After the Bubble: Where to Go from Here — 2026 Digital Asset Market Analysis Report

Preface: Certainty and Uncertainty in Crypto

At the start of 2026, as a new cycle of bull and bear markets begins, the entire market is filled with anxiety. After October 11th, market liquidity started to dry up; for a period, aside from a few top projects and companies still standing, most teams chose to shut down or pivot.

And after the emergence of Openclaw, swept up in a new wave of technological innovation, the huge uncertainty has only deepened everyone’s panic. As market liquidity shrinks, countless crypto workers are shifting toward AI. Originally focused solely on crypto, media outlets now feature more AI-related reports in their headlines, while OGs who have been battling in this space for over a decade are pessimistic, claiming “cryptocurrency is dead.”

With the crypto bubble burst, is crypto truly dead?

Ask AI this question, and it will give you countless answers. DeepSeek will tell you that the crypto market’s dividends have vanished, and now it’s the domain of professional, compliant players—ordinary people no longer have a chance; Grok will say that this is just a bull-bear transition, which will eliminate some people but also push crypto toward a better future; Gemini will say that AI development will drive crypto to develop in tandem.

The noise is overwhelming, so we want to find our own answer. There are no new stories under the sun; we vaguely remember that in 2001, when the internet bubble just burst, the market said the same thing—and every bubble has been described this way.

So this time, we choose to study bubbles.

Even if our answer is wrong, it is our own certainty.

Chapter 1: Exploring the Cyclical Laws of History—From Railroads to the Internet, How Do Tech Bubbles Repeat in Cycles

The Glory and Collapse of Railroads and Wireless Radio: The Cycles of Industrial Revolution Bubbles

On September 27, 1825, the world’s first railway built in Britain—the Stockton-Darlington Railway—was officially opened. Three years earlier, despite opposition from feudal aristocrats and religious groups, capitalists saw the future value of this steel beast and chose to bet on it, ultimately completing the project. They believed this technology would bring them profits, but they did not realize the profound impact it would have on the entire era.

Although the first railway was initially just a branch line for canal transportation, its convenience and cost-effectiveness led the entire industry to grow rapidly like mushrooms after rain. Investors flocked to participate. During the late 1824–1825 period, at the end of the South American mining speculation bubble, these risk investors shifted their funds into railway companies. In 1836–1837, as the overall stock market strengthened, railway stock prices doubled. The UK Parliament saw opportunity and approved 44 railway companies that year, raising more capital than all previous funding in the industry combined.

The Rise, Dissolution, and Rebirth of Bubbles

Like countless bubbles after, when a new technology gains market recognition, it quickly develops into a bubble that bursts just as fast. After infrastructure improves, a new, stronger bubble forms, eventually returning to normal.

After the establishment of these 44 companies, because the railway network was not yet fully built, rail transport seemed less convenient than traditional water transport, causing the railway stock index to decline during this period. By the early 1840s, valuations rebounded and approached previous peaks. Before 1843, annual capital investment in railway companies averaged about 1 million pounds (roughly $8B today). In 1844, this jumped to 20 million pounds (20x); in 1845, nearly 60 million pounds (60x); and by 1846, it reached 132 million pounds (about $120 billion today). That same year, total new railway length hit a record 4,538 miles. Everything looked prosperous.

The Bubble Bursts and Value Returns

Undeniably, early railways were successful commercial projects, but due to investor optimism, stock prices quickly exceeded what rational valuation would justify. The first movers had an advantage, but without entry barriers, this advantage disappeared. Ample market capital combined with low technical and market entry barriers created excellent opportunities for new competitors, compressing profits for existing companies and leading to a persistent decline in industry returns—what is called “involution.”

For investors at that time, the first sign that prosperity was ending was the disappearance of large premiums on new stock issues—only high-quality companies could sustain their stock prices. For surviving railway companies, expanding and occupying prime locations was the best way to maintain valuation and competitive advantage, often leveraging bank loans to accelerate growth. Worse, since the industry was emerging, most railway companies underestimated the difficulty of construction, causing actual costs to far exceed initial prospectus estimates. Over time, these stocks turned into financial games: dividends no longer came from profits but from capital and bank loans.

