Skip to content

The Data Scientist

the data scientist logo

The evolution of stablecoins from a tokenomist’s perspective

Wanna become a data scientist within 3 months, and get a job? Then you need to check this out !

Tokenomics 101: What are stablecoins?

Currencies like the U.S. dollar (USD) and the British pound (GBP) are fiat, meaning they are backed by nothing but the faith of their respective governments. The problem with these currencies is that they often fluctuate in value and can be worthless during periods of hyperinflation or financial crisis, which has led many people to assets such as cryptocurrencies.

Cryptocurrencies like Bitcoin (BTC) and Ethereum (ETH) offer a solution to this issue by being more decentralized than fiat currency and more liquid than gold; however, cryptocurrencies still experience price volatility that is counterproductive to their use as a stable medium of exchange. Cryptocurrencies have been on a rollercoaster ride for the past few years with highs and lows. A stablecoin is a cryptocurrency that has low volatility against the US dollar or other major world currencies.

As such, stablecoin projects strive to create cryptocurrencies that maintain stability relative to a single unit of a fiat coin, most commonly USD. As we will see in this article, this presents some unique tokenomics challenges.

Top 10 stablecoins by market capitalisation. Source: Coinmarketcap

Stablecoin types

The fundamental difference between stablecoins and traditional cryptocurrencies is that stablecoins are pegged to fiat currencies like the USD or EUR, and they use a variety of methods to ensure the peg holds. There are three main types of stablecoins:

  1. fiat-collateralized,
  2. crypto-collateralized
  3. non-collateralized (or algorithmic)

The most famous examples of fiat-collateralised stablecoins are USDC and USDT. The USDC token issued by Circle is backed by $1 worth of fiat currency for every USDC issued. The most important stablecoin (by market cap) is USDT, with a current USDT price of around $1, which is backed by a variety of assets. USDT, however, has been very controversial, as its liquidity has many times been disputed. Finally, another example of a stable coin is the BUSD, which has been created by Binance (which is the largest crypto exchange).

Crypto-collateralised use crypto instead of fiat assts as collateral. These projects usually follow under the umbrella of Decentralised Finance (DeFi). A great example is DAI, which has been created by MakerDAO. To create DAI, you need to provide some other crypto as collateral. This provides the necessary liquidity to the system. There is also one more token called MKR. The MKR token provides backstop liquidity in case the system accumulates bad debt, and holding MKR also entitles you to vote on how the Maker protocol is run.

Finally, algorithmic stablecoins are aiming to keep a peg using only algorithms. Algorithmic cryptocurrencies are still backed by crypto, like crypto-collateralised stablecoins. The difference is that in algorithmic stablecoins, this process takes place automatically. The best example of one such stablecoin is Frax.

The protocol ensures that Frax equivalent to a US dollar is always backed by at least $1 in reserves. The interesting part is that Frax is partially collateralised using both ETH but also its native FXS token to provide collateral for the FRAX stablecoin. The algorithms expand or contract supply in order to keep the peg stable.

Another great example of an algorithmic stablecoin is Terra. Much like in Frax, there are two tokens, Terra and Luna, that help stabilise each other, like the Earth and the Moon. Terra, however, aspires to create an ecosystem, whereas Frax is only aiming at being a stablecoin. Also, Terra is a pure algorithmic stablecoin, whereas Frax is also using ETH as a collateral.

Finally, I am very happy to declare that I am advising a partially-collateralised algorithmic cryptocurrency called BankX. I am the one responsible for the economic audit of the project. If you want to know more about it, make sure to visit the project’s page.

Comparing stablecoins

When comparing cryptocurrencies, it’s common to compare them using their marketcap. However, this metric alone doesn’t tell us enough about how useful a cryptocurrency is. After all, Dogecoin and Shiba Inu both witnessed huge market capitalisations, while being simply meme coins.

When comparing stablecoins, we need to keep the following parameters in mind:

Peg stability: This is probably the single most important factor. A stablecoin’s mission is to keep a stable peg. Therefore, this is the no1 criterion to determine a stablecoin’s performance.

Liquidity: Another extremely important factor. A stablecoin needs to be liquid. USDT has been riddled with controversy around its reserves, with many voices over the years saying that this could bring immense financial damage to the crypto world.

