06/19/22 Market Efficiency 1 Market Efficiency and Market Efficiency and Information Processing in Information Processing in Financial Markets Financial Markets Andrei Simonov
Mar 28, 2015
04/10/23Market Efficiency1
Market Efficiency and Information Market Efficiency and Information Processing in Financial Markets Processing in Financial Markets
Andrei Simonov
04/10/23Market Efficiency2
OutlineOutlineEfficient Markets HypothesisPredicting future returns from the past returns
– Value strategies– Momentum Strategies
Analysts & Information dissemination
04/10/23Market Efficiency3
Do security prices reflect information ?Why look at market efficiency
– Implications for business and corporate financeMitigation of agency problems: market sees through
– Implications for investment Impossibility to ”beat the market”
Efficient Market Hypothesis Efficient Market Hypothesis (EMH)(EMH)
04/10/23Market Efficiency4
Random Walk - stock prices are random– Actually
submartingaleExpected price is
positive over timePositive trend and
random about the trend
Random Walk and the EMHRandom Walk and the EMH
Security Prices
Time
04/10/23Market Efficiency5
Why are price changes random?Prices react to informationFlow of information is randomTherefore, price changes are random
Random Price ChangesRandom Price Changes
04/10/23Market Efficiency6
04/10/23Market Efficiency7
Case study:Challenger DisasterCase study:Challenger Disaster
Jan 28, 1986 at 11:39 am Shuttle explodes on live TV
Four companies involved: Rockwell, Martin Marietta, Morton Thiokol and Lockheed. Trading is suspended for 90 min, companies are in ”no comments” mode.
Feb. 2nd: first mention of faulty seals Feb. 5th: First time MT is mentioned as prime
suspect March 31st: Problems with O-rings reported by
Fortune.
04/10/23Market Efficiency8
What happen on NYSE?What happen on NYSE?Variable MT Lockheed MM Rockwell
Ret(28.02) -11.86% -2.14 -3.25 -2.48
3mo mean 0.21% 0.07% 0.14% 0.06%
3mo StdDev 1.86% 1.36% 1.80% 1.80%
Daily Volume, Kshares
1740 668 446 563
3mo mean 100 350 200 221
3mo StdDev 60 160 137 117
04/10/23Market Efficiency9
Stock prices fully and accurately reflect publicly available information
Once information becomes available, market participants analyze it
Competition assures prices reflect informationGrossman-Stigliz ParadoxForms of EMH
EMH and CompetitionEMH and Competition
StrongAll
available info,Incl.
private
Semi-Strong
All publicInfo
WeakOld stock prices
04/10/23Market Efficiency10
Econometric tool to study EMHEconometric tool to study EMH
Event studyMeasure abnormal returnProblem: what is normal return?Problem: How to measure event date?
04/10/23Market Efficiency11
Example: Takeover premium before Example: Takeover premium before and after introduction of insider trading and after introduction of insider trading
laws (Bris, 2000)laws (Bris, 2000)
BP oil spill
04/10/23Market Efficiency12
04/10/23Market Efficiency13
Predicting from past returnsPredicting from past returns
04/10/23Market Efficiency14
Buying losers and selling winnersBuying losers and selling winners(1)(1)
• "Contrarian" strategy; the opposite is a "price-momentum" or "relative strength" strategy
de Bondt and Thaler (1986) considered all NYSE stock return data from 1926 to 1983. They measured cumulative abnormal returns over 36-month periods and identified the 35 best and worst performers which they called winners and losers. The diagram shows the performance of these stocks over the subsequent months.
04/10/23Market Efficiency15
Buying losers and selling winnersBuying losers and selling winners (2)(2)
04/10/23Market Efficiency16
Buying winners and selling loosersBuying winners and selling loosersShort-run momentum effect: Select 6mo past
winners/loosers and create 0-cost portfolio. Hold it for 3-6 mo.– Industry effect (Grinblatt/Moskowitz) – essentially
missed macroeconomic trend– Cross-sectional variability in stock returns (Caul&
Condrad, but rejected by Grundy&Martin)Money are lost in 261 out of 828 mo.Feasible only by institutions (high transaction
cost)
04/10/23Market Efficiency17
Earnings announcementsEarnings announcements
Bernard and Thomas (1989) and many others before them document that, subsequent to the announcement of earnings, the stock price continues to drift up for "good news" firms and down for "bad news" firms [Price/earnings momentum].
The enclosed figures are based on a sample of approximately 85,000 observations for NYSE and AMEX stocks over the 1974-1986 period.
04/10/23Market Efficiency18
04/10/23Market Efficiency19
04/10/23Market Efficiency20
Event driven investingEvent driven investing
Extension of the same idea to other events:– Analysts’ forecasts and revisions– Share buy backs– Dividend continuations or not– Etc..