Under this vicious cycle, bank interest rates kept rising. At a critical point, railway companies could no longer sustain this capital cycle, and the shine of technological capital suddenly faded. Overnight, many investors went bankrupt, and public praise for railways turned into criticism.

Faced with this, the UK government was forced to pass legislation allowing industry consolidation and to abandon nearly 20% of approved new railway projects. Surviving companies regained profitability, and a wave of mergers began. Afterward, Britain’s railway glory was no longer dazzling but more like gentle morning sunlight slowly warming the land. Although those frantic bubbles could not be recreated, they truly nourished the growth of the Industrial Revolution.

The same story later played out on the American continent.

Marconi and Wireless Radio

As a footnote to this era, with the continuous development of transportation, the distance between the world’s regions shrank. People could travel farther or communicate via wired telephones and telegraphs without leaving their homes.

Of course, the limits of information transmission speed should not stop here.

In 1865, Scottish physicist James Clerk Maxwell systematically proposed electromagnetic wave theory. Some inventors began experimenting with various radio waves. Finally, in 1895, the goddess of luck favored Italian inventor Guglielmo Marconi. Using his self-developed transmitter, he successfully made a receiver ring a bell at a distance of 10 yards. He believed the range could be extended further.

Marconi keenly saw the future commercial value of this technology, applied for a patent in 1896, and started pitching to government agencies. Soon after, he founded the Wireless Telegraph and Signal Company to develop and sell radio telegraphy equipment. As a cost of relinquishing patent rights, Marconi received 15,000 pounds (about $600,000 today) in cash and 60k pounds (about $2.8 million today) in shares, removing financial worries. That year, Marconi was only 22 years old.

From War to Market

As a rising star, Marconi quickly gained attention from all sectors. In the early days, he seized the global communication needs of the British Navy and, in 1899, sold wireless equipment to both the British and Italian navies. The first order was worth 6,000 pounds (about $250,000 today), with annual revenue exceeding 3,000 pounds (about $125,000 today).

Despite securing national-level cooperation, market doubts remained about whether this technology could generate routine commercial value. After years of trial and error, Marconi shifted from direct sales to leasing. This approach emphasized ecosystem building: allowing any product or enterprise to use wireless equipment by paying a partial rent, with the restriction that all customers could only communicate with each other through Marconi’s network.

This strategy led to the emergence of numerous radio stations and competitors.

The Birth of Radio Concept Stocks

With Marconi and other tech competitors entering, the entire radio industry flourished, attracting massive capital. Early on, despite losses on financial statements, investors remained enthusiastic: the technology and business models were still in early development, and losses were acceptable. Later, Marconi’s company was renamed RCA, leveraging the technological advantages and business network built in the US. They pooled patents from AT&T, GE, RCA, and Westinghouse, forming an impregnable business fortress, which caused RCA’s sales and profits to explode.

One person’s success lifted all boats; related upstream and downstream companies also enjoyed this wave of technological dividends. At the peak, some simply registered a company related to “radio” and easily raised funds and listed stocks. The story then repeated the railway bubble: a flood of capital and companies entered, but as dividends waned, profits shifted to bank loans as dividends, and the market collapsed—dividends disappeared. Unlike railroads, radio technology was so epoch-making that this boom lasted nearly twenty years. Once the infrastructure was complete, from radios and broadcasting stations to televisions and wireless media, the potential was so vast that the market could stay prosperous for a long time.

Eventually, the Great Depression arrived, and capital games could no longer continue. People had to seek more difficult but more practical ways to increase actual sales and profits.

The Peak of the Internet Wave: A New Social Experiment in Technology

After IBM’s attempt at personal computers and Apple’s push, the mass adoption of computers reached a new height, marking the emergence of a technology previously confined to research labs— the internet.

From Academia to Business

The origin and birth of the internet are well known, so we won’t dwell on that. What’s more instructive is how the internet transitioned into commercialization.

