Markus K. Brunnermeier LECTURE 1: INTRODUCTION EMPIRICAL ...€¦ · Lecture 01 Intro: Empirical...

FIN501 Asset PricingLecture 01 Intro: Empirical Regularities (1)

LECTURE 1: INTRODUCTION EMPIRICAL REGULARITIES

Markus K. Brunnermeier


Money, Bonds vs. Stocks

$0

$5

$10

$15

$20

$25

$30

$35

$40

$45

$50

19

72

19

75

19

78

19

81

19

84

19

87

19

90

19

93

19

96

19

99

20

02

20

05

20

08

20

11

SP500

US 1y Tsy

US 10Y Tsy


Monthly Returns

-25.0%

-20.0%

-15.0%

-10.0%

-5.0%

0.0%

5.0%

10.0%

15.0%

20.0%

1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

SP500

US 10Y Tsy

US 1y Tsy

But what is the right horizon?


Compared to Recessions

10

100

1000

10000

4/1

/19

53

4/1

/19

55

4/1

/19

57

4/1

/19

59

4/1

/19

61

4/1

/19

63

4/1

/19

65

4/1

/19

67

4/1

/19

69

4/1

/19

71

4/1

/19

73

4/1

/19

75

4/1

/19

77

4/1

/19

79

4/1

/19

81

4/1

/19

83

4/1

/19

85

4/1

/19

87

4/1

/19

89

4/1

/19

91

4/1

/19

93

4/1

/19

95

4/1

/19

97

4/1

/19

99

4/1

/20

01

4/1

/20

03

4/1

/20

05

4/1

/20

07

4/1

/20

09

4/1

/20

11

Recession

SPX Index


Stock Market Predictability

• Random walk hypothesis:

– Stock market prices evolve according to a random walk, and therefore cannot be predicted.

• Changes in stock prices can only be attributed to:

– News on future cash flows

– Change in “Risk premia” – the “dark matter” of finance

– Shifts in Behavioral Bias


Testing Market Predictability• Cross Section vs. Time-Series

– Cross-Sectional studies refer to data collected at the same point in time, or regardless of differences in time

– Time-Series studies refer to a sequence of data points and look at how the data changed through time

• To test market predictability:

– Cross-Sectional studies look at whether some factors can explain the stock price changes, potentially in contradiction to the random walk hypothesis

– Time-Series studies look at the existence of time-related patterns (e.g. trends, seasonality) or event-specific behavior that would invalidate the random walk hypothesis


Dividend/Price Ratio and Stock Prices

A regression of the S&P 500 index price growth on the dividend/price ratio shows that the D/P ratio is a good predictor of future price growth

Campbell and Shiller (1986)


Stocks are more volatile than consumption growth

-50.0%

-40.0%

-30.0%

-20.0%

-10.0%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Nominal GDP

S&P500 (w/ Dividends)


In the cross section: return & risk

Source: Ang, Hodrick, Xing, Zhang 2008


Adding stocks in alphabetic order

0.00%

5.00%

10.00%

15.00%

20.00%

25.00%

30.00%

35.00%

40.00%

45.00%

50.00%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

S&P 100 Stocks from 1990-2011


Measuring Risk differently

• When combining assets to a portfolio

– Idiosyncratic component diversifies away

– Covariance captures contribution of asset to portfolio’s risk

• Use Covariance as risk measure


Covariance with market

0.00%

0.50%

1.00%

1.50%

2.00%

2.50%

3.00%

3.50%

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

AAPL ABT

AEP AMGN

APC AXP

BA BAX

BHI BK

BMY CAT

CL CPB

DD DELL

DIS DOW

EMC F

FDX GD

GE HAL

HD HNZ

IBM INTC

JNJ KO

LLY LOW

MCD MDT

MMM MO

MRK MSFT

NSC ORCL

monthly return


Covariance with consumption

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

AAPL ABT

AEP AMGN

APC AXP

BA BAX

BHI BK

BMY CAT

CL CPB

DD DELL

DIS DOW

EMC F

FDX GD

GE HAL

HD HNZ

IBM INTC

JNJ KO

LLY LOW

MCD MDT

MMM MO

MRK MSFT

NSC ORCL

annual return


Size and value effect of stock returns


Value vs. Growth Stocks

-60

-40

-20

0

20

40

60

19

80

19

81

19

82

19

83

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

Large Value-Growth (%)

