TechFlow: Sunny
Gyroscope: Ariah Klages-Mundt, co-founder
“Integrating decentralized stablecoins directly into DeFi applications can provide advantages that cannot be achieved with current centralized stablecoins.”
Ariah Klages-Mundt, co-founder of Gyroscope
“Many decentralized stablecoins are moving toward centralized reserve-backed designs, including decentralized stablecoins like Dai and Frax.”
The quote above, from Ariah Klages-Mundt, reflects the trend in stablecoin design choices in the market since major failures of non-custodial stablecoins, including MakerDAO’s “Black Thursday” in 2020, the Luna crash in 2022, and concerns about the Curve stablecoin in 2023. Since then, the stablecoin space has tended to favor a custodial design approach in order to ensure stability in volatile cryptocurrency market conditions.
Stablecoins can be divided into two categories: custodial and non-custodial (or centralized and decentralized). Custodial stablecoins are backed by centralized assets, such as USDT, USDC, treasury bills, or other real-world assets. Non-custodial stablecoins are fully backed by crypto-native assets, such as Ethereum and Bitcoin.
Klages-Mundt is the co-founder of Gyroscope, which aims to develop a new stablecoin design that challenges the current stablecoin landscape. Prior to co-founding Gyroscope, Klages-Mundt received a Ph.D. in applied mathematics from Cornell University, focusing on the design of decentralized finance (DeFi) protocols and economic networks. Before "Black Thursday", Klages-Mundt successfully predicted the event, believing that the original DAI design might experience a "deleveraging spiral."
The term “deleveraging spiral” describes feedback mechanisms in the structure of stablecoin assets that lead to liquidity shortages during crises, exacerbating the depletion of collateral. This phenomenon defines the dynamic stability of the system under certain conditions.
Today’s stablecoin market is dominated by Tether’s USDT, which currently has a market cap of $112.5 billion and a market share of 69.48% (see DefiLlama data below), followed by Circle’s USDC, MakerDAO’s DAI, Ethena’s USDe, and First Digital USD.
Market capitalization of Tether stablecoin with the largest market share (Source: DefiLlama, 2024)
In the case of USDT, only 4.87% of USDT backing reserves are crypto-native assets such as Bitcoin, rather than centralized assets backed by governments or real-world assets. In the case of DAI, MakerDAO governance also voted to change DAI's decentralized collateral reserves to a multi-collateral system (MCD), with more than 24% of DAI backed by USDC.
Tether’s USDT reserve breakdown (Source: Tether, 2024)
In a market where everyone aims to reduce the “deleveraging spiral” by backing their stablecoin designs with centralized assets, what are the advantages of crypto-native stablecoins that rely entirely on blockchain technology?
“Non-custodial stablecoins have been challenged to compete with USDC and USDT, but if you build them smartly, it is possible to surpass decentralized stablecoins,” Klages-Mundt said.
He believes that the current custodial stablecoin design not only has centralized credit risk problems, but also centralized liquidity pool problems. If savings assets such as USDT are decoupled, these problems may jeopardize the stability of current mainstream stablecoins. Therefore, diversifying supporting collateral and liquidity pools is a priority issue for solving the design of the next generation of stablecoins.
In Gyroscope, GYD is a stablecoin that draws lessons from previous “deleveraging spirals” and key economic events such as the British pound’s “Black Wednesday” in 1992. One similarity between the two situations is that they made overly optimistic assumptions about good market conditions without taking into account the cyclical nature of the markets they operate in.
Below is the full conversation between TechFlow and Klages-Mundt, which includes a comprehensive review of the current stablecoin landscape, Klages-Mundt’s proposed universal stablecoin design following his doctoral research, and how Gyroscope’s GYD is driving innovation in the field.
(Because the dialogue is long and informative, a table of contents is provided at the beginning to guide the reading.)
Table of contents
Klages-Mundt’s goal in founding Gyroscope began when he discovered a “deleveraging spiral” in the original MakerDAO stablecoin design
MakerDAO’s “Black Thursday” and the 1992 Pound Crisis: How do complex systems fail? What lessons can be learned about stablecoin design from economic history?
