Behavioral Finance 11232020 Behavioral Finance 1 Efficient Markets

Behavioral Finance 11/23/2020 Behavioral Finance 1

Efficient Markets Hypothesis • Security prices fully reflect available info (Fama, ‘ 70) -- identical securities should trade at identical prices = companies with identical cashflows and riskiness (discount factors, cost of capital) are worth the same -- it is impossible to make abnormal returns using only current information = since prices reflect all available information there’s no opportunity for anyone to price securities better then the market does -- one cannot consistently beat the market = abnormal performance is a chance event 11/23/2020 Behavioral Finance 2

Efficient Markets Hypothesis -2 • Key components of EMH -- investors are rational = investors do not make mistakes -- if not rational then trade randomly = even if investors do mistakes these are not correlated across investors, i. e. mistakes investors makes are “different” -- if not rational and trade in similar ways then mispricing is eliminated by arbitrageurs = even if mistakes made by some unsophisticated investors point in the same direction, e. g. investors may erroneously believe that security is worth more than it truly is, there is an army of sophisticated investors which will immediately spot the mispricing and correct it 11/23/2020 Behavioral Finance 3

Efficient Markets Hypothesis - 3 • [Textbook] Arbitrage: By simultaneously selling and purchasing identical securities at favorably different prices, the arbitrageur captures an immediate payoff with no up-front capital and no risk Example: -- market price of asset 1 > fundamental value -- strategy: sell short the expensive asset 1 and buy “identical”, but cheaper asset 2 (substitute security) Suppose that companies ABC and DEF are identical. Their fair value is 10$/share. However, ABC is currently trading at 12$/share. An arbitrageur would (short) sell the security ABC and buy DEF. 11/23/2020 Behavioral Finance 4

Efficient Markets Hypothesis - 4 • This strategy would lead to immediate net payoff of 2$/share. Since there will be many arbitrageurs pursuing this strategy prices of ABC and DEF will equate soon – value of short and long positions would soon be equal = no risk of losing money • Outcome 1: arbitrageur pockets immediate gain equal to the difference in price of assets 1 and 2, no risk as 1 and 2 have identical pay-offs in the future • Outcome 2: mispricing could only be short-lived as opportunistic behavior of arbitrageurs will bring price of asset 1 back to fundamental value 11/23/2020 Behavioral Finance 5

Market Efficiency: subsets of available information All Available Information including inside or private information EMH: “A market is efficient if it reflects ALL available information” [1] Strong-form All Public Information [2] Semi-strong form Informatio n in past stock prices [3] Weak-form • . 11/23/2020 Behavioral Finance 6

3 forms of market efficiency • If Weak-form of the hypothesis is valid: – Technical analysis or charting becomes ineffective. You won’t be able to gain abnormal returns based on it. • If Semi-strong form of the hypothesis is valid: – No analysis will help you attain abnormal returns as long as the analysis is based on publicly available information. • If Strong-form of the hypothesis is valid: – Any effort to seek out insider information to beat the market are ineffective because the price has already reflected the insider information. – Under this form of the hypothesis, the professional investor truly has a zero market value because no form of search or processing of information will consistently produce abnormal returns. 11/23/2020 Behavioral Finance 7

Early empirical evidence: semi-strong form Keown and Pinkerton (1981): target’s stock price reaction to the takeover bid Bodie, Kane, and Marcus, “Investments”, 7 th ed. , p. 359 11/23/2020 Behavioral Finance 8

Expected Utility paradigm • Rational utility maximizers – People aim to maximize their well being or satisfaction – Satisfaction is a function of the goods and services that the individual consumes and be measured using a utility function – Maximization occurs in the presence of budget constraints • Care about levels of goods – Individuals derive satisfaction from how much goods and services they have at a particular point in time, not how much quantities of goods and services have changed over time – Level of satisfaction at a certain point in time depends only on the levels/quantities of goods and services at the point in time; the path, i. e. previous quantities are irrelevant • Correctly estimate probabilities of events 11/23/2020 Behavioral Finance 9

Let’s do a simple reality check • How are you doing? • Here’s 20$ • How are you doing now? • Now give it back to me • How are you doing now? 11/23/2020 Behavioral Finance 10

What’s wrong? • Why not so happy? • Should you be not so happy? • You have the right to feel this way; economics should explain why 11/23/2020 Behavioral Finance 11

Challenges to EMH: individuals - 1 • Attitude towards risk – Assess gambles on changes rather then levels Kahneman and Tversky (1979) – Losing-and-gaining or gaining-and-losing equal amounts generally makes you feel worse off (even though your original level of wealth did not get affected) – Prefer to play safe when gaining, double up when losing 11/23/2020 Behavioral Finance 12

Challenges to EMH: individuals - 2 • Non-Bayesian formation of expectations – Do not form expectations about the likelihood of events correctly –. 1 1 0. 5 0 0 1 2 3 4 1 -0. 5 -1 -1 2 3 4 Example: see patterns where there’s none 11/23/2020 Behavioral Finance 13

Challenges to EMH: individuals - 3 • Representativeness – Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations. Which is more likely? – A: Linda is a bank teller – B: Linda is a bank teller and is active in the feminist movement • Many people chose answer B even though it is not possible 11/23/2020 Behavioral Finance 14

Deal or no deal Yellow box is the one you set aside, blue ones are on stage The boxes are worth 1$, 75 000$, 1 000 000$ You can accept the deal and go home or no deal and eliminate one of the blue boxes I offer you 11/23/2020 Behavioral Finance 410 000 $ 15

Deal or no deal Yellow box is the one you set aside, blue ones are on stage The boxes are worth 1$, 75 000$, 1 000 000$ You can accept the deal and go home or no deal and eliminate one of the blue boxes I offer you 11/23/2020 Behavioral Finance 603 000 $ 16

