---
type: Estimate
title: Independent estimate of Senior loss to 18 January 2027
description: Beta-Binomial estimate of the Senior's probability of loss, expected loss and ES99 from the on-chain Ikon book, under design layers, an economic-budget view and the 300k srLoanLimit cap. Not Ensuro's numbers.
tags: [independent-estimate, model, risk, senior]
claim_basis: [independent-estimate]
status: draft
generated: { by: grok-bot/lp-bundle-v0.2, at: 2026-11-17T15:30:00-03:00 }
stale_after: 2026-12-01T00:00:00-03:00
sources:
  - id: model
    resource: /metrics/model/senior-risk-model.json
    title: Output of /scripts/senior_risk_model.py (snapshot block 26,436,335; Ikon aggregates 2026-11-12)
  - id: code
    resource: /scripts/senior_risk_model.py
    title: Model script (reproducible)
  - id: ensuro-1117
    resource: Ensuro team answers to LP due-diligence questions, 2026-11-17
    title: Ensuro team statements (unverified)
    author: team:ensuro
---

> **Independent estimate.** Produced for this bundle from on-chain data and stated assumptions. It is not Ensuro's model and Ensuro has not reviewed it. Ensuro's own loss metrics are **Pending**.

# Set-up[^model][^code]

- Book: the **7,674** Ikon policies on-chain (sum insured $8.74M); horizon to expiry **2027-01-18 06:59:59 UTC** (61.55 days from the snapshot).
- Final refund rate R (share of sum insured refunded) ~ **Beta-Binomial**(n = 7,674, mean p, dispersion κ). Base case: p = **5.2%** (Ensuro's mean), κ = **99** (pairwise correlation 1%). Refund pays 100% of sum insured (Ensuro statement).[^ensuro-1117]
- Variants:
  - **A – design layers:** Junior 8.7145–10%, Senior 10–22% of sum insured.
  - **B – A with the $300k `srLoanLimit` cap** on Senior loss (separate scenario; a reverted payout is not a cancelled liability).
  - **C – economic budget:** loss capacity before the Junior = Spot surplus + Ikon active pure premiums + refunds already paid ($653,890), Junior capped by `jrLoanLimit` $200k → Junior attaches at R ≈ 7.48%, Senior at ≈ 9.77%. Non-Ikon Spot business assumed to break even. **C + cap** adds the $300k limit.
- Excluded: policies not yet on-chain, non-Ikon losses, MSV/asset losses, smart-contract or operational failure, later loan repayment.

# Base-case results (Senior supply $1,381,589)

| | P(Senior hit) | Expected loss | ES99 | ES99.9 |
|---|---|---|---|---|
| A design | 3.20% | $3,872 (0.28%) | $256.5k (18.6%) | 36.3% |
| B design + cap | 3.20% | $3,604 (0.26%) | $229.7k (16.6%) | 21.7% |
| C economic | 3.74% | $4,574 (0.33%) | $276.8k (20.0%) | 37.8% |
| C + cap | 3.74% | $4,250 (0.31%) | $244.3k (17.7%) | 21.7% |

Refund-rate distribution (base): P(R > 8.7145%) 7.35%; P(R > 10%) 3.20%; 99th percentile 11.65%; 99.9th 14.58%. VaR99 (A) $144k.

# Sensitivities (variant A)

| Case | P(hit) | EL (% supply) | ES99 (% supply) |
|---|---|---|---|
| κ = 49 (correlation 2%) | 8.1% | 1.22% | 44% |
| mean 7% | 12.6% | 1.36% | 35% |
| mean 8.7145% (priced rate) | 29.8% | 4.07% | — |
| mean 9.01% | 33.5% | 4.75% | — |

Deterministic Senior losses (A): R = 11% → $87k; 12% → $175k; 15% → $437k (B: $300k); 18% → $699k; 22% → $1.049M (76% of supply).

With $500k of new deposits the same dollar losses are spread wider: base EL 0.21%, ES99 13.6% of the larger supply.

# Reading

The estimate is dominated by two inputs that cannot be verified on-chain: the mean refund rate and policy correlation (for example season-wide events). The on-chain refund rate to date (2.88%) is below the 5.2% mean but covers about one third of policies before the end of the season. See [risk-adjusted return](/07-independent-estimate/risk-adjusted-return.md).

[^model]: Senior risk model output
[^code]: Model script
[^ensuro-1117]: Ensuro team statements, 2026-11-17 (unverified)
