SamΒ·2026-09-10Β·11 min readΒ·Reviewed 2026-09-10T00:00:00.000Z

The 2017 Bitcoin Bubble: From $1,000 to $19,783 and Back

Bitcoin ran from $998 to $19,783 in eleven months on Japan's 2017 legalisation, a $5.6 billion wave of initial coin offerings, and retail mania. Regulated futures launched the week of the peak; within a year the price had fallen 84 percent.

BitcoinCryptocurrencyInitial Coin Offerings2017 BubbleCrypto WinterBitcoin Futures
Source: Historical records

Editor’s Note

Bitcoin's 2017 run shows a bubble can strip away every piece of a market's usual plumbing and still trace the same arc from mania to collapse. β€” Sam

Contents

A Line Out the Door in Chicago

Shortly after five in the afternoon on 10 December 2017, traders on the floor of the Chicago Board Options Exchange watched a product that had not existed a month earlier trade for the first time. Bitcoin futures under the ticker XBT opened at $15,460, more than 5,000 dollars above the exchange's own price limits, and CBOE's trading system paused twice within the first half hour as circuit breakers tripped on both sides. Retail interest had grown so intense in the preceding weeks that Coinbase, then the largest US cryptocurrency exchange, had added roughly 100,000 new accounts a day through November and was warning customers that its servers might not keep pace with sign-ups. A week later, on 17 December, the CoinDesk Bitcoin Price Index touched $19,783, having started the year at $998. An asset that had traded for less than the price of a bicycle wheel twelve months earlier was now worth more than a used car, and a large share of the buyers had never owned a security of any kind before.

What followed was one of the fastest bubble collapses in modern financial history, unfolding almost entirely outside the regulatory architecture built for every prior mania this publication has covered. There was no central exchange, no single custodian of record, and β€” until the CBOE and CME launched futures that December β€” no venue where a professional investor could take a bearish position at any scale. The 2017 cryptocurrency bubble tested whether a market with none of the plumbing of the dot-com bubble or the Japanese asset bubble β€” no earnings, no dividends, no regulator of first resort β€” could still follow the same arc from euphoria to collapse. It did, and the record of exactly how illuminates what a bubble looks like when technology removes most of the frictions that used to slow one down.

The Slow Build, 2009 to 2016

Bitcoin had traded publicly since 2010, when a Florida programmer named Laszlo Hanyecz paid 10,000 bitcoin for two pizzas, a transaction crypto historians still cite as the first commercial use of the currency. Prices moved from fractions of a cent to roughly $1,100 by late 2013, driven partly by demand from Chinese investors working around capital controls, then collapsed after the Tokyo-based exchange Mt. Gox halted withdrawals in February 2014 and later revealed the loss of some 850,000 bitcoin to theft. The currency spent most of 2015 trading below $300, dismissed by mainstream finance as a curiosity for libertarians and dark-web merchants.

Two structural events set up the next cycle. Japan's Payment Services Act, amended in 2016 and effective from April 2017, formally recognised bitcoin as a legal method of payment and brought exchanges under Financial Services Agency licensing β€” a rare instance of a G7 regulator legitimising the asset class rather than restricting it. Bitcoin's third halving, on 9 July 2016, cut the block reward paid to miners from 25 to 12.5 bitcoin, a preprogrammed supply reduction that prior cycles had shown tended to precede price appreciation. Prices rose from around $450 at the start of 2016 to $998 on 1 January 2017, a gain that had already outpaced most conventional asset classes before the year commonly associated with the bubble had even begun.

ICO Summer and the Retail Rush

The defining feature of the 2017 boom was not bitcoin itself but the initial coin offering, a fundraising mechanism through which a company issued its own token on the Ethereum blockchain in exchange for existing cryptocurrency, typically without registering the sale as a securities offering anywhere. ICOs raised roughly $250 million globally in the first half of 2017 and then, as retail participation accelerated, closer to $5.6 billion for the year as a whole, according to tallies compiled by the research firm CoinSchedule. Some raises took minutes: the Brave browser's Basic Attention Token sold $35 million of tokens in twenty-four seconds in May 2017, and messaging app Kik's Kin token raised $98 million that September despite the company having never turned a profit in its decade of operation.

