What we know about markets and what we

What we know about markets, and what we think we know about markets Prof. Dr. Oliver Spalt spalt@uni-mannheim. de

Why are markets good? 2

The Classical Answer n Classical economic position is that “markets are good” q n n They efficiently allocate risks and resources Theory: q Adam Smith’s Invisible Hand (18 th century) q Arrow-Debreu: Welfare Theorems (20 th century) Does that settle the debate? 3

Purpose of this Talk n Discuss some problems with thinking about how well markets function by looking at an important special case: q n Illuminate the difference between q q n financial markets What we know, and What we think we know Get you started on thoughts like: q q q What are the boundaries of science? What should my null hypothesis be? How can we shape society in the absence of conclusive evidence?

The Efficient Market Hypothesis n Eugene Fama, Nobel Prize 2013 5

The Efficient Market Hypothesis n Fama (1970): An efficient market is one where prices always fully reflect available information n If markets are efficient, they provide informative signals for investors. Hence, efficient markets make sure funds are allocated to their best use q n The power of “capitalism” EMH has lead to several first-order academic break-throughs that changed the world (for better or worse) q q q CAPM Option pricing Passive investment industry 6

EMH Theoretical Justifications (Shleifer (2000)) 1. All investors are fully rational (Bayesian EU Maximizers) 2. Some investors are less than fully rational, but their effect cancels out in the aggregate q q Random mistakes No price impact 3. Some investors are non-rational in similar, correlated ways. However, rational arbitrageurs eliminate their influences on prices n Decreasingly strict assumptions Common misconception: EMH requires everybody is fully rational! n q Not true if arbitrage process works well 7

Evidence consistent with EMH – Example 1 Source: Ross, Westerfield, Jaffee, and Jordan (2008) based on Pastor and Stambough (2002) n n Annual returns US mutual funds 1963 – 1998 It is very hard to beat the market 8

Evidence consistent with EMH – Example 2 Source: Ross, Westerfield, Jaffee, and Jordan (2008) based on Szewczyk, Tsetsekos, and Zantout (1997) n n Abnormal returns around dividend omissions (event study) Information is often incorporated into prices quickly 9

Challenges to market efficiency n Challenges to theory of efficient markets have come mostly on two fronts: q Based on observed market “anomalies”… n q …and sustained by a large literature on limits to arbitrage Based on systematic deviations from rationality in individual decision making n In particular: the Kahneman and Tversky Nobel Prize winning research program 10

Daniel Kahneman & Amos Tversky n n Showed that standard ingredients used in economic models are often at odds with (or insufficient to describe) actual human choice behavior Nobel Prize in Economics 2002 q (Not bad for cognitive psychologists!) 11

12

The “Market-Based Challenge” n Robert Shiller, Nobel Prize 2013 13

Market anomalies (1) – tech stock bubble (? ) 14

Market anomalies (2) – housing bubble (? ) Source: Robert Shiller 15

Shiller’s Interpretation n Markets are “excessively” volatile Shiller (2013): “It is hardly plausible that speculative prices make effective use of all information about probabilities of future dividends. It is far more plausible that the aggregate stock market price changes reflect inconstant perceptions, changes which Keynes referred to with the term “animal spirits, ” changes that infect the thinking even of the most of the so-called “smart money” in the market. ” Markets are governed by fads, fashions, collective illusions, and stories q Think: “new era stories” about tech stocks 16

Alternative View n Start with Campbell-Shiller decomposition of D/P ratio. n D/P ratios are driven by: q q q Expected future returns (“discount rates”) Changes in future dividends “Rational bubbles” 17

Alternative View n n You can estimate how much each individual component contributes to explaining D/P Fraction of variance explained (see e. g. Cochrane (2011)): q q q Expected Returns: 100% Expected Dividends: 0% “Rational” Bubbles: 0% n Hence: “Prices are high because expected returns are low” n Cochrane (2011): “Prices being too high can only mean expected returns are too low relative to some theory” n Discount rates may vary systematically due to some underlying rational theory. No need for “Animal Spirits” q For example: Campbell and Cochrane (1999) (rational model? ? ) 18

The Heart of the Debate n At the heart of the debate is what Fama (1970) calls the Joint Hypotheses Problem: q q n If you want to claim that the price of an asset differs from its true fundamental value, you need a model to determine that price. Hence any test for mispricing is at the same time a test of the pricing model. Since we can never be completely sure that our model is right, we can also never be sure that there is mispricing. In other words: observing something that looks like mispricing can simply mean that we need a better model q That new model may well be a “rational” one 19

Example Tech Stocks n Many believe the “Tech Bubble” is a prime example of an event due to investor irrationality q q Shiller: “Irrational Exuberance” If you have seen it, it is hard to deny at least some irrationality… n But even for that episode there is room for alternative models if you are smart and creative enough n For example: Pastor and Veronesi (2006) q q q Key idea: uncertainty about future profitability growth was extremely high for tech stocks They show higher uncertainty translates into higher valuations They calibrate a model to show uncertainty levels needed to match Nasdaq “bubble” are not excessively high 20

Pastor and Veronesi (2006) 21

Who Owns the Null Hypothesis? “Before we relegate a speculative event to the fundamentally inexplicable or bubble category driven by crowd psychology, however, we should exhaust the reasonable economic explanations. . . “bubble” characterizations should be a last resort because they are nonexplanations of events, merely a name that we attach to a financial phenomenon that we have not invested sufficiently in understanding. ” Garber (2000, p. 124) 22

Who Owns the Null Hypothesis? n “Owning the null hypothesis” is a crucial issue in the debate q q n Summers (1986) makes this point as follows: q q n H 0: P(t) = P(t-1) + u(t), where u(t) is uncorrelated across time H 1: P(t) = P(t-1) + u(t), where u(t) is AR(1) Even under unrealistically favorable assumptions for H 1 being true, we lack the power in most empirical tests to reject H 0 q n Many disagreements can be traced back to differences in the null See Fox (2009) for fascinating account of how EMH became H 0 With monthly data, need 5, 000 years of data (even though everything is strongly rigged towards H 1) If we cannot reject H 0, should we view it as “true”? 23

Markets in Society n What do we really know about the efficiency of markets? q n If markets are not efficient (i. e. , prices do not correctly reflect all available information), then: q n The Joint Hypothesis Problem How inefficient will the resulting allocation of resources and risks be? What follows for how much markets should matter in society more broadly? q q q Role for regulation? How would you regulate? What should be our null hypothesis? 24

Questions for us in this Course n What is my implicit null hypothesis? n How do I know morals are important? n If you and I come from different priors, how can we productively move forward? n If we feel we need to act now (instead of waiting 5, 000 years until our t-stats are high enough)… q …How to decide on how to regulate markets? 25

“Homework” n Watch and think about: n https: //www. youtube. com/watch? v=fa. Sa 3 r 8 WIU 0 26