A decisive factor was the US National Science Foundation (NSF) deciding to relinquish control of the National Research and Education Network (NREN), transforming it into a privatized, profit-driven enterprise. During this process, many key elements emerged that enabled the internet’s widespread societal application: Apple’s PCs provided hardware, the World Wide Web offered a framework, and Mosaic provided an entry point. Coupled with the commercialization of NREN, a giant industry began its magnificent journey.

In the early days of open-source commercialization, not everyone saw the opportunity. Many companies remained conservative because they lacked awareness of the potential. Also, in the business environment of the time, industry giants preferred to expand their ecosystems through land grabbing and self-built platforms, resisting this highly open new environment. Nonetheless, this was not bad for industry development: the resistance from giants created ample market space and opportunities for new entrants.

Netscape: The First to Take a Bite

As one of the earliest companies to seize the opportunity, Netscape’s peak was a major boost for the market. In late 1994, Mosaic Communications was involved in a legal dispute over its name, leading to a name change to Netscape Communications Corporation.

Although the company still had $12 million in cash, spending $1 million monthly forced Netscape to consider a new business model. They shifted from their previous service model to a freemium approach: 30-day free trials plus a $49 subscription fee. Thanks to the superior performance of their product, they quickly gained a large market share. Their goal was to boost market valuation, but this tactic proved so effective that in their August 1995 IPO, Netscape raised $140 million, catapulting it to the peak.

However, success also brought downfall. The sales strategy made Netscape overconfident, and while basking in IPO joy, they failed to build a moat. They neither acquired upstream/downstream companies nor improved their products, even dismissing industry cooperation, choosing instead to do nothing.

The outcome was clear: once the market discovered this huge cake and saw that Netscape had validated its deliciousness, many competitors flooded in. Netscape was eventually acquired by AOL.

A Whale Falls, All Things Grow

Netscape’s story is lamentable, but overall, it was a meaningful event for market development. Countless profit-seekers and innovators joined the adventure, giving birth to many dazzling projects. Almost simultaneously, Jerry Yang and David Filo spent much time researching browsing needs and created an efficient information indexing system called “Yahoo,” while Sergey Brin and Larry Page at Stanford explored faster ways to find information online. These ideas spread globally, inspiring Jack Ma to start developing the “China Yellow Pages.”

The Extreme of Concept Bubbles

Compared to the railroad and radio tech bubbles, internet tech’s entry barrier was much lower. You didn’t need to hire workers to build railways or obtain government licenses. As long as you understood internet knowledge, you could do anything. The huge wealth effect combined with low barriers sparked a market frenzy.

Initially, markets were cautious, but when they saw that simple garage startups like Yahoo and Google could generate enormous profits with innovative business models, they realized the old valuation logic was failing. As internet tech stocks’ prices soared, investors threw all doubts aside. Eventually, for value investors, the valuation of the TMT sector was recklessly exaggerated, and everyone thought it was fine.

As company valuations approached extremes, analysis standards also distorted. Usually, higher stock prices led analysts to assign higher valuations based on profit models. To keep valuations reasonable, they shifted from profit-based metrics to revenue, then to “click-through rates,” “retention,” and other concepts to project future market potential. The logic seemed sound, but the fatal flaw was the lack of historical cases—how to ensure the validity of business model analysis? The only way was to listen to the founders’ stories.

In the end, people stopped paying for technology’s practicality and started paying for stories—whose business story is more convincing, with a broader outlook, and thus more likely to attract funding. A real FOMO (Fear of Missing Out) market began. Initially, people carefully designed business models, but as the market grew impatient, some found that even if their business had nothing to do with the internet, just registering a website could be classified as a TMT project and enjoy market benefits. Some pioneering projects emerged, like online shopping, food delivery, and even pet care. But the problem was, with infrastructure still incomplete, stories remained just stories.

The same ending played out again: for listed companies, only a few truly adapted to the times and survived; most relied on bank loans to sustain the false prosperity. When interest rates hit a critical point, the market collapsed.

