Nonlinear Identification and Early Warning of Stablecoin Price Depegging: Empirical Evidence Based on a Cusp Catastrophe Model

In the chaotic world of digital currencies, where prices can swing wildly from one moment to the next, there exists a special class of assets designed to remain calm. These are called stablecoins, digital tokens intended to hold a steady value, usually pegged to one dollar. They act as a bridge between the volatile crypto market and the real economy, allowing traders to move money in and out of risky investments without leaving the digital ecosystem. For the most part, these coins do exactly what they promise, trading very close to that one-dollar mark. However, history has shown that this stability is not guaranteed. When trust falters, when reserves are questioned, or when the market panics, these coins can suddenly break their peg, dropping in value or spiking unpredictably. Understanding why and how this happens is crucial, not just for investors, but for anyone watching how digital money evolves. The core question researchers face is whether these price breaks are just normal bumps in the road or if they are sudden, catastrophic jumps that signal a deeper failure in the system.

A team of researchers from Sun Yat-sen University and Fudan University set out to investigate this exact problem, looking at four of the most popular stablecoins: USDT, USDC, DAI, and FRAX. They wanted to move beyond simple price charts to understand the hidden mechanics that cause these coins to snap out of stability. To do this, they used a mathematical framework known as a catastrophe model. Think of this model as a way to map a landscape where a system can exist in a stable valley, but under certain pressures, it can suddenly tumble over a ridge into a completely different state. This approach allowed the scientists to distinguish between two very different types of price movement. One type is the everyday fluctuation, where the price wiggles slightly up or down due to normal trading. The other is a true break in the peg, where the coin's value fundamentally shifts away from one dollar, often due to a loss of confidence or a structural flaw. By applying this lens to daily data from 2019 to 2023, the researchers uncovered a more nuanced picture of risk than previously understood.

The study began by testing two different ways to measure the health of these coins. The first method looked at how far the price was from one dollar, a direct measure of the peg's integrity. The second method looked at the daily percentage change, which captures the speed and intensity of trading. The results revealed a striking difference between these two views. When measuring the distance from one dollar, the model found that true, catastrophic jumps were rare. This suggests that for most of the time, the mechanisms holding these coins together are working well, keeping the price tightly bound to its target. However, when the researchers looked at daily percentage changes, the model detected a flood of jumps. This indicates that while the coins rarely break their peg completely, they are constantly experiencing sharp, short-term shocks that look like jumps in the data. The researchers concluded that these two measures tell different stories: one reveals the rare moments when the system actually fails, while the other captures the constant, high-frequency turbulence of the market.

Perhaps the most surprising discovery was that the risk of a coin breaking its peg does not depend simply on what type of coin it is. For a long time, experts have categorized stablecoins into neat groups: those backed by real dollars in a bank, those backed by other cryptocurrencies, and those that use complex algorithms to maintain their value. The common assumption was that the algorithmic ones were the most fragile and the dollar-backed ones the safest. The researchers found this view to be incomplete. Instead, the risk is determined by how the coin is embedded in the market and how its supply and demand are managed. For instance, two of the dollar-backed coins, USDT and USDC, behave very differently despite having similar backing. USDT tends to see larger price deviations when its market size grows, likely because the new coins take time to distribute across exchanges, creating temporary imbalances. In contrast, USDC often becomes more stable as it grows, because its redemption process is more direct and transparent. This means that the specific design of how money enters and leaves the system matters far more than the broad label of "dollar-backed."

The study also traced how the sources of risk have changed over time. In the early days of the crypto market, around 2020, the prices of these coins were heavily influenced by traditional financial markets, such as stock indices and global interest rates. As the ecosystem matured, the drivers shifted inward. By 2022 and 2023, the biggest threats came from within the crypto world itself, such as the collapse of other digital assets or the failure of crypto exchanges. The researchers observed that during major crises, like the bankruptcy of the FTX exchange or the failure of Silicon Valley Bank, the coins reacted in complex ways. Some coins acted as safe havens, drawing in capital, while others suffered from the panic. The study highlighted that the risk is no longer just about external shocks but about how different parts of the digital financial system are interconnected. A problem in one area, like a bank holding reserves for a stablecoin, can quickly ripple through to affect other coins, even those with different designs.

One of the most practical findings of the research is the ability to predict trouble before it happens. The team tested whether they could spot warning signs in the days leading up to a price break. They found that by watching specific structural indicators, they could forecast risk with a high degree of accuracy. For moderate price deviations, which are small but noticeable, the warning signs appear quickly, often within a day or two. These are usually caused by sudden shifts in trading volume or liquidity. However, for severe depegging events, where the coin loses a significant amount of value, the warning signs build up more slowly. The researchers found that looking at a seven-day window of data provided the best early warning for these major crises. Indicators such as the concentration of holdings among a few large investors, the flow of money on the blockchain, and the overall size of the coin's market capitalization all served as reliable signals. This suggests that regulators and market participants do not need to wait for a coin to crash to know it is in trouble; they can see the cracks forming in the structure days in advance.

The implications of these findings are significant for how we monitor and manage digital assets. The research suggests that we need to stop looking at stablecoins as a single, uniform group and start treating them based on their specific behaviors and structures. A one-size-fits-all approach to regulation or risk management will miss the unique vulnerabilities of each coin. Instead, monitoring systems should track both the long-term stability of the peg and the short-term turbulence of trading. By paying attention to the underlying mechanics, such as how quickly new coins can be redeemed or how concentrated the ownership is, it is possible to identify which coins are most likely to fail under stress. The study provides a clear path forward: by understanding the nonlinear nature of these risks and using the right tools to detect them early, we can better protect the stability of the digital financial system. The goal is not just to react when a coin breaks, but to understand the forces that push it toward the edge before it ever gets there.