Mid Value-Growth (%)

Small Value-Growth (%)


Winner vs. Losers

• All shares are ranked each yearon basis of five-year holdingreturns

– Includes companies that de-list

• Assigned to portfolio by decileof performance

– Graph shows average performanceof equally weighted portfolios adjusted by equally weighted market portfolio

• Momentum or mean-reversion?– Short-run

– Intermediate

– Long-term

• UK Stocks 1960-2002 – see paper http://papers.ssrn.com/sol3/papers.cfm?abstract_id=998418.

• Original article: DeBondt & Thaler 1985

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=998418


Market Efficiency in Event Studies

T

t

tT AARCAAR

30

-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30

Over-reaction

Efficient Reaction

Under-reaction

T

Important: Information has to become public at a single moment


Event StudiesObjective: Examine if new (company specific) information is incorporated into the stock price in one single price jump upon public release?

1. Calculate the daily excess returns 𝐴𝑅𝑡 = 𝑅𝑖𝑡 − 𝑅𝑚𝑡relative to the market or benchmark for 30 days prior and after release

𝑡 = −30,−29,… ,−1,0,1,… , 29,30

2. For each relative date 𝑡, calculate average returns and cumulative returns across events

𝐴𝐴𝑅𝑡 =1

𝑁 𝑖=1𝑁 𝐴𝑅𝑖𝑡

𝐶𝐴𝐴𝑅𝑇 = 𝑡=−30𝑇 𝐴𝐴𝑅𝑡


Event Study: Earning Announcements

Event Study byBall and Brown (1968)Pre-announcement drift prior to earnings due to insider trading

! against strong-form

Post-announcement drift

! against semi-strong form


Event Study: Earning Announcement

(MacKinlay 1997)

Cumulative abnormalreturns around earningannouncements


Event Study: Stock SplitsEvent Study on Stock Splits byFama-French-Fischer-Jensen-Roll(1969)

Split is a signal of good profit

Pre-announcement drift can be dueto selection bias (only firms whose price rose) or insider trading.

! inconclusive

No post-announcement drift! for weak form

Selection bias or Insider trading


Event Study: Take-over Announcement


• The Expectations Hypothesis is the proposition that the long-term yield is determined only by the market’s expectations of future short-term yield

• Let 𝑍(𝑠, 𝑡) and 𝑦(𝑠, 𝑡) be the discount factor (zero-coupon price) and the yield for a zero-coupon price bought at 𝑠 that matures at 𝑡.

• The 1-year return on a 1-year bond

𝑟1,1 = 𝐸𝑡 ln1

𝑍(𝑡, 𝑡 + 1)= ln

1

𝑒−𝑦 𝑡,𝑡+1 1= 𝑦(𝑡, 𝑡 + 1)

• The 1-year return on a m-year bond bought at t and sold at 𝑡 + 1.

𝑟1,𝑚 = 𝐸𝑡 ln𝑍(𝑡 + 1, 𝑡 + 𝑚)

𝑍(𝑡, 𝑡 + 𝑚)= 𝐸𝑡 ln

𝑒𝑦(𝑡+1,𝑡+𝑚)(𝑚−1)

𝑒−𝑦 𝑡,𝑡+𝑚 𝑚

= 𝑚𝑦 𝑡, 𝑡 + 𝑚 − 𝑚 − 1 𝐸𝑡[𝑦(𝑡 + 1, 𝑡 + 𝑚)]

Government Bonds


• By the EH, single-period holding on bonds of all maturities are equal in expectation. Therefore setting 𝑟1,1 = 𝑟1,𝑚and rearranging we get

𝑦 𝑡, 𝑡 + 𝑚 − 𝑦 𝑡, 𝑡 + 1 = 𝑚 − 1 𝐸𝑡 𝑦 𝑡 + 1, 𝑡 + 𝑚 − 𝑦 𝑡, 𝑡 + 𝑚

• According to the EH, the yield spread 𝑦 𝑡, 𝑡 + 𝑚 − 𝑦(𝑡, 𝑡 + 1) has a positive relation with short-term changes in the long-term bond yield. However, this does not hold empirically

• Source: Campbell 1995

Government Bonds


Corporate BondsA very popular model to value

corporate debt is the Merton Model,

which postulates that the assets of a

firm follow a geometric Brownian

motion process.