Deal or no deal Yellow box is the one you set aside, blue ones are on stage The boxes are worth 1$, 75 000$, 1 000 000$ You can accept the deal and go home or no deal and eliminate the remaining blue box I offer you 11/23/2020 Behavioral Finance 416 000 $ 17

Deal or no deal Yellow box is the one you set aside, blue ones are on stage The boxes are worth 1$, 75 000$, 1 000 000$ You are walking home with a yellow box Yes, you are correct, there’s 1 dollar inside 11/23/2020 Behavioral Finance 18

male nurse Richie Bell, “Deal or No Deal”, episode 407 (October 22 nd, 2008) People believe that they got hands, can influence the outcome etc 11/23/2020 Behavioral Finance 19

Challenges to EMH: individuals - 4 – Overweighting extreme probabilities Kahneman and Tversky (1973) Highly unlikely events are either ignored or overweighted • Example: – We overestimate the risk of ’spectacular’ risks • Plane crashes • SARS – We underestimate the risk of common risks • E. g. Cancer 11/23/2020 Behavioral Finance 20

Slovic, Fischhoff, Lichtenstein (1982) 11/23/2020 Behavioral Finance 21

Challenges to EMH: individuals - 5 • Example: imagine that the U. S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Program A: 200 people will be saved Program B: 1/3 likelihood of 600 saved, 2/3 that none saved 72 28 Program A: 400 people will die Program B: 1/3 likelihood of nobody dying, 2/3 that 600 die 22 78 Sensitivity of decision making to framing of the problems Ø Different choices as a result of a different presentation of the same problem Benartzi and Thaler (1995) 11/23/2020 Behavioral Finance 22

Do we care about the mistakes of individuals? • It is all extremely interesting… People are making a lot of mistakes. May be, by knowing its origin, one can avoid some…” • But does it matter for big picture? – Errors individuals are making may tend to cancel each other without any effect on aggregate market behavior – If not, arbitrageurs should eliminate those deviations fast – But do they? (next classes) 11/23/2020 Behavioral Finance 23

Challenges to EMH: markets • Stocks are more volatile then classical theory can justify – Shiller (1981) • Past performance can predict future – Mean-reversion (contrarian): De Bondt and Thaler (1985) – Momentum Carhart (1997) • And much more … 11/23/2020 Behavioral Finance 24

De Bondt and Thaler (1985) 11/23/2020 Behavioral Finance 25

Traditional vs. Behavioral • Behavioral • Traditional – Rational Investors • Correct (Bayesian) Updating of the Likelihoods of Events Happening • Choices Consistent with Expected Utility – Some are Not Fully Rational • Investor Sentiments – Limits to Arbitrage – Arbitrage • Investment horizon • Risk-bearing capacity • Always corrects shortterm mispricing 11/23/2020 Behavioral Finance 26

Challenges to Behavioral Finance • Collection of stories, not a unified theory • (Mostly) does NOT provide magnitudes of the effect, only a direction • Behavioral puzzles often can be explained within rational framework if one gives a try … 11/23/2020 Behavioral Finance 27

Pontiac Allergic to Ice Cream • Pontiac gets a crazy complaint – Family eats ice cream after dinner every night – Family votes; father drives to get one – Vanilla ice cream – car won’t start – Any other – just fine – Customer admits that it is silly, but still … • Engineer was sent to make a check – Responsible person; many days spent 11/23/2020 Behavioral Finance 28

Did the car turn behavioral? • First evidence: – Vanilla the most popular ice cream; sold at the front for quick pick up – Other types – at the back, more time to find and check out – The problem about timing! not vanilla • A little bit of critical thinking – Vapour lock – Engine was still hot when then man got vanilla 11/23/2020 Behavioral Finance 29

Roadmap Behavioral Finance Limits to Arbitrage Psychology Violation of Exp. Utility 11/23/2020 Behavioral Finance Violation of Bayesian Rules 30

Roadmap-2 • Limits to arbitrage: – Examples of limits to arbitrage – What prevents arbitrage from correcting the prices quickly – Noise Trader risk Note: take investor irrationality for granted analyze the effect on prices • Behavioral Biases – Prospect Theory – Psychological Biases: evidence from experiments – Market Anomalies Note: try to understand the causes of investor irrationality • Behavioral Trading strategies – Is there money to make? – Persistence of behavioral trading strategies – Free lunch or another source of risk? 11/23/2020 Behavioral Finance 31

Topics • Limits to Arbitrage : securities may not be fairly priced – Negative stubs: subsidiary is worth more than subsidiary+parent – Renaming of stocks – Closed-end fund puzzle • Case on Closed-end fund Puzzle • Certainty, Probability and Possibility – Humans do not behave like they were supposed to – Prospect theory – Behavioral biases • Asset Pricing: Failure of CCAPM and Behavioral Response – Equity Premium puzzle – Excess Volatility Puzzle – Excess Correlation puzzle • Existence of Cash Dividends – Self-control, Mental Accounting and Regret Avoidance 11/23/2020 Behavioral Finance 32

Topics -2 • Framing – Affecting choice through different representation of information • Attention Anomalies – Decision making when attention is a scarce resource – Limited attention and stock returns • Stock Return Predictability – Are Stock returns predictable? – Overreaction and Underreaction in financial markets 11/23/2020 Behavioral Finance 33

Topics -3 • Relativity and Anchoring – People make judgments in relative rather than absolute terms – People use (at times arbitrary) anchors / reference points to make judgments • Group Process – “One brain is good, two brains is better “ is not necessarily true – Why? Ask Enron • Herding – Blind leading blind: how could it happen? – How come that herding could be a) rational; b) individually efficient; c) overall inefficient? • … • Happiness 11/23/2020 Behavioral Finance 34