In my opinion, most promising form of active (tactical) investing
04/10/23Market Efficiency21
NewslettersNewslettersfrom Graham and Harvey, from Graham and Harvey, FAJFAJ, , Nov/Dec 97Nov/Dec 97
As a group, newletters do not appear to possess any special information about the future direction of the market (see picture),
Nevertheless, investment newsletters that are on a hot streak (have correctly anticipated the direction of the market in previous recommendations) may provide valuable information about future returns.
04/10/23Market Efficiency22
Over the past decade, I have attempted to exploit many of the seemingly most promising ”inefficiences” by actually trading significant amount of money... Many of these effects are surprisingly strong in the reported empirical work, but I have never yet found one that worked in practice.
Richard Roll
04/10/23Market Efficiency23
How information is reflected in How information is reflected in prices?prices?
Via trading and price discovery Via public announcements Via investment research
– Analyzing analysts’ “Task Accuracy” just how good are analysts at what they do? we’ll examine estimates and recommendations
– Evaluating their Decision Making are they better at some things than others? why do they make the errors they make?
– Debiasing and Using Analyst Information we do get their info, how best can we use it?
04/10/23Market Efficiency24
What do (sell-side) What do (sell-side) security analysts security analysts dodo??
Tasks?– Estimate industry & company models & EPS
– Recommend the best securities
– Sell new securities to investors Ultimate Goal?
– Find the “right” price for MSFT or Amazon.com? Not a trivial task . . .
Incentives?– “On the folly of rewarding A, while hoping for B”
04/10/23Market Efficiency25
Task #1: forecasting EPSTask #1: forecasting EPS
EPS: a virtual sub-industry in investing– IBES, First Call, others
– A separate II All-American category: “EPS accuracy”
How good are analysts at forecasting EPS?
– An opinion: “somewhere between mediocre and bad”
– Dreman and Berry (1995):
off by more than 10% over 55% of the time
04/10/23Market Efficiency26
Dreman and Berry, FAJ, ‘95Dreman and Berry, FAJ, ‘95
04/10/23Market Efficiency27
… … and Montier (DKW)and Montier (DKW)
04/10/23Market Efficiency28
Task #1: forecasting EPS Task #1: forecasting EPS (cont.)(cont.)
Why “between mediocre and bad” ?– In spite of “help” from the companies– not yet adept at “Games Firms Play”
Incentives of firms “Earnings Manipulation to Exceed Thresholds”
Why?– A discretionary component in EPS?– Listening too carefully to the firms? Or not carefully
enough?– Missing the macro-economic trends! See Chopra (FAJ,
Nov/Dec ‘98) Great article!
04/10/23Market Efficiency29
Task #2: recommending stocksTask #2: recommending stocks
Predicting relative stock price performance is “hard”– Recommendations: a pure test of market “skill”
– “Efficient markets” argues against success
The Catch 22 of market efficiency– The markets need information “snoopers”
– Grossman and Stiglitz: An “equilibrium level of inefficiency” is needed in markets. The inefficiency is needed to pay the “snoopers”.
04/10/23Market Efficiency30
Do Brokerage Analysts’ Recommendations Have Do Brokerage Analysts’ Recommendations Have Investment Value?”, Womack, JF ‘96Investment Value?”, Womack, JF ‘96
Buy Recommendations
– Lots of them
– Do move stocks, about 3% on average
– Stocks continue to go up 4-6 more weeks
Sell Recommendations
– 1 for every 15 Buys
– Taken seriously, -5%
– Stocks drift lower for 6 more months, -9%
Ç Ç Ç Ç Ç Ç Ç
ÇÇ Ç Ç Ç Ç Ç
ÇÇ
ÇÇ
ÇÇ Ç
Ç
ÜÜ Ü Ü
ÜÜ Ü
Ü Ü
ÜÜ
Ü Ü ÜÜ
ÜÜÜ Ü
ÜÜ
Ü-6 -5 -4 -3 -2 -1 0
0.8
0.85
0.9
0.95
1
1.05
1.1
1.15
1.2
Pric
e R
elat
ive
to R
ecom
men
datio
n Ev
ent D
ate
Months Relative to Recommendation Event Date
ÇBuy Recs
ÜSell Recs
Months Relative to Removal from Rec.1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
Ç Ç Ç Ç Ç Ç Ç
ÇÇ Ç Ç Ç Ç Ç
ÇÇ
ÇÇ
ÇÇ Ç
Ç
ÜÜ Ü Ü
ÜÜ Ü
Ü Ü
ÜÜ
Ü Ü ÜÜ
ÜÜÜ Ü
ÜÜ
Ü-6 -5 -4 -3 -2 -1 0
0.8
0.85
0.9
0.95
1
1.05
1.1
1.15
1.2
Pric
e R
elat
ive
to R
ecom
men
datio
n Ev
ent D
ate
