Contractibility is a less talked about stablecoin characteristic - it's not strictly necessary to construct something stable, but it can have drastic effects on end users. What is contractibility? 🧵

This is why, for ESDv2, we ultimately chose to use a deterministic model, while additionally providing a framework to allow governance to safely experiment with nondeterministic models over time. 💔

This isn't to say nondeterminism is bad - this is probably the most exciting area of study in DeFi 🧠 - but it's still early and it's risky.

However Dai, Rai, and Frax are nondeterministic, additionally susceptible to black swan events. MakerDAO famously went temporarily insolvent during Black Thursday last year, and Frax had a short period of under-collateralization (incl. share value) earlier this year.

Now, even within this bucket, there's diversity in risk profiles. USDC, USDT, and ESDv2 will be deterministically over-collateralized - meaning their primary risk is traditional smart contract bugs (or bank account bugs). 🐛

Despite its branding - @fraxfinance also fits into this bucket, over-collateralized by the sum of its reserves and the market value of its seignorage shares. And finally, @emptysetdollar V2 will be over-collateralized by its reserves directly at launch. 🏦

Over-collateralization. The market needs to be able to exit a majority of the stablecoin all at the same time. ✅USDC and USDT allow this by centrally redeeming to USD ✅@MakerDAO and @reflexerfinance enable this by allowing users to pay back and close their vaults at any time

These both create a stable asset - but one that can rarely be exited at par. So, how do you fix it? 🔧

Similarly this problem plagues direct incentive mechanisms like @feiprotocol that rely on disincentivizing the most common behavior in the system to smooth out supply and demand.

Low contractibility was a major downside of all un-collateralized stablecoins (especially ESDv1), as these used circulating supply as the main lever for stability.

🪢Contractibility allows the supply of the stablecoin to significantly contract in a short period of time. Or in user-centric terms: can a large amount of users exit the stablecoin at the same time, or must the protocol restrict this to maintain stability?

With ESD, our goal was simply to create the best stablecoin. We knew it wouldn't be perfect on the first try, it was more important to build an open platform for collective experimentation. Couldn't be more excited about the vision Scott and Will have put forward. 🏦

🔗 https://t.co/b6RpilAqu9 now links directly to the latest IPFS build.

introducing døllar and the {ess}. https://t.co/miCRoC2xNZ

to translate to ethereum, we incorporate ideas from 0x staking, rho, and build on top of uniswap v2 oracles. 3/3

its construction is an iteration on basis and its predecessors, aiming to solve core issues within their mechanics. 2/3

døllar is a fully decentralized elastic-supply stablecoin. 1/3

the mechapunks will decentralize finance.