Let’s talk about CAPM is market efficient Rate of return (cost of capital) depends only on systematic risk and that systematic risk can be measured precisely by security’s beta What is missing here? All work and no play make Jack a dull boy. 11/23/2020 Behavioral Finance 35

11/23/2020 Behavioral Finance 36

Fine print (we tend to forget about): All investors are price takers No taxes and no transaction costs (that includes no short-selling constraints) Unlimited access to funds: can borrow and lend unlimited amounts at the risk-free rates Investors attempt to construct efficient porfolios (i. e. The rational utility maximizers) Investors are homogenous: a)Share same believes about distributions of returns b)They know the same about all assets When you happen to live in the economy blessed with this set of assumptions everyone holds market portfolio (which appear also to be on the efficient frontier), passive investment strategy is optimal and prices of securities reflect fundamentals 11/23/2020 Behavioral Finance 37

Limits to arbitrage What do we mean by this: 1. Price of security does not reflect its fundamental value 2. This mispricing is not corrected within short period of time Question: what is the fundamental value of an asset? Suppose I give you a company to evaluate: what is likelihood that you all come to the same conclusion about the company value? Valuation is not easy, a lot of uncertainy involved (model risk) Textbook arbitrage: two identical assets are traded at different prices. Sell expensive, buy cheap – no money involved and no risk Reality: There’s no two identical assets Wurgler and Zhuravskaya (2001): at best you can achieve R 2 of 25% when trying to explain a return on asset with a matched security (as many of 3 of them) Real arbitrage is not riskless! 11/23/2020 Behavioral Finance 38

Joint Hypothesis Problem • In order to claim that the price of security differs from its fundamental value one needs a proper asset pricing model • Fama (1970): any test of mispricing is therefore a joint test of mispricing and of an asset pricing model – Is it that security is mispriced? – Or that your model is wrong? _____________ • In a number of cases mispricing can be established beyond any reasonable doubt. We are going to analyze some of them 11/23/2020 Behavioral Finance 39

Can market add and subtract? • Palm, the maker of Palm. Pilot used to be a division of 3 Com • 4. 1% of Palm equity was issued at $38 on March 1, 2000. • The shares of Palm opened at $145, peaked at $165 and closed at $95. 06 • At close, this implies a negative value of $21 bn put on the remainder of 3 Com’s business • The mispricing persisted for several months 11/23/2020 Behavioral Finance 40

Negative stubs • Definition: negative stub: market value of the parent is less than market value of parent’s ownership stake in a publicly traded subsidiary • Where can it come from? : – Equity carve-outs (3 Com/Palm) – Acquisitions of stakes in publicly listed companies 11/23/2020 Behavioral Finance 41

Value of Palm, 3 COM and stub 11/23/2020 Behavioral Finance 42

3 Com-Palm stub timeline 11/23/2020 Behavioral Finance 43

Negative stub: how was this possible? Why did the mispricing occure in the first place? • A lot of investors bullish on the prospects of Palm; drive price of Palm up But then, this creates a risk-free arbitrage opportunity! Buy cheap (3 Com), short sell expensive (Palm), pocket the difference Short selling Palm is risky and virtually impossible (why? ) Small Palm float We need a reason why arbitrageurs cannot step in and correct mispricing immediately: Short selling overpriced stock was difficult We need a reason why prices deviate from fun-da-mentals: mistakes of investors are correlated Why lack of short-selling allows for mispricing to exist/persist? How limits to short-selling affect prices Why real-life short selling may be difficult 11/23/2020 Behavioral Finance 44

• Short-selling allowed pessimists – Optimistists and pessimists balance each other out – Correct price is achieved optimists • Short-selling not allowed – Optimists are there, but pessimists can not make their opinion count – The price is biased upwards 11/23/2020 optimists Behavioral Finance 45

Mechanics of Real Life Arbitrage • Textbook arbitrage: buy cheap, short sell expensive, no capital involved • Real life arbitrage requires capital This is how much money we should make when value of parent will become equal to the value of its stake in subsidiary Example: • • Let if the market value of the parent equal to the market value of its stake in the subsidiary 1 share of a parent be worth 0. 7154 shares of a subsidiary Currently parent trades at $26. 25/share, subsidiary at $48. 00. Investor wishes to buy 1 share of a parent and sell short 0. 7154 shares of a subsidiary – Value of subsidiary less value of parent=48*0. 7154 -26. 25*1=$8. 09>0, subsidiary overpriced, negative stub Then she has to contribute at least $30. 29 (50 % of both long and short position) to comply with Fed’s requirements (Regulation T), initial margin – 26. 25*1*0. 50+48. 00*0. 7154*0. 50=$30. 29 • In addition to posting the required amount, investors may choose to allocate additional precautionary capital This would correspond to 8. 09/30. 29=26. 70% return 11/23/2020 Behavioral Finance 46

Mech. of Real Life Arbitrage-2: Shortrebates • If investor A has a long position in company C and his shares are held through a brokerage acount B then it is a brokerage B who lends shares to another investor D who wishes to sell C’s shares short. – A can demand shares of C to be physically delivered to it. In such a case, shares that A owns would not be possible to be shorted (unless directly negotiated with A). • A broker who lends shares to investors pays them interest on the amount posted as a collateral to borrow shares, ”short-rebate” – Short-rebate is typically 25 to 50 bp less then Fed funds rate • However, short-rebate may ocasionnaly be lower than that or even NEGATIVE = borrowers pay interest on the collateral they posted 11/23/2020 Behavioral Finance 47

