What Terra/Luna Taught Policymakers: The Algorithmic Stablecoin Reckoning

In the first week of May 2022, TerraUSD (UST) was the third-largest stablecoin by market capitalization, with approximately $18 billion in circulation. Its sister token Luna had a market capitalization exceeding $40 billion. The Terra ecosystem, built by Terraform Labs and its founder Do Kwon, was the most watched experiment in algorithmic stablecoin design: a system that maintained UST’s dollar peg not through dollar reserves but through an algorithmic relationship with Luna, where arbitrage incentives were supposed to keep the peg stable regardless of market conditions.

By May 13, UST had lost its dollar peg, trading at $0.10. Luna had fallen from above $80 to fractions of a cent, a decline of more than 99.9%. More than $40 billion in combined market value had been destroyed. Hundreds of thousands of retail investors — concentrated especially in South Korea, where Luna had been heavily promoted and widely held — had lost life savings, retirement funds, and in some documented cases, everything they owned. Do Kwon would eventually be arrested in Montenegro on fraud charges and extradited to South Korea.

The collapse was the most consequential event in the history of stablecoin regulation, generating more rapid and comprehensive legislative responses than any other single event in crypto history. Understanding what happened, why it happened, and what policymakers concluded is essential for understanding the stablecoin regulatory landscape that exists today.

The Mechanics of Collapse

TerraUSD’s peg mechanism rested on the relationship between UST and Luna. The system was designed so that $1 of UST could always be exchanged for $1 worth of Luna, creating an arbitrage incentive: if UST fell below $1, arbitrageurs would buy cheap UST and exchange it for Luna at face value, burning UST supply and (in theory) restoring the peg. If UST rose above $1, the opposite arbitrage would apply.

The design had one fatal vulnerability that critics had identified for over a year before the collapse: the stabilization mechanism depended on Luna having positive market value. If confidence in Luna fell, the exchange rate mechanism itself became a vehicle for panic: UST holders could convert to Luna, immediately sell the Luna they received, crashing Luna’s price, which further undermined confidence in UST, prompting more conversions, further crashing Luna. This feedback loop — called a “death spiral” in the analysis circulated by skeptics at the time — is exactly what happened.

The trigger is still debated. Large coordinated selling of UST by unknown actors beginning on May 7 appears to have destabilized the peg below the level where arbitrage incentives were sufficient to restore it. Once the death spiral began, the Anchor Protocol’s role accelerated it: Anchor had been offering 20% annual yield on UST deposits — an unsustainable rate funded by reserve subsidies — attracting tens of billions of UST in deposits. As UST de-pegged, Anchor depositors withdrew simultaneously, creating a run that the stabilization mechanism could not absorb.

The Anchor Protocol yield had been the mechanism through which Terraform Labs drove UST adoption. The 20% yield was explicitly unsustainable, funded by the Luna Foundation Guard’s reserves, and served primarily to create demand for UST. Retrospectively, it is clear that the system was designed with the appearance of stability rather than the substance of it.

The Human Cost: Korean Retail

South Korea’s retail investor exposure to Terra/Luna was exceptional. Terraform Labs had deep Korean roots — Do Kwon was Korean, the company was founded in Seoul, and its marketing was heavily concentrated in the Korean market. Luna had been promoted by Korean retail brokerages and fintech apps as a blue-chip investment, not a speculative bet on a novel financial mechanism.

Korean retail investors’ losses were enormous. Parliamentary testimony in the months following the collapse included accounts of individuals who had invested retirement savings, taken mortgages against their homes, or borrowed from family members to buy Luna. Suicide reports were documented in Korean media in connection with Luna losses. The human scale of the disaster was qualitatively different from previous crypto crashes, which had been largely confined to more sophisticated investors who understood the risks.

This human context is essential for understanding the speed and aggressiveness of the Korean legislative response. Korean politicians could not credibly explain to devastated constituents that regulating algorithmic stablecoins was too complicated or that the industry needed more time to self-regulate. The political pressure was immediate and specific.

Japan’s Response: The First Stablecoin Law

Japan moved first. The Act on Settlement of Funds, amended in June 2022 just weeks after the Terra collapse, created the world’s first comprehensive stablecoin regulatory framework. Japanese law now defines stablecoins as “electronic payment instruments” and restricts issuance to licensed banks, registered money transmitters, and trust companies.

Critically, Japanese law implicitly excludes algorithmic stablecoins: the definition of a lawful stablecoin requires that holders can redeem at par value in fiat currency. An algorithmic stablecoin whose value is maintained through market mechanisms rather than actual reserves cannot guarantee par redemption. Japan did not need to create a separate “algorithmic stablecoin ban” — its reserve-backing requirement achieved the exclusion by definition.

