Investors' optimal response to stock price bubbles

Table of Contents

1. Introduction

2. Asset overvaluation

3. The efficient market hypothesis

4. The role of arbitrage traders

5. The behavior of the crowd

6. The role of institutional investors

7. Conclusion

8. References

9. Appendix
A. Overreaction to a news event at the example of Deutsche Bank AG
B. DJIA - Illustration of price fluctuation around the true value
C. Deviation from fundamental value (Example: Shell and Royal Dutch)
D. Synchronization and coordination risk

1. Introduction

According to the efficient market hypothesis (later referred to as: EMH) there should not be an asset overvaluation. Nevertheless, bubbles appear from time to time in the real world. In a financial bubble, the price of a security deviates grossly from its fundamental intrinsic value (Watanabe, Takayasu & Takayasu, 2007). Fundamentals or fundamental value refer to economic variables such as discount rates or future cash flows (Siegel, 2003). Depending on the valuation technique one can define an asset’s intrinsic or fundamental value, based on economic variables and assumed growth. A financial bubble is defined as a price run-up, where an initial price rise generates positive expectations of higher future prices, which attracts new buyers that are rather interested in reaping profits by trading the assets than using its earnings capacity (Siegel, 2003). There is a long history of bubbles such as the 1720 South Sea bubble, 1929 the Great Crash, in the mid-1970s the REIT bubble, in 1987 the housing crash, in 1991 the banking crisis, in 2002 the NASDAQ technology bubble and just recently the housing bubble in the United States, just to name a few.

This capstone assignment deals with the question of how investors should act in the case of asset overvaluation in financial markets. In particular, it tries to answer how investors should behave. The central question asks whether investors should step aside and wait until the bubble bursts, whether they should ride the bubble or trade against it. Of course, there is support for all three, albeit contradicting theories. The different trading and investment strategies are reviewed, thereby touching upon various asset bubbles, financial concepts and empirical evidence in the academia. Moreover, it is elaborated on positive feedback trading and rational speculations, as well as behavioral finance concepts such as herding or overconfidence.

The remainder of this paper describes different concepts outlined in the empirical literature, starting with asset overvaluation, followed by the efficient market hypothesis and the random walk phenomenon. The role of arbitrage traders is explored, and their impact on efficient markets and bubbles discussed. A review of behavioral traits during bubbles and the impact of human behavior on asset prices is included. Further, there is an examination of mutual fund strategies and their success in exploiting profit opportunities during bubbles. Finally, it is summarized which arguments support each of the different viewpoints and how investors can react to bubbles in order to exploit opportunities.

2. Asset overvaluation

The definitions of bubbles are various, but all agree on the fact that something is too expensive relative to the provided benefits or utility. Siegel (2003, p.12) writes “The increasing price is unjustified by earnings capacity and is amplified by momentum investors who buy with the purpose of selling quickly to other investors at a higher price”. Bubbles mostly develop in settings with high price volatility and in markets where it is difficult to calculate the true value of a security. In fixed income markets, arbitrage is remarkably easy and nearly riskless. On the contrary, in equity and housing markets it is much harder to detect the true value of a security. The higher volatility increases fundamental risk, which deters risk arbitrageurs. Therefore, asset overvaluations are most often observed in volatile, and hard to calculate markets (Shleifer & Vishny, 1997).

In a momentum environment, which characterizes a bubble setting, market participants increase the price for assets due to higher demand. This can lead to large and long-lasting deviations from fundamental value in the intermediate term before the bubble ultimately bursts and the price converges to its true value (Abreu & Brunnermeier, 2003). Synchronized and coordinated actions are required to burst a bubble, meaning that a sufficiently large number of investors have to realize that the security or class of securities is overvalued and start to sell out their holding. For example, at the end of 1720, during the South Sea bubble, investors finally realized with the announcement of a 5% dividend that it was time to get out since the “even greater fools” were coming to an end. The dividend announcement could be viewed as a synchronizing event because the 5% dividend level was the broadest hint that the stock was overvalued. Based on the dividend discount model a dividend of 40 £ per 100 £ par value would have been required to justify the maximum price of 1,000 £ in march 1720. Thereby investors would not have demanded a risk premium (Temin & Voth, 2004). A later section touches upon the topic of positive feedback and momentum trading in more detail because it plays a crucial role in this context.