Mech. of Real Life Arbitrage-3: Margin Calls • investor is supposed to keep a deposit which corresponds to a certain % of the value of the position : maintenance margin • NYSE and NASDAQ require minimum margin of 25% of long positions and 30% of short positions what if security prices move such that the investor's position has less then required maintenance margin? • he receives a margin call • either post additional collateral or liquidate part of the position to meet margin requirements • bad thing: you have to post additional funds when you are losing money rather than making them. . . 11/23/2020 Behavioral Finance 48

Margin Calls : example cont’d Let’s go back to the previous example: • Maintenance margin on NASDAQ is 25% of longs and 30% of shorts 26. 25*1*0. 25+48. 00*0. 7154*0. 30=$16. 87 • What do we have to worry about? we put $30. 29 in initial margin • Let prices of both shares double • Maintenance margin goes up to 52. 5*1*0. 25+96. 00*0. 7154*0. 30=$33. 73>$30. 29 11/23/2020 Behavioral Finance 49

Margin Calls : example cont’d Post more collateral • We put additional money in the amount of 33. 73 -30. 29=$3. 43 • If convergence occurs next day we realize 8. 09/(30. 29+3. 43)=23. 98%<26. 70% we originally expected On more invested capital Same dollar return 11/23/2020 Behavioral Finance 50

Margin Calls : example cont’d Partially liquidate the position • Let’s do it proportionally x*(52. 5*1*. 25+96. 00*0. 7154*. 3)= 3. 43 x=10. 17% of both long and short positions has to be liquidated • If convergence occurs next day we realize 0. 8983 bought @ 26. 25 liquidated @ P 1 0. 6426 sold short @ 48. 00 liquidated @ P 2 (such that the value of long position @ P 1 was equal to value of short position @ P 2) Dollar profit = -0. 8983*26. 25+0. 6426*48=$7. 27 0. 1017 bought @ 26. 25 liquidated @ 52. 5 0. 0728 sold short @ 48 liquidated @ 96 Dollar profit =0. 1017(52. 5 -26. 25)+0. 0728*(48 -96)=$-0. 82 Total return (7. 27 -0. 82)/30. 29=21. 27%<26. 70% originally expected 11/23/2020 Behavioral Finance Lower dollar return on the remaining positions Loss on part of the position terminated due to a margin call 51

Mech. of R. L. Arbitrage: Shares Recall • Once a short seller has initiated a position by borrowing stock, the borrowed stock may be recalled at any time by the lender. • If the short seller is unable to find another lender, he is forced to close his position. 11/23/2020 Behavioral Finance 52

Some facts about short-selling constraints • Short Selling: (US market): – – Most stocks can be borrowed – – Only 16% of the stocks are impossible to short • (mainly small cap stocks) – – Only 7% of the loan capacity is currently utilized – – 10% of stocks are never shorted – – For 91% of the stocks the borrowing costs are less than 1% per annum – – Only 9% have average loan fees of ~4% – – Recall is rare, only 2% of the stocks on loan • Source: D’Avolio, “The market for borrowing stock” • In short, market for borrowing stocks usually works very well, except when you want to use, in which case it works terribly. – It can be difficult or expensive to short stocks that many people belieave are overpriced and many people want to short. 11/23/2020 Behavioral Finance 53

How does the convergence happen: • 1. favorable movements in the prices of the parent and subsidiary – Price of parent may increase relative to the price of subsidiary so that the value of the parent will exceed the value of its stake in the subsidiary • 2. distribution of shares of subsidiary to parent shareholders – This creates a more liquid market for subsidiary shares – More opportunities to short, hence arbitragers could correct prices 11/23/2020 Behavioral Finance 54

We need reasons why arbitrageurs were not taking large positions Fundamental risk Buy-in risk Financing risk Horizon risk Margin risk Uncertainty over the nature of anomaly, little ability to hedge 11/23/2020 Behavioral Finance 55

Fundamental Risk • Convergence of arbitrage spread is by no means assured! • About 30% of times the link between parent and subsidiary disappears without converge of the arbitrage spread: – Acquisition/delisting of the parent/subsidiary • Remember: – 1. The arbitrage trade involves holding a long position in the parent firm and a short position in the subsidiary firm. – 2. The long position in the parent firm gives the arbitrageur an indirect holding of the subsidiary firm, which can be shorted out, leaving a net position in only the stub assets. – 3. The key to the trade is the link between the parent and the subsidiary firm created by the parent’s substantial ownership of the subsidiary. 11/23/2020 Behavioral Finance 56

Converge is not clear ahead of time • Daisytek (the parent of PFSWeb) recently announced that it had received an unsolicited offer to acquire all of Daisytek’s outstanding shares. After considering a variety of factors, Daisytek’s board determined that the offer was inadequate and inconsistent with Daisytek’s previously disclosed plans to complete the spinoff. If, however, the bidder decides to begin a tender offer for the outstanding shares of Daisytek without the approval of Daisytek’s board, such an offer, or stockholder litigation in connection with such an offer, could significantly divert our attention away from our operations and disrupt or delay our proposed spinoff from Daisytek. In addition, if the bidder is successful in acquiring control of Daisytek prior to the proposed spin-off, it would control a majority of our shares and the spin-off would likely not occur. 11/23/2020 Behavioral Finance 57

Financing risk: horizon + margin risks • The path to convergence can be long and bumpy – arbitrageurs have to deal with the possibility of interim liquidations even in the case when convergence is certain. • Substantial variability in the time to termination, even for negative-stubvalue investments that eventually converge: – the average time between initial mispricing and a terminating event is 92 days, – the minimum is 1 day – the maximum is 2796 days • as a result: even if convergence is eventually achieved, the negative-stubinvestment often underperforms the risk-free rate • discourages investments by arbitrageurs who are uncertain of the time to convergence and unable to close the arbitrage spread of their own 11/23/2020 Behavioral Finance 58

