> For the complete documentation index, see [llms.txt](https://docs.aquaflux.pro/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.aquaflux.pro/the-primitives/risk-waterfall.md). # Risk Waterfall How returns are paid and how losses are absorbed ### TL;DR * **Returns flow top-down:** cash first fulfills **P (principal)**, then **C (coupons)**, and any **surplus** flows to **S**. * **Losses flow bottom-up:** **S** absorbs losses **first** (the airbag), then **C** is reduced, and **P** is only touched in **extreme** cases.

*** ### Two flows, One System #### 1) Distribution of Returns (when money comes in) 1. **P-Token** – Principal repayments are honored first. 2. **C-Token** – Scheduled interest payments are distributed next. 3. **S-Token** – eceives its **coupon/fee share** and, at maturity, any **unused safety buffer** (residual). > Think of it like a waterfall flowing over three ledges. The top pool (P) fills first. Overflow goes to the second pool (C). Whatever is left cascades down to the third pool (S). #### 2) Protection Against Losses (when default) When the underlying asset loses value or defaults, losses are absorbed in the **opposite** direction: 1. **S-Token loses first** – The first-loss buffer absorbs defaults and shortfalls, up to its full coverage. 2. **C-Token is next** – If losses exceed S, C holders take a pro-rata haircut (everyone takes the same percentage cut). 3. **P-Token last** – Only if losses remain after S and C are both exhausted would P be affected. > Think of S as the **airbag** in your car. In a small accident, the airbag absorbs the impact and you walk away fine. In a bigger crash, the airbag still helps, but the car (C) takes some damage. Only in a catastrophic wreck does the cabin (P) get compromised. ### Examples **Setup:** $100 pool. S coverage is 10%, meaning there is a $10 first-loss buffer. #### Scenario 1: No Loss Everyone is happy. * **P** receives principal at maturity ($1.00 per token). * **C** receives scheduled coupons. * **S** receives its coupon share, fee share, and any unused portion of the $10 buffer at maturity. #### Scenario 2: $6 Loss A manageable loss -- the airbag handles it. * **S** absorbs **$6** of its $10 buffer. * **C** is **untouched**. * **P** is **untouched**. #### Scenario 3: $15 Loss A bigger hit -- the airbag is used up and C takes some damage. * **S** absorbs **$10** (fully used up). * **C** takes a **$5 haircut** (pro-rata across all C holders). * **P** is still **untouched**. #### Scenario 4: $28 Loss A severe event -- damage reaches beyond S and C. * **S** absorbs **$10** (fully used up). * **C** absorbs **$18** (pro-rata haircut). * **P** would be affected by any remaining loss beyond that. ### Is This a Tranche? Short answer: yes, but it’s more than that. In a traditional securitization, tranches only determine **who takes losses first**. They are static slices that you buy and hold. In AquaFlux, **P / C / S** are cash-flow layers with loss priority **that you can compose**. In other words, the Tri-Token model is not just risk slicing, it lets users **shape their payoff curve**, not merely their place in the waterfall. Check out the [Composable LEGO strategies](https://docs.aquaflux.pro/the-primitives/pages/nU3HJcfBRXvvNSA8Puy7#id-3.-composable-lego) for more. ### FAQ **What does “haircut pro-rata” mean?**\ Every holder in that token takes the **same percentage** reduction. **Can P be affected?**\ Only in **severe** loss scenarios **after** S and **after** C are reduced. **Does S always get residuals?**\ Only if the asset repays in full and the safety buffer is **unused** (per pool rules). **Do these rules ever change?**\ Each pool defines its own parameters (coverage, fee shares, schedules). Always check the pool’s settings in the dApp.