The Outlook for Decentralized Stablecoins

While the centralized stablecoins prove to be much safer than their decentralized counterparts, they leave a large chunk of the crypto ecosystem exposed to the risks of centralization. The lack of transparency coupled with the dependence on centralized entities could prove to be problematic for the industry in the long run.

Many people realized this early on and began experimenting with many different mechanisms to create a truly decentralized stablecoin. So let’s go over the most popular mechanisms created to date, as well as the various risks and benefits that each mechanism presents.

The 4 types of decentralized stablecoins:

• Algo-Stables (algorithmic stablecoins)
• Non-pegged stablecoins
• Overcollateralized stablecoins
• Fractional algorithmic stablecoins

Algorithmic Stablecoins are among the most controversial stablecoin mechanisms. They typically operate through a dual-token model. One token is the stablecoin, which is pegged to a fiat currency, and the second token is the native token, which is designed to absorb the volatility of the stablecoin. Suppose the stablecoin is pegged to $1. When one stablecoin is minted, $1 worth of the seignorage token is burned. When one stablecoin is burned, $1 worth of the seignorage token is minted. So the supply is adjusted deterministically using an algorithm.

The reason they are controversial is because the mechanism has historically not been sustainable. The entire mechanism is dependent on consistent demand being there for the stablecoin. Once demand and liquidity dissipate, they go into a so-called ‘death spiral’ where the mechanism collapses the stablecoin and the seignorage token by design. We have seen many algo-stables suffer this fate, the most popular of which were basis cash, TITAN, and UST.

Non-Pegged Stablecoins are very rare. The core idea is to have stablecoin that isn’t pegged 1:1 to any fiat currency. If one of the purposes of crypto is to create an alternate financial system, then relying on fiat pegged stablecoins is counterintuitive. The most popular non-pegged stablecoin is RAI, made by the team at Reflexer Finance. Such stablecoins are collateralized with a target price and interest rate that’s algorithmically adjusted. Using PID controllers, they aim to dampen volatility rather than peg the asset to a certain price. This allows the stablecoin to stay within a certain band of price points, which allows them to be relatively stable while mitigating the centralization risks of being pegged to a fiat currency.

Moving on to Overcollateralized Stablecoins. This was the first iteration of a decentralized stablecoin championed by the team at MakerDAO. The premise is fairly simple: a user who wants to borrow a stablecoin can only borrow through overcollateralization. A user deposits accepted collateral into the protocol and receives stablecoins in return at a value lower than the collateral deposited. Therefore, the protocol has sufficient collateral in its reserves to mitigate the risks of defaults and depegs.

The final mechanism is Fractional Algorithmic Stablecoins, which were pioneered by the team at FRAX Finance. Fractional Algorithmic stablecoins try to combine the powers of algorithmic stablecoins and collateralized stablecoins by blending them together. In these systems, a portion of the stablecoin will be algorithmically pegged while the other portion will be collateralized by eligible collateral. Depending on market conditions, the collateral ratio is adjusted. The collateral ratio is higher in more adverse market conditions, while in preferable market conditions, the collateral ratio will be lower.

Despite there being so many different options for decentralized stablecoins, none have been able to surpass their centralized counterparts. To understand why, we need to look at the varying risk profiles as well the benefits that each mechanism provides. Through this we can potentially a identify a decentralized stablecoin that can break the dominance of centralized stablecoins.

To analyze the risk profiles, I will look at the following factors:

• Sustainability of the mechanism
• Performance under different market conditions
• Potential attack vectors
• Previous stress tests

Starting with algo stables. This category of stablecoins can by and large be considered unsustainable. The lack of sustainability can be attributed to its reflexivity. Given that its stability is highly demand driven, algo stables depend on momentum and favorable market sentiment. As soon as the tide turns, the reflexivity shifts the other way just as hard, which leads to the feared “death spiral”. Coupled with the stablecoin collapsing, the seignorage token also collapses which not only spells the end for the protocol but also leaves investors hurt.

Therefore, it is clear that under bullish market conditions they perform really well, while in bearish market conditions they collapse just as fast. But you may be wondering, if the likelihood of a collapse is so high, then why do people opt to create it? It’s primarily due to the ease of scalability. The reflexive nature allows for quicker growth and adoption of the stablecoin. As you may have seen in the case of UST, it reached very high levels of adoption to the point where so many people were dependent on it that they attempted to create a BTC reserve in order to reach a certain escape velocity. This did not work out, but it goes to show how quickly an algo stable can grow in favorable market conditions.

In terms of attack vectors, the most common one would be simply inducing fear into the market. When liquidity is thin, it only takes a few large sell orders to cause a momentary depegging. A slight depeg will create fear amongst participants as they begin to lose trust in the system. When fear starts to spread and people rush to the exit, that’s when the stablecoin can completely collapse - it simply takes a slight push over the edge to create market-wide panic. Therefore, algo stables don’t tend to fare very well through stress tests. Momentary depegs have been restored in the past, but in the event of large sells and market panic, most algo stables fail to recover from the large stress tests.

Moving on to non-pegged stables. To understand the risk profile, I will be looking specifically at RAI, given that it is the most prominent non-pegged stablecoin in the market. Looking purely at the stability of the mechanism, it can be considered sustainable mainly due to the fact that an unpegged stablecoin can’t de-peg. Floating within a range of prices rather than needing to stay hard pegged to a certain price reduces the chance of people losing faith in the system. This system uses a mix of collateral and algorithmic measures to dampen volatility. ETH is used as collateral to mint the stablecoin, while the interest rates and target prices for RAI are algorithmically adjusted to incentivize users to keep price within a certain band.

