Case #1

If Powell prints $20T (or roughly 1x GDP, 1.3x M1 or roughly 1x M2), or about 4.4x the number of Treasuries held by the Fed right now, inflation would skyrocket, treasuries would sell off (massively), yields would presumably increase (as investors would start to refuse to hold treasuries which are paying out in inflated USD).

A few things would happen. USDT would face some liquidity problems because yields would fall reducing face value of treasuries. USDT holders would presumably exit to Bitcoin or something similar (unless USDT hiked yields to holders), because dollar inflation would be Argentina level, so USDT would have to process a lot of redemptions. But they would have duration issues because their bond portfolio would have sold off, so they would not have the cash immediately to make large redemptions. If they redemptions took place over 90 days, they would have sufficient funds to give investors back their (devalued) dollars. Note: this is the same thing that happened to SVB, but in Tether’s case, we’d be dealing with a bigger shock and so even 3-month treasuries would be affected.

Inflation would become extremely high and variable. The US government would have to engage in Yield Curve Control because the market rate of interest that holders of debt would demand would render the government immediately insolvent. Interest payments on the debt would climb to 30-50% of GDP. It’s complex to determine what inflation would be exactly, the problem is inflation expectations would become unanchored (because dollar holders would lose trust in the Fed), so it’s likely that inflation would be higher than 200% for a one year period. Let’s hope Case 1 doesn’t happen.

Case #2

In this case, half the money supply is backed by ETH (via Liquity), half is backed by treasuries. I would agree that Liquity holders would be largely immune from inflationary pressures, however the Liquity protocol still tracks the dollar, so it’s worth exploring whether Liquity would choose to devalue their UoA as dollar inflation skyrocketed, or simply move to a flatcoin/inflation indexed Unit of Account. My expectation is that Liquity if faced with a dollar devaluation, would move towards a different UoA. The inflation discussion is obscure again, because these things are nonlinear and can’t be derived like normal equations. Just look at how hard a time the Fed has prediction inflation. Even in our simplified model it’s hard to predict.

However, because the economy in this case would have a liquid alternative to Treasuries as a collateral asset, in this example, I think capital would flee to ETH/Liquity, making inflation worse for dollars stuck in the legacy system (by driving up velocity). Because ETH in this example would be big enough to absorb this capital inflow, this would leave dollar holders worse off inflation-wise.

Case #3

The difference between case 2 and 3 is the scalability of Liquity and Ethena. Even though Ethena is a more interesting protocol to me (because it’s based on a tokenized funding trade plus staking versus mere overcollat ETH), it’s less scalable, since the yield is based on the demand for long-side leverage on ETH. One Ethena grows “too much” it suppresses yields on ETH staking and pushes ETH funding rates negative. So Ethena is fundamentally less scalable as a stablecoin than Liquity (keep in mind we’re talking about tens of trillions of dollars here though). But it certainly true that if Ethena becomes a total juggernaut (and I’m talking 50b+ supply), it would be stacking on a ton of short-side liquidity, meaning funding rates would go negative. Of course, normally, this would causes USDe yields to fall, reducing demand for Ethena, but in your example, Ethena is super massive. So in conclusion, in the scale of trillions, Ethena would have a negative interest rate, since it would have to pay to be short (assuming Ethereum also isn’t in this case worth $10T+ with abundant demand for leverage or staking). So Ethena doesn’t look as attractive relative to Treasuries/USD in this case relative to Liquity. I think it would also be perceived as a less safe asset than an overcollat stablecoin, so I think less capital would flee the dollar system == lower inflation.