1. Via MakerDAO: DAI’s initial design
2. George Soros shorted the British pound
A general design for a new type of non-custodial stablecoin
1. Basic principles
2. Luna’s Ponzi scheme lies in its endogenous issuance mechanism
Design review of DAI, Frax and Ethena
Gyroscope: Beyond PSM and Automated Monetary Policy Adjustment
Route liquidity through GYD instead of USDC or USDT
Automated governance and off-chain governance
The first batch of GYD users 12
Popular products Vs. ideal products: When will centralized stablecoins surpass centralized stablecoins?
Klages-Mundt’s goal to create Gyroscope began when he discovered a “deleveraging spiral” in the original MakerDAO stablecoin design.
TechFlow: I saw your interview with Economic Design and one of your PhD talks. Your background is very interesting, especially your work in mathematics and graph networks for modeling complex systems like financial systems.
Can you start by talking about how your background in complex financial systems design led you to focus on stablecoin design?
Klages-Mundt:
I explored a variety of topics in both theory and practice with the ultimate goal of bringing everything back to real-world application.
I started out in pure mathematics, and did quite a bit of in-depth research in this area during my undergraduate studies. After that, I moved on to practical work in financial and economic modeling. This experience inspired me to create better models, solve more complex problems, and highlight the gap between real applications and complex models. I had some experience in ETF design and mortgage-backed securities modeling, but found that my real interest was developing new models to understand the complex dynamics that affect the real world. This led me to pursue a PhD focused on creating such models.
During my PhD, I started developing graph-theoretic models of complex financial systems. These models involve networks of relationships between different types of firms, and aim to understand how risk is transmitted through the system and the dynamics within it. This area straddles economics and computer science because of the computational challenges involved in this research.
MakerDAO came to prominence in the crypto space around 2017-2018, around the same time I published my first paper in this area. While I was already interested in cryptocurrencies, this development became a foundational point, showing another area that needed new models to understand complex dynamics with practical consequences. I incorporated this into my PhD research, writing my first paper on DAI dynamics and describing a “deleveraging spiral”, similar to the short squeeze effect. This theoretical work later proved to be predictive, predicting the occurrence of “Black Thursday” in 2020.
This experience solidified my interest in stablecoins. I saw that this was an open field that needed new models to better understand and have direct practical impact. My partner and I, both PhD students working on related papers, were inspired to build Gyroscope as a new type of stablecoin designed to avoid the problems we identified in existing models.
TechFlow: Before we dive into stablecoins and your thoughts on non-custodial stablecoin design, I want to understand your way of thinking. Can you define what complex systems are? How do you think these systems can be controlled so that their risks can be buffered through artificial systems, interventions, or new stablecoin designs?
Klages-Mundt:
I don't think there is a clear definition of a complex system, but you can think of it as something that is computationally complex and requires new infrastructure, especially computer science infrastructure, to understand its behavior. There are many types of complex systems, but I focus on financial systems.
For example, consider a complex network of financial companies. You can see similar situations in the cryptocurrency space, especially in DeFi.
In DeFi, there are many interconnected protocols, often described as "financial Lego blocks" where different components are stacked on top of each other. The innovative potential of DeFi allows for a lot of flexibility and creativity in financial products, but this also stacks risks together.
Understanding how these different DeFi protocols interact with each other is a complex system in itself, and it is a challenging problem that requires grasping all potential outcomes.
MakerDAO’s “Black Thursday” vs. the 1992 Pound Crisis: How do complex systems fail? What lessons can be learned about stablecoin design from economic history?
Via MakerDAO: DAI’s Initial Design
TechFlow: In a previous interview, you compared traditional fiat currencies to non-custodial stablecoins like DAI. Can you elaborate on how DAI represents a complex system, but how its components differ from traditional currencies?
Klages-Mundt:
In non-custodial stablecoins, there are various agents or protocols that perform different roles and connect to each other. For example, consider the original DAI system, which connects borrowers (or single coin holders) and people seeking to leverage crypto assets.