Decentralisation: This seems to be getting a more and more important requirement over time. With the possibility of governments cracking down on cryptocurrencies and imposing more and more regulations, true decentralisation seems to be the only way to achieve the original vision of Satoshi Nakamoto who was behind Bitcoin’s creation.

Peg stability

Based on those factors we can start our research into some popular stablecoins. Below you can see a table summarising peg stability. The maximum deviation was calculated over the whole history of a project and describes the maximum deviation from a peg ever recorded (up or down). The standard deviation was calculated, in a similar manner, over the lifetime of the project.

StablecoinMaximum deviationStandard deviation
Binance USD0.050.00351076

What can be seen is that all popular stablecoins have very low standard deviation. This is to be expected, since this is their job. Amongst the full list of stablecoins on coinmarketcap there are many that failed, and you can see this from their broken peg. Something that stands out is that Tether seems to have worse volatility than Frax. Also, Terra seems to have had the single worst incident out of all major cryptos.

This was an incident that the Terra Luna team discussed extensively on Twitter. Apparently, the protocol managed to hold up well.

The TerraUSD crash in January 2021. Source: Coinmarketcap

UPDATE: Apparently, while TerraUSD managed to survive the first major crash, it didn’t manage to survive the second one that took place in May 2022, which led to both TerraUSD and Luna to $0 value. It looks like the previous crash was a precursor to what was to follow.

Cryptocurrency liquidity

Liquidity is a common tokenomics problem. In theory, there are two ways to ensure a protocol is liquid:

  1. A protocol is centralised, but audited by an authority that can testify it has appropriate liquidity.
  2. The protocol itself runs in a way that maintains liquidity at all times.

We already mentioned the problems that USDT seems to be having with liquidity. USDC seems to attest, but not audit, its reserves, something which provides greater confidence to its users. However, in one way or another, centralised stablecoins do not seem to be solving the problems of the fiat financial system, if they still have to undergo through the same processes, as mainstream financial institutions.

In theory, algorithmic stablecoins should present the ultimate solution to this problem, but they’ve had their own set of tokenomics-related challenges.

A good example is the case of Iron Titanium. This is the first bank run in the history of DeFi. Coindesk and CNBC had covered the story. Iron Titanium was an algorithmic stablecoin. The stablecoin was stabilised through reserves consiting of 75% USDC and 25% Titan token. When the price of the Titan token spiked, the whales started dumping the coin, leading to other users to follow suit. The token’s value went from $60 to below $0.1, and the project’s marketcap evaporated.

Therefore, while traditional centralised stablecoins have their own issues, algorithmic stablecoins have still to prove themselves. Frax and Terra/Luna, however, seem to be doing great so far, and seem to have learned from the tokenomics mistakes of previous algorithmic stablecoin protocols.

It looks like DAI is a solution that manages to be transparent, while also being collateralised in a safe way, since its reserves are in Ethereum. However, algorithmic stablecoins are not going to stop trying and improve their protocols, given that the concept is very attractive from the perspective of tokenomics and decentralisation.


This is a bit of a more abstract metric. Decentralisation depends upon factors which are difficult to measure. In principle, every DeFi project is decentralised. However, there can be centralised points of failure. For example, while Uniswap is decentralised, there were rumours about the authorities regulating it. The Uniswap website exists somewhere on a traditional server, and nothing prevents the authorities from bringing it down.

So, decentralisation has to be tackled on a case-by-case basis. In general, DeFi projects have the upper hand in this regard, against Tether and USDC. However, judging how decentralised a project is, can be more complicated than the project simply being described as a DeFi project.

Conclusions: Stablecoins

Creating a stablecoin cryptocurrency is both an art and a science. In this article we examined the different categories of stablecoins, and compared a few different protocols against one another. My prediction is that within the next 2-5 years, there will be up to 10 (perhaps up to 5) stablecoins, dominating the crypto scene. Algorithmic stablecoins seem to be having a strong userbase, and in-spite of some failed protocols, the wave of decentralisation in stablecoins is still strong. What is very interesting is also some synergies that are emerging in the space of tokenomics between what is called DeFi 2.0 and stablecoins.

It’s definitely an exciting time for any tokenomics expert. If you have any questions or comments, feel free to get in touch. Also, make sure to check out the BankX stablecoin project where I am actively involved as an advisor.

Wanna become a data scientist within 3 months, and get a job? Then you need to check this out !