Often it is only possible to analyze a bubble and draw conclusions from it after its burst. The majority of market participants does not realize the bubble or does not estimate it to persist in the foreseeable future. During the 1720 South Sea bubble and also during the 1929 stock market crash, tighter requirements on margin loans and higher interest on them indicate that sophisticated, well-informed market participants, namely brokers and banks, expect the bubble to burst in the near future (Temin & Voth, 2004).

3. The efficient market hypothesis

The EMH assumes that investors cannot achieve risk adjusted excess returns above the market, as long as financial markets are information efficient. Readily available information should be included instantly in market prices, due to arbitrage traders and investors. Theory suggests that an anomaly such as a bubble should not exist. With this consistent, the random walk hypothesis states that one cannot predict asset prices based on successive price changes or trading volume (technical analysis). Besides other empirical studies, the work of Eugene F. Fama (1965) delivers empirical results of no serial correlation that support the random walk hypothesis and thereby negate the value of technical analysis to stock market investors. Serial correlation measures if the returns of one period are correlated with the returns of a successive period. From the random walk and EMH point of view, one might expect that only the superior analyst can achieve superior returns in financial markets, due to the fact he can distinguish between undervalued and overvalued assets based on fundamental factors (Fama, 1965). Moreover, the superior analyst can better predict and interpret the occurrence of political and economic events. In this notion, technical analysis is useless. Nevertheless, Fama and French examine an overreaction to news events in a multiyear study (Bodie, Kane & Marcus, 2010). In the short run, markets show positive serial correlation, also known as momentum, but no overreaction in the long run because the price converges eventually (Bodie, Kane & Marcus, 2010).

Figure A1 displays the chart of Deutsche Bank (Appendix A). On the 12/12/2012, the news of a CO2 fraud came out, and the stock reacted strongly negative. The following days it increased slightly, a correction to the overreaction, and continued to trade around 33 EUR. The huge increase at the first trading days in 2013 that increases the stock to old levels is attributable to the temporary solution of the US fiscal cliff, a market-related (β) factor as illustrated by the DAX index. Looking at the chart in figure A1, one can identify the overreaction to the news event.

De Long, Shleifer, Summers and Waldmann (1990) cite the above outlined phenomenon as positive feedback effect where rational investors purchase the asset based on a positive news event. In anticipation of even higher future prices, less well-informed traders buy the asset because of the stark price increase. They extrapolate a series of price increases, rather than determining the fundamental value justified with the news incorporated (De Long, Shleifer, Summers & Waldmann, 1990). In this context and counter to the EMH, resistance and support levels, as often considered by professional traders practicing technical analysis, are crucial (Bodie, Kane & Marcus, 2010). In addition, these movements are amplified by stop-loss or stop-buy orders. This is followed by negative serial correlation in the long-term because the market converges eventually, since those who bought the earliest, based on the news event, might sell at the peak of the overreaction. From a helicopter view, it seems that asset prices fluctuate around its true value while the period of time of the over- or undervaluation is unspecified. Consequently, it can last from few seconds until several years (Fama, 1965). Figure B1 illustrates this phenomenon based on the data of the Dow Jones Industrial Average Index (later referred to as: DJIA) dating back to 1930. The linear black line shows the hypothetical intrinsic value, while the red line illustrates the natural history of the DJIA (Appendix B).

4. The role of arbitrage traders

Arbitrageurs are traders that capture profits while reversing the asset mispricing, due to short or long positions and their engagement in different markets. The existence of arbitrageurs is one assumption of the EMH. However, the theoretical model of riskless arbitrage by making a profit without a capital requirement and without risk due to simultaneously buying in the low price market and selling in the high price market, as described by Tirole, is not feasible in reality. In practice arbitrage traders have margin requirements and the pressure to deliver profits (Shleifer & Vishny, 1997). Shleifer and Vishny (1997) define the term performance based arbitrage (PBA), as professional, highly specialized risk arbitrageurs with complicated and difficult to understand arbitrage strategies. They are dependent on their performance because otherwise investors would withdraw the money and force arbitrageurs to liquidate their positions early. The withdrawal of funds due to unsatisfactory performance takes place when the expected return from arbitrage is the highest, namely when prices move further away from their true fundamental value (Shleifer & Vishny, 1997). To meet the withdrawals the trader has to liquidate the position at a loss.

[...]