page 1 important: this document is for internal use only. this document presents the ideas and the...

20
Page 1 Important: this document is for internal use only. This document presents the ideas and the views of the Strategy team and does not reflect CAAM Investment Committee's decisions. This document may not be reproduced or copies circulated without Sovereign Bonds and Socially Responsible Investment Bastien Drut Fixed Income, FX and Volatility Strategy CAAM Université Paris Ouest Nanterre La Défense Université Libre de Bruxelles AFG 17 December 2009

Post on 18-Dec-2015

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1

Important: this document is for internal use only. This document presents the ideas and the views of the Strategy team and does not reflect CAAM Investment Committee's decisions. This document may not be reproduced or copies circulated without authority.

Sovereign Bonds and Socially Responsible InvestmentBastien Drut

Fixed Income, FX and Volatility Strategy CAAMUniversité Paris Ouest Nanterre La DéfenseUniversité Libre de Bruxelles

AFG 17 December 2009

Dec. 2009 Page 2 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Sovereign Bonds and Socially Responsible Investment

Crisis resulted in huge stimulus packages

Very high level of public debt (in the USA: from 65% of GDP in 2007 to 90% in 2009) on the capital markets

How to construct a responsible sovereign bond portfolio?

Dec. 2009 Page 3 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Sovereign Bonds and Socially Responsible Investment

Portfolio construction in the SRI framework

Data and methodology

Is there a cost to be SRI in the government bonds market?

Dec. 2009 Page 4 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Portfolio construction in the SRI framework

SRI empirical literature is mainly focused on:Performances of ethical or “SRI” funds: do SRI funds underperform or outperform? Renneboog et al. (2008)

Performances of “SRI” stocks portfolios: do well-rated companies perform better? Derwall et al. (2005)

Very few papers on the links between sovereign bond returns and extrafinancial features

Erb, Harvey & Viskanta (1996): bonds of countries with decreasing political risk performs better than bonds of countries with increasing risk

Connolly (2007): SP credit ratings strongly correlated with corruption indexes

Little literature on the construction of govt bonds portfolios with SRI “constraint”…

Dec. 2009 Page 5 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Portfolio construction in the SRI framework

According to the efficient markets theory, imposing a constraint on the portfolio implies a diversification cost

So, we want to answer to the questions:

When does the increase of the SRI value of a sovereign bond portfolio imply a significant cost of diversification?

In other words, is the efficient frontier significantly modified when we increase the SRI value of a portfolio?

Dec. 2009 Page 6 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

Sample: 20 developed countries

Sovereign bonds returns: Citigroup World Government Bonds Indices (WGBI) all maturities in US dollars hedged for FX variations

Socially responsible index: VIGEO Sustainability Country Ratings

For 162 countries, Vigeo rates from 0 to 100 the countries on 3 SRI dimensions:

The Environmental rating

The Governance rating

The Social and Solidarity rating

Dec. 2009 Page 7 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

CountryVigeo

Sustainability Country Rating

E S G

Sweden 86.89 71.05 91.18 98.45Norway 86.27 68.30 92.89 97.64Finland 82.51 65.18 84.68 97.67

Switzerland 81.77 74.24 79.48 91.58Denmark 81.20 60.94 84.86 97.80Austria 80.71 67.14 77.60 97.40

United Kingdom 80.63 64.94 81.98 94.98Netherlands 80.56 56.80 87.71 97.18

Germany 77.64 61.71 76.65 94.56France 77.38 60.29 80.27 91.58

Belgium 75.79 52.44 85.54 89.39Ireland 75.66 51.25 82.84 92.89Spain 74.57 52.84 77.91 92.95

Australia 74.11 57.74 72.93 91.67New Zealand 73.55 54.20 80.46 86.00

Italy 72.33 54.14 77.09 85.76Portugal 71.27 51.67 68.54 93.60Canada 70.60 48.91 78.95 83.92Japan 67.41 52.69 72.20 77.34

United States 59.46 47.75 67.89 62.83

Dec. 2009 Page 8 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

i

ii ratingSRIcountryweightratingPortfolioSRI

As measure of portfolio responsibility, we introduce the SRI portfolio rating:

High SRI portfolio rating means that we invest more on well-rated countries

The idea is to measure the effect on the efficient frontier of imposing a constraint on the SRI portfolio rating

Dec. 2009 Page 9 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

Computation of the “true” efficient frontier (without constraint)

Computation of a set of constrained efficient frontiers “Portfolio rating > threshold” with increasing thresholds

For each efficient frontier, we measure whether the diversification cost is significant or not

0.58%

0.59%

0.60%

0.61%

0.62%

0.63%

0.64%

0.65%

0.66%

0.67%

0.68%

0.69%

0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20%

Monthly volatility %

Mo

nth

ly r

etu

rns %

True eff icient frontier Constrained eff icient frontier

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 10 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

The Basak, Jagannathan and Sun (2002) test allows to measure the mean-variance efficiency of a given benchmark to a given efficient frontier:

BJS (2002) derives the asymptotic distribution of λ:

),0()( 2 NT T

0.58%

0.59%

0.60%

0.61%

0.62%

0.63%

0.64%

0.65%

0.66%

0.67%

0.68%

0.69%

0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20% 1.30%

Monthly volatility %

Mo

nth

ly r

etu

rns %

True eff icient frontier Benchmark

Efficiency measure λ

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 11 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Data and methodology

Ehling & Ramos (2006) procedure

For each threshold on the SRI portfolio rating, we process:BJS test for the minimum variance portfolio

BJS test for the tangency portfolio

0.58%

0.59%

0.60%

0.61%

0.62%

0.63%

0.64%

0.65%

0.66%

0.67%

0.68%

0.69%

0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20%

Std Deviation %/month

Mean

%/m

on

th

True eff icient frontier Constrained eff icient frontier

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 12 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Is there a cost to be SRI in the government bonds market?