However, this model

implies a credit spread to

Treasuries that is

consistently lower than

the observed credit

spreads.


Fixed income: Commercial paper


Rollover risk: Composition of ABCP

0

50

100

150

200

250

300

350

Ma

r-00

Se

p-0

0

Ma

r-01

Se

p-0

1

Ma

r-02

Se

p-0

2

Ma

r-03

Se

p-0

3

Ma

r-04

Se

p-0

4

Ma

r-05

Se

p-0

5

Ma

r-06

Se

p-0

6

Ma

r-07

Se

p-0

7

Ma

r-08

Se

p-0

8

$ B

illi

on

s

Other

Non-U.S. ResidentialMortgages

Student Loans

Credit Cards

Autos

Commercial RealEstate

Home Equity(Subprime)

ABS issuance

Source: JPMorgan


Important spreads


Derivatives

What’s a Credit Default Swap?


ABX index


DerivativesIn the Black-Scholes model for options pricing, a European

call option on a stock worth S with strike K and maturity T

today is worth:

back out the volatility for

different strikes with observed

option prices.

• Implied volatility

• Model ⇒ constant

• Reality ⇒ smile

Strike

Imp

lied

Vola

tilit

y


The VIX

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

12/26/03 12/26/04 12/26/05 12/26/06 12/26/07 12/26/08 12/26/09 12/26/10 12/26/11


Frictions

• In Physics

– Drop of a stone = drop of a feather

– Aviation wouldn’t work without frictions

• In Finance

– Walrasian auctioneer

– Asymmetric information

• Margins/haircuts/

• Limited risk bearing capacity, limited (risky) borrowing, …


Beyond prices: Margins/Haircuts

• Margin/haircut/LTV determines max leverage ratio

Gorton-Metrick (2011)

“Repo Run”

But …


Copeland,Martin, Walker (2011)

Margins very stable in tri-party repo market

– contrasts with Gorton&Metrick (2011)

– no general run on certain types of collateral

• http://www.ny.frb.org/research/staff_reports/sr477.pdf

Run (non-renewed financing) only on select counterparties

– Bear Stearns (anecdotally)

– Lehman (in the data)

Like 100% haircut…(counterparty specific!)

http://www.ny.frb.org/research/staff_reports/sr477.pdf


Liquidity mismatch: sensitivity of wealth shifts

• Macro: real projects techno + market liquidity– Total supply in economy irreversibility + specificity

• Finance: financial claims market liquidity– How liquidity is distributed in economy – intermediation chain

• Shifts in wealth distributions affects costs of liquidity (amplification)

Technological liquidity Reversibility of investment

Market liquidity Specificity of capital

Price impact of capital sale

A L

Maturity mismatch

Funding liquidity Maturity structure of debt

Can’t roll over short term debt

Sensitivity of margins

Margin-funding is recalled


Conclusion

• Risk & Return

– What is risk?

• Fixed income – risky borrowing/lending

• Derivatives (CDS)

• Market efficiency

– Asymmetric information

• Friction Finance

– Margins/haircuts/LTV


Overview

• Frictionless Finance

One period model Part I

Multi-period model Part II

• Friction Finance Part III

Markus K. Brunnermeier LECTURE 1: INTRODUCTION EMPIRICAL ...€¦ · Lecture 01 Intro: Empirical...

Documents

Transcript of Markus K. Brunnermeier LECTURE 1: INTRODUCTION EMPIRICAL ...€¦ · Lecture 01 Intro: Empirical...