In this simplest form, someone who wants to speculate on the price of ETH would act as a borrower, provide ETH as collateral, mint DAI through overcollateralized lending, and sell DAI to someone who wants to use it as a stablecoin.
Typically, borrowers use DAI to buy more ETH, thereby leveraging it. This is a basic setup, but as more people seek leverage, the system becomes complex, forming a web of decisions and potential problems. In addition, understanding the demand side of DAI - how much demand there is for holding DAI as a stablecoin and what it is used for - is also critical.
MakerDAO’s initial mechanism: The initial failure of the MakerDAO system can be summarized into the following four reasons:
(1) With only two collateral options (ETH and BAT), the lack of diversification could lead to a “deleveraging spiral” if the entire crypto market collapses;
(2) the split-second decision-making process in governance;
(3) Insufficient demand for DAI;
(4) Oracle failure.
For example, during a "deleveraging spiral" such as what happened to DAI on Black Thursday, DAI's liquidity and usage significantly affects the system.
Modeling all of these components is very complex because you need to consider the decisions of various agents and the overall demand dynamics for the stablecoin.
George Soros shorts the British pound
TechFlow: Traditional currencies also sound very complicated by comparison. What is the “Black Friday” case for traditional currencies compared to the “Black Friday” case for DAI? What are the triggers in both cases, and how are they similar in some ways?
Klages-Mundt:
Finance is inherently very complex. Most people don't fully understand how banking or monetary systems work; they only know how to interact with these systems through banks. Many people may think that traditional finance is simpler than cryptocurrency, but it is just as complex, if not more so. The difference is that the complexity of traditional finance is hidden, while in cryptocurrency, especially in DeFi, it is transparent. You can see all the connections, which is an important innovation. However, for newcomers, this transparency can also lead to information overload and make cryptocurrency appear more complex. In reality, all financial systems are complex.
This complexity was a key motivation for us to create Gyroscope. There is already a large body of research in finance and economics, and we should learn from these experiences. While some models need to be adjusted for cryptocurrencies, there is still a solid foundation to build on. For example, consider models of currency attacks such as George Soros's on the British pound. There are already mature models to understand the mechanics and profitability of such attacks, which can help design resilient monetary policies.
Unfortunately, many stablecoins, such as Terra, did not learn these lessons when they were launched. Terra’s system appears ostensibly vulnerable to depegging attacks because it is undercollateralized. Applying traditional models, one would predict that such a system could spiral to zero. Terra’s design led to significant problems, highlighting the importance of integrating established economic principles in stablecoin development to avoid future financial disasters.
On September 16, 1992, Black Wednesday occurred and the British pound was forced to withdraw from the European Exchange Rate Mechanism (ERM). Although the British government tried to keep the pound above the ERM minimum by raising interest rates and buying pounds, under the short selling pressure of George Soros and other institutions, the pound depreciated and withdrew from the ERM. The UK lost about 3.3 billion pounds in the process of trying to stabilize the currency, but ultimately failed.
A general design for a new type of non-custodial stablecoin
Basic decentralized stablecoin design principles
TechFlow: Before Gyroscope, you mentioned a general design for non-custodial stablecoins, where you mentioned two terms: tangible and intangible. You pointed out that Terra UST only focused on the tangible part and did not address the intangible part. Can you briefly introduce this general design for non-custodial stablecoins and apply it to the current environment?
Klages-Mundt:
When discussing decentralized or non-custodial stablecoins, the two most important factors are:
What assets actually back stablecoins?
What is the issuance mechanism - how are new stablecoins generated and how are they withdrawn from circulation? Is there a redemption process, and when do stablecoins withdraw from circulation?
Most of the problems with existing stablecoins can be traced back to problems with these two mechanisms.
When it comes to the backing assets of stablecoins, the key distinction is between exogenous collateral and endogenous collateral.
Exogenous collateral refers to assets that exist outside the system and have value.