Horizon risk (illustration): • increasing the length of the path reduces the arbitrageur's return: "horizon" risk • illustration: if investment strategy yields 15% return over 92 days (sample median) annualized return is 47% • a decrease in the number of days to 25 th percentile increases the return to 238% annualized • an increase to 75 th percentile would decrease annualized return to 14% 11/23/2020 Behavioral Finance 59

Margin risk • If the arbitrageur faces a margin call, he will be forced to post additional collateral or partially liquidate • Example: Creative Computers/Ubid – Creative Computer (parent) ; Ubid (subsidiary). – On December 4 th 1998 parent carves out 20% of its online auctions subsidiary in an IPO. – At the time of IPO announces intention to distribute, after a minimum of 6 month, the remaining shares of Ubid to Creative Computer's shareholders – At the end of first trading day Ubid's total equity was 349$ mln – The implied value of Creative Computer's 80% stake in Ubid was greater then Creative Computers' total market value by appr 80$ mln 11/23/2020 Behavioral Finance 60

What could have been in ideal world • Since shares at IPO companies are often not available shorting for a few days following IPO let's assume that arbitrageur's initial trade was placed on December 9 th, 4 days following the IPO. By the time the value of the stub was -28$mln (mispricing has decreased) • Arbitrageur trying to profit from this opportunity should have shorted 0. 72 shares of Ubid for every 1 long share of Creative Computer • In six months, if the remaining Ubid shares were distributed , the value of the stub would turn positive, anticipated return at the end of six months ~45% 11/23/2020 Behavioral Finance 61

How it happened in reality • By December 18, 1998, the discrepancy between Creative Computers and Ubid stock prices had increased substantially—the value of the stub assets had decreased from negative $28 million to negative $94 million. Using margin maintenance requirements specified by NYSE and NASD, the arbitrageur would have faced a margin call and would have been forced to partially liquidate his position to satisfy maintenance margin requirements. The arbitrageur would have lost 26 percent in seven trading days. 11/23/2020 Behavioral Finance 62

How it happened in reality-2 • On December 21, 1998, the value of Creative Computers’ stub assets decreased to negative $254 million. For a second trading day in a row, the arbitrageur would have faced a margin call and been forced to reduce his position even further, incurring an additional one-day loss of 84 percent. • Bad luck continued when, on the following trading day, the value of Creative Computers’ stub assets fell to negative $505 million, causing a one-day loss of 91 percent. 11/23/2020 Behavioral Finance 63

How it happened in reality-3 • On December 23, 1998, the value of Creative Computers’ stub assets reached its minimum level of negative $766 million. The arbitrageur received his fourth and final margin call and an additional one-day loss of 63 percent. • After December 23, 1998, the prices of Ubid and Creative Computers become to converge. 11/23/2020 Behavioral Finance 64

Creative Computers/Ubid timeline 11/23/2020 Behavioral Finance 65

How did it come out eventually? • The portion of the arbitrageur’s capital that was not liquidated returned 150 percent between the peak mispricing on December 23, 1998, and the spinoff on June 7, 1999. However, because the arbitrageur lost most of his capital prior to December 23, 1999, his overall return from the Creative Computers/Ubid investment was negative 99 percent. • To avoid the costly margin calls, the arbitrageur would have had to post $4. 53 of excess cash for every $1 of long position. Doing so would have generated a return of 8. 7% between December 9, 1998, and June 7, 1999. This is significantly lower then 45. 9% return he could have obtain with the same initial investment had he not been required to liquidate to meet margin calls. • Note: setting so much capital aside either requires perfect foresight or very deep pockets and nerves of steel 11/23/2020 Behavioral Finance 66

What if we create portfolios? • A bit better, but still not so well: abnormal performance in excess of 1% monthly, but barely statistically significant • diversification may alleviate the problem of margin calls to some extent, but not solve it completely • Graph: at one point the spread on most of the negative stub spreads widened 11/23/2020 Behavioral Finance 67

Buy-in risk • if arbitrageur is unable to maintain his short position he will be forced to terminate the trade. • When shares available for shorting are most scarce, brokers cannot maintain their clients’ short positions no matter what interest rate the investor is willing to pay. This situation, which arises when owners of the stock demand that their loaned-out shares be returned, is often referred to as being “bought-in. ” • The possibility of being bought-in at an unattractive price provides a disincentive for arbitrageurs to take a large position and represents a substantial friction to executing the arbitrage trade. • Overall: returns to a specialized arbitrageur would be roughly 50% higher if the path to termination was smooth 11/23/2020 Behavioral Finance 68

Buy-in risk -2 : negative short-rebate • • Consider the case of the most extreme negative short rebate, Stratos Lightwave. An arbitrageur wishing to exploit the relative mispricing of Methode/Stratos Lightwave would have been charged a 40% annual interest rate on short proceeds from short selling Stratos Lightwave. The arbitrageur would have invested in the deal on July 11, 2000, and would have still been invested at the end of the year. Over this period, the equity value of the position increased 21. 1 percent before including the effects of the negative short rebate. However, after paying nearly six months of negative short rebate, the arbitrageur’s return is reduced to -0. 6%. 11/23/2020 Behavioral Finance 69

Why did negative stub value persist? • 1. enormous uncertainty over the economic nature of the mispricing and time to learn about it. • 2. Uncertainty over the distribution of returns (after all there are not so many cases when mispricing was so apparent). Remember that in about 30% of cases arbitrage opportunity terminated without convergence. Events which are causing disadvantageous termination are idiosyncratic, hence could not be hedged against. • 3. uncertainty over the course of actions when mispricing widens. Remember Creative Computers 11/23/2020 Behavioral Finance 70