Retail behaviour mirrored, and in places exceeded, the day-trading culture of the dot-com years. Companies unrelated to blockchain technology renamed themselves to capture the enthusiasm: Long Island Iced Tea Corp, a beverage maker, rebranded as Long Blockchain Corp on 21 December 2017 and saw its stock rise as much as 289 per cent intraday despite disclosing no actual blockchain business, a move the SEC later cited when it delisted the company in 2021 for filing false and misleading statements. South Korean exchanges traded bitcoin at a persistent markup over global prices β€” the so-called kimchi premium reached 40 to 50 per cent at points in December 2017 β€” reflecting capital controls that trapped won-denominated demand inside domestic venues (Corbet, 2018). JPMorgan chief executive Jamie Dimon told an investor conference on 12 September 2017 that bitcoin "is a fraud" and "worse than tulip bulbs," a remark that generated headlines for weeks and did nothing to slow the rally that followed it.

Academic work published during and after the run identified the same speculative signature seen in earlier bubbles: price momentum trading by investors with limited fundamental basis for valuation, herding reinforced by social media and message-board enthusiasm, and a feedback loop between media coverage and new-account growth (Cheah and Fry, 2015). A widely cited later study dated the most explosive phase of the 2017 rally to the roughly six weeks between early November and mid-December, using statistical tests for explosive price behaviour originally developed to identify equity and housing bubbles (Corbet, 2018).

Bitcoin Price (CoinDesk BPI), 2016–2018

Source: CoinDesk Bitcoin Price Index

Futures, the Fork, and the Peak

Two events in the final quarter of 2017 changed the market's structure just as prices reached their highest levels. On 1 August a contentious dispute among developers and miners over how to scale bitcoin's transaction capacity produced a hard fork, splitting the network and creating Bitcoin Cash as a separate asset β€” the second major chain split of the year after a smaller one in March, and a reminder that the "bitcoin" retail investors were buying was itself a contested and evolving protocol. Then, in December, the CBOE and CME Group each launched cash-settled bitcoin futures contracts, giving institutional investors a regulated venue to short the asset for the first time. The CME contract launched on 18 December, one day after bitcoin's all-time CoinDesk peak of $19,783.

Whether the futures launch caused the top or merely coincided with it remains disputed among researchers, but the timing was conspicuous enough that it became a standard reference point in subsequent academic work on the episode (Corbet, 2018). What is not disputed is the composition of the rally's final leg. Trading volume on South Korean exchanges Bithumb and Upbit, and on smaller Southeast Asian platforms with minimal identity verification, rose sharply relative to volume on regulated US venues through November and December, consistent with a market increasingly dominated by inexperienced retail money chasing a vertical chart rather than by informed capital assessing a thesis.

DateEvent
9 Jul 2016Third bitcoin halving cuts block reward to 12.5 BTC
1 Apr 2017Japan's amended Payment Services Act takes effect, licensing exchanges
12 Sep 2017Jamie Dimon calls bitcoin "a fraud, worse than tulip bulbs"
4 Sep 2017China's central bank bans initial coin offerings and orders exchange closures
1 Aug 2017Bitcoin Cash hard fork splits the network
29 Nov 2017Bitcoin closes above $10,000 for the first time
10 Dec 2017CBOE launches the first regulated bitcoin futures contract
17 Dec 2017CoinDesk BPI peaks at $19,783
18 Dec 2017CME Group launches its own bitcoin futures contract
6 Feb 2018Price falls below $7,000 amid a broad equity sell-off
15 Dec 2018Bitcoin bottoms near $3,191, down roughly 84% from the peak

