The radio stock bubble of the 1920s remains one of the most instructive episodes in financial history, and its lessons have never been more relevant. A century ago, investors poured their savings into a revolutionary technology that promised to transform daily life. They were right about the technology; they were catastrophically wrong about the timing and valuations. The parallels to today’s artificial intelligence frenzy are striking enough to give any thoughtful investor pause.
When Radio Was the Future
In 1920, owning a radio receiver was an exotic hobby. By 1923, household radio adoption stood at roughly 1%; by 1937, it had reached 75%[s]. This was genuine, world-changing adoption. Radio brought news, entertainment, and music into homes that had never experienced anything like it. The number of radio stations in the United States exploded from 5 in 1921 to 556 by 1923[s]. Sales of radio equipment surged from $60 million in 1922 to $843 million in 1929[s].
The technology was real. The business opportunity was real. And Wall Street noticed.
The Rise and Fall of RCA
The Radio Corporation of America became the emblematic stock of the radio stock bubble. Formed in 1919 as a government-backed entity to consolidate American wireless technology[s], RCA quickly became the hottest stock on Wall Street. In January 1926, a share of RCA cost $43. By September 1929, that same share peaked at $568[s]. Over the full decade, RCA stock rose 200-fold, one of the largest increases in stock market history[s].
Then came the crash. From its September 1929 peak of 114.75 (after a 5-for-1 split), RCA collapsed to 2.625 by May 1932, a 98% decline[s]. The dividend-adjusted price did not recover to 1929 levels until the 1960s, by which time the company was making most of its money from television rather than radio[s].
The technology succeeded beyond anyone’s dreams. The investors who bought at the peak lost nearly everything.
The Anatomy of Speculation
The radio stock bubble exhibited classic signs of speculative maniaA period of irrational market behavior where investors buy assets not for their fundamental value but to resell at higher prices to others.. People sold their Liberty Bonds and mortgaged their homes to pour cash into the stock market[s]. Margin buyingPurchasing stocks by borrowing part of the cost from a broker, using the investment as collateral. Amplifies both gains and losses. became rampant; investors borrowed up to 75% of a stock’s purchase price, often paying 20% interest rates, convinced that rising prices would cover their debts[s]. By midsummer 1929, some 300 million shares were being carried on margin[s].
The mania extended beyond RCA. Simply including “radio” in a company name could send its stock soaring. Kolster Radio, a radio receiver manufacturer, rose from 10 to 95 between 1927 and 1929, then crashed below 1 by 1930 when it filed for bankruptcy[s]. The total value of the New York Stock Exchange tripled from $27 billion in 1925 to $87 billion by September 1929[s].
The AI Parallel
Today’s artificial intelligence investment boom bears uncomfortable resemblance to the radio stock bubble. By the fourth quarter of 2025, AI and machine learning accounted for 52% of global venture capital deal value, the first time the sector made up more than half of total investment[s]. Global private AI investment reached $252.3 billion in 2024, with generative AI funding soaring to $33.9 billion, more than eight times higher than 2022 levels[s].
The valuation dynamics echo the 1920s. OpenAI’s valuation surged beyond $100 billion in August 2024 despite reports of $8.5 billion spent on AI training and staffing, with projections pointing toward a $5 billion annual loss[s]. Much as radio companies could boost their stock prices by name alone a century ago, today’s AI startups attract capital based on technical metrics like model parameters and benchmark performance rather than traditional measures like profitability[s].
The Pattern Repeats
The radio stock bubble fits a broader historical pattern. Great technological innovations lead to great bubbles. Railroads in the 19th century, electricity, radio, the internet: each triggered similar cycles of euphoria, over-investment, and severe decline[s]. Historical analysis shows that most significant bubbles last approximately six years, and their gradual development makes them difficult to identify in real time[s].
The dot-com bubble offers a more recent reference point. When it burst, venture capital spending dropped 80%, falling from peaks exceeding $100 billion in 2000 to approximately $20 billion by 2002-2003[s]. Yet companies that survived the carnage, like Amazon, which dropped 92% from its peak before recovering to become a dominant company[s], demonstrated that the technology’s success does not guarantee investor success.
What History Teaches
The lesson of the radio stock bubble is not that transformative technologies are bad investments. Radio genuinely changed the world. RCA eventually recovered and thrived. But investors who bought at the peak waited three decades to break even. The technology’s success was never in doubt; the valuations were the problem.
Today’s AI moment may or may not be a bubble. The technology is certainly impressive enough to justify significant investment. But when half of all venture capital flows into a single sector, when companies achieve $100 billion valuations while projecting $5 billion losses, and when adding “AI” to a business plan can unlock funding, the echoes of the radio stock bubble grow hard to ignore.
History does not repeat exactly, but it rhymes. The investors who survived the 1929 crash were those who remembered that even revolutionary technologies can be overpriced. That lesson, a century old, remains the most valuable insight the radio stock bubble has to offer.