Punchline • At times some crazy things happen • Mispricing can persist for a long period of time due to limits to arbitrage – Costs and risks for arbitragers are substantial • Additionally, a lot of uncertainty about – The economic nature of the phenomenon (there are not so many of these events to be prepared to act in advance) – The profit opportunities it offers • Distribution of returns is very non-normal and though on average there are some abnormal profits to make at times arbitrageur’s entire wealth would be wiped out before – Transaction costs also have a huge bit on arbitrageur’s profits • Bright side: fewer then 100 cases in 15 years. 11/23/2020 Behavioral Finance 71

Short-selling constraints again. . . • Negative stubs: is the only time when short-selling constraints lead to inflated prices of assets? • Greenwood (2009): – Examines a series of stock splits in Japan in which firms restrict the ability of their investors to sell their shares for a period of approximately 2 months. – By removing potential sellers from the market, the restrictions have the effect of increasing the impact of trading on prices. – The greater the desire of investors to trade, and the greater the restrictions, the larger the impact of the restrictions. Particularly severe restrictions are associated with returns of over 30% around the ex-date, • most of which are reversed when investors are allowed to sell again. – Firms are more likely to issue equity or redeem convertible debt during the restricted period, suggesting strong incentives for manipulation. 11/23/2020 Behavioral Finance 72

SS constraints: spillover to other markets • So far: constraints on short-selling equity lead to inflated prices of equity, may persist for a long time • How about other markets? • Do constraints on short-selling equity have effect on equity derivative markets? Robert Battalio and Paul Schultz “Hastily Implementing Rules that are not Well Thought Out: The Impact of the 2008 Short Sale Ban on Equity Option Markets“ – In the early morning hours of September 19 th, the SEC issued a ban, effective immediately, on short selling for 797 financial stocks. – The ban was set to expire in 10 days, but could be extended to 30 days at the SEC’s discretion. 11/23/2020 Behavioral Finance 73

SS constraints: cont’d 1. Were options markets used to avoid the short selling restrictions? Find no evidence of a migration of trading to the options market. 2. How did the ban affect trading costs and liquidity in the options market? Disastrously. By all measures, trading costs increased dramatically for options on banned stocks. 3. Did biases in option prices emerge during the short sale ban? Yes, it did. Put call parity was grossly violated 4. Did the short sale ban produce price gaps that would allow arbitrage profits to be earned with short selling? Yes, when the arbitrage required selling the shares short and buying stock synthetically. In practice, these opportunities could not be exploited. 11/23/2020 Behavioral Finance 74

Companies Fighting Short Selling • Many corporate managers do not like short-sellers as they believe that their actions are directed toward artificially reducing the price of the stock • “Your activities are mean, shameful and loathsome. They are motivated by appalling avarice and greed, and they will not be permitted to go unanswered”. • Firms can take a variety of actions to impede short-selling of their stocks – Coordinate with their shareholders to withdraw shares from short-selling market – File lawsuit against short-sellers as those spreading false information about the company – Devise technical actions which force shareholders not to lend • Require shareholders to send their stock certificated to the firm’s transfer agent in order to receive a distribution. An owner cannot send in the certificate unless he is in physical possession of it. 11/23/2020 Behavioral Finance 75

Companies Fighting Short selling -2 • Solv-Ex, a firm that claimed to have a technology for economically extracting crude oil from tar-laden sand. • Short sellers claimed Solv-Ex was a fraud. • Managers of Solv-Ex faxed a letter to brokers and shareholders – “To help you control the value of your investment … we suggest that you request delivery of the Solv-Ex certificates as soon as possible”. – Any shareholder which follow Solv-Ex suggestion would have withdrawn his shares from the stock lending market, making short-selling more difficult/expensive. • Additionally, Solv-Ex hireda private investigator to find out who was spreading misinformation about the firm and subsequently filed suit against a well-know short-seller, claiming he had spread false information. 11/23/2020 Behavioral Finance 76

Companies Fighting Short selling -3 • It appears that companies fighting short selling do in fact have a reason to be worried about • Their shares appear to be overpriced – 4 -factor adjusted returns are -2. 40%/month! – Actions, that firms undertake make short-selling more expensive, lead to distorted prices 11/23/2020 Behavioral Finance 77

Companies Fighting Short selling -4 • In 1991 SEC official testified: “many of the complaints we receive about alleged illegal short selling com from companies and corporate officers who are themselves under investigation by the Commission or others for possible violations of the securities or other laws”. • During the same testimony officials from three firms testified as well. • Subsequent to this testimony, the presidents of two of these three firms were prosecuted for fraud. – For third firm, the SEC determined the company had made materially false and misleading statements, but that the evidence was insufficient to prosecute. • What happened to Solv-Ex? • Within a year Solv-Ex delisted and shortly thereafter entered bankruptcy. In 2000 the court ruled that the firm defrauded investors. 11/23/2020 Behavioral Finance 78

A rose is a rose. . . If calls himself does it make him more attractive? – No, common sense tells us it is not (unless it is a credible signal about one’s future looks) – In fact, if you chose a “hot” name it does help • See Cooper et al. (2003): Measuring the effect of renaming a company to “. com” during the bubble years – Roughly a 80% announcement effect • Measuring the effect of removing “. com” after the bubble years – Roughly 70% cumulative abnormal return 79 Behavioral Finance 11/23/2020

Cooper et al ”A Rose by Any Other Name” 11/23/2020 Behavioral Finance 80

Money on the table? • What caused the mispricing? • Errors by individual investors, often confused about what they buy ”When App. Net Systems filed for IPO under symbol APPN, investors began buying shares of Appian Technology (inactive circuit manufacturer) even before the IPO of App. Net systems”. Appian technology earned return of 142. 757% in two days after filing with over 7. 3 mlns shares traded, compared to 200 the day before filing. Micro stocks – Very small size and trading volume – A buy or sell order moves price up or down (price taking) – No arbitrageur would mess up with a profit opportunity of 300$. 11/23/2020 Behavioral Finance 81

