Loss Aversion and Incentives Omission Bias Evaluating Talent Momentum vs. Luck

Preview:

DESCRIPTION

Loss Aversion and Incentives Omission Bias Evaluating Talent Momentum vs. Luck. Loss Aversion and Incentives. In 4 th down situations where (statistically) it’s better to go for it, NFL coaches punt/kick 96% of the time There are some exceptions . . . But didn’t used to be this way. - PowerPoint PPT Presentation

Citation preview

Loss Aversion and IncentivesOmission Bias

Evaluating TalentMomentum vs. Luck

Loss Aversion and Incentives

• In 4th down situations where (statistically) it’s better to go for it, NFL coaches punt/kick 96% of the time

• There are some exceptions . . .

• But didn’t used to be this way

2009 Patriots vs. Colts

• 4th and 2 on own 28 yard line, up 34-28

Deviating from convention can be risky

• “My vocabulary is not big enough to describe the insanity of this decision”—Trent Dilfer, former NFL QB and ESPN analyst

• “Ghastly . . . Too smart for his own good this time. The sin of hubris”—Dan Shaughnessy, Boston Globe

• “Fourth-and-jack-ass. That’s our name for a now-infamous play in New England Patriot’s history”—Pete Prisco, CBS

Omission Bias(3 ball vs. 2 strike counts)

Omission Bias(3-0 vs. 0-2 counts)

Omission Bias

• Super Bowl XLII: Giants vs. Patriots

Omission Bias

• 2009 U.S. Open Tennis Championship

Omission Bias

• 1993 NCAA Basketball ChampionshipMichigan vs. NC

Evaluating Talent

1998 NFL Draft: Two “franchise” QBs

Evaluating Talent

Less than 10 years later . . .

Estimated Trade Value of Draft Picks0

.2.4

.6.8

1V

alue

Rel

ativ

e to

#1

Pic

k

1 33 65 97 129 161Overall Draft Pick

Top pick acquired (all possible)Picks exchanged for top pick (observed)

14

Evaluating Talent

Early picks are very highly valued:#1 = #10+#11 = #29+#30+#31+#32

You have to pay about the same

Does this match future performance?

15

Probability “Better Than The Next Guy”0

.1.2

.3.4

.5.6

.7P

roba

bilit

y

1 2 3 4 5 6 7By draft round

Average=.52

Evaluating Talent is Risky• 9 of last 12 #1 picks have been QB’s

Who Else Could You Have Had?

= + a 3rd round pick

= +

Evaluating Talent in the NFL

• What were the #1’s worth?

1x 6,000x 2x 1.5x 3x

Evaluating Talent

How do you hire and fire people?

How confident are you in your ability to judge performance?

The Myth of Momentum

• Quiz #1: Which of the following is the best predictor?

a) Performance over last 5 attemptsb) Performance over last 5 gamesc) Performance over last monthd) Performance over seasone) Performance over multiple seasons

The Myth of Momentum

• Quiz #2: Heading into playoffs, which of the following is the best predictor of success?

a) Performance in most recent gameb) Performance in last weekc) Performance in last monthd) Performance over entire regular season

Fooled by Randomness

• Flip a coin 10 times . . . Which sequence is more likely to be random?1. HTHTHTHTHT2. HHHHHHHHHH3. HHHTHHTTHT

• Which Lotto number is more likely to win?1. 999992. 13679

Damned Statistics

• What does “4 out of his last 5” really mean?

• Is a player 0 for 12 “due” to hit his 13th or doomed to miss it?

• Situational or conditional statistics

24

Momentum in Financial Markets

Momentum is the phenomenon that stocks

which have performed well in the past

relative to other stocks (winners) continue

to perform well in the future, and stocks

that have performed relatively poorly

(losers) continue to perform poorly.

24

25

Academic Discovery

Jegadeeshand

Titman

1993

Asness

1994 1996

Chan, Jegadeesh, and Lakonishok

Fama and French: “…[T]he main embarrassment of the three-factor model, [is] its failure to capture the

continuation of short-term returns documented by Jegadeesh and Titman and Asness…[T]he continuation

anomaly exposes one of its shortcomings.”

1999

Moskowitz and Grinblatt

2004

Grinblatt and Moskowitz

2008

Asness, Moskowitz

and Pedersen

6 new working

papers from AQR

2011

26

Historical Evidence from AcademiaAverage Returns for Portfolios Grouped by Momentum

1927 – 2009

* Source: CRSP Database. Data is based on monthly returns from overlapping portfolios. Momentum is calculated by ranking stocks based on their past 12-month return excluding the most recent month. Returns are in excess of the beta-adjusted CRSP Value Weighted Index. Past performance is not an indication of future performance.

6.0%

9.7%11.0%

13.6%

18.0%

-4.5%

1.1%3.2%

5.8%

9.1%

-0.1

-0.05

0

0.05

0.1

0.15

0.2

P120% of stocks withWorst Momentum

P2 P3 P4 P520% of stocks withBest Momentum

Ann

ual R

etur

n in

Sub

sequ

ent 1

2 M

onth

s (%

)

Average Return of Each Quintile Excess Return of Each Quintile*

27

Momentum is the Fourth Factor

SmB Factor is a long-short portfolio composed of small stocks minus big stocks by market capitalization. HmL Factor is a long-short portfolio composed of stocks with high book to price valuations minus stocks with low book to price valuations. UmD is a long-short portfolio composed of stocks with positive momentum minus stocks with negative momentum by looking at the last twelve months price return excluding the last month. Past performance is not an indication of future performance.

Since the mid-1990s, the standard for any asset pricing study is to adjust for 4 factors (betas):

Size Value Momentum

Annu

alize

d Re

turn

(%)

27

Out of Sample Evidence

• Small and large cap stocks• Industries• Bonds• Currencies• Index futures• Commodities• Corporate bonds• Emerging markets (all asset classes)• REITS

Momentum in 58 Different Contracts

Why?

1. Why does momentum exist in financial market prices, of all assets?

2. Why doesn’t momentum seem to exist in sports, all sports?

Recommended