The 2017 ICO Bubble: What Happened and Lessons Learned

The setup: a token, a whitepaper, an Ethereum address

An Initial Coin Offering (ICO) is the sale of a newly issued blockchain token to the public, usually in exchange for ETH or BTC. The format became possible at scale because of Ethereum: any project could deploy an ERC-20 token contract for a few dollars and then sell that token to anyone with a wallet. By early 2017 the formula crystallised. A project published a whitepaper, ran a sale at a public Ethereum address, distributed tokens to buyers, and used the raised funds to build (in theory). Anyone could participate from anywhere. There was no underwriter, no exchange listing requirement, no audit and — for most of 2017 — no clear regulatory framework.

This is educational, not financial advice. The 2017 ICO bubble matters less for its size than for how cleanly it demonstrated what happens when retail capital meets minimal friction, minimal disclosure and maximal narrative.

What actually happened: the numbers and the names

The bubble unfolded fast and produced both the largest individual raises and the most notorious failures in the period.

• 2014-2016: precursors. Ethereum itself ran one of the earliest token sales in 2014, raising 31,000 BTC. A handful of other projects followed. None of this hit retail scale.
• Early 2017: the curve steepens. Projects begin to raise multiples of expectations within minutes. Bancor raised $153 million in three hours in June 2017. EOS began a 12-month rolling sale that would eventually raise around $4.1 billion. Tezos raised $232 million in July.
• Q3 2017: full bubble. Hundreds of projects launch each month. Whitepapers are heavily reused. Telegram groups become primary marketing venues. Anonymous teams raise eight-figure sums.
• 25 July 2017: the SEC DAO Report. The US Securities and Exchange Commission publishes a Section 21(a) Report concluding that The DAO tokens, sold in 2016, were securities under the Howey Test. The implication: many ICO tokens fit the same definition.
• Late 2017 - early 2018: the peak. Despite the DAO Report, ICO volume continues to rise into early 2018, with raises topping $11 billion in the first half of 2018 alone.
• Mid 2018 onwards: the unwind. The 2018 bear market hits hard. Most ICO tokens collapse 90-99% from peak. Enforcement actions accelerate. Many projects quietly disappear or produce minimal product.

Notable raises included Filecoin ($257M), Polkadot ($145M in 2017 plus a later round), Tezos ($232M), EOS ($4.1B over 12 months) and The DAO ($150M in 2016 before its exploit). Many of these projects eventually shipped working products; many others did not.

Who was involved

• Project founders. A mix of credible technical teams and opportunists. Some, like the Tezos founders, faced years of disputes about governance and fund control. Others were anonymous and disappeared after the sale.
• Retail buyers worldwide. Hundreds of thousands of individuals with no professional background in early-stage investing bought tokens through Telegram links and Etherscan addresses. Many bought at all-time-high prices in late 2017 and early 2018.
• Exchanges. Centralised exchanges, particularly outside the US, listed new tokens aggressively, often within days of the sale. The listings produced liquidity for early buyers — and the first exits for many founders.
• Promoters and "influencers". A new ecosystem of paid promoters emerged. Several were later sanctioned by the SEC for failing to disclose payment for promotion.
• Regulators. The SEC, CFTC, FINMA (Switzerland), MAS (Singapore) and others issued guidance through 2017-2018. The SEC was the most consequential. By 2019 it had brought dozens of enforcement actions against ICO issuers.

The aftermath: enforcement, vesting and the move to IDOs

The legal and structural consequences of the ICO bubble continued for years.

• SEC enforcement. Major cases included Telegram's TON sale ($1.7B), which was halted by a US court and the funds returned; Block.one's EOS sale ($24M settlement); Kik ($5M plus rescission); Tezos (settled for $25M); and many more. The legal pattern: even where projects shipped, the original sale was often deemed an unregistered securities offering.
• The end of large public US ICOs. By late 2018 most credible projects had stopped selling to US retail buyers in the ICO format. Subsequent token launches shifted to IDO (Initial DEX Offering), IEO (Initial Exchange Offering), private sales with US accredited investors, or KYC-gated public sales outside the US — covered in ICO vs IDO vs IEO.
• The rise of vesting and unlocks as a structural concern. Many 2017 ICOs distributed tokens with no team or insider lockups. As prices rose, insiders were able to sell into retail demand. Later token launches almost universally adopted multi-year vesting schedules — which became a separate analytical concern as scheduled unlocks created predictable selling pressure cycles after.
• The fate of capital. A large share of the funds raised in 2017-2018 was held in ETH. When ETH fell from over $1,400 to under $100 in 2018, many project treasuries lost most of their runway. This produced a wave of layoffs and project shutdowns through 2019.
• A small number of survivors built real things. Filecoin, Polkadot, Tezos, Aave (formerly ETHLend), 0x, Decentraland and others shipped working products and persisted through multiple cycles. Their success does not retroactively validate the format; it shows that some teams executed despite the surrounding environment.

The lessons

The 2017 ICO bubble was not just "a mania that ended". It produced clear structural lessons that have repeated, in different costumes, in every later launch cycle.

• Securities law is not optional. The Howey Test existed before 2017 and continued to apply afterwards. Projects that raised on the assumption that crypto was outside securities law mostly faced enforcement. That risk has not gone away in later cycles.
• Hype outruns due diligence. The combination of FOMO, anonymous teams, and easy on-chain participation produced massive raises for projects that had no working product. Every later launch format — IDO, fair launch, memecoin season — has reproduced that pattern in adapted form.
• Founder economics matter. When insiders hold large unvested allocations, the post-launch incentive is to sell into early buyer demand. Vesting and lockups exist not just for project credibility but for incentive alignment.
• Treasury risk is real. Projects that held raised capital in ETH instead of stablecoins were exposed to a 90%+ drawdown that killed many of them. Treasury management is not a side issue; it is core to project survival.
• "This time it's different" is rarely different. The structural shape — public sale, retail capital, minimal disclosure, narrative-driven valuation — has recurred in NFT mints, meme coin manias, restaking points farming, and other formats. Naming changes; mechanics persist.

It is also worth being honest about what the 2017 cycle did produce: a generation of developers who learned to ship under public scrutiny, several projects that became durable infrastructure, and a body of regulatory experience that shaped every subsequent token sale. The bubble was costly. It was also, in part, formative.

Watch how the next launch cycle structures itself

New token launches — under any acronym — continue every cycle. The structural questions are the same: who controls the tokens, when do they unlock, who is paid to promote, what is the legal posture, and what working product exists. Zippfeed tracks token-launch headlines across many sources with sentiment and importance scoring, so you can watch the launch ecosystem honestly — separate from the marketing language any single project uses. This is educational, not financial advice.