Skip to content
Digital Asset Summit New York 2026

Multichain Collateral & Risk Management: A Demo from Espresso & AWS | DAS NYC 2026 | Day 1 | Inst.

Ben Fisch, Ellie Davidson, Everton Fraga

Session summary

Ben Fisch of Espresso Systems presents a live demo of a multichain collateral and risk management setup built on the Espresso Network in collaboration with AWS. Fisch argues that institutional finance will be inherently multichain and that fragmentation and slow settlement finality are key obstacles to interoperable on‑chain markets. He describes Espresso as a fast, chain‑agnostic settlement layer that today offers roughly three‑second finality, with an upgrade in development targeting sub‑second, around 500‑millisecond finality, while supporting customized execution environments across many chains.

Ellie Davidson explains that the demo simulates five chains: four host tokenized assets and one acts as a collateral engine chain evaluating around 10,000 collateral agreements and about 1.6 million eligible assets, with continuous risk recalculations. In a simulated 20% equity price crash, the engine reads settled state from Espresso and initiates atomic cross‑chain actions such as pledging, substituting, or freezing collateral in real time. Everton Fraga details how AWS compute, including C‑family instances and Nitro Enclaves, is used to run millions of price and position recalculations roughly every 300 milliseconds at an estimated infrastructure cost of about $300 per day in this demo, which Fisch contrasts with much higher projected costs on other chains.

Key takeaways

  • Ben Fisch claims the Espresso Network currently achieves around three‑second transaction finality and is being upgraded toward sub‑second (about 500‑millisecond) finality while remaining chain‑agnostic and compatible with multiple virtual machines and custom chains.
  • According to Fisch and Ellie Davidson, the demo runs five chains on Espresso, with about 10,000 simulated collateral agreements and roughly 1.6 million eligible assets, and a dedicated collateral engine chain performing millions of price and position recalculations approximately every 300 milliseconds.
  • Fisch and Everton Fraga state that the high‑performance AWS‑based setup for this demo costs on the order of $300 per day, and Fisch contrasts this with a projected cost of at least $26,000 per day for a similar workload on Base, while suggesting it would be impractically slow on Ethereum.