# \[RFC\] Arbitrum USDC Vault (OLD) User Reimbursement **URL:** \ **Category:** Request for Comment \[RFCs\]\ **Tags:** usdc, arbitrum, rfc, sumr, reimbursement\ **Created:** [December 18, 2025, 2:48pm UTC](https://forum.summer.fi/t/rfc-arbitrum-usdc-vault-old-user-reimbursement/582 "2025-12-18T14:48:20Z")\ **Posts on this page:** 1\ **Showing post:** 42
**Author:** ![jensei](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.summer.fi/jensei/32/69_2.png) [@jensei](https://forum.summer.fi/u/jensei)\ **Post date:** [February 2, 2026, 3:48pm UTC](https://forum.summer.fi/t/rfc-arbitrum-usdc-vault-old-user-reimbursement/582/42 "2026-02-02T15:48:21Z")
Thanks everyone for continuing to engage here; I hear the frustration clearly, and I want to respond constructively to a few recurring points, especially around _“acknowledge the debt and pay it over time”_. > **I want to be very explicit:** > > 1. I am **NOT** going to argue that users expected “risk-free” outcomes, nor that this vault failure should be ignored. > 2. I **DO** disagree with the framing that curated vaults imply a standing guarantee of reimbursement regardless of cause or feasibility. * * * `Risk management ≠ risk-free` Every DeFi user pays fees everywhere (Maker, Aave, Euler, Silo, Yearn, Balancer) and there is **very little precedent** where users were reimbursed when losses stemmed from external integrations and cascading failures, rather than a direct exploit or insolvency of the protocol itself (also rare). This incident was: - a third-party stable failure (USDX), - combined with oracle design limits at Silo, - not a direct Lazy Summer Protocol exploit. That distinction matters when we talk about precedent and long-term protocol design. * * * Several comments suggest allocating a small % of protocol revenue to reimburse users over time. This sounds reasonable, until we look at actual numbers. **Current protocol metrics (approx):** - Annualized revenue: **$419,660** - Current TVL: **~$63.4M** - Loss to cover (84% baseline): **~$1.26M** - Allocation scenarios: **1–3% of protocol revenue** ## Scenario A: Current TVL (~$63M) Annual revenue: **$419,660** - **1% allocation:** ~$4,200 / year → ~ **300 years** to reach $1.26M - **2% allocation:** ~$8,400 / year → ~ **150 years** - **3% allocation:** ~$12,600 / year → ~ **100 years** > _At current scale, this is clearly non-viable._ ## Scenario B: $120M TVL TVL ~1.9× current Annual revenue ≈ **$794,000** - **1%:** ~$7,940 / year → **~160 years** - **2%:** ~$15,880 / year → **~80 years** - **3%:** ~$23,820 / year → **~53 years** > _Still multiple decades, even at ~2× TVL._ ## Scenario C: $240M TVL TVL ~3.8× current Annual revenue ≈ **$1.59M** - **1%:** ~$15,900 / year → **~79 years** - **2%:** ~$31,800 / year → **~40 years** - **3%:** ~$47,700 / year → **~26 years** > _Strong growth, still measured in generations._ ## Scenario D: $480M TVL TVL ~7.6× current Annual revenue ≈ **$3.18M** - **1%:** ~$31,800 / year → **~40 years** - **2%:** ~$63,600 / year → **~20 years** - **3%:** ~$95,400 / year → **~13 years** > _Now we’re talking decades, but only after ~8× TVL growth._ ## Scenario E: $940M TVL TVL ~15× current Annual revenue ≈ **$6.22M** - **1%:** ~$62,200 / year → **~20 years** - **2%:** ~$124,400 / year → **~10 years** - **3%:** ~$186,600 / year → **~6–7 years** > _This is the **first scenario** where a revenue-based recovery becomes plausible within a human timeframe, and it requires ~ **$1B TVL** , sustained._ Even under **optimistic growth assumptions** , “programmatic reimbursement via fees” is measured in **decades** , not months or even a few years and this is why I keep pushing back on the idea that there is an **“obvious middle path”** being ignored. The middle path collapses under basic math. So when people say “just stream it over time”, what the math actually says is: > _Streaming only works if the protocol becomes much larger first._ That doesn’t mean **“never”**. But it does mean **“not now, and not without explicit growth conditions.”** It is important to clarify and highlight here that I was not assuming SUMR Token and its potential price increase in these TVL growth assumptions - but very briefly: | SUMR Price (USD) | SUMR Needed | % of DAO Treasury (~123.43M SUMR) | | --- | --- | --- | | 0.005598 | 224,200,143 | 182% → impossible | | 0.01017 | 123,425,821 | 100% → just feasible | | 0.03 | 41,849,667 | 34% → feasible | | 0.09 | 13,949,889 | 11% → very feasible | | 0.15 | 8,369,933 | 6.8% → minimal impact | Here it becomes quite clear that the potential for any reimbursement could be made only paired with the protocol growth and treasury SUMR value growing. * * * ### On acknowledgment vs. false certainty I fully agree with one thing some of you said: _“Postponement without structure feels like abandonment.”_ Where I disagree is that **committing to an unpayable obligation today** restores trust. In my view, that replaces one failure with another, promising something the DAO cannot realistically deliver. A credible approach must imo: - be conditional, - be explicitly feasibility-gated, - and not rely on revenue streams that mathematically cannot close the gap. * * * ### What a realistic framework _could_ look like If this discussion continues (and I believe it should), the only paths I see that don’t mislead anyone are **trigger-based** , not schedule-based. For example (illustrative, not prescriptive): **Level 1:** SUMR price sustains \> X Liquidity \> Y Treasury stable holdings \> Z → partial reimbursement unlock **Level 2:** Treasury reaches N% of loss amount in _non-SUMR assets_ → additional tranche **Level 3:** External recovery (e.g. Silo / USDX) materializes → debt receipts become withdrawable > _Until conditions like these exist, any “timeline” is unfortunately symbolic rather than executable._ ### On Silo responsibility (not deflection) Pointing to Silo is not about shifting blame, it’s about **where the debt actually lives**. > _Silo holds the debt. The DAO holds the receipts, and is actively scanning for any liquidity/backing restored._ If those positions are ever restored, that is the cleanest and least destructive recovery path for users. Pretending otherwise does not help affected lenders; it just concentrates risk onto a DAO that did **NOT** custody those funds. * * * I understand why hope is fading, that’s on all of us to acknowledge. But I don’t believe honesty about constraints is abandonment, nor do I believe that promising repayment paths that require 15–300 years restores trust. I remain open; genuinely; to **non-destructive mechanisms** that close this gap _within a realistic horizon_. If someone has numbers that show otherwise, I’m very willing to review them. What I don’t want is for this DAO to replace a painful loss with a second failure: committing to something that cannot be honored. I appreciate everyone who continues to engage, even when emotions are understandably high. --- _[View the full topic](https://forum.summer.fi/t/rfc-arbitrum-usdc-vault-old-user-reimbursement/582)._