long-term global market correlationshenry lowenfeld, 1909 ... 1900 1950 2000 ... venezuela # # #...

29
Yale School of Management Long-Term Global Market Correlations Global Linkages Pre-onference IMF, April 26,2002 William N. Goetzmann, Lingfeng Li K. Geert Rouwenhorst

Upload: others

Post on 17-Oct-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Long-Term Global Market Correlations

Global Linkages Pre-onferenceIMF, April 26,2002

William N. Goetzmann, Lingfeng LiK. Geert Rouwenhorst

Page 2: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860
Page 3: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Henry Lowenfeld, 1909

“It is significant to see how entirely all the rest of the Geographically Distributed stocks differ in their price movements from the British stock. It is this individuality of movement on the part of each security, included in a well-distributed Investment List, which ensures the first great essential of successful investment, namely, Capital Stability.”

From: Investment and Exact Science, 1909.

Page 4: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

History of Diversification

! First Mutual Fund: Eendracht Maakt Magt (1774)! Danish and Viennese banks! Danish Tolls and Holstein! Russia and Sweden! Brunswick and Mecklenburg! Postal services of Saxony! Spanish Canals of Taouste and Imperial! British Colonies ! Essequebo! Berbice! Danish American Islands

Page 5: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Diversification: First Mutual Funds

! In the portfolio construction the fund “will observe as much as possible an equal proportionality”

! “Because nothing is completely certain, but subject to fluctuations, it is dangerous to allocate all capital to a single security”

! “Nobody will have reason to believe that all securities will stop paying off at the same time thereby losing the entire invested capital”

Page 6: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Page 7: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Determinants of Diversification

! Correlation between the assets in the portfolio" Take a long-term look at market correlations over time

! The number of assets in the portfolio" Markets have disappeared and (re-)emerged over time

! What is the relative contribution of changing correlations and evolution in the investment opportunity set for diversification benefits?

Page 8: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Preview of Results

! Major shifts in correlations through time.! Correlations are lower during periods of capital market

segmentation than during integration.! Benefits of diversification among core markets are

currently lower than in the 1930’s.! Decomposition shows that currently about half the benefits

come from opportunity set growth and half from correlation structure of markets.

! Recent globalization has not diminished the diversification benefits for the “average” investor

Page 9: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Data

! Monthly equity market index returns for about 50 countries, converted to USD.

! Sources: Jorion and Goetzmann (1999), Global Financial Database, Ibbotson Associates, IFC.

! Simulate perspective of the U.S. based global investor.

Page 10: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Data Issues

! Missing countries, missing data" Little data before 1850" Russia in 19th century, China, Japan" Dividend information often unavailable

! Which markets were investable and when?" Portfolios assume that foreign markets investable

! Transactions costs?! Prices accurately recorded?! Cap-weighted vs. equal-weighted?

Page 11: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Founding Dates For World Equity Markets

1600

1650

1700

1750

1800

1850

1900

1950

2000

Net

herla

nds

Fran

ceBe

lgiu

mAu

stria UK

USA

Irela

ndIta

lyD

enm

ark

Pola

ndR

ussi

aSw

itzer

land

Spai

nPe

ruC

uba

Swed

enH

unga

ryTu

rkey

Cze

chAu

stra

liaN

ew Z

eala

ndAr

gent

ina

Can

ada

Braz

ilJa

pan

(1)

Nor

way

Egyp

t (1)

Sout

h Af

rica

Sing

apor

eH

ong

Kong

Gre

ece

(1)

Chi

leVe

nezu

ela

Mex

ico

Yugo

slav

ian

stat

esZi

mba

bwe

Sri L

anka

Portu

gal

Chi

naKo

rea

Indo

nesi

aFi

nlan

dLe

bano

nSl

oven

iaU

rugu

ayPh

ilippi

nes

Mor

occo

Rou

man

iaLu

xem

bour

gC

olom

bia

Mal

aysi

aPa

kist

anIs

rael

Keny

aTa

iwan

Nig

eria

Kuw

ait

Iran

Tuni

sia

Thai

land

Jam

aica

Jord

anTr

inid

ad-T

obag

oIc

elan

dBa

hrai

nM

aurit

ius

Bots

wan

aG

hana

Swaz

iland

Chi

naSl

ovak

Nam

ibia

Egyp

tZa

mbi

aM

alaw

iM

alta

Tanz

ania

Page 12: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Year

1700

1750

1800

1850

1860

1870

1880

1890

1900

1901

1902

1903

1904

1905

1906

1907

1908

1909

1910

1911

1912

1913

1914

1915

1916

1917

1918

1919

1920

1921

1922

1923

1924

1925

1926

1927

1928

1929

1930

1931

1932

1933

1934

1935

1936

1937

1938

1939

1940

1941

1942

1943

1944

1945

1946

1947

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

Argentina # "Long period of decline"Australia # #

Austria #Belgium # V -XII

Brazil # Predomonantly BondsCanada # #

Chile # ? ?China #

ColumbiaCzechoslovakia #

Denmark # IX-XEgypt

Finland # #France ?