Japan’s response was informed by its earlier experience with Mt. Gox and Coincheck: the JFSA had developed institutional muscle memory for rapid legislative response to crypto failures that most regulators lacked.

Korea: VAUPA

Korea’s Virtual Asset User Protection Act (VAUPA), enacted in July 2023 and effective July 2024, is the most comprehensive legislative response specifically to the TerraUSD collapse. Its provisions address the specific vulnerabilities Terra revealed: stablecoin reserve requirements that mandate 1:1 fiat backing for coins marketed as stable; exchange licensing requirements that mandate client fund segregation and insurance; prohibition on unsustainable yield offerings that function as Ponzi-like recruitment mechanisms; and criminal penalties for market manipulation, unfair trading, and fraud — directly targeting the conduct alleged against Terraform Labs.

VAUPA’s mandatory reserve requirement for stablecoins served as a direct Template for subsequent legislation in other jurisdictions, including provisions in the GENIUS Act and MiCA. Korea’s legislative timeline — eighteen months from collapse to enacted law — was rapid by democratic standards and reflected the exceptional political pressure created by the scale of retail losses.

MiCA: Categorical Exclusion

The EU’s Markets in Crypto-Assets Regulation, finalized after the Terra collapse and reflecting its lessons, takes an elegant approach to algorithmic stablecoins: it excludes them entirely from its e-money token and asset-referenced token frameworks. MiCA’s definition of an e-money token requires that the token is backed by a single fiat currency and that holders have a redemption right at par. An algorithmic stablecoin — which does not guarantee par redemption and is not backed by fiat reserves — cannot qualify as a MiCA e-money token or asset-referenced token.

An issuer can attempt to offer an algorithmic stablecoin in the EU as a generic “crypto-asset,” subject to the white paper and disclosure requirements of MiCA’s basic tier, but without the stablecoin-specific regulatory protections or the legal marketing as a stable asset. The practical effect is a prohibition on marketing algorithmic stablecoins as stable instruments to EU consumers.

ESMA’s implementing guidance explicitly addresses the TerraUSD scenario, noting that stablecoins with redemption mechanisms dependent on market-based arbitrage rather than actual reserves are not e-money tokens regardless of how they are marketed.

The GENIUS Act: 1:1 Reserve Requirement

The United States, which was still processing the political and legal consequences of TerraUSD and FTX when the GENIUS Act was being negotiated in 2024-2025, was explicit about its lesson from the collapse. The GENIUS Act’s central stablecoin requirement is a 1:1 reserve mandate: payment stablecoins must be backed at all times by an equivalent amount of permitted payment stablecoin assets (US dollar deposits, Treasury bills, reverse repos, and similar high-quality liquid instruments).

The legislative history includes explicit Congressional references to TerraUSD as the failure mode the reserve requirement is designed to prevent. The 1:1 backing requirement is not technically novel — it was always the obvious design for a sound stablecoin — but TerraUSD demonstrated that without a mandatory reserve requirement, market participants would create and promote unbacked stablecoins to retail investors who did not understand the distinction.

The GENIUS Act also prohibits payment stablecoin issuers from offering yield on customer stablecoin balances — a direct response to Anchor Protocol’s 20% yield mechanism that was central to TerraUSD’s fraudulent adoption campaign.

The Global Policy Consensus

Six jurisdictions — Japan, Korea, the EU, the United States, Singapore, and the United Arab Emirates — have now enacted or implemented stablecoin frameworks that share a common core: reserve backing requirements that preclude algorithmic stabilization mechanisms, redemption rights for holders, and prohibition on marketing unbacked tokens as stable assets. The convergence across these frameworks was not coordinated through a single international body but emerged independently from the shared experience of TerraUSD’s collapse.

This is the global policy consensus that Terra/Luna produced: algorithmic stablecoins are categorically different from reserve-backed stablecoins. The technological innovation is not in dispute — the algorithmic mechanism is clever — but clever is not the same as safe, and safe is what a stablecoin must be. The policy conclusion drawn worldwide is that “stablecoin” is a regulated term that can only be used for instruments meeting reserve requirements, and that market mechanisms cannot substitute for actual assets.

Do Kwon is facing fraud charges. The Anchor Protocol is defunct. TerraUSD trades at negligible value. The policy legacy is lasting: a global consensus that reserve backing is not optional, codified in law across the jurisdictions that collectively represent the majority of global digital asset activity.