Without reliable economic design, under stressful conditions stablecoins are at risk of entering a death spiral when they are unable to defend their peg and win back market confidence.
This is especially severe if the stablecoins are only backed by themselves or other cryptocurrencies.
While fiat-backed stablecoins with appropriate capital buffers have low volatility, only CBDCs have zero risk.
Core Research Question
Are the economic designs of existing stablecoins reliable? What are effective ways to avoid economic risks?
When well designed, the technology behind stablecoins may lower the cost of payments and increase interoperability. This is because they are able to trade on par with a reference asset (usually the US dollar).
Stablecoins differ from each other by the following qualities:
economic design
quality of the backing asset
stability assumptions
legal protections for coin holders
The economic design of stablecoins has two critical dimensions:
The volatility of the reserve assets against the reference asset
The degree of risk exposure to a death spiral
Lack of robust economic design can undermine a stablecoinâs usefulness. In the worst cases, it can even turn into a threat to financial stability.
High-quality, liquid assets perform better in stressed market conditions, making it easier for stablecoins to maintain their redeemability. This is especially true when combined with capital buffers.
The volatility of stablecoins refers to how much reserve is needed to support the peg at any point in time.
The current economic risks of stablecoins:
Given the size of their issuing amounts, it is unclear how many stablecoins would be redeemable at par in a crisis (the top five USD stablecoins account for $110B of worth)
Stablecoins that rely on investment tokens to back them suffer from this problem even more severely.
If expectations change these stablecoins would enter a death spiral, making them worthless.
Decentralized stablecoins rely on algorithms by design. They have a big challenge maintaining stability against a real dollar.
Summary
1 Introduction
As seen in the previous section (Background).
2 Key Design Choice: The Reserve Risk Profile
The key challenge stablecoins face is to ensure they trade close to par with their reference asset.
When a stablecoinâs price trades above par, it can restore the peg by issuing new coins.
When the price trades below par, it needs to liquidate reserve assets and buy back coins.
However, if the reserve assets have dropped in value after coins have been minted, the issuer cannot buy coins back and defend the peg.
2.1 Volatility against the reference asset
At one extreme, the value of the reserve and the reference asset move in perfect synchrony. This kind of reserve has zero risks, and is only true for CDBCs, as the banks control both the digital assets and its reference assets.
Though fiat-backed stablecoins have low volatility, they still need appropriate capital buffers to offset potential losses coming from credit risks, market risks, liquidity risks, and operational risks.
At the other extreme, stablecoins are backed by non-fiat-like assets, such as risky types of commercial paper or cryptocurrencies. Their required amount of reserve assets for buying back a given amount of stablecoin supply will fluctuate dramatically over time. In this case, over-collateralization is expensive and cannot protect coin holders from extreme events.
2.2 Self-fulfilling prophecies and stablecoin death spirals
Death spirals are likely to occur whenever the value of a stablecoinâs reserve is tied to the future success of the stablecoin itself. This is true for stablecoins reserved by their investment coins
They can fail even when initially over collateralized.
Stress is sufficient to cause loss of confidence and more redemptions, which can then lead to death spirals even in a healthy market.
During a death spiral, a stablecoin backed by its investment coins needs to be shut down. The reserveâs value will fall to zero easily and cannot be recovered.
Features that are intended to stave off death spirals
Solution group 1
Partially winding down the stablecoin as the likelihood of a death spiral increases.
This is effective while the stablecoin is fully backed, and there will be no need to sell its investment coins.
Solution group 2
Stopping a run after it materializes.
Has not been tested in the context of stablecoins, but likely to make matters worse based on historical evidence on bank runs in the United States.
The risk of death spirals is inherently tied to the first design choice, which is the volatility of the reserve assets to the reference asset.
3 Conclusion
Decentralized stablecoins are either exposed to death spirals or capitally inefficient due to a lack of reliable on-chain, fiat-pegged assets.
Decentralized protocols rely on centralized stablecoins as part of their reserve, and/or over-collateralization to maintain their market price at par. This shows that fiat-backed stablecoins may be the only way to ensure long-run stability.
Method
The paper first goes through the natures of stablecoins and arrives at its conclusion by logical reasoning and comparison with historical events in a particular case. Some explanations are supported by diagrams which can be found in the next section.
