The 2018 Stablecoin Census and the Variable Nobody Scored

Forecasts are usually judged on whether the number came out right. That's the least interesting test. A more useful one is whether the forecaster was tracking the variable that ended up doing the work, because a model can miss the number by an order of magnitude and still be sound, while a model that misses the mechanism was never going to be right except by accident.

Which makes Blockchain's The State of Stablecoins worth revisiting. It was the first serious empirical survey of the sector, built on 1,600 data points gathered from 57 projects, with market data running through July 2018 and launch entries as recent as September 2018. Most of its descriptive work has aged well. Its central forecast did not. And the gap between those two facts points at something that still matters for anyone reading a 2035 outlook today.

We focused on the report's taxonomy, competitive assessment, and forward-looking sections rather than the individual project primers.

The scoreboard: right about concentration, wrong about everything underneath it

Start with the raw comparison, because it sets the scale of what happened.

In 2018 the entire stablecoin market was worth $3 billion, which the report noted was 1.5% of all cryptoassets. Tether held 93% of that value and roughly 98% of daily trading volume. Every stablecoin project in existence had collectively raised $350 million in venture funding.

As of late July 2026, total stablecoin market capitalization sits at roughly $310 billion according to DefiLlama data, or about a hundredfold increase, and stablecoins now account for something closer to 13% of total crypto market capitalization. Tether's supply share has fallen to about 59%, with USDC at roughly 24%, per DefiLlama figures reported in June 2026. Statista and CoinGecko were tracking 213 distinct stablecoins as of July 2026.

Set against that, the report's headline prediction reads oddly well: it argued that "space may exist for approximately 5-8 significant stablecoins in the short to medium-term." Two hundred coins exist and two of them hold about 83% of supply. The concentration call was right.

The reasoning behind it was not. The report attributed the expected concentration to design uncertainty and regional regulatory variation, the idea being that a handful of different technical models would each find a niche. What actually concentrated the market was distribution and liquidity depth, a dynamic we looked at in more detail in the 2026 landscape. The report even identified this, listing exchange listings and liquidity as "the two most important success factors" for stablecoin teams. It just didn't connect that observation to its own forecast about which designs would survive.

The design thesis that capital backed and law killed

The report's clearest forward-looking claim was about architecture. Asset-backed stablecoins were treated as the pragmatic starting point and algorithmic designs as the destination: "many believe the long-term solution will likely be a digital-native, fully-algorithmic stablecoin, and we expect over time that many asset-backed stablecoins will attempt to convert to an algorithmic design."

This wasn't a fringe position. The report's own funding data shows the market agreed with it. Algorithmic projects had raised $174 million, half of all stablecoin venture funding, against $144 million for off-chain asset-backed models and $33 million for crypto-collateralized ones. The most sophisticated money in the sector was betting on the algorithmic thesis.

Two events settled it.

Basis, the report's top-funded pre-launch project at $133 million and the case study for the algorithmic model, announced in December 2018 that it was shutting down and returning capital to investors. Coverage at the time, including CoinDesk's reporting on 13 December 2018, attributed the decision to the conclusion that its bond and share tokens could not avoid classification as securities under US law, which would have required transfer restrictions and accredited investor whitelisting. That was roughly three months after the report was compiled. Not a market failure, not a broken peg. A legal characterization.

TerraUSD, described in the report as an algorithmic design with strong e-commerce partnerships, reached a market capitalization of more than $18 billion before losing its peg in May 2022. Congressional Research Service documentation records UST trading at $0.12 by 16 May 2022. Founder Do Kwon was sentenced to 15 years in a US federal prison in December 2025.

The category the report ranked highest on transparency and automation is now largely absent from regulated markets, and the frameworks that govern the sector are built around reserve backing. The GENIUS Act, signed in July 2025 and taking effect in January 2027, is structured around backed payment stablecoins and prohibits yield payments on them outright.

The column missing from the comparison table

Here is the part worth sitting with.

The report's central analytical artefact is Table 1, which scores 24 stablecoin projects on three axes: transparency, automation, and complexity of the stability mechanism. It's a genuinely useful frame, and building it from primary data across two dozen projects was real work.

It contains no column for regulatory tractability. And that single missing column would have predicted survival better than the other three combined.

Tether scored low on transparency, low on automation, low on complexity, the worst combination in the table on the two axes the report treated as virtues. It went on to grow roughly seventy-fold and still leads the market. Basis scored high on automation and high on transparency. It never launched. The correlation between the report's chosen metrics and eight-year outcomes is close to nil. The correlation between "can this structure be licensed" and survival is close to total.