Closed-end fund puzzle Mutual fund industry: • Broadly speaking the industry is divided into three types of funds – Mutual Funds (open-end funds) – Closed End Funds – Hedge Funds • Closed-end fund: – Holds publicly traded securities – (Unlike open-end fund) issues a fixed number of shares that are traded on a stock market – To liquidate a holding in a fund investor must trade with other investors, cannot redeem shares with a fund itself. 11/23/2020 Behavioral Finance 82

Closed-end fund puzzle -2 • Lee, Shleifer and Thaler (1991) define the puzzle as: – Closed-end funds are usually issued at about 10% premium to their NAV, more often than not start trading at a premium NAV, and then decline. – On average, closed end funds trade at a discount of 10 to 20 % relative to their NAV – The discount is subject to wide variation over time and across funds. – Discounts disappear as the fund approaches the open end date – Discount co-varies with return on small stocks 11/23/2020 Behavioral Finance 83

Closed-end fund puzzle -3 11/23/2020 Behavioral Finance 84

Earlier rational explanations • Taxes – Reported NAV of a closed-funds does not reflect the capital gains tax that must be paid by the fund if the assets in the funds are sold. – The tax liability associated with assets which have appreciated in value would reduce the liquidation value of the fund’s assets • Problems: – The discounts are too large to be explained by taxes – The discounts should widen when the stock market raises which is opposite to what is observed in reality – At the fund’s closure (open ending) NAV should fall down to the fund’s share prices, however it is funds prices go up to be closer to NAVs 11/23/2020 Behavioral Finance 85

Earlier rational explanations -2 • Assets Illiquidity • – If closed-end funds hold substantial amounts of illiquid (and/or private) stocks the market value of these securities would be lower then that of similar liquid stocks. However, these illiquid securities may be overvalued in calculation of NAV Problems: – large closed-end funds (like TRICON) hold almost exclusively very liquid stocks. The effect is likely to be less then 0. 5% – Additionally regulation required the funds to discount illiquid securities when calculating NAV to reflect their true market value What if closed-end funds hold large blocks of securities and their discounts merely reflect the fact that such blocks could only be liquidated at a discount? – But then we would expect the price of the fund to go down at the announcement of its closure (anticipation of discounted selling) – Instead, the price of the fund goes up, gets closer to NAV. 11/23/2020 Behavioral Finance 86

Earlier rational explanations -3 • Agency costs (management expenses) – Management fees are relatively stable: fixed % of NAV, little fluctuations – Closed-end funds’ discount exhibit wild swings – Additionally, can not explain why closed-end funds initially sell at a premium 11/23/2020 Behavioral Finance 87

Beh. ’l Explanation of closed end fund puzzle � We draw from the model of noise trader risk discussed before � Suppose that the safe asser s is the portfolios of a closed end fund, and the unsafe asset u is the fund itself � Suppose that noise’ traders’ beliefs about the return on u relatively to the return on s are subject to fluctuating sentiment � The risk from holding a closed end fund (and any other security subject to the same sentiment) consists of two parts � The risk of holding fund’s portfolio itself � We abstract from that � Risk that noise trader sentiment about the funds changes � The latter leads to a discount relatively to NAV �Recall: more risky security is traded at a lower price � Since sentiment is stochastic (random) this leads to time-variation in a discount Note: think of price of u as expressed relative to s. In that way the price of s (relative to s) ≡ 1. Now you could fit closed-end fund puzzle into the model of investor sentiments without any adjustments. 11/23/2020 Behavioral Finance 88

Behavioral Explanation -2 • The noise trader approach to closed end fund puzzle explains why fund mispricing relatively to its portfolios is not eliminated by arbitrage • The hedge in which an arbitrageur buys an underpriced closed end fund, and sells short its underlying portfolio, even if feasible and costless, is not a pure arbitrage opportunity unless the arbitrageurs – Have an infitinte time horizon – Never forced to liquidate their positions 11/23/2020 Behavioral Finance 89

Behavioral Explanation -3 Why wouldn’t a threat of takeover push prices back to fundamentals (NAV)? • • Free-raiding funds’ investors would not tender their shares unless they get NAV; no gain for arbitrageur Some of closed-end fund successfully fought multiple restructuring attempts 11/23/2020 Behavioral Finance 90

Behavioral Explanation -4 Why would funds go public at a premium? • • • When investors are particularly overexcited, entrepreneurs can profit by putting assets into closed end funds and selling them to noise traders. In fact, most of the funds are started when discounts are relatively low LST: ”It seems necessary to introduce some type of irrational investors to be able to explain why anyone buys the fund shares at the start when the expected return over the next few months is negative”. 11/23/2020 Behavioral Finance 91

11/23/2020 Behavioral Finance 92

Behavioral Explanation -5 Why is discount eliminated at the announcement of the fund closure? • • • Noise trade risk is being eliminated (reduced) and so the discount You know the terminal date = when the convergence between fund’s price and NAV would be achieved, i. e. when there would be no mispricing The latter means that investing in a close end fund becomes a riskless strategy (subject to the promise of fund’s liquidation to come true) 11/23/2020 Behavioral Finance 93

Behavioral Explanation -6 Securities/portfolios affected by the same sentiment should comove • • Co-movement of discounts of closed-end funds Since closed-end funds are held primarily by individual investors it is not surprising that the discount comoves with returns on small stocks (which are also dispropportionally held by small investors) 11/23/2020 Behavioral Finance 94