The authors include an appendix and Figure.5 to discuss why algorithmic stablecoins find it hard to defend their peg in a crisis by reducing stablecoin supply.
Suppose the current price of the stablecoin is p < $1, and there is S amount of excess coins in circulation
The protocol needs to raise $(1+p)S/2 funds to burn S coins (Because the average price is (1+p)/2, and the quantity is S)
The additional coins to be issued is Q, with P as the price of the investment token, Pâ as the price of investment token of the Qth coin is sold, and (P+Pâ)Q/2 = (1+p)S/2
During a death spiral, the amount of new investment tokens needed to be minted and coins to be burned rises dramatically. The cost of defending the peg quickly becomes infinitive.
@Twan thank you so much for contributing a fascinating summary. A couple of quick questions: Do you know of any recent examples of stablecoins failing to maintain their peg and entering the death spiral? Are there ways that stablecoin projects can mitigate the risk by conducting audits?
Iâm also curious whether there is a point at which a stablecoin has so much money behind it that it becomes âtoo big to fail.â It also seems interesting to consider what might happen if more countries like El Salvador adopt Bitcoin as legal tender. Seems like it would blur the difference between CBDCs and crypto-backed assets a bit.
Thatâs interesting to think if stablecoins become âtoo big to fail.â
[Fitch Ratings (2021). Stablecoins Could Pose New Short-Term Credit Market Risks. Retrieve from https://www.fitchratings.com/research/fund-asset-managers/stablecoins-could-pose-new-short-term-credit-market-risks-01-07-2021] In a recent report, Fitch Ratings seems to regard the reserve of Tether as a high-risk asset allocation. âA sudden mass redemption of USDT could affect the stability of short-term credit markets if it occurred during a period of wider selling pressure in the CP market, particularly if associated with wider redemptions of other stablecoins that hold reserves in similar assets.â (Fitch Ratings, 2021) However, it also points out that âWe believe authorities are unlikely to intervene to save stablecoins in the event of a disruptive event, partly owing to moral hazardâ. (Fitch Ratings, 2021) From this perspective, they donât think stablecoins reach the quality of âtoo big to failâ so far. Maybe stablecoins need more capital and breakthrough moral concerns by some approach such as FIDF to reach it?
I think there will not be an international trend for countries to adopt Bitcoin as legal tender as El Salvador in the recent cryptocurrencies development phase. If they do so, it will be hard to adjust the exchange rate and currency supply. This may cause a severe problem on loan liquidity, then affect the economic activities in the country. So I think most countries may prefer to preserve their currency sovereignty rather than use this kind of non-sovereign, decentralized cryptocurrencies, which may also be influenced by unknown third parties. However, if there come more countries to do so, the power of the USD may be weakened. I also wonder what will happen in other aspects.
On the other hand, as I know, many people are expecting to see a whole new financial system can be built, and this is prospective to begin from the cryptocurrencies-world. I think this is an interesting issue too.
It seems that exchange rates and currency supply are less pressing problems in their country, which is why they adopted USD in the first place. President Bukele of El Salvador says he believes this will encourage investors with cryptocurrency to spend more of it in his country, but I wonder if people holding Bitcoin are looking to spend them at all.
Thank you for contributing. Interesting and clarifying read!
I am curious. What is the reason that over-collateralization still occurs in the status quo if it is such an expensive option? And, why have protocols utilizing this option not switched over to fiat-backed stable coins if this is the most stable option?
Thanks, shoule. Although over-collateralization could further ensure stability, switching to fiat-backed stablecoins could be financially devastating and thus is not an available option for everyone.
Thanks Twan, really great writeup. Would love to hear your thoughts on the below in regards to the users of CBDCs
Have you seen any literature/insight on what goes into the consumer choice of which stablecoin to use? Of course there are some specific points of differentiation (ex. why someone would use a crypto-backed stablecoin vs. a Fiat-backed one)
The authors state that fiat-backed stablecoins appear to be the only way to ensure long-run stability - with the eventual introduction of CBDCs what might be the use cases / demand for fiat-backed stablecoins that are not CBDCs?
Thanks for appreciating the work put into the summary. To your questions:
The first one sounds more like a topic that belongs to business or marketing, not economics. The resources that may be helpful to gain âinsightsâ are no more likely to show in academic papers than industry reports or user studies.
