Conversation

3) "sure", you say, "but that's only because people have to!" This doesn't turn out to be true. FTX gets constant requests to have futures priced in USDT rather than USD. We don't--for many of the reasons

@MonetSupply

says--though we do accept USDT collateral.

4) And if you travel to Asia, you'll see an interesting pattern: lots of clients actually trust USDT _more_ than USDC, or TUSD, etc. Why? Because rather than focusing on the banking part of it--something that's tough for a lot of the world anyway--they look at liquidity

5) It's a *lot* easier to get market liquidity for USDT than other stablecoins. If you don't trust your *own* banking to create/redeem, you might end up prefering USDT's liquidity to other stablecoins.

6) For a while FTX only had USD spot markets. We recently added USDT ones and on day one they traded as much as our USD markets did.

7) Now for

's purpose it makes sense to care primarily about crash risk. And on that

@MonetSupply

is basically right--USDT's crash risk is probably greater than USDC's. And that is a real argument for taking down the collateral factor for it.

8) But this has to be traded off against utility, and the truth is that by not supporting utility you're not supporting the majority of the crypto ecosystem. (And, not a huge deal, but USDT is way more decentralized than USDC.)

9) So I think a reasonable compromise here is to set USDT's collateral factor to 0.6 or so.

10) Note that if USDT starts to depeg, 40% is still a long way to go! There's lots of space for liquidations, and USDT has lots of liquidity on market. And the truth is, it *also* has real liquidity off-market. There is a bank account, with $ in it; you can redeem for $.

11) It's not as cheap or fast or easy as redeeming USDC/TUSD/PAX. I really appreciate a lot of what the folks at BFX and USDT have done for the ecosystem--supported most other exchanges' banking for years!--but if I had one suggestion it would be to furthermake redemptions chill

2:50 AM · Jun 28, 2020Twitter Web App

12) But you can do it if you need to, and some have in fact done it, and done it for size. Which doesn't mean it's always perfect, but it's better than the horror stories that are sometimes rumored.

Replying to

13) I've edited my proposal to suggest setting the collateral factor for USDT to 0.60--I think that probably is a compromise worth making. Huge thanks to everyone--particularly

@MonetSupply

--for all the feedback; this is the third time my post has gotten better because of it.