Closed-end Fund: Rational Re-thinking Elton, Gruber and Busse (1998): • • • If investor sentiment is a market wide phenomenon (systematic risk factor like size, book/market or momentum) than it should be reflected in stock returns/ values: – Stocks which are more subject to investor sentiment should be priced lower (extra risk) – Relate investor sentiment (proxied by the discount on the closed-end funds) to stock returns and portfolio returns and find no evidence of a relationship Presumably investor sentiment should affect larger stocks (held primarily by large institutional investors) to a lesser degree and have more pronounced effect for smaller stocks – Find no evidence of this after controlling for usual risk factors EGB conclusion: closed-end fund discount is not related to a systematic risk 11/23/2020 Behavioral Finance 95

Steve Ross’s critique: • Consider a fund whose manager is paid a fraction c of the fund’s value at the end each year (say 1%) • What is the value of manager’s claim if investor leaves money in fund forever, and all dividends are reinvested? – PV of first years’ compensation = c x current value of fund – PV of second year’s compensation = c x (1 -c) x current value • The current manager gets everything! This holds for any c>0 {any management fee} 11/23/2020 Behavioral Finance 96

Steve Ross’s critique: (cont’d) • In general, if the fund pays a constant fraction of assets every year, , and if the manager charges a proportional fee , then Fee based discount: Df = /( + ) – Plug in actual numbers and we obtain that average discount is about 7. 7%; just like in reality • Discounts are positively correlated with NAV’s and negatively correlated with market returns – When NAV goes up less dividends are paid: dividend + (unrealized) capital gain – When stock market goes up more dividends are paid: keeping up with a benchmark 11/23/2020 Behavioral Finance 97

Steve Ross’s critique: (cont’d)-2 Why do we observe that returns on small stock and returns on closed end funds are correlated? This holds even for country funds (which invest in foreign securities)? • As a fund manager you want to keep your clients happy: – Once again: small stocks and closed end funds (domestic or country) are held primarily by individual investors – When returns on small stocks do well you want live a good feeling with your clients which also hold small stocks in their portfolios – It is true both for domestic and country funds as they serve the same clientele 11/23/2020 Behavioral Finance 98

Steve Ross’s critique: (cont’d)-3 But why do these funds start at a premium? • • Let’s go back to our nice picture on short-selling constraints Only optimistic investors buy closed end funds at the inception; pessimistic investors could not make their opinions count – no shorting ability; arbitrage is limited optimists 11/23/2020 Behavioral Finance 99

Jonathan Berk’s critique Fee based discount: Df = /( + ) is enough to explain the average magnitude of discounts, but there’s not enough variation in fees to explain the cross-sectional variation in discounts • Managerial ability: – In the absence of fees, good managers should trade at premia and bad managers should trade for a discount – Problem: • Investors must expect that manager is good or average at the IPO • After the IPO investors must expect most managers to be poor • LST: Logic suggests that it is impossible for both predictions to be rational 11/23/2020 Behavioral Finance 100

Jonathan Berk’s critique -2 • Competitive capital markets – Investors always receive a fair return • Berk and Green (2002): tradeoff -- fees (-) vs ability (+) – If managers add more in ability than they charge in fees the funds must trade at premium – If managers add less in ability that they charge in fees the fund must trade at a discount 11/23/2020 Behavioral Finance 101

Jonathan Berk’s critique -3 Inferences: • Uncertainty exists on managerial ability – Neither investors nor the manager himself knows the ability of the manager, they have the same priors and update based on the same info. • What happens? – Bad managers are entrenched and so these funds trade at discounts – Good managers leave so these funds do not trade at premia. 11/23/2020 Behavioral Finance 102

Jonathan Berk’s critique -4 Closed end fund IPO • • • Pick a fee such that a fund trades at par Investors understand that they are providing employment insurance for the manager, the fee is reduced to take this option into account So this means that for the first period, at least, investors expect managers to make more than they charge in fees Post IPO • • Investors expect good managers to leave (or get a pay raise) and bad managers to become entrenched, so they rational expected the average fund to fall into discount They still get a fair return, because in each period, the discount adjusts to ensure this 11/23/2020 Behavioral Finance 103

Jonathan Berk’s critique -5 Discounts • • • Since discounts adjust to ensure that investors get a competitive return, they reflect the cross sectional variation in management ability, so they have wide cross sectional and time series variation Since discounts are the capitalized value of the expected cost of entrenchment, they shrink to zero as the open end date approaches Aside: It’s not cross sectional variation in fees that drives variation in discounts --it is variation in perceived ability. 11/23/2020 Behavioral Finance 104

Summary of rational critiques The conclusion that the behavior of closed end funds is a hard proof evidence of irrationality is premature 11/23/2020 Behavioral Finance 105

Limits to arbitrage: summary • Markets are efficient (mostly) • Prices may not reflect fundamental values Mistakes of investors may not cancel out However, this is not sufficient for the prices to deviate from fundamentals Cause arbitrageurs have to be unable to do their job This may happen due to their limited risk-taking ability • Risk of mispricing widening & performance based arbitrage • Liquidity shortage • Short-selling constraints • At times profit opportunites are also too small for arbitrageurs to be bothered – Real-life arbitrage is risky; taking into accounts convergence risk and transaction costs profit opportunities are usually not that great – – 11/23/2020 Behavioral Finance 106

Limits to arbitrage: summary -2 “prices are right” “no free lunch” “prices are right” Even though the security is mispriced returns are not necessarily predictable Risk of forced liquidation means that many arbitrageurs effectively face short horizon a) may not be able to wait until mispricing disappears = possibility of a loss b) Hence, do not take large positions 11/23/2020 The road to the disappearance of mispricing may be very bumpy; Implies that arbitrageurs’ returns adversely affected by forced liquidations due to capital shortages, margin calls etc; model risk Behavioral Finance 107