Bubbles and Crashes
The Boom and Bust of Technological Innovation
Brent Goldfarb and David A. Kirsch


Chapter 1


What Is a Bubble?

In January 1926 a share of the Radio Corporation of America (RCA), the leading radio manufacturer, patent holder, and broadcaster of the day, could be had for $43 on the New York Stock Exchange. The same share peaked at $568 in September 1929 but cost only $15 in 1932. RCA’s dividend-adjusted price did not recover to 1929 levels until the 1960s. By that time, the company was making most of its money from television rather than radio. By any measure imaginable, investors were better off avoiding RCA stock in 1929.

The RCA story is not unique. In early 1637, prices for some tulip bulbs in Amsterdam were briefly on the order of seventeen years’ wages, before collapsing by 99.99%. For a time in 1998, plush toys known as Beanie Babies sold for $5,000 each, and trading in such toys accounted for 10% of eBay transactions.1 Today, most of these toys can be had for $10, a 99.998% collapse. Were these bubbles? If not, it is hard to imagine what would qualify as one.

This question of whether an event is a bubble is confusing without some discussion of what we mean by bubble—and we will have difficulty answering the central question of our book, “When are there not bubbles?” without careful attention to this. Therefore, we take a moment here to define what we mean by a bubble. This becomes even more interesting when we consider that bubble is a loaded term in some academic circles. Some financial economists contend that bubbles do not exist at all!

Academics who study financial markets traditionally define a bubble as a deviation from fundamental value. A subset of these scholars, the true believers in efficient markets, deny that prices can deviate from fundamental value. They do not deny that prices rise and fall; rather, they contend that prices always reflect nothing but investors’ reasoned beliefs about the asset’s fundamental value. But the prices must summarize the beliefs that investors have about future profits or value. Most economists call price expansions and collapses bubbles only if investors are paying more than future profits or expected share-price increases would justify—or, more specifically, more than a reasoned or “rational” investor would expect the market to price the asset in the future. If investors purchase an investment asset that they know will be worth less when they expect to sell it, this is clearly foolish.2 For example, paying full price for a Christmas tree on December 26 with the purpose of reselling it is foolish, or in academic speak, “irrational.” Following common practice, we define bubbles as extreme price fluctuations associated with fools and foolish behavior.

To stay true to this terminology, we refer to a rise in prices followed by a sharp decline as a boom and bust episode.3 Bubbles are those boom and bust episodes in which investors drive up prices and get fooled. Perhaps the investors are newcomers or naïfs, but however we identify them, their essential behavior consists in getting fooled. If reasonable beliefs justified the high prices, then boom and bust episodes are better thought of as examples of investors making reasonable bets that turn out poorly.

The distinction between a foolish investor and one making a bad bet may seem arbitrary. Were the ill-fated investors who bought RCA stock in 1929 buying because they were irrational (and therefore foolish), or did they simply overestimate the expected value of future cash flows accruing to the company? Both look the same in hindsight—a high price followed by a collapse, with many people losing money. However, for financial economists and the policy makers they advise, assumptions about investor beliefs are critical: if the cause was foolishness, perhaps we should consider a policy response to prevent such events in the future. In contrast, if the rearview mirror is showing us just a reasoned bet gone bad, then there is no room for any policy response. In this view, the result is not a bug but a feature of capitalism. We need to experiment to find the correct solution, and capitalist incentives are funneling money to such experiments. To summarize, a boom and bust episode is a bubble if it is associated with an increase in investors who are unlikely to have the tools or experience to understand whether any given price is reasonable, or if we have compelling evidence that investors were justifying their investments on the basis of particularly foolish arguments.

Although bubbles can arise in many different asset classes, this book is about technologies. It is simply much more interesting for us, and intellectually compelling for you, to learn about the history of airplanes as we study bubbles than, say, the history of Beanie Babies. Airplanes shape our experience today to a much greater extent than cute stuffed animals.4 In Chapters 5 and 6, we argue that because the causes of bubbles are based in fundamental characteristics of markets, human psychology, and the interaction of those two things, our conclusions generalize to many different asset classes, including Beanie Babies, real estate and mortgage-backed securities, and other objects of speculation.

