Realized loss among current constituents is 0.00 percent. survivorship filtered. Impaired names are screened out before this is computed, so it reads near zero by construction. The cohort's real loss record is below.

The rulebook, run on screen

Every defaultable vault is screened against these. We mark each one as it actually stands, including the two that fail, rather than show a row of green ticks.

Constituents

The 30 loans that passed, each weighted by size, with the coupon each borrower pays. Loans below the base rate are relationship priced, flagged.

References are loan addresses, not yet mapped to borrower entities. One borrower is known to hold two of these loans.

Reconstitution watchlist

Live candidates under review for the next series. Adding any second venue clears the diversification rule this series fails. The full eligible universe, and why a name is in or out, is in the methodology.

Loss track record

What this market lost in the last cycle. One event verified on chain, the rest from public reporting, marked, not yet checked deal by deal. This is the loss the headline 0.00 percent does not show.

What this is, and what it is not

• One protocol. Every name is a Maple loan, so this carries Maple protocol and admin risk, not diversified protocol risk.
• 6 loans paying at or below the risk free rate, about $368M, were excluded as subsidized relationship paper. The published 216 basis point spread is the arm's length credit price, not a blend with zero coupon teasers.
• Defaults are read from Maple's own impairment flag. The independent, bonded determination layer is phase two.
• Names are loans, not deduplicated to borrowers, so borrower level breadth and concentration are not yet exact.
• Live book, coupons, and collateral status are verified on chain. The historical loss record is mostly public reporting, marked, not yet checked deal by deal.
• Below the 100 name breadth target a broad index would carry, because the live market is one protocol deep.