Technological Revolutions and Stock Prices for presentation at UESTC, June 2011

1. Technological Revolutions and Stock Prices for presentation at UESTC, June 2011

2. According to Malkiel’s (1999) A Random Walk Down Wall Street • “What electronics was to the 1960s, • biotechnology became to the 1980s... • Valuation levels of biotechnology stocks reached levels previously unknown to investors... • From the mid-1980s to the late 1980s, • most biotechnology stocks lost three-quarters of their market value.”

3. The Economist, Sep 21, 2000 • “Technological revolutions and financial bubbles seem to go hand in hand.”

4. Technological Revolutions and Stock PricesP´astor and Veronesi (2009, AER) • Technological revolutions tend to be accompanied by bubble-like patterns in the stock prices of firms that employ the new technology. • Recent examples of such price patterns include the “Internet craze” of the late 1990s, the “biotech revolution” of the early 1980s… • The bubble-like stock price behavior during technological revolutions is frequently attributed to market irrationality. • This paper proposes an alternative explanation, without appealing to irrationality.

5. Basic idea • New technologies are characterized by • high uncertainty about their average future productivity. • The time-varying nature of this uncertainty can produce the observed stock price patterns.

6. A general equilibrium model of a finite-horizon representative-agent economy • Two sectors: the “new economy” and the “old economy.” • The old economy implements the existing technologies • in large-scale production whose output determines the representative agent’s terminal wealth. • The new economy, • which is created when a new technology is invented, • implements the new technology in small-scale production that does not affect the agent’s wealth. • By observing the new economy, • the representative agent learns about the average productivity of the new technology • before deciding (as a utility-maximizing social planner) whether to adopt the technology on a large scale.

7. A technological revolution • Under simple assumptions, • it is optimal for the new technology to be initially employed on a small scale because its future productivity is uncertain. • The irreversible adoption takes place • if the agent learns that the new technology is sufficiently productive. • They define a technological revolution • as a period concluded by a large-scale adoption of a new technology.

8. Risk dynamics and technological revolutions • Initially, the risk associated with new technologies is mostly idiosyncratic • due to the small scale of production and a low probability of a large-scale adoption. • The risk remains largely idiosyncratic for those technologies • that are never adopted on a large scale. • For the technologies that are ultimately adopted, however, • the risk gradually changes from idiosyncratic to systematic: • As the probability of adoption increases, • the new technology becomes more likely to affect the old economy and • with it the representative agent’s wealth, • so the systematic risk in the economy increases.

9. This time-varying risk has interesting implications for stock prices. • Initially, while uncertainty about the new technology is mostly idiosyncratic, • the new economy stocks command high valuation ratios. • As the adoption probability increases, • the resulting increase in systematic risk pushes up the discount rates and • thus depresses stock prices in both the new and old economies. • The new economy stock prices fall deeper because • their discount rates rise higher due to an increase in the new economy’s market beta. • In short, stock prices begin falling during technological revolutions • when it becomes likely that • the new technology will eventually be adopted on a large scale.

10. Stock prices are affected not only by discount rates but also by expected cash flows. • The technologies that are ultimately adopted must turn out • to be sufficiently productive before the adoption. • This positive cash flow news pushes stock prices up, • countervailing the effect of the higher discount rate. • The cash flow effect prevails initially, pushing the new economy stock prices up, • but the discount rate effect prevails eventually, • pushing the stock prices down. • The resulting pattern in the new economy stock prices looks like an irrational bubble • but it obtains under rational expectations through a general equilibrium effect.

11. Discussion on realized returns • The bubble-like pattern in stock prices arises • due to an ex post selection bias. • Researchers study technological revolutions • with the ex post knowledge that the revolutions took place, • but investors living through those periods did not know • whether the new technologies would eventually be adopted on a large scale. • The representative agent never expects stock prices to fall; • she always expects to earn positive stock returns commensurate to the stocks’ riskiness, and • she subsequently earns those fair returns, on average. • However, in those rare periods • that are recognized as technological revolutions ex post, • the agent’s realized returns tend to be initially positive • due to good news about productivity and • eventually negative due to unexpected increases in systematic risk.

13. Return volatility • The high stock return volatility observed during technological revolutions • can also be explained by uncertainty about new technologies. • Due to this uncertainty, the new economy stocks are more volatile than the old economy stocks. • After an initial decline, the new economy’s volatility rises sharply • when the stochastic discount factor becomes more volatile as a result of a higher probability of a large-scale adoption. • The same effect also pushes up the new economy’s market beta and the old economy’s volatility, • two different measures of systematic risk in the economy.