Endogenous collateral is an asset whose value circulates within the system. Its value depends on the success of the system. For example, Luna only has value while the Terra system is functioning and growing. If confidence in the system is lost, its value could collapse to zero, causing the entire system to fail. This circular value derivation carries inherent risks, making stablecoins susceptible to devaluation when confidence in the system is shaken.
Many of the problems with existing stablecoins stem from neglecting or mishandling these fundamentals. By taking these two points into account during the design process, the stability and reliability of stablecoins can be greatly improved. This is also the basic principle we followed when designing Gyroscope. We hope to create a more stable and secure non-custodial stablecoin system by learning from the lessons of our predecessors and combining existing knowledge in economics and finance.
Luna’s Ponzi scheme lies in its endogenous issuance mechanism
TechFlow: Are you saying that stablecoins fully backed by crypto-native assets are inherently flawed in the current climate of uncertainty and volatility?
Klages-Mundt:
Not exactly. For example, some stablecoins may be backed by assets like ETH, while others, like Terra, are backed by endogenous assets like Luna. In the case of ETH, its value is independent of the operation of the stablecoin system. For example, DAI is backed by ETH, and although ETH is more volatile, it has intrinsic value outside of the stablecoin system.
On the other hand, the value of Luna is inherently tied to the success of the Terra system. If the Terra system is functioning and growing, Luna has value. However, if confidence in the system is shaken, the value of Luna can spiral to zero, causing the entire system to collapse, as we saw with the Terra crash. This circular value derivation is why we caution against using endogenous collateral, especially as a primary mechanism.
A mechanism like Terra could resemble a Ponzi scheme. When you break down the system, new money coming into the system is used to pay off early investors. This makes it very dangerous because the system relies on continued growth and new investment to maintain its value. In contrast, reserve-backed stablecoins are a better option because they are backed by assets with independent value, providing more stability and transparency.
TechFlow: How does exogenous collateral avoid this circularity?
Klages-Mundt:
If you think about ETH, it has a variety of uses outside of being collateral in a stablecoin system. In any monetary system, you eventually reach a point where value is based on a synergy game, whether it's the dollar, gold, ETH, or Bitcoin. The point is that certain assets have a stronger and established synergy game that gives them intrinsic value.
For stablecoins, the relevant factor is the strength of this coordination game. Assets with established external value are less vulnerable because their value is already recognized and is likely to persist. In contrast, assets whose value is only derived internally within the system have newer and weaker coordination games and are more likely to break down.
Technically, while all monetary systems involve some degree of coordination, systems that are built on stronger, already established foundations are more robust. Therefore, using assets with external value as collateral makes the system more resilient.
Current Design Review of DAI, Frax, and Ethena
TechFlow: This is very interesting. In the current stablecoin space, we have seen DAI, Frax, and now Ethena. What are your thoughts on these three stablecoins? How did these different design spaces influence your decision to build Gyroscope?
Klages-Mundt:
DAI itself has gone through an evolution that’s worth highlighting, and we can also show how Frax fits into that evolution. Ethena is a separate case, so I might come back to that later.
For DAI, if you start with the original DAI design, which is something I studied in an early paper, we found a "deleveraging spiral" that eventually occurred. After that happened, the design of DAI went through a significant evolution.
DAI no longer relies solely on the leverage backing system and adds the Peg Stability Mechanism (PSM). PSM allows convertibility between Dai and USDC. Introducing this mechanism overcomes the short squeeze effect (deleveraging effect). If the price of DAI rises during a short squeeze, people who want to deleverage can redeem Dai with USDC at a 1:1 ratio. This prevents the deleveraging spiral effect because the price has a hard cap at $1 due to this convertibility feature. However, this also introduces centralization risk and sometimes DAI may be mostly backed by USDC. This leads to another crisis for DAI when people question why they hold DAI since it is more than 60% backed by USDC.
DAI evolved further to include real-world assets like Treasury bills. Now, DAI is a hybrid stablecoin, partially backed by Treasury bills and partially backed by crypto leverage mechanisms. This is not to say that this approach is bad, but it does introduce different risks. When assessing the risk of DAI, the risks of all backing and issuance mechanisms need to be considered.