Months Relative to Recommendation Event Date
ÇBuy Recs
ÜSell Recs
Months Relative to Removal from Rec.1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
04/10/23Market Efficiency31
Do Brokerage Analysts’ Recommendations Have Do Brokerage Analysts’ Recommendations Have Investment Value?”, Womack, JF ‘96Investment Value?”, Womack, JF ‘96
Removals of Buy Recs– Analysts pick stocks that
have recently outperformed by 5%ish
– Stocks have negative abnormal returns for 3-4 months after removal
– Total underperformance of stocks after a “buy” removal: - 7%
Ç Ç Ç Ç Ç Ç Ç
ÇÇ Ç Ç Ç Ç Ç
ÇÇ
ÇÇ
ÇÇ Ç
Ç
ÜÜ Ü Ü
ÜÜ Ü
Ü Ü
ÜÜ
Ü Ü ÜÜ
ÜÜÜ Ü
ÜÜ
Ü-6 -5 -4 -3 -2 -1 0
0.8
0.85
0.9
0.95
1
1.05
1.1
1.15
1.2
Pric
e R
elat
ive
to R
ecom
men
datio
n Ev
ent D
ate
Months Relative to Recommendation Event Date
ÇBuy Recs
ÜSell Recs
Months Relative to Removal from Rec.1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
Ç Ç Ç Ç Ç Ç Ç
ÇÇ Ç Ç Ç Ç Ç
ÇÇ
ÇÇ
ÇÇ Ç
Ç
ÜÜ Ü Ü
ÜÜ Ü
Ü Ü
ÜÜ
Ü Ü ÜÜ
ÜÜÜ Ü
ÜÜ
Ü-6 -5 -4 -3 -2 -1 0
0.8
0.85
0.9
0.95
1
1.05
1.1
1.15
1.2
Pric
e R
elat
ive
to R
ecom
men
datio
n Ev
ent D
ate
Months Relative to Recommendation Event Date
ÇBuy Recs
ÜSell Recs
Months Relative to Removal from Rec.1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
04/10/23Market Efficiency32
Brokerage Analysts’ Recommendations, Brokerage Analysts’ Recommendations, Womack, JF ‘96Womack, JF ‘96
Other conclusions– Smaller stocks respond more, and drift more after
recommendations, too– Are the abnormal returns from “stock picking” or
“market timing”?
– Very substantial asymmetry between the value of 1) (the large amount of) positive new and 2) the small amount of negative news
When they say “sell” or “remove from buy”, watch out!
04/10/23Market Efficiency33
What are analysts good at?What are analysts good at?
A study of Womack:– Reasons analysts give for their
recommendations– Then, categorizing them into four or five
broad categories, then sub-categories– Two very common categories of reasons
“its really cheap” by relative or historical valuation standards
“something new is or will happen”, new news
04/10/23Market Efficiency34
““Can Investors Prophet from the Prophets?”Can Investors Prophet from the Prophets?” by by Barber, Lehavy, McNichols & Trueman ‘99Barber, Lehavy, McNichols & Trueman ‘99
In a non-event study context, they find that the consensus recommendation “average” has value
– Uses top and bottom quintiles of “averages”
– But, not if you wait 30 days to act on the “revisions”
– Therefore, the important finding of both studies: value of buy recs is gone in a month, but, value of negative recs lasts longer
04/10/23Market Efficiency35
Task #3: Selling new securitiesTask #3: Selling new securities
Underwriting transactions are highly profitable for firm and analyst– The changing role of security analysts in last
decade: they’re now the main “pitch” people
– Compensation to analyst for this is very big (double or triple the salary)
2nd year analyst at M.S offered $500K vs. corporate finance at $250K
04/10/23Market Efficiency36
““Why do firms switch underwriters?”Why do firms switch underwriters?”Krigman, Shaw, and Womack ‘99Krigman, Shaw, and Womack ‘99
Examined issuers switching to a new lead underwriter for second offering (30% of second-time issuers)– Conducted a survey of switching CFOs– Analyzed other empirical data for switchers vs.
non-switchers “Research coverage” and “influential analyst”
were top reasons to switch– Along with “trade up” to higher reputation
04/10/23Market Efficiency37
““Conflicts of Interest and the Credibility of Conflicts of Interest and the Credibility of Underwriter Analyst Recommendations”, Michaely Underwriter Analyst Recommendations”, Michaely and Womack, RFS ‘99and Womack, RFS ‘99
Are analysts “truth telling” or “rent seeking”?– MW examine recommendations by the lead
underwriter vs. all other analysts during the first year after the IPO
Their behavior is quite “suspicious”– underwriters’ recommendations are often called
“booster shots” are more likely to be made by underwriter when stock
is doing poorly; not so for non-underwriters!