Data-Driven Bubble Indicators: How Internet Valuations Went Off the Rails

The story is simple, but to find more valuable insights, we need to convert these narratives into quantifiable macro-financial indicators and identify patterns. This section uses the internet bubble (1995–2002) as the core sample, supplemented by data from the 1929 Great Depression, analyzing four macro dimensions—valuation metrics, monetary environment, capital flows, and real economy—to systematically trace the evolution of macro data during the bubble lifecycle. These regular patterns will serve as a “constant” benchmark for the subsequent analysis of crypto market cycles.

Extreme Price-to-Earnings Ratios (P/E)

The most direct bubble signal is in valuation metrics. In every tech bubble, market optimism about new tech drives valuation multiples higher, eventually detaching from any rational fundamentals. This is a gradual “anchoring drift”: investors accept increasingly absurd valuations until the entire valuation system fails.

During the internet bubble, the NASDAQ Composite’s P/E ratio peaked at about 200x in March 2000, far exceeding Japan’s 60–80x during the 1989 bubble. This meant investors were willing to pay $200 for every $1 of current earnings—implying it would take 200 years to recover costs if profits remained flat. Notably, over half of the tech companies listed on NASDAQ at the peak were unprofitable, making meaningful P/E calculations impossible.

Meanwhile, the S&P 500 hovered around 29–33x in 1999–2000, with over 45x appearing around 2002—long-term average being about 15–20x. In March 2001, NASDAQ’s P/E was still as high as 175x, indicating that even as the bubble burst, valuation normalization was far from complete.

Shiller CAPE Ratio: A Century-Long Warning

Nobel laureate Robert Shiller’s cyclically adjusted P/E (CAPE, also called P/E 10) smooths short-term fluctuations by using the average inflation-adjusted earnings over the past ten years. It’s widely regarded as one of the most reliable long-term valuation indicators. Over 140+ years of data since 1881, the median CAPE for the S&P 500 is 16.04x, with a mean of about 17.17x.

At three iconic bubble moments, CAPE exceeded the “danger threshold” of 30x: before the 1929 crash, it hit 32.56x, leading to an 89% market decline and a recovery only by 1954; during the 2000 dot-com bubble, it peaked at 44.20x, with the S&P 500 falling 49% and NASDAQ dropping 78% from 2000–2002. Over the decade, investors earned an annualized real return of about -1.4%. Historically, when CAPE exceeds 30x, the following ten years’ annualized real return averages only 0–3%, well below the long-term average of about 7%.

It’s important to note that CAPE is not a “timing tool”—a high CAPE does not predict when a crash will happen, but it effectively signals low returns over the next decade. As the Minneapolis Fed’s research states, after the 2000 tech bubble burst, the impact on the real economy was mild, but the destruction of wealth for investors was profound.

Extreme Divergence in Price-to-Sales (P/S) Ratios

Because many companies listed at the bubble’s peak were unprofitable (over half of NASDAQ tech firms in March 2000), P/E ratios lost their meaning. Therefore, the price-to-sales ratio (P/S) became a more reliable indicator of bubble severity.

CFA Institute’s research shows that in March 2000, the median P/S for “Internet Content” companies reached 32.44x, while by September 2020, the same category’s median P/S was only 3.15x—a gap of over 10x. The median P/B ratio for semiconductors also fell from 13.85x in 2000 to 3.32x in 2020.

The Double-Edged Sword of Monetary Policy: Loose Policy Fuels Bubbles, Tight Policy Bursts Them

Behind every major asset bubble, loose monetary policy is almost always involved. The level of interest rates determines the opportunity cost of capital. When risk-free yields are very low, funds naturally flow into high-risk, high-return assets, fueling speculation. When central banks tighten, raising rates and increasing borrowing costs, bubbles become vulnerable.

Loose Cycle: The Bubble Catalyst. The internet bubble’s monetary environment in the mid-1990s was characterized by loose policy. From 1995 to 1998, under Greenspan, the Fed kept rates around 5.25–5.5%. The critical event was the LTCM crisis in fall 1998, which prompted three rate cuts, lowering the federal funds rate from 5.5% to 4.75%. Goldman Sachs reviewed this period, noting that the rate cuts “released abundant liquidity,” directly fueling NASDAQ’s rise from 11% of NYSE trading volume in 1990 to 80% of total market cap in 1999. The “insurance cut” in 1998 greatly boosted investor confidence—The Globe.com’s IPO in November saw a 600% first-day surge, a record on Wall Street.