On the other hand, Frax was initially designed to be similar to Terra, which is worrisome. The original Frax design was partially backed by USDC and partially backed by endogenous collateral, bringing decoupling risks similar to Terra. However, after the Terra debacle and many Frax clones, Frax is moving towards a more fully collateralized model, similar to a reserve-backed design. This is a positive development.
Many decentralized stablecoins are moving toward centralized reserve-backed designs, including DAI and a large portion of Frax.
With Gyroscope, we hope to introduce the most innovation by building what we believe to be the most secure stablecoin, automating risk controls and achieving this by managing a reserve-backed system as well as liquidity in and out of the system.
As for Ethena, it is a different design because it operates as a centralized system. Ethena uses a perpetual leverage mechanism that pairs long positions with short positions, hoping that the net interest payment is positive and paid out as yield. Ethena does this by executing trades on centralized exchanges with higher liquidity, but this introduces centralization risk and the risk that the net interest rate could be negative, affecting the stability of the peg.
Gyroscope: Beyond MakerDAO PSM and Automated Monetary Policy Adjustment
TechFlow: From what you just said, DAI is not completely non-custodial since Black Thursday because it has a centralized reserve mechanism. Frax is not completely non-custodial either, while Ethena is centralized. It seems that current stablecoins have both centralized and decentralized risks. How does Gyroscope avoid these risks? I believe you also have a reserve mechanism backed by some centralized stablecoins such as USDC and USDT.
Klages-Mundt:
At Gyroscope, our approach is to design from the ground up with a clear understanding of the risks involved, and to manage those risks as effectively as possible. We don’t claim to eliminate all risk—that’s impossible—but we believe they can be managed better than simpler solutions like Maker’s original PSM. Our motivation at Gyroscope is to build infrastructure that can better manage these risks.
A particular focus of Gyroscope’s design was to prevent concentration of risk in the system’s backing assets. Maker’s simplistic PSM ultimately led to excessive USDC exposure, which proved detrimental when USDC depegged, prompting Maker to take the next step in its evolution away from relying solely on USDC.
Another problem is that governance in simple designs is both slow and responsive to system changes. For example, during last year’s USDC decoupling, maintaining DAI price stability required split-second decisions by Maker governance. We believe decentralized governance is not suited to such rapid responses. Instead, we prefer automation of monetary policy adjustments. Governance should pre-agree on policies that the system automatically executes when certain events occur, eliminating the need for immediate human intervention. This principle is a key aspect of Gyroscope’s design.
USDC decoupling highlights several key issues that Gyroscope aims to solve.
One issue is oracle risk — the challenge of accurately pricing the assets backed by the system. Current oracles often rely on a multi-signature security model. We believe in a more robust approach that includes multiple sources of information and security models. Our system layers additional checks to ensure the reliability of pricing data and prevent extreme manipulation when the data is inconsistent.
Another issue is the decentralized liquidity problem. It is difficult to initiate liquidity, and many decentralized liquidity channels are routed through USDC or USDT. If these assets decouple, it may disrupt most liquidity channels, even if the assets are locked in an automated market maker (AMM). We aim to make Gyroscope's trading infrastructure more robust to prevent a full liquidity stagnation when an asset decouples.
Independent Liquidity Pool: Route liquidity through GYD instead of USDC or USDT
TechFlow: Let’s start with the last point you mentioned: the centralization risk issue of USDT and USDC. How does Gyroscope plan to make the system more robust against the decoupling of USDT and USDC?
Klages-Mundt:
A good example of this happened during the USDC decoupling last year. At the time, the main liquidity route was usually through Curve 3pool (including DAI, USDT, and USDC). When USDC decoupled, the liquidity pool was filled with USDC, causing liquidity to dry up for transactions involving other stablecoins. This affected the entire liquidity system for stablecoin transactions, even if those stablecoins had nothing to do with USDC. The ability to route transactions through the liquidity pool collapsed.
