Our primary objectives while designing xBTC as a crypto-native stablecoin were as follows:
(1) Compute a stable USD redemption rate for a coin backed by 1 BTC, such that the value of the collateral (bitcoin) never falls below the redemption rate (with >99.9% certainty in backtests).
(2) Subject to (1), we want to optimise the value appreciation in the redemption rate, to the maximum extent possible — so that users can be safely exposed to the long term positive appreciation seen with Bitcoin, with minimal volatility.
Achieving the first goal independently is trivial — just setting the redemption rate to something like $10 basically guarantees collateral sufficiency. However that would be an extremely inefficient utilisation of capital.
Trying to achieve both goals (1) and (2) is a fine balancing act — increase the redemption rate too fast, and you risk the asset becoming under-collateralised during bitcoin dips. On the other hand, if we increase the redemption rate too slowly, asset holders miss out on asset price appreciation that they could have otherwise been exposed to. Thus, finding an algorithm that ensures collateral sufficiency while maximising value appreciation is the most economically efficient solution for a stable reserve asset. After backtesting a number of algorithms, we found that the SMA1458 achieves this balance of price appreciation and collateral sufficiency very effectively.
Why SMA1458?
First, in over 13 years of trading history, Bitcoin has never dropped to less than 90% of the SMA1458. However SMA1458 isn’t just a random backtested algorithm — it corresponds to the natural halving cycles hardcoded in the Bitcoin Code.
In the diagram below you will see that in every 4-year-cycle, Bitcoin has posted new all-time highs. Notably, despite all the volatility, the bitcoin price in a cycle has never dropped below, or even approached, the previous all-time high. Thus the 4 year cycle has extremely strong correlations to price action, that can be factored in to build a probabilistic model for future price action. In particular, if the goal is to simply calculate a minimum floor of bitcoin price that holds with >99% certainty, that is possible using regular regression methods. However, such pricing tends to be capital inefficient and tends to underutilise Bitcoin’s dramatic and fast-paced growth rate.
In SMA1458 we found an algorithm that matches the above mentioned 99.9% certainty, but also shows significant appreciation over the short to medium term — often posting gains of as much as 50% per year in USD terms.
Algorithms build narratives around pricing, and these narratives are often self-reinforcing. The longer a particular heuristic for price holds, the more likely people are in the future to bank on that heuristic, making it more likely to hold in the future. Thus, a heuristic tested over 13 years of trading history, which included over 10,000x price appreciation, multiple dips of as much as 80% -90% of value, and three separate halving cycles and consequent bull runs, is a very strong heuristic indeed, and we should expect it to continue to hold in the future.
An additional constraint on our design, was that the redemption rate must be computable *on-chain* from oracle prices, to ensure full decentralisation of the protocol. Thus, computationally complex approaches like Machine Learning and regression were ruled out on this basis alone.
Secondary Stability Mechanisms
The redemption rate serves as the floor of the trading price for xBTC. To
prevent upward volatility and speculation beyond the redemption rate, the
following mechanisms are in place:
First, Arbitrageurs expecting the price to drop down to the redemption rate, can exploit the difference in price by borrowing more xBTC when the market price is above the redemption rate, and returning the debt when the market price drops to the redemption rate. In this way they will expand xBTC supply and quickly drop the price to the redemption rate.
Second, in case of sustained market prices above the redemption rate, EVI Token holders may vote to reduce the xBTC savings rate, making it cheaper to borrow xBTC and consequently expanding the supply. Notably, in many cases, just the anticipation of such an action may be sufficient to motivate arbitrageurs to bring down the price as described above.
Conversely, if the xBTC token is trading at or close to the redemption rate for sustained periods, EVI holders may vote to slowly increase XSR to maximise system revenue as well as provide a cushion for future rate drops in case of upwards volatility. The function of EVI DAO in this way mirrors that of a central bank.