04/10/23Market Efficiency38
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12
Months (Before) / After Buy Recommendation
Buy Recommendations by Non-Underwriters N=102
All Buy Recommendations N=214
Buy Recommendations byUnderwriters N=112
3-Day Event Return: +2.7% (Underwriters) vs. +4.4% (Non-Underwriters)
Michaely and Womack (1999) "Conficts of Interest . . . ", forthcoming, RFS
Figure 1: Cumulative Mean Size-Adjusted Event Return for Firms Receiving New Buy Recommendations within One Year of their IPO, Conditional Upon the Source of Recommendation
04/10/23Market Efficiency39
Underwriters’ Conflicts of Interest: Underwriters’ Conflicts of Interest: Does the market “Does the market “understand understand ”?”?
Investors do discount the initial info in the underwriter’s buy recommendation– But, stock still goes up +2.7% (vs. +4.4%)
On average, stocks recommended by non-underwriters increase 13% in next six months, market adjusted– But, for underwriters’, stocks decrease 5%!
Interestingly, most brokerage firms’ analysts do better on “other people’s stocks”– For 12 of the 14 large firms, their recs on their own
underwritings do worse than their recs on others’ !
04/10/23Market Efficiency40
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0 3 6 9 12 15 18 21 24 27 30
Months after IPO
Firms withRecommendations by Non-Underwriters Only N=44
Firms withRecommendations by bothUnderwriters and Non-Underwriters N=41
All Firms conducting IPOs(overall average) N=391
Firms with NoRecommendations N=191
Firms withRecommendations byUnderwriters Only N=63
IPO First Day Return
Michaely and Womack (1999) "Conficts of Interest . . . ", forthcoming, RFS
Figure 2 : Cumulative Mean Buy-and-Hold Size-Adjusted Return for Companies Conducting Initial Public Offerings in 1990-1991 Conditional Upon Source of Brokerage Recommendations. Cumulative Return begins at the IPO Price.
04/10/23Market Efficiency41
Scandals:Scandals:
Spinning: Allocating hot IPOs to the personal brokerage accounts of top executives in return for company business Global settlement bans spinning by major underwriters
Laddering: Requiring the purchase of additional shares in the aftermarket in return for IPOs
Analyst conflicts of interest: Giving “buy” recommendations in return for underwriting and M&A business
Commission business in return for IPOs: Underwriters allocated IPOs primarily to investors that generated a lot of commissions on other trades
04/10/23Market Efficiency42
04/10/23Market Efficiency43
Conflicts of Interest Case: Lehman Conflicts of Interest Case: Lehman Brothers and RealNetworksBrothers and RealNetworks
07/11/2000
Stock drops from $52 to $38 in 10 days. Analyst issues report calming investors and reiterating the strong buy
07/18/2000
Stock bounces back. Analyst sends a message to one investor saying “RNWK has to be a short big time”. He explains its inconsistency regarding public reports: “We bank these guys”
07/19/2000
One day later, analyst issues report describing quarter results as “stellar” and reiterating strong buy
Lehman co-managed SEO for RealNetworks in June 1999 and kept a strong buy on the stock until June 2001
Source: Global Settlement Letter of AWC for Lehman Brothersdays (Jun/2000-Jun/2001)
Jan/2001
Stock is already at $9. Analyst explains privately to an investor that RNWK “is a short”, but does not changeits strong buy rating
9
04/10/23Market Efficiency44
Where were regulators?Where were regulators?
04/10/23Market Efficiency45
Survey of IPO InvestorsSurvey of IPO Investors
“Do you think that investors expect reputable underwriters to take some account of true investment value in deciding the offering price in an IPO, rather than just the price the market will bear on the day of the offering?”
84% agree
04/10/23Market Efficiency46
Survey of IPO InvestorsSurvey of IPO Investors
Have you done any calculations of what the true fundamental value of a share in the company was, and compared the price of a share with this value?– 80% no.
04/10/23Market Efficiency47
Is underwriter bias intentional?Is underwriter bias intentional?
Analysts are conflicted between two goals: their long-term reputation (“truth telling”) and profit generation for their firm– Survey results suggest “intentional”
Kahneman and Lovallo, MS, ‘93– The “inside view” vs. “the outside view”
a very important behavioral concept The most important issue w.r.t. analysts
04/10/23Market Efficiency48
Other possible conflicts & Other possible conflicts & problemsproblemsBetween sell-side and proprietary trading
sideAnalysts invests their own moneyOther pressure from the managementHerding
– Follow the crowd– Career concern
04/10/23Market Efficiency49
Using analyst information: conclusionsUsing analyst information: conclusions
The “on average” value of analysts’ info is short-lived and only modestly positive– especially for the positive news
My rules-of-thumb:– When hearing a new recommendation:
“Is the analyst’s firm the investment banker? “What in this is out of consensus?” “Is is a “news” or “valuation” story?
– When hearing a negative report:“It’s probably true.”