To compare episodes across time periods and industries, we also need to take into account the fact that some stocks will be naturally more volatile than others. When ventures are inherently hard to assess because they are trying to do something that is actually new, investors will find it difficult to find similar ventures to which to anchor their assessments of value—but unless investors’ analogies are close, the resulting assessments of value will rarely be accurate. Because of this, every bit of new information causes investors to reassess. If the information is ambiguous or the analogy imprecise, investors will arrive at different opinions about the value, which will lead to volatile trading patterns: there will be numerous price movements as investors trade on their various guesses. That is, we will see volatility. Contrast this to, say, a stable, boring utility business that is well understood. New information is unlikely to be weighted as heavily given a reliable and predictive track record. Different opinions about value will be rare, and volatility lower. With more uncertain stocks, it will take more severe swings to raise our scholarly suspicions that something is amiss. Alternately, better-understood assets will trigger our interest with more modest deviations in value, simply because such patterns are likely to be more exceptional with respect to historical price trends.

Over the course of the next few pages, we introduce several cases, each an instance of a significant technological innovation that either was or was not associated with the formation of a boom and bust episode. Not only do we find the cases interesting in their own right; these examples also provide context for us to introduce the sample of technologies that we will analyze throughout the book. For each one, we spend some time contextualizing how the technology and investment opportunity was viewed in the day and documenting the price fluctuations. We then consider how we might compare episodes across time and technology.

“Talking at a piece of sheet iron”: The Telephone, 1878–1889

“The very idea of talking at a piece of sheet iron was so new and extraordinary that the normal mind repulsed it. Alike to the laborer and the scientist, it was incomprehensible. It was too freakish, too bizarre to be used outside the laboratory and the museum. No one, literally, could understand how it worked; and the only man who offered a clear solution to the mystery was a Boston mechanic, who maintained that there was ‘a hole through the middle of the wire.’


Historically, the conditions necessary to create a bubble have been rare simply because the markets for assets have been limited. The investment history of the telephone illustrates this clearly. When Alexander Graham Bell demonstrated the telephone on August 4, 1876, the ability to transfer voice in real time across distance was incredible, if rudimentary. The telephone business model was neither obvious nor uncontested, and success of the platform was not ensured. Even Bell’s inner circle was unsure of how to build a profitable business around the invention. In mid-1879—three years after Bell first demonstrated the device—when National Bell sought additional working capital to support geographic expansion, investor demand was weak. Par-value shares priced at $100 fetched only $50 in the contemporary equivalent of a failed IPO.5 We do not know whether investors were uncertain about the basic functioning of the technology or doubted demand for the initial value proposition or worried about potential competition. All three concerns would have been valid. The feasibility of the telephone as a network connecting any subscriber to any other subscriber was still very much a work in progress. The exchange-based architecture that supported any-to-any communication had only been introduced in New Haven, Connecticut, in 1878. Telephone numbers were introduced in 1879. Challenges to the enforceability of Bell’s patent were looming, as was competition from the leading telegraph operator of the time, Western Union. Investors likely debated all these issues, resulting in the first tranche of Bell shares being overpriced relative to market demand.

Soon thereafter, however, Western Union and National Bell Telephone Company settled outstanding legal claims and agreed not to compete in each other’s respective lines of business. Investor interest strengthened, and as Stehman describes it, “a mad rush for the stock” followed. Within a matter of months, the final tranche of the same $100-par-value shares that had been hard to place at $50 sold for $600 each. The fact that the agreement with Western Union calmed investors’ fears suggests that investors had doubted Bell’s ability to successfully compete with a well-capitalized incumbent. As noted, Western Union and Bell would soon come to occupy distinct lines of business, but at this early stage, basic features of the telephone industry were not yet fixed, and investors saw the resolution of the potential competitive threat from Western Union as an important positive signal for Bell.6

Was there a bubble in Bell stock? On the one hand, the reported stock price surged 1,200% in less than a year, and it is unlikely that the investors who purchased shares in the final tranche at $600 profited from their investment for some time.7 But who were the buyers, and were they investors participating in a bubble or simply rewarding the firm for having successfully resolved underlying uncertainties and thereby paving a path for successful growth? Unfortunately our sources do not enable us to definitively answer these questions.