The second one is asking for a prediction of the future. So far my favorite author on this topic is Nassim Taleb. If you know what he advocates, then you probably can understand that I think in a world dominated by unpredicted black swans, it is risky to assume CDBCs would take over. Needless to say what role stablecoins would play at that time.
@Twan and @Astrid_CH do you see any connections between your two recent summaries (Wildcat Stablecoins discussion post and this one)? Thereâs also a bit of anxiety in the market right now about a Chinese real estate company called Evergrande, which has liabilities of about $300B and seems to be on the brink of failure. The company has issued many low-quality bonds, which are similar to the ones said to be backing some the larger stablecoins. Is this the sort of volatility trigger that could trigger a death spiral (or more legislation from regulators)?
Evergrande is concerning to investors for many reasons. Analysts also have diverging opinions on how things may turn out. Yet I would say that the threat is different from the threat of a death spiral.
Death spiral happens when the market loses confidence in the promise from the institution, and starts panic selling. Panic selling is something that happens during the death spiral, but it can also happen somewhere else, such as a market crash, which could also lead to panic selling.
One of the main concerns with Evergrandeâs impact on crypto is that investors that lost money may need to sell coins to raise funds. If bulk selling happens, prices may drop very quickly because of the surplus in supply. Price drops could also lead to problems with leveraging, hitting the market even harder.
In this case, the price drop from panic selling is only part of the process. Other factors play significant roles too.
So far issuers of some stablecoins have assured that they are not backed by the bond. However, if a stablecoin backs itself with bad bonds, and loses the value of the reserve due to that, then yes, it could enter a death spiral. That would be very relevant to this paper.
Stablecoin design and mechanisms is a constant work-in-progress, and having writing such as this provides such valuable insight and clarity over the topic.
Why thank you. There is plenty of interesting work in the forum. I hope you stay around and find more summaries that you would enjoy.
If youâre particularly interested in stablecoins, you can check out Discussion Post: Taming Wildcat Stablecoins by @Astrid_CH.
Thank you so much for pointing out the Taming Wildcat Stablecoins discussion post. @Twan can you tell us a little bit about how some of the regulations being considered in the US (such as Taiwan) might affect the design of future stablecoins? And how might that shape decentralized finance?
Regulations in Taiwan and the US are quite different, and they are going to change as the market is still in an early stage of development. But if cryptocurrency continues to go mainstream, the government is likely to set up regulations that protect general investors against large risks.
Iâd also like to follow up with your question on the similarities between the Taming Wildcat Stablecoins post and this one.
I think that post might be a more general overview of how stablecoins could develop over time, and this post is about a specific type of tail risk.
Thank you for sharing that discussion post, which was also very insightful!
Iâm curious, you write in your conclusion, âfiat-backed stablecoins may be the only way to ensure long-run stabilityâ. Can you expand on your thoughts pertaining to how crypto-backed and algorithmic stablecoins can be a long-run solution with significant risk mitigation against a death spiral scenario? Or if you donât ever see these types of stablecoins being a viable option, why that is so? You write that stablecoins backed by high quality assets perform better under stressed market conditions, and I would love to hear more about what you determine as the criteria for high quality.
Lastly, if given the perfect scenario, what are your thoughts on the ideal economic design of a stablecoin?
IMAO crypto-backed and algorithmic stablecoins are less reliable because of the nature of that type of asset.
Because the crypto doesnât pay dividends, the only price gain comes from the expected value in the future.
This easily leads to a vicious cycle when a death spiral happens, because the cause of that is exactly a significant drop of the expected value in the future.
Ideally, a stablecoin should be collateralized by fiat or highly liquid assets. The reserve should also be higher than 1:1 to buffer other financial risks.
I keep thinking about the discussion we had about the digital eNaira and how struggling governments might use CBDCs to enforce spending controls or prevent withdrawals during a bank run. Smart Contract Summit 2021: Central Bank Digital Currency (CBDCs) & Blockchain Panel. Do you think stablecoins based on hard currencies might be used as a âsaferâ substitute for a CBDC in some cases? Perhaps instead of using Bitcoin or pegging a currency to the dollar.
If it is a legal tender, rather than a means to store value in the digital world, then instead of issuing stablecoins backed by the dollar, they might as well just used the dollar.
But what are the safety issues on CBDCs are you referring to?
As for Bitcoin, I think the country that adopts that as their legal tender usually have different goals than what expectations we have from more traditional options.
