Technobubbles: Ancient, recent, and future Andrew Odlyzko Preliminary version, March 13, 2005 History does not repeat, but it does rhyme, as a common saying goes. The wisdom of this claim is strongly reinforced by the history of technology. When we compare the recent Internet bubble to the British "Railway Mania" of the 1840s (the greatest technobubble ever by many measures), there are a number of striking similarities. Both collapsed because of gross overinvestment based on foreseeably foolish overestimates of demand for the new technology. But post-crash, most of the public attention in both cases was devoted to the frauds that were uncovered (and which did play important roles in inflating those bubbles). The final collapse of the Railway Mania came when it was discovered that George Hudson, the "Railway King," the dominant figure in British railroad industry at that time, had been massaging financial reports for his lines and paying high dividends out of capital. The Internet bubble had the final nail driven into its coffin when Bernie Ebbers' WorldCom was discovered to have committed gigantic accounting fraud and went bankrupt. The similarities between old and recent bubbles are instructive, but so are the differences. As just example, George Hudson's responsibility for the railroad fraud was easy to establish, as everything was done by his command. (This does not mean that he could not defend himself on the grounds of carelessness, overwork, and good intentions, nor that he did not have his defenders, of course.) Not so with the Internet bubble. Bernie Ebbers (along with some of the key players at Enron and other disasters of recent years) was able to make at least a semi-plausible case that he was not involved in any accounting details. This figure, previously lauded as a technological visionary, the prophet of the new telecommunications future, did not use email, and managed to leave very few signatures on paper documents. He now claims that he didn't "know about technology," and didn't "know about finance and accounting." The greater ability of modern business leaders to keep at distance from the actual criminal acts they induce may be a reflection of what they have learned from the experience of their predecessors in white-collar crime in the last century and a half. But it also likely reflects the greater complexity of business, which has led to separation of functions into sales, legal, accounting, and many other specialized areas. On the technology side, there is the added complication that the basic products and services are often becoming harder to comprehend by both the public and the supposedly more expert managers, Wall Street analysts, and the press. Everybody knew what railroads did, and it was not possible to exaggerate traffic too much. But counting bits on a telecommunications network, or the "mind share" that visits to a Web site represent, are much more subjective judgments. And while the total volume of information is growing explosively, what information is available can be manipulated, as can ways this information is presented. And of course all the usual reasons, rooted in human psychology and economics, why speculative bubbles develop still operate. So we can look forward to many future bubbles, and many more Bernie Ebbers-type defenses when those implode. For there are many areas, for example in nanotechnology, biomedicine, and even in IT (which appears to be far from exhausted) that are likely to offer breakthroughs that will stimulate inventors, innovators, and investors, and occasionally stimulate them sufficiently to produce new bubbles. What we will need is ever greater scrutiny to make sure promoters' claims do not run too far away from reality.