The radio stock bubble of the 1920s provides a canonical case study in technology-driven speculation, and contemporary scholars have increasingly drawn parallels to present-day AI investment patterns. Understanding these parallels requires examining both the structural dynamics of the 1920s mania and the mechanisms that drove ordinary investors into catastrophic losses.
The Radio Industry’s Explosive Growth
Radio’s commercial adoption followed a compressed timeline that few technologies have matched. Household penetration went from approximately 1% in 1923 to 75% by 1937[s]. The supply side expanded with similar velocity: radio station count in the United States grew from 5 in 1921 to 556 by 1923[s]. Equipment sales increased fourteenfold, from $60 million in 1922 to $843 million in 1929[s].
This growth was not speculative froth. Radio represented a genuine general-purpose technologyA technology with broad applications across many economic sectors, capable of driving widespread innovation and productivity gains over decades. with applications across entertainment, news, and eventually advertising. The Radio Corporation of America, formed in 1919 as a government-sanctioned consolidation of American wireless interests[s], emerged as the dominant player in manufacturing, patents, and broadcasting through NBC.
RCA’s Trajectory: The Radio Stock Bubble in Numbers
RCA’s stock price trajectory exemplifies the radio stock bubble’s extremes. Stanford University Press research documents that a share purchased for $43 in January 1926 peaked at $568 in September 1929, then collapsed to $15 by 1932[s]. Measured from early-decade lows to the 1929 peak, the stock rose 200-fold[s]. The subsequent crash wiped out 98% of value: from 114.75 in September 1929 (post-split) to 2.625 by May 1932[s].
The dividend-adjusted price did not recover to 1929 levels until the 1960s, by which time television had replaced radio as RCA’s primary business[s]. Investors who purchased at the peak and held through the crash waited approximately 35 years to break even in real terms.
Speculation Mechanisms
The radio stock bubble was sustained by structural features of 1920s financial markets. Margin buyingPurchasing stocks by borrowing part of the cost from a broker, using the investment as collateral. Amplifies both gains and losses. allowed investors to borrow up to 75% of a stock’s purchase price, with interest rates reaching 20%[s]. By midsummer 1929, approximately 300 million shares were being carried on margin[s]. Retail participation reached unprecedented levels; contemporary accounts describe people selling Liberty Bonds and mortgaging homes to invest[s].
The speculation extended to companies with tenuous radio connections. Kolster Radio, a receiver manufacturer, saw its stock rise from 10 to 95 between 1927 and 1929, only to crash below 1 by 1930 when the company filed for bankruptcy[s]. The total market capitalization of the New York Stock Exchange increased from $27 billion in 1925 to $87 billion by September 1929[s].
Contemporary Parallels: AI Investment
Current AI investment patterns display structural similarities to the radio stock bubble. By Q4 2025, AI and machine learning accounted for 52% of global venture capital deal value, crossing the 50% threshold for the first time[s]. Stanford research reported global private AI investment at $252.3 billion in 2024, with generative AI funding reaching $33.9 billion, more than eight times 2022 levels[s].
Valuation dynamics echo the 1920s pattern. OpenAI reached a $100 billion valuation in August 2024 despite $8.5 billion in annual spending and projected $5 billion losses[s]. Just as adding “radio” to a company name could boost valuations in the 1920s, contemporary AI companies are increasingly valued on technical metrics such as model parameters and benchmark performance rather than traditional financial measures[s].
Historiographic Framework
GMO’s research situates the radio stock bubble within a recurring pattern: transformative technologies generate extreme speculation before experiencing severe corrections. The sequence moves from railways to electricity to radio to the internet, with AI representing the latest iteration[s]. Historical analysis suggests significant bubbles typically last approximately six years, making real-time identification difficult[s].
The dot-com bubble provides a more recent comparison point. Venture capital spending dropped 80% after the 2000 peak, falling from over $100 billion to approximately $20 billion by 2002-2003[s]. Yet Amazon, despite a 92% drawdown from its bubble-era peak, ultimately emerged as a dominant company[s]. This underscores that technological success and investor success can diverge dramatically.
Assessment
The radio stock bubble offers a precise historical parallel for evaluating contemporary AI investment. Both episodes involve genuinely transformative technologies generating legitimate excitement. Both feature capital concentration in a single sector, extreme valuations detached from near-term profitability, and retail participation driven by fear of missing out. Available dry powderUninvested cash or liquid assets held by an investment fund, kept in reserve and available to deploy into future opportunities. in venture funds now exceeds $300 billion[s], providing fuel for continued investment regardless of fundamental valuations.
Whether current AI valuations constitute a bubble remains contested. What the radio stock bubble demonstrates is that transformative technology does not guarantee investor returns at any price point. RCA survived and thrived for decades after the crash, but investors who bought at the 1929 peak waited a generation to recover their capital. The technology’s ultimate success was irrelevant to those who overpaid.