15. Predictions • Their model makes many empirical predictions for technological revolutions: • The “bubble” in stock prices should be much stronger in the new economy than in the old economy; • stock prices in both economies should reach the bottom at the end of the revolution; • the new economy’s market beta should increase sharply before the end of the revolution; • the new economy’s volatility should also rise sharply and it should exceed the old economy’s volatility; • the old economy’s volatility should rise but less than the new economy’s volatility; • the new economy’s beta and both volatilities should all peak at the end of the revolution; and • the productivity should begin rising at the end of the revolution.

17. All of these predictions are supported by the empirical evidence from the recent Internet revolution. • According to the model, this revolution ended in 2002 • (i.e., the probability of a large-scale adoption of the Internet technology reached one). • The “bubble” pattern was much stronger • in the NASDAQ index (our proxy for the new economy) than in the NYSE/AMEX index (the old economy); • both stock price indexes reached the bottom in 2002; • NASDAQ’s beta doubled between 1997 and 2002; • NYSE/AMEX’s return volatility also doubled and • NASDAQ’s volatility tripled over the same period; • NASDAQ’s volatility always exceeded NYSE/AMEX’s volatility; • NASDAQ’s beta and both volatilities peaked in 2002; and • the productivity growth of the U.S. economy accelerated sharply after 2002.

19. The welfare of the representative agent • the adoption (or even an increasing probability thereof) • increases systematic risk and thus reduces the new economy’s market value. • It appears that the adoption is not favored by the new economy shareholders. • However, in the model, there is only one shareholder, • the representative agent, who employs infinitely more capital in the old economy than in the new economy. • This agent wants the adoption to take place because the utility gain from making the old economy more productive outweighs the (negligible) loss of market value in the new economy.

20. Conclusions • This paper offers a rational explanation for the bubble-like patterns in stock prices observed during technological revolutions. • Stock prices of innovative firms initially rise due to good news about the productivity of the new technology, • but they ultimately fall as the risk of the technology changes from idiosyncratic to systematic. • The rise and fall in stock prices are observable only in hindsight – • this pattern is unexpected while investors are uncertain • whether the new technology would be widely adopted, • but we observe it ex post when we focus on technologies that eventually led to technological revolutions. • These “bubbles” should be most pronounced • in revolutions characterized by high uncertainty and fast adoption.

21. Conclusions_2 • To formalize the intuition, this paper develops a general equilibrium model that features • a real option decision and • Bayesian learning about the average productivity of the new technology. • The model makes many empirical predictions. • They find substantial support for these predictions in the evidence from 1830–1861 and 1992–2005 • when the railroad and Internet technologies spread in the United States.

22. Future research • Future research can also test their model against alternatives that involve behavioral biases. • To construct a fair horserace, it would be useful to develop • a behavioral model of technological revolutions that produces testable predictions. • Some predictions of our model, such as those involving market beta, • are unlikely to follow from behavioral models, • in which there is typically no role for systematic risk.

23. A Bayesian’s BubbleLI and XUE (2009, JF) • THE U.S. STOCK MARKET EXPERIENCED DRAMATIC MOVEMENTS from 1998 to 2001: • In less than 4 years, it skyrocketed, fluctuated, and then plummeted. • This paper examines this phenomenon • in the context of U.S. economic growth in the late 1990s. • They argue that the observed stock market movements • can be largely explained by a rational investor’s uncertainty about the future of the economy.

25. motivation • Their motivation is based on the unusual growth that • the American economy experienced in the second half of the 1990s. • What distinguished this growth from other economic expansions in the previous two decades was • the acceleration of productivity growth, • which suggested a significant advance in technological innovation.

26. Questions • This paper aims at addressing the questions: • In what way can macroeconomic conditions cause a bubble? • Why did the stock market not skyrocket until 1998? • Why did the market crash begin in 2001?

27. Their model • P´astor and Veronesi demonstrate that, • due to the convex relation between valuation and the growth rate, • the uncertainty about the long-term growth potential of a firm can lead to substantially higher valuation. • Adopting the modeling technique of Pastor and Veronesi, • this paper develops a simple valuation model • to examine the impact of changes in investor beliefs about the arrival of a new economy • on the stock market, especially during the market boom and crash of 1998 to 2001.

28. Investor learning is facilitated by investors’ observation of the economic data. • Technological innovation is the fundamental force • that can shift the economy into a new regime. • They use total factor productivity (TFP) data, • the measure of technological innovation in the United States, • to calibrate the model of investor learning. • The estimation results show that • investor beliefs about switching to a new economy • increase gradually from 1995 to the third quarter of 1998, • accelerate and peak in the third quarter of 1999, and then • drop substantially after the second quarter of 2000.

