gap here (axes 4 and 6).
The daily supplement. A discretionary rate set under an unpublished policy is the first thing to move under stress (axis 6).
Secondary depth on Plasma. 133 holders is the exit for anyone who did not come through the primary gate (axis 3).

Revision history: 2026-08-29 — first publication of this asset, staged. yzPP is Yuzu’s third product and this coverage had none. ⚠️ Written so the tranche’s purpose lands before any figure, because every number here reads as deterioration if the reader does not already know that absorbing losses first is what the instrument does — a −3.68% NAV move on a first-loss tranche is the structure functioning, not failing. The finding worth leading on is about the senior asset: yzPP is 116.8% of yzUSD’s entire surplus, so yzUSD’s 6.92% overcollateralization is redeemable third-party capital rather than retained earnings — a liability to holders who can ask for it back — and no cap, attachment or detachment point is documented, making 8.09% an observation rather than a commitment. The April 2026 loss-absorption event is attributed to the issuer’s published history and is explicitly not chain-confirmed; the live NAV of 1.184 is. Structural is 4.0 — better than syzUSD’s 2.0 — because owner() and the ProxyAdmin both sit behind the 48-hour timelock and the syzUSD bridge EOA holds no role here; the identical 2.5 overall reflects subordination and a gated exit rather than a key problem. Held at production: false pending the site owner’s decision on whether this page shares the Yuzu dashboard gate.