What we do know, however, is that it was difficult to invest in telephone stocks until at least 1881. Shares of the early Bell companies like New England Bell Telephone Company (1877), National Bell Telephone Company (1879), and American Bell Telephone Company (1881) were placed privately, with scant secondary trading on the Boston Stock Exchange. There were 338 investors who owned the 8,500 shares of National Bell Telephone and 12 who owned 4,795 shares (56.4%), but even the remaining “small” investors had average holdings larger than 11 shares each (more than $1,100 in 1879 at par value).8 We note that at this point in time, $1,100 was three times the amount an average American worker earned in an entire year.9 Thus, direct, public access to the Bell stock book was limited.

Access to the market was easier after the American Bell Telephone Company’s private placement was complete. An index of telephone stocks from the 1880s was not available, so we dug into old Boston newspapers to reconstruct one on our own. We pick up the price trend on October 22, 1881, when the price was $145 per share. The price rose to $296 by March 1883. In late 1883, Bell’s general manager Theodore Vail announced to shareholders, “The telephone business has passed its experimental stage.”10 Nevertheless, the price dropped by 50% to $150 by the summer of 1884, perhaps because patent litigation continued to cast a shadow on the company’s prospects. It wasn’t until 1887 that the Supreme Court granted Bell expansive rights over all transmission of voice using wires and electricity.11 Thereafter, the price rebounded, hovering around $200–$230 through late 1889. The stock was volatile. Should we call this a bubble?


1. While it is not disputed that Tulip prices collapsed in this way, it is disputed whether there were many, if any, transactions completed at these asking prices. See Peter M. Garber, “Famous First Bubbles,” Journal of Economic Perspectives: A Journal of the American Economic Association 4 (1990): 35–54; Earl A. Thompson, “The Tulipmania: Fact or Artifact?,” Public Choice 130 (2007): 99–114. On the Beanie Baby Bubble, see Zac Bissonnette, The Great Beanie Baby Bubble: Mass Delusion and the Dark Side of Cute (Portfolio, 2015).

2. Alternatively, “rational” bubbles occur when investors understand that the asset is overvalued, but also believe that others will purchase the asset at an inflated price. Thus, it is rational to buy, since a profit will be made by selling to some “greater fool.” Financial economists use various strategies to determine if prices at a particular point in time actually reflected future profits (measured after the fact). This strategy has two shortcomings. First, beliefs may have been well informed but wrong. Or beliefs may have been crazy but right.

3. Bubbles unfold over time. While we have recently witnessed the rise of so-called “flash crashes,” these are a new phenomenon and likely differ in fundamental ways from the larger, macro-economic events in which we are interested. Stock bubbles expand (or to keep the metaphor precise, “inflate”) over time as more and more investors (a) become aware of the existence of a new investment opportunity and (b) decide to invest in it.

4. Bissonnette, Great Beanie Baby Bubble, chap. 3.

5. J. W. Stehman, The Financial History of the American Telephone and Telegraph Company (Houghton Mifflin, 1925).

6. Overall, the 2,000 shares yielded $430,000 for the National Bell treasury. The implied average price of $215 per share was well in excess of par; see Stehman, Financial History of the American Telephone and Telegraph Company.

7. Stehman (Financial History of the American Telephone and Telegraph Company, 19) observes that “the $600 per share paid for last 500 shares of the National Bell’s stock was probably the highest price at which any considerable amount of the stock changed hands.”

8. Stehman, Financial History of the American Telephone and Telegraph Company, 19.

9. Average annual earnings were $347 in 1880. Clarence D. Long, “Wages and Earnings in the United States, 1860–1890,” NBER Books, 1960, https://ideas.repec.org/b/nbr/nberbk/long60-1.html.

10. Claude S. Fischer, America Calling: A Social History of the Telephone to 1940 (University of California Press, 1994). 41.

11. Christopher Beauchamp, “Who Invented the Telephone? Lawyers, Patents, and the Judgments of History,” Technology and Culture 51, no. 4 (2010): 854–78.