On Stablecoins, AFTs, and Product-Market Fit
As the Terra Classic Community thinks about what kind of stablecoin or stablecoin-like asset it wants to underwrite, it helps to understand the current competitive landscape for stablecoins and stablecoin-adjacent token types, like algostables and Algorithmic Fungible Tokens (AFTs), and how USTC would fit into the existing competitive landscape.
I’m a stablecoin nerd whose day job revolves around blockchain tokenization and financing of real-world assets. I spent a year studying all the different stablecoin models so you don’t have to. (One of the reasons I became so interested in Terra was because of how perfectly suited it is for real-world asset borrowers, but we’ll leave that for later.)
The market for stablecoins and stablecoin-adjacent fungible tokens is has at least 3 major types. Each has its own strengths, weaknesses, and product-market fits.
100%-USD-collateralized stablecoins (USDC, USDT, BUSD, others)
Feature set: 1 USD gets deposited <=> 1 USDC/T token gets minted; closely regulated; umbilical cord to a regulated bank to facilitate large-scale USD movement; 100% token-to-fiat collateralization requirement; publicly marketed as 1:1 redeemable with USD
Pros: bulletproof from bank runs (assuming they’re managed honestly), well-suited for regulated use cases in which fiat law has high leverage over the transaction (e.g., underwriting a mortgage), total convertibility (no restrictions on amounts of money that can be moved), most-tolerated by major regulators
Cons: zero censorship resistance, 100% centralization, 100% dependency on an umbilical cord with a US-approved bank which could revoke that permission at any time; no real difference from fiat USD, aside from the ability to send it cross-chain instead of via bank wire.*
*Tether does have significant censorship-resistance street cred by virtue of surviving a bloody battle with US regulators several years ago and having its USD umbilical cord with a non-US bank, and has permanent patronage from Asian CEXs who want more distance from the US government’s regulatory unpredictability.
2. Overcollateralized crypto- or asset-backed private stablecoins (DAI, RAI, USK, Aave’s GHO and Curve’s stablecoin soon)
Feature set: Very high (160%+) exogenous collateralization requirements (BTC, ETH, or some other asset not under that protocol’s control); near-total convertibility with USD; often but not always marketed as 1:1 convertible with USD
Pros: Fairly censorship-resistant (unless fiat-backed stablecoins are in the collateral pool, in which case their censorship resistance quickly evaporates), proven ability to maintain a stable USD currency peg even under very high macro/crypto stress
Cons: Very limited growth potential; requires someone to vault $1.60 or more of crypto-assets for every $1 of overcollat-stable in circulation, or in essence, be given a free lunch by depositors. As a result, these stablecoins never really scale.
3. Undercollateralized algorithmic stablecoins (UST, FRAX, AMPL)
Feature set: <100% exogenous collateralization requirements, algorithmic capital controls; historically, heavily marketed as being 1:1 convertible with USD. Zero umbilical cord with a USD bank.
Pros: very high censorship resistance (if collateral != USDC or other fiat stablecoin), very powerful growth potential because collateralization requirements are below 100 percent, ideal for borrowers (because loans could be redeemed at a discount during stress)
Cons: highly experimental, poor track record, hunted by regulators, difficult to predict how the system can handle stress as it becomes truly decentralized/distributed; weaker convertibility with USD than other stablecoin classes. No truly distributed algostable has maintained its original model and survived
4. Algorithmic Fungible Tokens (USTC, under this plan)
Feature set: similar to algostables, but never marketed as 1:1 convertible with USD (even though it will do its best to have that capability at all times, we must be extremely clear that this cannot be the case 100% of the time); probably accepts slightly more short-term peg volatility for higher longer-term resilience than algostables historically have. Zero umbilical cord with a regulated bank.
Pros: maximum censorship resistance, far more systemic resilience than UST, higher growth potential than any stablecoin currently on the market, an innovative reserve system which makes catastrophe extremely unlikely
Cons: on the experimental bleeding-edge; will take years to rebuild trust
So, we can make a few conclusions.
We shouldn’t try to compete with the 100%-USD-collateralized regulated stablecoins. We have no competitive advantage in that arena and wouldn’t be filling any unmet market need in doing so. We also shouldn’t try to be “wrapped USDC” (i.e. an algostable that is basically 1:1 exchangeable with a stablecoin like USDC) as we would simply be a much more inefficient derivative of that fiatcoin, with idiosyncratic Terra-specific risks.
The market for overcollateralized asset-backed stablecoins is crowded, is getting more crowded, and has weak structural growth characteristics. These coins meet a significant need for whales’ capital-gains tax arbitrage (borrowing at a very conservative LTV against a lot of crypto to access some liquidity, as opposed to selling that crypto and being liable for high capital gains tax) but otherwise are extremely capital-inefficient.
The market for undercollateralized algostables and AFTs remains wide open. Yes, it’s open because many have tried and failed before. But it’s also where the innovation alpha lies, and no brand better captures the highs and lows of that ambition than Terra Classic’s does. Succeeding here would be a true innovation, bringing to market something that’s completely missing from today’s competitive landscape.
Conclusion: Rebuilding USTC as an undercollateralized AFT with purely decentralized collateral represents the best product-market fit for USTC, and thus the best underwriting decision for LUNC holders — if we can dramatically improve resiliency and also render a failure non-catastrophic.
Postscript: Stablecoins and taxes
Stablecoins are an insanely competitive market.
The biggest stablecoin market is a decentralized exchange called Curve.
Curve introduced the concept of leveraged liquidity pools for assets that “should trade at or near 1:1 at all times” — stablecoins-to-stablecoin pairs and token-to-its-liquid-staking-derivative (eg, ETH-stETH) pairs. This allowed Curve to offer insanely slow spreads on stablecoin swaps, and basically set the benchmark for competitive stablecoin exchange rates.
In the above USDT <> USDC transaction, you can swap $100,000 USDT for $99,990 USDC on Curve, including fees (but not including ETH gas). This is how competitive the stablecoin market is; every single basis point counts, and there’s simply no room for a tax of any size (let alone 1.2%) at that level. Doing so would make USTC a completely uncompetitive product in the stablecoin market, and would thus be an extremely poor underwriting decision for LUNC holders today.
Therefore, the Quant Team believes that any launch of a re-pegged stablecoin is mutually exclusive with the proposed USTC burn tax, and future Quant Team proposals will assume that the USTC portion of the burn tax does not exist.
Now, that doesn’t mean there’s no room for taxes at all. Terra had a 2% UST-LUNC minimum spread until 9/2/2020, when the rate decreased to .5% to facilitate more transactions. (The .5% swap fee served a very useful role of preventing oracle price manipulations and in my opinion should be kept.) Terra also had a .25% Tobin tax (~1/5 the proposed 1.2% rate) on all transactions, which they experimentally raised to .35% later, and eventually abolished, in favor of facilitating more on-chain liquidity. TFL’s internal experiments, validated by third parties within the ecosystem, seemed to yield superior outcomes from low (but very surgical, on-chain-only) transaction taxes as opposed to high transaction taxes.