Tight Cycle: The Bubble’s End. Starting June 1999, the Fed recognized the risk of overvaluation and began raising rates. Over ten months, the federal funds rate increased six times from about 4.75% to 6.5%—a level not seen since January 1991. The discount rate also rose to 6%, one of the highest since August 1991. These hikes increased borrowing costs, making bonds and other fixed-income assets more attractive relative to high-risk tech stocks, leading to capital outflows from speculative assets.

It’s crucial to understand that rate changes are not the sole trigger but a key catalyst. External shocks, like Japan’s recession announcement on March 13, 2000, and the Barron’s cover story on March 20 warning that internet companies were running out of cash, combined with aggressive accounting by firms like MicroStrategy, created a perfect storm. The confluence of rising rates, external shocks, and waning confidence triggered the bubble’s collapse.

After the bubble burst, the Fed quickly responded. In 2001, it cut rates 11 times, bringing the federal funds rate from 6.5% down to 1.75%, one of the fastest easing cycles in history. However, the labor market’s deterioration persisted—by June 2003, unemployment hit 6.3%, well after the bubble’s end in November 2001. The lag in monetary policy’s transmission to the real economy is a key factor in understanding the aftermath.

Capital Flows and Leverage: From VC Frenzy to Retail Margin Trading

If valuation metrics are the “thermometer,” monetary policy is the “fire,” then venture capital (VC), IPO markets, and margin debt are the “fuel” that fuels the bubble’s growth.

Venture Capital: From Selective to Spreading Money. During the bubble, VC investment exploded. According to NVCA data, US VC funding rose from about $8 billion in 1995 to a peak of roughly $105 billion in 2000 (nominal). Over five years, it grew 13-fold. In 1999, 39% of US VC went to internet companies. This capital frenzy led to a sharp decline in project quality—many startups with no clear path to profitability could get huge funding just for a “.com” domain.

Post-bubble, VC funding dried up. In 2001, investments plunged to about $36.5 billion—still the third-highest year but down over 67% from the peak. CFA research shows that the internal rate of return (IRR) for VC funds in 1999 was -4.29%, and in 2000, -2.51%. This means that institutions that entered at the peak generally lost money.

IPO Market: From Boom to Freeze. IPO volume is a sensitive indicator of speculative sentiment. The peak was 677 IPOs in 1996. It then declined to 474 in 1997, 283 in 1998, before rebounding to 476 in 1999. In 2000, 380 companies went public. After the crash, IPOs plummeted to just 80 in 2001—less than a quarter of the previous year. Over 280 VC-backed companies went public in 1999, many with first-day gains over 100%. VA Linux’s IPO in December 1999 surged 698% on the first day, still one of the most extreme IPO performances.

Margin Debt: Leverage at Its Peak. Margin debt measures market leverage and speculative sentiment. In the late 1990s, retail investors flooded into stocks, causing margin debt to surge. In March 2000, it hit a peak of about $300 billion (nominal), roughly $500 billion in today’s dollars. Margin debt as a percentage of nominal GDP reached 2.6%, close to the 2.5% before the 2007 subprime crisis, and soared past 3.97% in 2021.

During the 2000 crash, retail investors did not withdraw but increased their participation. Data shows they invested about $260 billion net in stocks in 2000, more than in 1998 and 1999. By the end of 2002, about 100 million retail investors had lost around $5 trillion in market value. Vanguard’s research indicates that by 2002, 70% of 401(k) accounts had lost at least 20%, illustrating the typical retail behavior during bubbles—while institutions and insiders cashed out, retail investors often became the last bagholders.

Lagging Transmission to the Real Economy: GDP Contraction, Employment Collapse, and No Jobless Recovery

The impact of asset bubbles on the real economy is often delayed, spreading along a transmission chain from financial markets to corporate investment and then to labor markets. After the internet bubble burst, the US economy experienced a mild GDP decline but deep, lasting scars in employment and corporate investment, exemplifying a “jobless recovery.”