Since its inception in 2021, RAI has only faced one major stress test, when price dropped from $3.52 to $2.89 in its first month. Ever since then, price has remained stable within the $2.90 to $3.10 range. Today, getting past the current market conditions would be considered a major success for the protocol. It’s fared well through bearish conditions so far, given that there haven’t been any major price swings. Additionally, the protocol has a total of 39,861 ETH locked with a system surplus of 507,776 RAI, signifying a healthy system.

The main attack vector against RAI would come from its reliance on the PID controller. In terms of stability, the PID controller enforces a negative feedback loop to ensure that RAI remains relatively stable. However, if the market does not react to the algorithmically adjusted redemption rates, it can cause the redemption prices to become unstable, thereby making RAI itself unstable. Other than the market not responding to incentives, any compromise, interference, or exploit of the PID controller itself could prove to be problematic for the stability of the protocol, given its importance in dampening volatility.

The other potential attack vector is from governance. Even though Reflexer Finance is a governance minimized protocol, there are still some parameters which are left up to governance and that can lead to suboptimal outcomes for the protocol.

Overcollateralized stablecoins have proved to be the least risky mechanism for decentralized stablecoins. The fact that every stablecoin in circulation is backed by more collateral in the reserves makes a depegging event very unlikely. To gauge the sustainability of overcollateralized stablecoins, we can take the example of DAI. DAI has been around for years, and it’s maintained its peg through periods of bearish market conditions. It has been through the most extreme of stress tests and has come out stronger. This builds further trust in the system, which allows it to grow and further maintain its stability, kicking in a long-term flywheel of adoption. This low-risk approach implies that the stablecoin is likely to remain agnostic to market conditions.

While this low-risk approach provides sustainability and certainty of stability, it also proves to be the root of the problems faced by this mechanism. The overcollateralized approach tends to be very capital inefficient when compared to algorithmic stablecoins. Lower collateral allows access to larger leverage and also simplifies the borrowing experience from the users’ perspective. This brings on more usage, which is what allows a system to scale much faster.

Other than the lack of capital efficiency, overcollateralized stablecoins are also susceptible to crises depending on the quality of collateral that they accept. Most overcollateralized stablecoins accept collateral in the form of USDT, USDC, and ETH. Some protocols also accept other forms of crypto collateral. This collateral is often exposed to the market, and in volatile market downturns, there is a significant liquidation risk if the value of the collateral falls below a certain amount. If the liquidation mechanism set by the protocol does not conduct the liquidation in an orderly manner, then the stablecoin could lose a portion of its backing leading to a depegging event. This risk becomes further pronounced when the accepted collateral is a high-risk crypto asset.

The key takeaway from overcollateralized stablecoins is that the tradeoff made here is favoring security & sustainability over scalability & capital-efficiency.

Lastly, fractional algorithmic stablecoins. This type of stablecoin combines the security and sustainability of collateralization with the scalability and capital efficiency of the algorithmic model. Regardless of this, since it is pegged to a fiat currency, the system is still dependent on arbitrage to maintain its peg. Hence, if the market loses confidence in the system’s solvency, then it’s likely to lose its peg. Even though the reflexive feedback loops typically seen by algorithmic stablecoins are minimized, there is still some level of risk involved. Therefore, it can be considered sustainable but with some reservations.

When it comes to different market conditions, this model is likely to fare well in all market conditions because the collateral ratio is adjusted depending on market conditions. The only risk is the quality of collateral - if the collateral accepted is higher risk than in times of high volatility, there are risks of large liquidations. So far, this hasn’t been a major issue. If you look at the example of FRAX, it has faced only two major depegs. It touched $0.96 once and $1.07 once. It recovered from both of those depegs and has maintained its stability ever since.

The primary risks or attack vectors that this model can face come from the protocol’s solvency. In the event of large liquidations or a collapse in the seignorage token, the protocol can face liquidity issues. This can lead to a loss of faith in the system, thereby causing a cascade of panic and fear and leading to an eventual collapse of the system.

As with everything in crypto, it is incredibly difficult to predict a winner. The main determinants of a successful stablecoin are sustainability and adoption. Adoption tends to come through bull cycles and sustainability keeps a stablecoin relevant through bear cycles. Being able to stand the test of time is what builds trust in the system, which leads to further adoption. So far, only overcollateralized stablecoins have stood the test of time. Non-pegged stablecoins are a worthy experiment but they struggle to gain adoption primarily due to the mindset of users. It will take users time to adapt to the notion of transacting with a non-fiat pegged stablecoin.

This brings us to fractional algorithmic stablecoins. They seem to have it all in terms of sustainability, capital efficiency, growth, and security but they still face issues when it comes to the eligible collateral accepted. For example, 90% of the collateral on FRAX is USDC essentially making it a proxy USDC at times.

I believe the winner in the long run would be either a non-pegged stablecoin or a fractional algorithmic stablecoin. Overcollateralized coins seem favorable in adverse market conditions but are not efficient enough to support usage from the masses, and algorithmic stablecoins are simply too reflexive to be a viable long-term solution. Non-pegged stablecoins and fractional algorithmic stablecoins are relatively new experiments but as time passes, the systems will grow stronger. Both these types of stablecoins check most of the boxes when it comes to finding a viable solution. To live up to the true ethos of decentralization, we need a decentralized stablecoin. Out of the two mechanisms – whichever one survives the bear markets and aggressively grows during the bull market – will someday come out as the big winner.