At Gyroscope, we aim to solve this problem by putting GYD in a central location and enabling independent trading pools that can move in and out of GYD. This approach is designed to be as resistant to the impact of individual depegging as possible. If a stablecoin (such as USDC) depegs, it may shut down liquidity for one trading pool, but other trading pools will be relatively unaffected. This way, liquidity can bypass USDC instead of going through it.
For example, during the USDC depeg, I looked up how to swap BUSD for USDT. Even though these stablecoins are not tied to USDC, the best on-chain quote market influence is over 10%, making liquidity between stablecoins almost impossible. With Gyroscope’s design, we aim to prevent this by maintaining more robust and independent liquidity channels.
TechFlow: Just a random question - I know Gyroscope has a dual-coin system: one is GYD, and the other is SPY which is more governance-focused. We can talk about this later. Regarding GYD, why does adding another token to the system help stabilize or prevent instability?
Klages-Mundt:
You can think of GYD as a tokenized way to hold stablecoins in an optimal risk-diversified manner. It is a stablecoin in itself, but also a means to manage the risks of holding stablecoins. This dual perspective explains the value of GYD.
You mentioned the question of how GYD helps stabilize the trading system. By putting GYD in a central location, we create a system where independent trading pools can move in and out of GYD. This setup is designed to handle risk efficiently and be as resistant to individual decouplings as possible. If a certain stablecoin (such as USDC) decouples, it may shut down liquidity for one trading pool, but other pools will not be affected. This allows liquidity to bypass USDC instead of passing through it, maintaining a stronger trading system.
TechFlow: What does it actually mean to tokenize other components into GYD? If I understand correctly, how does this contribute to the stability of the system?
Klages-Mundt:
GYD is a stablecoin designed to provide the best risk diversification of holding stablecoins and managing the risk of underlying assets. If you want to hold stablecoins, we believe GYD is the best choice because it effectively manages risk allocation and diversification.
To get GYD, you provide other stablecoins or specific assets that the GYD system wants to hold. The system has rules about what risks it can take, so it tells users what assets are needed to stay diversified. You listen to the protocol, provide the required assets, and get newly minted GYD. This process helps you manage your risk exposure by holding a well-diversified stablecoin.
Once you own GYD, you can use it in various DeFi protocols, as integrations expand. This allows you to focus on using your stablecoins without having to worry about managing different risk exposures between different stablecoins. GYD handles these challenges for you.
In addition, GYD helps manage yield risk. The system holds assets on the balance sheet and deploys them to different yield-generating opportunities, applying the same diversification rules as the assets. This ensures that the yield on the balance sheet is the best risk-diversified yield in DeFi.
Soon there will be a sGYD (savings GYD) version as a tool for anyone to access this risk-diversifying income potential. Therefore, holding GYD not only helps manage your asset risk, but also optimizes the source of income, making your stablecoin holdings more efficient and secure.
Automated governance and off-chain governance
TechFlow: Is this where decentralized governance comes into play in understanding risk exposure and tokenizing other stablecoins into GYD?
Klages-Mundt:
I mentioned automation before, but there are limits to what can be automated. You can foresee certain scenarios and decide how the system should respond, but some developments are unpredictable. One unpredictable factor is how the DeFi and stablecoin asset space will evolve.
At this point, we believe that the primary role of governance is to continually make those decisions that cannot be automated. However, those decisions should be made as slowly and thoughtfully as possible. Our argument is to automate what you can, ensuring automated responses to predictable events. For unpredictable developments, governance should be a slow, thoughtful process.
Decentralized governance is inherently messy, and that’s okay. The decision-making process needs to be slow to ensure careful consideration and adaptation to changes in the space. We aim to prevent governance from impacting the immediate health of the system, while still allowing it to make thoughtful decisions to keep up with the evolving cryptocurrency landscape.
The first batch of users
TechFlow: Who are the main users of Gyroscope stablecoin at present?
Klages-Mundt:
Currently, our largest users are liquidity providers, but this may change soon with the launch of sGYD. sGYD will help people access diversified sources of yield. Currently, we provide a liquidity provider tool that we believe is risk-diverse.