The Crash and the Crypto Winter

The decline that followed was steep even by the standards of the asset's history. From the 17 December peak, bitcoin fell 32 per cent in five days, then continued grinding lower through the first half of 2018 as regulatory pressure mounted on multiple fronts simultaneously. South Korea's Financial Services Commission announced plans in January 2018 to ban anonymous cryptocurrency trading accounts, a move markets read as a prelude to an outright exchange ban even though none materialised. The US Securities and Exchange Commission, under chairman Jay Clayton, issued a formal statement on 11 December 2017 warning that "merely calling a token a 'utility' token or structuring it to provide some utility does not prevent the token from being a security," building on the agency's earlier DAO Report of July 2017, which had concluded that tokens sold in that project's 2016 ICO qualified as unregistered securities (SEC, 2017).

Enforcement followed through 2018: the SEC brought fraud charges against the founders of two ICOs, PlexCoin and AriseBank, within weeks of each other in January, and by year-end had opened investigations touching dozens of token issuers. Trading volume across the roughly 1,600 tokens tracked by CoinMarketCap fell by more than half between January and December 2018, and total cryptocurrency market capitalisation, which had briefly exceeded $800 billion in January, closed the year near $130 billion. Warren Buffett, asked about bitcoin at Berkshire Hathaway's annual meeting on 5 May 2018, called it "probably rat poison squared," extending a running theme of skepticism from established finance that had done little to blunt the rally and did little to soften the crash.

A subsequent body of research examined whether the 2017 rally itself had been partly manufactured. Griffin and Shams (2020), analysing blockchain transaction data, found that purchases of bitcoin using Tether, a dollar-pegged stablecoin whose reserves were never independently audited during the period studied, were followed by statistically significant price increases in a pattern consistent with the token being used to prop up prices during periods of market stress β€” a finding Tether's issuers disputed but one that reframed how economists understood the rally's mechanics after the fact. Liu and Tsyvinski (2021), studying cryptocurrency returns more broadly, found they were driven overwhelmingly by investor attention and momentum rather than by any exposure to conventional macroeconomic or equity risk factors, a result consistent with a market trading almost entirely on sentiment.

What the Episode Left Behind

Unlike the railways financed by Victorian speculation or the fibre-optic networks laid during the dot-com years, the 2017 cycle left comparatively little durable physical infrastructure. What it left instead was market structure and regulatory precedent. The CME bitcoin futures contract, dismissed by some observers as the proximate cause of the crash, survived the crypto winter and became the foundation for a regulated derivatives market that institutional investors used in every subsequent cycle. The SEC's ICO enforcement actions, modest individually, established the legal theory β€” that a token sold to fund a common enterprise with an expectation of profit from the efforts of others met the decades-old Howey test for a security β€” that shaped every US crypto regulatory dispute for the following half-decade.

The comparison to the South Sea Bubble is closer on mechanics than the comparison to any twentieth-century episode: both involved an asset with no cash flow to anchor valuation, a wave of copycat schemes riding the coattails of the primary mania, and a retail public entering near the top through channels β€” coffeehouse subscription lists in 1720, mobile trading apps in 2017 β€” that had only recently become available to non-specialists. The comparison to tulip mania is more often invoked than earned; seventeenth-century Holland's futures contracts on bulbs traded within a market small enough that no systemic consequence followed its collapse, while cryptocurrency's total addressable market by December 2017 had grown large enough to draw the CME, the SEC, and the Bank of Korea into direct engagement within weeks of the peak.

Bitcoin itself did not disappear the way Pets.com or the bulk of 2017's ICO tokens did. It spent most of 2018 and 2019 range-bound between $3,200 and $13,000, then entered a second, larger cycle in 2020 and 2021 built on a different investor base β€” corporate treasuries, futures-based exchange-traded funds, and institutional allocators who had watched the 2017 episode from the sidelines. The asset that Jamie Dimon called worse than tulip bulbs in September 2017 was, by the time regulators finished writing the rules the crash had made necessary, already being custodied by some of the banks whose executives had dismissed it.

Educational only. Not financial advice.