We can reject mean-variance efficiency with different levels of confidence

0.58%

0.59%

0.60%

0.61%

0.62%

0.63%

0.64%

0.65%

0.66%

0.67%

0.68%

0.69%

0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20% 1.30%

Monthly volatility %

Mo

nth

ly r

etu

rns

%

True eff icient frontier p-value=10% p-value=5% p-value=1%

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 13 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Is there a cost to be SRI in the government bonds market?

The SRI constraint reduces the weight of badly-rated countries and increases the weight of well-rated countries

True Frontier

10% 5% 1%True

Frontier10% 5% 1%

AUS 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

AUT 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

BEL 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

CAN 0.00% 0.00% 0.00% 0.00% 5.62% 0.00% 0.00% 0.00%

CHE 15.28% 25.14% 25.46% 25.72% 18.21% 25.49% 25.63% 25.88%

DEU 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

DNK 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

ESP 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

FIN 0.00% 5.96% 5.09% 3.82% 0.00% 12.91% 12.23% 11.02%

FRA 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

GBR 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

IRL 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

ITA 1.14% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

JPN 39.36% 28.29% 26.50% 22.98% 41.04% 28.24% 26.33% 22.95%

NLD 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

NOR 12.03% 29.03% 30.75% 32.17% 0.00% 7.84% 8.61% 9.97%

NZL 6.89% 0.68% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

PRT 25.30% 0.00% 0.00% 0.00% 27.66% 0.00% 0.00% 0.00%

SWE 0.00% 10.90% 12.20% 15.31% 7.47% 25.52% 27.20% 30.18%

USA 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

Minimum variance portfolio Tangency portfolio

Null hypothesis rejection probability Null hypothesis rejection probability

Dec. 2009 Page 14 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Is there a cost to be SRI in the government bonds market?

We can clearly increase the portfolio rating without a significant cost of diversification

50

55

60

65

70

75

80

85

90S

WE

NO

R

FIN

CH

E

DN

K

AU

T

GB

R

NLD

DE

U

FR

A

BE

L

IRL

ES

P

AU

S

NZ

L

ITA

PR

T

CA

N

JP

N

US

A

Su

sta

inab

ilit

y C

ou

ntr

y r

ati

ng

Sustainability Country rating Mean p-value=5%

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 15 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Is there a cost to be SRI in the government bonds market?

However, the result depends on the ESG rating chosen:

40

45

50

55

60

65

70

75

80

CH

E

SW

E

NO

R

AU

T

FIN

GB

R

DE

U

DN

K

FR

A

AU

S

NLD

NZ

L

ITA

ES

P

JP

N

BE

L

PR

T

IRL

CA

N

US

A

En

vir

on

men

tal ra

tin

g

Environmental rating Mean p-value=5%

60

65

70

75

80

85

90

95

100

SW

E

DN

K

FIN

NO

R

AU

T

NLD

GB

R

DE

U

PR

T

ES

P

IRL

AU

S

FR

A

CH

E

BE

L

NZ

L

ITA

CA

N

JP

N

US

A

Insti

tuti

on

al ra

tin

g

Institutional rating Mean p-value=5%

60

65

70

75

80

85

90

95

100

NO

R

SW

E

NLD

BE

L

DN

K

FIN

IRL

GB

R

NZ

L

FR

A

CH

E

CA

N

ES

P

AU

T

ITA

DE

U

AU

S

JP

N

PR

T

US

A

So

cia

l an

d S

olid

ari

ty r

ati

ng

Social and Solidarity rating Mean p-value=5%

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 16 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Conclusion

Our contributions:

First paper to use the Vigeo Sustainability Country Ratings

Extension of the SRI research to sovereign bonds asset class

It is possible to build a socially responsible sovereign bond portfolio without a significant diversification cost

Investors should not be worried about SRI in sovereign bonds

Limits: study focused on developed countries

Further developments: emerging countries

Dec. 2009 Page 17 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Further developments

Constraint on the Environmental Performance Index (EPI)Emerging countries

5%

10%

15%

20%

25%

30%

5% 10% 15% 20% 25%

Volatility

True Efficient Frontier EPI 65 EPI 75 EPI 85 EPI 95

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 18 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Further developments

Constraint on the Environmental Performance Index (EPI)Developed countries

5%

7%

9%

11%

13%

15%

17%

19%

2% 4% 6% 8% 10% 12% 14%

Volatility

True Efficient Frontier EPI 65 EPI 75 EPI 85 EPI 95

Sources: Datastream, VIGEO, CAAM Strategy

Dec. 2009 Page 19 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Further developments

ULB Working Papers available at:

“Sovereign bonds and socially responsible investment”

http://www.solvay.edu/EN/Research/Bernheim/documents/wp09014.pdf

“Nice guys with cold feet”

http://www.solvay.edu/EN/Research/Bernheim/documents/wp09034.pdf

Dec. 2009 Page 20 STRATEGY CAAM: SOVEREIGN BONDS & SRI - AFG 17/12/2009

Crédit Agricole Asset ManagementSociété Anonyme au capital de 546 162 915 eurosSociété de Gestion de Portefeuille agréée par l’AMF n° GP 04000036Siège social: 90, boulevard Pasteur 75015 Paris - France - 437 574 452 RCS Paris