International Capital Markets

and Investment Practice

Part 4
Rational asset pricing and multifactor models

Dr. Peter Oertmann

[email protected]

This lecture is largely congruent in content with the video

recording of the 2020 lecture available on lecturio.de
Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 2

Part 4: Positioning in the lecture

Major fields in asset management Knowledge base

Diversification Asset allocation Stylized facts on Sources of risk

global markets and return
Empirical research Models and empirical
research

Security Investment Asset pricing Portfolio

selection strategy Theories, models and construction
empirical tests
implementation Models and empirical
tests

Risk Trading and Financial Time series

management execution economics analysis
Models and empirical Models and empirical
research procedures

© 2018-2022 Peter Oertmann 3

Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 4

What moves the capital markets?

© 2018-2022 Peter Oertmann 5

The idea of rational asset pricing

Experience
§ Asset prices react to economic news
§ A variety of unanticipated events influence asset prices, and some events have a
more pervasive effect on asset prices that do others
§ The comovements of asset prices suggest the existence of common exogenous
driving forces
Theory
§ Asset prices depend on their respective exposures to (latent) state variables that
describe the economy and are likely sources of systematic investment risk
§ Foundation: Merton (1973); Cox, Ingersoll and Ross (1985); Ross (1976)
Empirical approach
§ Economic variables are used as proxies for the latent variables
§ Seminal work: Chen, Roll and Ross (1986)

© 2018-2022 Peter Oertmann 6

Asset prices are exposed to variables describing the economy

Latent state
variables

Economic proxy
variables

© 2018-2022 Peter Oertmann 7

Chen, Roll and Ross (1986) – Starting point

© 2018-2022 Peter Oertmann 8

Chen, Roll and Ross (1986) – Macroeconomic state variables

§ Industrial production
– Monthly growth rate industrial production (MP)
– Yearly growth rate in industrial production (YP)
§ Inflation
– Unanticipated inflation (UI)
– Change in expected inflation (DEI)
§ Risk premium: Unanticipated change in the risk premium (Baa and under return
minus long-term government bond return) (UPR)
§ Term structure: Unanticipated change in the term structure (long-term government
bond return minus Treasury-bill rate) (UTS)
§ Stock market
– Return on the value-weighted NYSE index (VWNY)
– Return on the equally weighted NYSE index (EWNY)
§ Innovation in real per capita consumption
§ Oil price

© 2018-2022 Peter Oertmann 9

Chen, Roll and Ross (1986) – Results

Priced sources of risk Not priced

Industrial production Stock market returns

Changes in the risk Strong Innovation in real per

premium evidence capita consumption

Changes in the term Oil prices changes

spread of interest rates

Unanticipated inflation
Weak
evidence
Changes in expected
inflation

© 2018-2022 Peter Oertmann 10

Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets

© 2018-2022 Peter Oertmann 11

Introduction

Derivation of the „domestic“ APT *

Core concept

* Following Oertmann (1996): Strands of the Arbitrage Pricing Theory

© 2018-2022 Peter Oertmann 12

Introduction

Starting point
§ Variation of asset returns on international capital markets is driven by common
global systematic risk factors
§ Long-term expected returns of internationally traded assets include premiums for
global factor risk
§ Differences in the global factor risk profiles across assets account for differences
in expected returns across assets

Theoretical frameworks to derive multifactor pricing restrictions

§ IAPM of Adler and Dumas (1983)
§ International APT (IAPT) of Solnik (1983) and Ikeda (1991)

© 2018-2022 Peter Oertmann 13

International Arbitrage Pricing Theory

What might disturb APT in the international environment?

Fluctuations of FX rates FX hedging might affect

add additional risk to an arbitrage activities on
internationally diversified international capital
portfolio markets

Idea of arbitrage pricing

In a large capital market, it is possible to construct arbitrage
portfolios that do not have any systematic or unsystematic risk
in the sense of a given factor structure
i.e. the cross-section of available assets is large enough to
permit diversification of idiosyncratic risk

© 2018-2022 Peter Oertmann 14

International Arbitrage Pricing Theory (cont.)

Starting point:
k-factor model generating international asset returns

Rii = E(Rii ) + bii1 × d1 + bii2 × d2 + ... + biik × dk + eii

where Rii return on an asset in country i in terms of local currency

dj common global factors driving asset returns

biij factor betas measured on the basis of local currency returns

eii idiosyncratic return

© 2018-2022 Peter Oertmann 15

International Arbitrage Pricing Theory (cont.)

Crucial conditions for arbitrage pricing to hold in an international setting

Fluctuations of FX rates FX hedging might affect

add additional risk to an arbitrage activities on
internationally diversified international capital
portfolio markets

Prerequisites
1. Technical: The risk stemming from FX rate shifts must be
diversifiable like any other unsystematic risk
2. Technical: An arbitrage portfolio that is riskless in any given
currency must be riskless in any other currency
3. Economic: The factor structure assumed to drive asset returns
must be invariant to the choice of a currency

© 2018-2022 Peter Oertmann 16

International Arbitrage Pricing Theory (cont.)