GDP: Shallow Recession, Deep Wounds. The NBER defined the 2001 recession as lasting from March to November—about 8 months, relatively short. Real GDP grew only 1.0% in 2001, far below the pre-bubble expansion (4.8% in 1999). In 2002, it rebounded to 1.7%, driven by consumer spending and housing. With loose monetary policy and fiscal stimulus, growth accelerated to 2.8% in 2003 and peaked at 3.8% in 2004, showing a gradual recovery.

Employment: Severely Lagging. The unemployment rate rose from 4.0% (a 30-year low) in late 2000 to 6.3% in June 2003—more than a year after the recession ended. This “jobless recovery” pattern is key to understanding the aftermath. The Labor Department reports that in 2001 alone, about 1.735 million jobs were lost, and in 2002, another 508k. The tech sector was hit hardest—estimates suggest Silicon Valley lost about 200k jobs from 2001 to early 2004. The unemployment rate only returned close to 2000 levels by late 2006, making it one of the longest recoveries since WWII.

Summary: A Four-Stage Macro Model of Bubble Evolution

Synthesizing the data from these four dimensions, we can outline a macro four-stage model of tech bubble development and burst, validated repeatedly in the railroad bubble, 1929 crash, and 2000 internet bubble:

• Stage 1: Mild Valuation Deviations (Emergence). New tech appears, early adopters and professional capital lead the way. Valuations slightly exceed historical averages (e.g., CAPE > 20), but remain explainable. Rates are low, VC grows steadily, market sentiment is optimistic but rational.

• Stage 2: Accelerated Valuation Rise (Frenzy). Loose monetary policy plus positive narratives create positive feedback. CAPE exceeds 30x “danger threshold,” P/E and P/S reach extreme levels (e.g., NASDAQ P/E > 100x). VC investment surges, IPOs and first-day gains hit record highs. Retail participation and margin debt soar. Market participants start using new frameworks to justify extreme valuations.

• Stage 3: Valuation Collapse and Liquidity Dry-up (Crash). Rate hikes or external shocks trigger confidence collapse. Asset prices plummet 50–80% within months. VC funding dries up (down over 60%), IPO market freezes, margin calls cascade. Assets with extreme valuations (loss-making, conceptual projects) first go to zero.

• Stage 4: Real Economy Transmission and Long Recovery (Adjustment). GDP contracts mildly, but employment losses are severe and prolonged (unemployment peaks 2–3 years late). Corporate investment shrinks sharply. Market shifts from speculative growth to focus on profits and cash flow. Regulations tighten (e.g., Sarbanes-Oxley). Surviving companies (like Amazon) undergo valuation re-rating over years, eventually becoming new growth engines. Full index recovery may take 5–15 years or more.

The core insight is that, despite different variables (technologies, market structures, participants), the underlying macro patterns are remarkably consistent:

This is the “invariant” basis we will rely on when mapping this model onto the crypto market in subsequent chapters.

Chapter 2: Multiple Bull-Bear Cycles in Crypto—Unique Economic Trends of Blockchain

Bitcoin’s Independent Evolution: From Cryptography Experiment to Institutional Risk Asset

Most past bubbles are now history, but we are in a new one. If you’ve experienced multiple crypto bull-bear cycles, you might resonate with these bubble patterns.

In this chapter, we use Bitcoin as the core metric, analyzing the crypto cycle’s similar yet distinct lifecycle. These paths reflect common human speculative behaviors but also show clear differences due to blockchain’s decentralization, global immediacy, and tokenomics. After bubbles, some tracks survive as foundations for new cycles, some are eliminated, and others transform—this is the core of crypto’s cyclical nature.

On May 22, 2010, programmer Laszlo Hanyecz posted on Bitcoin Talk, offering 10,000 BTC for two Papa John’s pizzas. This became the first real-world purchase with Bitcoin. At the time, those 10,000 BTC were worth about $41; today, they are worth over $1 billion. This event is forever remembered as “Bitcoin Pizza Day,” vividly recording Bitcoin’s transformation from an almost worthless tech experiment to a core asset held by global institutions.

