So we are investors in Ethena and I’ve known Guy for a while now. I did a podcast with him before we invested (https://onthebrink-podcast.com/ethena/) and one more recently (https://onthebrink-podcast.com/ethena-ii/). Additionally, on the matter of synthetic stables, I was an outspoken Luna critic, I diligenced Basis Cash (basecoin) for an investment back in 2017, and we’ve invested in various previous iterations of the ‘nakadollar’ on Bitcoin. So I have been thinking about synthetic stables forever. (You can see I reference them back in 2020 in my whitepaper cryptodollars: https://www.castleisland.vc/cryptodollars)

Previous experiments with synthetic USD failed because expectations were wrongly conditioned (i.e. founders wanted to hold the token at $1 at all costs, and were forced to abandon the system or take a loss when the basis went negative). Ethena is somewhat different in that the long side of the trade also pays a yield, via ETH staking (previous iterations focused on Bitcoin). StETH yields vary, but they are in the 4-6% range (although you might expect that Ethena’s entry into the market would further suppress that). So that does give you some margin of safety if funding goes negative.

The other thing that’s different about Ethena is that it’s envisioned as a crypto native interest rate that grows and shinks its supply in line with rates. So when funding rates fall, Ethena supply shinks (since it offers a less attractive yield), thus lifting the shorts (and unstaking stETH), both of which have the effect of pushing yields back up. Ethena is not trying to target a fixed yield, and so because they are willing to allow the yield to fluctuate, the system can be adaptive to market conditions.

Historically, the basis trade yielded 18% in 2021, -0.6% in 2022, and 7% in 2023 (and it’s tracking much higher in 2024). With stETH yields overlaid on the funding, Ethena would have paid a positive yield in all historical phases, even the bleakest portions of 2022. Historically (and this is a little questionable because Ethena’s presence changes yield dynamics, so you can’t just plug in historical data and get a reliable estimate), Ethena would have paid 21% in 2021, 5% in 2022, 13.3% in 2023, and right now it’s paying 33%.

The actual reason the yield exists is as follows. On the ETH side, there’s a yield because ETH is a vibrant place to transact with plenty of MEV opportunities. That won’t change. On the funding side, the yield exists because there’s generally in crypto a long-biased demand for leverage on futures and derivs exchanges, and not enough supply to accommodate this demand. This is in my view a market structure issue. If there was a single extremely trustworthy futures or perps exchange, and traders relied on a single large prime broker giving them credit lines at a few points above SOFR, I don’t think you’d see the same basis we see today. So in a sense, the persistent positive funding in crypto is due to counterparty risk assessments of these derivs exchanges. No one really wants to leave a ton of cash laying on one of these exchanges forever, and so the big lenders aren’t arbing away the basis (by going short futs and long spot), even if it’s paying 10 points+ more than SOFR. Crypto is a uniquely fragmented global market, with very disparate regulatory approaches, and a still significant “offshore” market for futs and perps, so I don’t see this changing any time soon. However, if things do become more formalized and come more onshore, then I expect the basis does over time collapse.

Ethena currently has an insurance fund (https://ethena-labs.gitbook.io/ethena-labs/solution-design/insurance-fund) , but it’s relatively small ($12m) relatively to the supply of USDe ($828m). (https://app.ethena.fi/dashboards/solvency) In theory, Ethena could ‘fill up’ the insurance fund in this high growth portion of the protocol and grow its capital ratio by skimming off a portion of the net interest margin. It’s yet to be determined how exactly the insurance fund would be utilized and in what conditions. The token issuance would presumably bolster the insurance fund, and Ethena itself can raise equity.

On negative funding and risks: as Ethena notes in their documents (https://ethena-labs.gitbook.io/ethena-labs/solution-overview/risks/funding-risk), they have only seen one quarter in the last three years where the stETH yield and the funding trade aggregated to a negative yield, and that was during the ETH PoW issuance period which polluted the data (by persuading a lot of people to go long spot and short futs, flipping funding).

Another risk worth considering is depegs in stETH derivatives that Ethena uses to collateralize its short positions. If stETH depegs (like it did in mid 2022), even though Ethena isn’t using more than 1x leverage, its collateral could decline in value, causing its positions to be liquidated. However, post Shapella, stETH doesn’t meaningfully depeg (because unstaking is possible).

Based on my investments in the space, I’ve long felt a ‘nakadollar’ was possible. The introduction of the positive carry on ETH with staking is the unlock that I felt made it meaningfully possible. That way holders could benefit from a somewhat uncorrelated portfolio of long stETH and long funding, with the stETH yields covering up shortfalls in funding. Long term, I expect yields to compress, both because more credit will make its way into futures and perps exchanges, as crypto prime brokerage gets more sophisticated, and because Ethena itself (and perhaps competitors) will grow, adding demand to stETH and funding markets, thus suppressing yield. However, there’s an equilibrium where USDe holders refuse to accept xx% yields above SOFR (let’s say it’s 500 bps), and so they simply stop creating units of USDe if yields aren’t sufficiently appetizing. This acts as a natural ceiling on the supply of the asset. And regarding the funding markets, I think there will always be a risk premium associated with leaving collateral on an exchange, and I think most traders will generally continue to be long-biased, so I do see a residual positive funding rate in the long term.