Germany Late 18th Century 1211/IV/32Greece #

Hong Kong #Hungary #

India # #Indonesia

Ireland #Israel

Italy #Jamaica

Japan #JordanKenyaKorea

KuwaitLebanon

LuxembourgMalaysia

Mexico #Netherlands # V -IX VIII VI

New Zealand #NigeriaNorway # # #

Pakistan # #Philippines

Poland #Portugal #

Roumania ?Singapore #

South Africa #Spain #

SriLankaSweden #

Switzerland #Taiwan

ThailandTrinidad

Turkey #UK # #

UruguayUSA #

Venezuela # # #Yugoslavia #

goslavia (Slovenia

Page 13: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

0

10

20

30

40

50

60

1860 1880 1900 1920 1940 1960 1980 2000

Num

ber o

f Cou

ntrie

s

Core Markets Total Available Markets

Sample Market Entry and Exit

Page 14: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860
Page 15: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Basu -Taylor (1999) Episodes

! 1872-1889: Early Integration

! 1890-1914: Turn of the Century

! 1915-1918: World War I

! 1919-1939: Between the Wars

! 1940-1945: World War II

! 1946-1971: Bretton Woods

! 1972-2000: Present

Page 16: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Pairwise Correlations Core Countries

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1872-89 1890-14 1915-18 1919-39 1940-45 1946-71 1972-00

Cor

rela

tion UK-US

UK-FUK-GUS-FUS-GF-G

Page 17: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Average Correlation US UK Germany France

-0.2

0

0.2

0.4

0.6

1872

-8918

90-14

1915

-1819

19-39

1940

-4519

46-71

1972

-00seg

mentat

ioninteg

ratio

n full

Cor

rela

tion

Page 18: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Decomposing diversification benefits

! Benefits of diversification extends far beyond core markets

! Recent era shows high average correlation among core markets, but there are currently many more markets available for diversification.

! Does increase in opportunity set compensate for increase in average correlation?

Page 19: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860
Page 20: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Diversification measure - 1

( )

( )

( )

( )

( )

1

,1

1

1

1

/

1

2

1

12

1

1

=

=

=

=

= +=

n

ii

n

jiji

n

ii

n

ii

n

ii

n

ii

xVarn

xxCovn

xVarn

xVarn

xVarn

nxVar

( )( )

,11

i

ji

xVarxxCov

nn

−+=

Page 21: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Diversification measure - 2

! Diversification benefits measured by the ratio of :" return variance of a diversified portfolio" average return variance of the assets in the portfolio

! Ratio lies between 0 and 1 : " 1: perfect correlations" 0: uncorrelated returns

Page 22: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Page 23: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Decomposing the Benefits of International Diversificationequally-weighted portfolio variance / average market variance

0.0

0.2

0.4

0.6

0.8

1.0

1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Rat

io p

ortf

olio

vol

atili

ty /

aver

age

mar

ket v

olat

ility

Core Countries (limited diversification)

Average four countriesAll Countries (unlimited diversification)

Page 24: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Current diversification benefits -1

! Variance of a portfolio of 4 core markets is currently about 70% of the average individual country variance (30% reduction)

! Core markets have higher correlation than emerging markets: average portfolio of 4 markets provides 50% risk reduction

! Maximum global diversification offers 65% risk reduction

! Expansion of opportunity set contributes about half of the total diversification benefits

Page 25: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Current diversification benefits -2

! Effect of increase in the investment opportunity set has been twofold:" “Emerging” markets have lower correlation than

developed markets: risk reduction rises from 30% to 50%

" The increase in the number of markets: risk reduction rises from 50% to 65%

! Expansion of opportunity set doubles diversification benefits (from 30 to 65%): about half comes from lower correlations have from increase in the number of markets

Page 26: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Equilibrium diversification benefits -1

! Not every investor can hold an equally-weighted portfolio

! Assets have to be held in proportion to market weights

! Weight of smaller/emerging markets needs to be reduced

! What do equilibrium benefits of diversification look like?

Page 27: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Recent Benefits of International Diversificationvalue-weighted portfolio variance / market variance

0

0.2

0.4

0.6

0.8

1

1977 1980 1982 1985 1987 1990 1992 1995 1997 2000

Rat

io p

ortfo

lio v

olat

ility

/ av

erag

e m

arke

t vol

atili

ty

Core Countries (limited diversification)

All Countries (unlimited diversification)

Page 28: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Equilibrium diversification benefits -2

! Benefits are generally lower because value-weighted portfolios are less diversified.

! Despite the slow rise in return correlations, the benefits to international diversification have been remarkably stable over last 25 years

! “Emerging markets” double the benefits to diversification across core markets.

Page 29: Long-Term Global Market CorrelationsHenry Lowenfeld, 1909 ... 1900 1950 2000 ... Venezuela # # # Yugoslavia # goslavia (Slovenia. Yale School of Management 0 10 20 30 40 50 60 1860

Yale School of Management

Conclusions

! Longer view reveals shifts in correlations.! Integration is associated with high market correlation! While correlations are at highpoint in history, investors

benefit from an expansion of the investment opportunity set

! About half of current benefits stem from expansion of opportunity set.

! Diversification benefits have been remarkably stable for “average investor” over last 25 years.