ArkStream Insights: The Unstable Algorithmic Stablecoins
Olympus is a new generation of algorithmic stablecoins. Its unique (3,3) community consensus and the generous financial returns of early investors have enabled it to obtain massive market exposure and is regarded as a typical model of Defi 2.0.
Many people have talked about Olympus. So instead of giving a basic introduction, we will analyze three design points that are not widely discussed, but we believe are crucial.
I. Rebase Correction
For the first generation of algorithmic stablecoins, a positive cycle occurs because of a lagged price correction. The timing mismatch between price and quantity leads to a misunderstanding of both price and quantity growth among token holders, which attracts token holders, expands the mismatch, and forms a positive cycle.
The Rebase design of the first generation algorithmic stablecoin, AMPL, is that Ample on the entire network will rebase, which includes Uniswap’s Liquidity Pool, and in the pricing formula of AMM DEX, xy=z. This results in an immediate price correction, which eliminates the mismatch.
Because of the Defi bull market in 2020, many community members were unaware of the price correction mechanism but simply believed that the price had corrected and they could start buying. This is another cognitive mismatch that helps it return to the positive cycle of a constant or slight decline in coin price and increase in quantity.
During the entire first generation of algorithmic stablecoins (mid-summer 2020), there were a lot of Forks. All these Forks had a moment of brilliance but died out eventually, and this confirmed our thoughts.
OHM is also a Rebase mechanism. The difference is that it rebases to Stakers instead of Liquidity Pool, which fixes the failure of the volume and price mismatch under the AMM DEX model. This helps Olympus consume less kinetic energy to maintain a positive cycle than Amplforth.
This kind of rebase also causes non-Staking holders to suffer inflation losses, thereby increasing their willingness to Stake. Thus, OHM has always maintained a relatively high Staking rate.
II. (3,3) Prisoner’s Dilemma
Olympus’ success is largely attributed to the widespread (3,3) consensus.
According to the official explanation:
(3,3) is the idea that if everyone cooperated in Olympus, it would generate the greatest gain for everyone (from a game theory standpoint). Currently, there are three actions a user can take:
- Staking (+2)
- Bonding (+1)
- Selling (-2)
Staking and bonding are considered beneficial to the protocol, while selling is considered detrimental. Staking and selling will also cause a price move, while bonding does not (we consider buying OHM from the market as a prerequisite of staking, thus causing a price move). If both actions are beneficial, the actor who moves price also gets half of the benefit (+1). If both actions are contradictory, the bad actor who moves price gets half of the benefit (+1), while the good actor who moves price gets half of the downside (-1). If both actions are detrimental, which implies both actors are selling, they both get half of the downside (-1).
At first glance, it makes sense — as long as everyone stakes, we will all get huge benefits. But there are actually huge cognitive traps.
1. Who defines Staking(+2)，Bonding (+1)，Selling (-2), benefits (+1), and downside (-1)? Are the values reasonable?
2. In a simple prisoner’s dilemma, why is the sum of the game not 0, but (3,1)+(1,3)=8?
When we try to make this chart again, we see a lot of issues:
1. Does the chart reflect floating profit/loss or real profit/loss?
2. The chart only discusses the (x, y) of A and B, but the text discusses the protocol as well, so shouldn’t the chart reflect (x, y, z) of A, B, and the protocol?
3. Are the conditions behind Stake, Sell, and Bond the same? Stake is not set to hold OHM-DAI LP. If under the same setting, Bond only holds OHM like Sell, then does Bond have to sell part of OHM to form OHM-DAI LP? Shouldn’t (Bond, Bond) be (-1, -1)?
However, these seem to be unimportant, and we do not need to draw another chart, because just like us, they don’t like doing math problems, and they don’t like to think so tirelessly.
The greatest value of the (3,3) chart is not to convey the whole truth, but to convey the truth that readers believe in with their limited resources of thinking. Truth does not become truth because it is the truth, but because it is easy to spread and understand.
From the popularity of (3,3), we see Olympus’s master understanding of human nature and communication.
III. Bond and APY
Stakers’ motivation comes from high APY, yet high APY is a common illusion in Defi. High APY needs time to compound interest, while Defi projects often only have a short period of ultra-high yields (due to the short-term FOMO in the market). Therefore, the basis of high APY — time — does not exist.
The Bond mechanism is used to build PCL in Olympus’s discourse system. But in fact, Bond and Rebase under high APY can combine. In theory, Rebase is going in the direction of 1OHM=1DAI.
After 316 days at the current rate, the number of OHM is exactly equal to the number of reserves.
The growth of reserves has a cost — the bond discount. But the cost is not high, and for Bond buyers, their implicit motive is that they are bullish on the project. Thus they exchange for OHM to Stake. People who are not optimistic about Bond would sell OHM or LP directly in the market.
Therefore, the Olympus protocol uses a small amount of capital that does not need to be fully redeemed to purchase a large number of reserves. The reserves support the high APY, and the high APY attracts people to purchase Bond. This is a complete mechanism.
The PCL service provided by Olympus Pro to other project parties does not have the effect of Olympus Bond. The reason is that Bond is part of Olympus’s entire cycle. Without Olympus Rebase mechanism and (3,3) consensus, Bond alone cannot create much value.
There is no doubt that Olympus has certain Ponzi characteristics, which is the commonality of algorithmic stablecoins we have seen so far. And a Ponzi design also helped them achieve great success in growth.
We are not warning risks of OHM. After all, IQ50 has already made a fortune. These projects are full of gaming nature and risks (returns). We just want to discuss the interesting mechanism design behind them. Since we did not participate in the Olympus community at a very early stage, we may also have cognitive biases.
Welcome to discuss further with us!
Author: Moc
Translator: Sisi