The liquidity structure I mentioned earlier, with separate pools for various stablecoins in and out of GYD, has two benefits. First, it makes routing challenges in potential stablecoin decoupling scenarios more robust. Second, it provides benefits to liquidity providers by reducing exposure to decoupling of a single stablecoin.
For example, if you are a liquidity provider for a stablecoin-to-stablecoin trading pair and want to reduce your exposure to a single stablecoin depeg (like USDC), you can allocate funds across multiple pools, moving in and out via GYD. In the worst case, if one or some of these pools are affected by a stablecoin depeg, the other pools remain independent and continue to operate, enabling you to earn trading fees.
This is a key advantage over being a liquidity provider in Curve 3pool. In the event of USDC depegging, liquidity providers in Curve 3pool were forced to hold USDC and their portfolio value took a significant hit. In contrast, managing your exposure by spreading your allocation across GYD pools will reduce this risk as the pools remain as independent as possible. Your portfolio will be less impacted and you will continue to benefit from trading fees.
Popular Products Vs. Ideal Products: When Will Decentralized Stablecoins Surpass Centralized Stablecoins?
TechFlow: Your explanation is very detailed. I think it would be very interesting to go deeper into how decentralized monetary policy interacts with governance, but I guess we have limited time today. So, my last question is a bit more high-level. I believe that stablecoins like USDC and USDT are currently the best products, especially for people who are affected by geopolitics and need stable currencies. I also appreciate the idea of non-custodial stablecoins and think it is a very ideal product. How do you define your motivation for building non-custodial stablecoins? How should ordinary people think about the future of non-custodial finance, rather than a market dominated by USDT and USDC?
Klages-Mundt:
That’s a good question. What I want to emphasize is the potential for innovation. There are many things you can do with crypto-native infrastructure that are not possible with centralized stablecoins. Decentralized stablecoins have been competing with USDC and USDT, but if you build them smartly, you may surpass centralized stablecoins.
A good example is risk management. Centralized stablecoins always carry issuer risk, which is unavoidable. However, you can mitigate this risk, which is why we built a diversified infrastructure for GYD and implemented various policy automations. We believe this approach can go beyond centralized stablecoins.
Another area is liquidity. If you launch a new centralized stablecoin, it is difficult to build liquidity that is sufficient to compete with USDC and USDT.
In addition, the issuance process of centralized stablecoins is rigid and relies on holding U.S. dollars or Treasury bills. In contrast, crypto-native infrastructure allows for greater flexibility in initiating liquidity. For example, innovation around protocol-owned liquidity could inject liquidity directly into the market, as would central bank action. This has significant advantages, although it also introduces risks, which we will consider carefully.
One of our main motivations is to harness the potential of innovation and create better solutions than centralized stablecoins. This includes features unique to DeFi that resonate with us, such as creating non-custodial currencies, reducing counterparty risk, and potentially becoming independent currencies.
Integrating decentralized stablecoins directly into DeFi applications provides benefits that centralized stablecoins cannot match. While this increases risk exposure and system complexity, we approach this problem with a research-driven perspective to create competitive and secure DeFi systems. A super simple solution may not be competitive, and ignoring risks is not feasible. We believe that complex DeFi systems are inevitable when competing with centralized stablecoins, and we are committed to building these systems as securely as possible.
TechFlow: When do you think is the breakthrough point for non-custodial stablecoins to become true crypto-native currencies (i.e. reserves can be crypto-native assets)?
Klages-Mundt:
At Gyroscope, we have assessed the crypto asset landscape, accepted the current state of the stablecoin market, and categorized the various risks that exist. One of the main categories is decentralization risk factors, which come from assets that are very close to or completely decentralized. Another category is centralization risk factors, which come from centralized stablecoins.
We see complementary value in these different risks because they provide diversification. We are not advocating to remove custodial assets entirely; rather, we advocate balancing the risks of custodial assets with the risks of decentralized assets. By diversifying these risks, you can achieve a more robust and resilient system.