29. Beliefs about the arrival of a new economy • The results suggest that as high TFP growth continued, • investors became increasingly confident that • the growth indicated a regime shift to a new economy, • not simply a sequence of positive shocks. • At the same time, beliefs about • important aspects of the new economy also took form through learning. • In late 2000, however, as investors observed new TFP data • inconsistent with their recently formed beliefs about the new economy, • they quickly and rationally abandoned their beliefs about the arrival of a new economy. • This shift in beliefs • can take place rapidly in a Bayesian updating framework and • result in a dramatic price adjustment in a short period of time.

30. Conclusions_1 • A rational investor’s uncertainty about • the future of the U.S. economy can potentially explain • the stock market bubble of the late 1990s. • The ex post-observed stock market movements in 1998 to 2001 • seem to be driven, to a large degree, • by the evolution of investors’ ex ante beliefs about a new economy. • Their results suggest that • the macroeconomy affects the stock market in a way that may not be captured by a simple linear regression. • Recently, financial economists have been trying to • get a better understanding of the relationship between the real economy and financial markets. • Their work complements these attempts by providing direct evidence on this relationship.

31. Conclusions_2 • Based on P´astor and Veronesi’s learning model, • this paper incorporates the dynamics of learning • about whether and when a shift to a new economy takes place, • based on macroeconomic data. • Thus, they provide an alternative direction to extend P´astor and Veronesi’s model.

32. Technological innovations and aggregate risk premiumsHsu (2009, JFE) • Technological innovations are able to • predict market returns and premiums in recent decades. • The predictability can be attributed to several reasons: • First, technological innovations raise the expected productivity and profitability of the representative firm. • Second, technological innovations improve overall efficiency and reduce investment costs. • Lastly, technological innovations work as options with returns more volatile than physical investments. • Since the representative firm’s expected stock returns • equal expected investment returns, • they rise with more technological innovations.

33. measuring technological innovations • use aggregate patent data and R&D data • to measure technological innovations in the U.S. • Main findings: • patent shocks and R&D shocks have positive and distinct predictive power for U.S. market returns and premiums. • Similar patterns are also found in international data • including other G7 countries, China, and India. • These findings are consistent with previous empirical studies • based on firm-level data, and • call for further theoretical explanations.

35. Financing Innovation and Growth: Cash Flow,External Equity, and the 1990s R&D BoomBROWN, FAZZARI, and PETERSEN (2009, JF) • Does finance cause growth? • The literature has establishes a strong connection • between broad measures of financial development and economic growth. • Questions remain, however, about • the channels through which finance may matter for growth. • One potentially important channel is the financing of R&D, • a critical input to innovation and growth in modern economies. • R&D is particularly interesting • not only because of the knowledge spillovers it creates • but also because R&D may be difficult to finance with external sources.

36. Financing R&D • Young publicly traded firms in high-tech industries • finance R&D investment almost entirely • with internal or external equity (i.e., cash flow or public share issues). • For these firms, information problems, • skewed and highly uncertain returns, and • lack of collateral value • likely make debt a poor substitute for equity finance.

37. Furthermore • If these firms face binding financing constraints, • then exogenous changes in the supply of either internal or external equity finance • should lead to changes in R&D. • If such firms undertake a large fraction of aggregate R&D, • then changes in the availability of finance • may have macroeconomic significance. • In particular, booms (or busts) in the supply of equity finance • should lead to booms (or busts) in R&D.

38. U.S. experience_R&D • the United States has recently experienced • a finance-driven cycle in R&D. • From 1994 to 2004, there was a dramatic boom, and subsequent decline, in R&D: • just seven high-tech industries • (drugs, office equipment and computers, electronic components, communication equipment, scientific instruments, medical instruments, and software) • accounted for virtually all of the 1990s U.S. R&D boom. • More important, virtually all of the boom was accounted for • by young firms (publicly traded for less than 15 years) in these industries.

39. U.S. experience_financing • From 1994 to 2004, there was also a dramatic boom and bust • in both cash flow and external equity finance in these industries. • Internal finance (cash flow) for publicly traded firms • increased from $89 billion in 1993 to $231 billion in 2000, • and then collapsed in 2001 and 2002. • External public equity finance • rose from $24 billion in 1998 to $86 billion in 2000, • but then plummeted 62% in 2001.

44. The central question and results • The central question in this paper is • whether supply shifts in both internal and external equity finance • can explain a significant part of the 1990s boom and subsequent decline in aggregate R&D. • For mature firms, the results are insignificant. • For young firms, however, the equity finance variables • are quantitatively large and highly significant. • Furthermore, the financial coefficients are large enough that • the financial cycles for young high-tech firms alone can explain • about 75% of the aggregate R&D boom and subsequent decline.

45. Implications • finance affects growth. • While the large literature on finance and economic growth • has good reasons to focus on debt and credit constraints, • more attention should be given to equity finance, • particularly for models that emphasize innovation. • stock markets can be an important source of funds, • which has implications for the debate about the relative merits of • bank-based versus market-based financial systems.