Data shows that each peak market cap growth rate since 2013 has gradually converged—from about 88x in 2013 to a collapse amplitude decreasing from 93% to about 47–48% in the 2024–2025 cycle (as of March 2026). Behind this trend is the ongoing divergence between Bitcoin and altcoin cycles: in 2025–2026, Bitcoin dominance remains stable around 58.6%. Since the approval of US spot Bitcoin ETFs in 2024, net inflows have exceeded $55 billion, with products like BlackRock’s IBIT leading.

This number far exceeds any previous crypto product and directly reflects that institutions now treat Bitcoin as an independent allocation asset, not just an altcoin derivative. In contrast, the overall crypto market remains highly speculative: in early bull markets, narratives are dense, and structural opportunities abound; later, homogenous projects flood the market, diluting liquidity.

Most altcoins die or slowly exit after the bull, mainly due to lack of real users and products, narrative disproof, and liquidity exhaustion after sharp price declines. This pattern was most thoroughly demonstrated by Terra-Luna in 2022: Luna’s market cap once hit $40 billion, but after UST’s de-peg, the core narrative was discredited, and market cap collapsed to near zero within days. On-chain data shows TVL dropped from a peak of $180 billion to less than $10 billion, with no real user recovery. According to DefiLlama, from 2021 to 2025, over 70% of DeFi and meme projects’ TVL fell more than 90%, with most entering chronic death—low trading volume, halted developer activity, and market oblivion.

From zero to $13.6 trillion, a simple overview of Bitcoin’s evolution helps us intuitively understand how its consensus value develops:

Unique Bubble Mechanics: Decentralized Speculation, Tokenomics, and Network Effects Amplification

Crypto bubble formation is similar to the internet bubble but with differences. The latter was mainly driven by VC, while crypto is amplified through decentralized speculation, tokenomics, and network effects. The 2017 ICO boom is a prime example: about $5.3 billion raised that year, with many projects funded solely by whitepapers, and failure rates of 46–59%.

During 2020–2022 macro liquidity expansion, stablecoin supply grew from about $5 billion to over $150 billion (stabilizing around $310 billion in 2026), further fueling leverage and speculation. When external liquidity recedes, tokens lacking self-sustaining mechanisms reveal problems: high inflation designs and short-term incentives dominate, relying on continuous external funding.

This cycle’s core can be explained via Everett Rogers’ diffusion of innovations (S-curve). The bull market is essentially the narrative spreading from early adopters—programmers, crypto VCs, tech enthusiasts—who enter driven by faith; then influencers, traders, and Web2 players amplify the story via social media; then the mass—ordinary workers, students, small businesses—join due to wealth effects; finally, laggards—those with low tech skills, relying on short videos and leverage—become the last wave. When adoption nears 80–90%, marginal participation drops sharply, and the bull ends. This was evident in Solana meme coins in 2024, with over 5,000 new tokens in a month, diluting attention and funds across many projects, most of which saw trading volume collapse within months.

When narratives reach society’s tail and early investors start cashing out, the upward price structure collapses, ending the bull. From a behavioral perspective, the end of prosperity begins when most of the population has been mobilized by speculative narratives.

Cycles and Differences: Internet Bubble vs. Crypto Bubble

In the internet bubble, the NASDAQ P/E ratio peaked at about 200x in March 2000, with many tech firms unprofitable (over 50%). The median P/S in “Internet Content” was 32.44x, while crypto valuations are more directly reflected by TVL/market cap ratios and fully diluted valuation (FDV):

In 2021, DeFi TVL peaked at about $180 billion, while total crypto market cap approached $3 trillion. Some meme and emerging projects’ FDV/TVL ratios exceeded 100x. From 2021–2025, over 70% of altcoins maintained high FDV despite large TVL declines, causing valuation systems to break down—much more extreme than the money-burning rate in the internet era.

To understand why, we can analyze differences in market participants and macro factors.

Fundamental Participant Structure: Grassroots Victory

The internet bubble was led by institutions—VCs and investment banks—with retail mainly participating indirectly via stocks. In contrast, crypto’s decentralized ethos means it’s mostly driven directly by global retail investors. In 2021, new addresses surged by tens of millions, spreading from tech geeks to the broader society.