This isn't a criticism of the analysts, and that's precisely the point. In 2018 regulation was genuinely a risk item, something you tracked and mitigated. The report handles it that way, in a five-paragraph section that correctly identifies the securities question and money transmission licensing, then moves on. Nobody was building against a stablecoin rulebook because no stablecoin rulebook existed. MiCA came into full effect in late 2024. The GENIUS Act was signed in 2025. Today the licensing regime is a design input before a line of code is written, which is why we maintain the Stablecoin Regulation Tracker across 200+ jurisdictions as a live reference rather than a periodic snapshot.

The forecast didn't fail on technology. It failed because it scored the technology and not the governance.

What held up better than the forecast did

Three things in the report look sharper now than they did then, and all three sit in sections the report itself treated as secondary.

The shortest one is a paragraph on how to improve Tether: use a reputable custodian, hold reserves in cash or safe government bonds with no fractional reserve banking, make the issuing entity a deposit-accepting bank, and give holders legally enforceable rights. That is, more or less, the reserve and redemption architecture that MiCA and the GENIUS Act now require. The report described the fix accurately and assumed the market would supply it voluntarily. Legislation supplied it instead.

The second is "Type 1b: E-Money," a half-page describing EU electronic money institution licensing as a variation on the fiat-backed model, complete with the €350,000 capital requirement. Monerium, the one project in the sample pursuing that route, received its e-money licence from the Financial Supervisory Authority of Iceland in June 2019, the first such licence issued for blockchain issuance under EU rules. It issues EURe today as a MiCA-compliant e-money token. The report's own assessment of Monerium flagged a "lack of buzz and online activity" as a concern. The quiet regulatory application was the durable asset.

The third is the question the report raised and left open, quoting Agustín Carstens and asking "how big is too big?" for a private currency competing with fiat. That question is now the substance of BIS and IMF working papers on deposit migration and monetary transmission, and remains unresolved.

What this means for different readers

For banks. The report's assumption that stablecoins would evolve away from the banking system inverted. The surviving models depend on bank custody, licensed issuance, and fiat redemption, which places banks inside the architecture rather than adjacent to it. This is the reversal we examined in the 2025 definition refresh: the intermediaries were upgraded, not removed.

For regulators. The Basis outcome is a data point on the reach of existing law. No enforcement action was taken and no new rule was written. Counsel read the securities framework, reached a conclusion, and a $133 million project unwound itself. Regulatory perimeter shaped the market before regulatory text addressed the market.

For issuers. The 2018 funding split suggests capital allocation and eventual market structure can diverge sharply when the binding constraint sits outside the domain investors are evaluating.

For anyone reading forecasts. The variables a forecast treats as background are where the risk sits. In 2018 that was legal characterization. The equivalent question for a 2026 forecast is which currently-stable assumption is being held constant, whether that's dollar dominance, the shape of settlement infrastructure, or the tax and accounting treatment of tokenized instruments.

The takeaway

The State of Stablecoins got the direction of travel right. Stablecoins did become infrastructure, they did expand the user base well beyond traders, and they did draw a competitive response from central banks. The empirical work was solid and the scepticism about algorithmic designs, expressed as a list of trade-offs rather than a prediction, was better calibrated than the prediction that followed it.

What it missed is that the market wasn't going to resolve on which stability mechanism worked best. It resolved on which structures could be licensed, supervised, and redeemed at par. Design questions the report treated as central turned out to be downstream of a legal question it treated as peripheral.

That's a durable lesson about this sector, and it applies just as much to the 2035 forecasts being published now. Technology forecasts in stablecoins tend to fail on the governance axis, and the governance axis rarely appears in the table.

Source: "The State of Stablecoins," Blockchain research team, 2018, covering 57 stablecoin projects with market data through July 2018. Market capitalization and supply-share figures for 2026 attributed to DefiLlama data as reported in June and July 2026; stablecoin count attributed to Statista and CoinGecko, July 2026; Basis shutdown details attributed to CoinDesk reporting, 13 December 2018; TerraUSD pricing attributed to Congressional Research Service documentation, May 2022; Monerium licensing details attributed to the company's June 2019 announcement and subsequent Central Bank of Iceland supervision. Methodology note: this analysis works from the report's taxonomy, competitive comparison table, and forward-looking sections, comparing its stated predictions against verifiable outcomes, rather than summarizing the report's structure.

This article is for informational purposes only and does not constitute financial, investment, or legal advice.