This structural difference amplifies volatility: retail leverage is higher (perpetual contract holdings often over 60%), making black swan events more impactful than in the institutional era. The FTX collapse in 2022 is a prime example: over $20 billion in liquidations in a month, with altcoin prices dropping over 60%.

The Devastating Impact of Macro Tightening

The Fed’s tightening cycles are the common end for bubbles, but their impact on crypto is more rapid and thorough. During the internet bubble, six rate hikes from 1999–2000 (from 4.75% to 6.5%) caused NASDAQ to crash 78%. In crypto, the 2022 tightening was the most aggressive: rates soared from 0% to 5.25–5.50% (a 525bp increase), and QT drained about $2.4 trillion from liquidity (balance sheet shrank from $8.9T to $500B). Higher rates raised the opportunity cost of speculation, and balance sheet reduction drained on-chain funds. High-beta, leveraged, liquidity-dependent altcoins bore the brunt, with market cap falling from about $2.5 trillion to under $400 billion—far more than Bitcoin’s decline.

Black Swans and Liquidity Drain Chain Reactions

Crypto’s global immediacy and cross-chain features amplify black swan events and internal liquidity drain effects. In the internet bubble, external shocks (like MicroStrategy’s accounting restatement in 2000) mainly affected individual sectors; in crypto, the FTX event triggered a chain reaction across the entire chain: massive liquidations + rapid listing of homogeneous projects (over 5000 new tokens on Solana in a month), diluting funds and collapsing trading volumes, leading to long-term stagnation. Data from 2022–2023 shows altcoin trading volume shrank over 85% from peak, developer activity (GitHub commits) remained low, confirming systemic liquidity exhaustion.

Crypto Bubble Liquidity Evolution

Token issuance in crypto is growing exponentially, diluting overall liquidity and making bubbles hard to sustain. According to Bobby Ong of CoinGecko, tracking 209 chains, 1450 DEXs, and nearly 5.5 million tokens, the creation rate is astonishing.

He predicts that in five years, about 1 billion new tokens could be created. This flood results from low-entry tools like pump.fun on Solana, allowing anyone to create tokens in seconds at minimal cost. This astronomical number confirms that such mechanisms greatly dilute liquidity across memes: speculative funds are constantly attracted to new projects, preventing concentration and fostering proliferation.

On-chain data further supports this dilution: a year ago, the “State of Crypto 2026” report by 21Shares showed that by 2025, Layer-2 activity would decline 61%, with most L2 ecosystems turning into “zombie chains.” Only Base remains profitable at $55 million, while others are deeply loss-making. This reflects the flood of new projects diverting funds from old ones, shrinking bubble size and expectations, accelerating decay.

Most projects lack stable cash flows—only a few generate occasional income, and mechanisms to directly boost token value (buybacks, burns, dividends) are rare. In this environment, project competitiveness depends almost entirely on narrative appeal and attractiveness, not underlying economic sustainability.

Thus, in highly speculative markets: Narrative itself becomes the core asset for attention and liquidity. Projects within the same narrative track are hard to outcompete with “dimensionality reduction,” but lack mechanisms for value capture and technological differentiation. This causes speculative funds to flow repeatedly among competitors, preventing sustained capital concentration. The result is fragmented liquidity: attention is diluted, narrative fatigue accumulates rapidly, investor expectations decline, and large-scale withdrawals trigger a shift from boom to bust.

In crypto, a new narrative often follows a similar path:

Take Ethereum Layer-2s and Pump.fun as examples, where this mechanism is especially evident.

First, the L2 narrative: centered on Ethereum scalability, Vitalik Buterin’s rollup-centric approach laid the foundation. The main chain handles security and settlement, while L2s handle execution. Initially, L2s attracted attention, fueling the sector’s prosperity. But the narrative became highly homogeneous—Optimism, Arbitrum, zkSync all focus on throughput and cost reduction, with no real moat. As a result, capital kept flowing out:

• Optimism’s OP