How do FX rate changes enter the model?

§ IAPT of Ikeda (1991)

– FX rates changes do not include a systematic component

sid = E(sid ) + uid

§ IAPT of Solnik (1983)

– FX rate changes are driven by a k-factor model

sid = E(sid ) + bid1 × d1 + bid2 × d2 + ... + bik

d
× dk + eid

© 2018-2022 Peter Oertmann 17

IAPT of Solnik 1983

k-factor model generating international asset returns

Rid = E(Rid ) + bid1 × d1 + bid2 × d2 + ... + bik

d
× dk + eid

Characteristics
§ Assumption that all FX rates follow the same k-factor model as do international
asset returns
§ Factor betas embody both the factor risk of local asset returns and the factor risk
of the respective FX rate change
§ Factor structure is invariant to the currency that is chosen to denominate asset
returns

© 2018-2022 Peter Oertmann 18

IAPT of Solnik 1983 (cont.)

International multifactor asset pricing restriction

E(Rid ) = Rdf + bid1 × ld1 + bid2 × ld2 + ... + bik

d
× ldk

where E(Rid ) expected return on the ith asset

Rdf domestic interest rate for riskfree investment
bijd exposures to global risk factors (factor betas)
ldj global factor risk premiums

Cov(Rid, d j )
and bijd =
Var(d j )

© 2018-2022 Peter Oertmann 19

IAPT of Solnik 1983 (cont.)

Components of expected returns

E(Ri ) = Rf + bi1 × l1 + bi2 × l2 + ... + bik × lk

! %$# %$# %$#
time factor 1 factor 2 factor k
premium risk premium risk premium risk premium
%"""""""
"$""""""""
#
premium for systematic risk of asset i

§ Time premium - Reward investors expect for investing capital over time
§ Factor risk premiums - Rewards investors expect for taking exposure to global
systematic factor risks

© 2018-2022 Peter Oertmann 20

Structure of the international pricing restriction

æ E(R1t ) ö æ b11 ö æ b12 ö æ b1k ö

ç ÷ ç ÷ ç ÷ ç ÷
ç! ÷ ç! ÷ ç! ÷ ç! ÷
ç E(R ) ÷ = R + ç b ÷ × l + ç b ÷ × l + ... + ç b ÷ × l
ç it ÷ f
ç i1 ÷ 1 ç i2 ÷ 2 ç ik ÷ k
ç! ÷ ç! ÷ ç! ÷ ç! ÷
ç E(R )÷ çb ÷ çb ÷ çb ÷
è nt ø è n1 ø è n2 ø è nk ø

Mechanics
§ Global risk premiums (factor prices) l are the same across international
investments
§ Cross-sectional differences in expected returns (left-hand side) are explained by
cross-sectional differences in factor betas (right-hand side)

© 2018-2022 Peter Oertmann 21

Application 1: Exploring asset return volatility

Volatility drivers
§ capture the variance of asset returns (on an asset-by-asset basis)
§ can be identified by an analysis of the significance of the beta coefficients in the
multifactor model

Rid = E(Rid ) + bid1 × d1 + bid2 × d2 + ... + bik

d
× dk + eid

Cases
§ beta significant: volatility driver, factor with a measurable impact on the return
variation of an asset class
§ beta not significant: factor with no systematic influence

© 2018-2022 Peter Oertmann 22

Application 2: Exploring long-term asset returns

Value drivers
§ explain the cross-sectional differences (variation) of asset returns
§ can be identified by an analysis of the significance of the lambda coefficients in
the asset pricing constraint

æ E(R1t ) ö æ b11 ö æ b12 ö æ b1k ö

ç ÷ ç ÷ ç ÷ ç ÷
ç! ÷ ç! ÷ ç! ÷ ç! ÷
ç E(R ) ÷ = R + ç b ÷ × l + ç b ÷ × l + ... + ç b ÷ × l
ç it ÷ f
ç i1 ÷ 1 ç i2 ÷ 2 ç ik ÷ k
ç! ÷ ç! ÷ ç! ÷ ç! ÷
ç E(R )÷ çb ÷ çb ÷ çb ÷
è nt ø è n1 ø è n2 ø è nk ø
Cases
§ lambda significant: value driver, factor with a measurable impact on the cross-
sectional variance of returns (priced factor)
§ lambda not significant: unsystematic factor (...but it may be a volatility driver)

© 2018-2022 Peter Oertmann 23

Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 24

Identifying common factors for stocks and bonds:
Setup of the study *

§ An unconditional beta pricing model is used to examine the structure of returns

and expected returns across international stock and bond markets.
§ The valuation framework is consistent with the APT developed by Ross (1976),
Huberman (1982), Chamberlain (1983), Chamberlain and Rothschild (1983),
Ingersoll (1984), and others. A theoretical foundation for using the APT in an
international context is given by Solnik (1983).
§ To set up the model it is assumed that
1. multiple risk factors have an impact on variances as well as on long-term
averages of returns of international assets
2. only global factors are sources of systematic risk
3. a set of observable economic factors is a valid representation of the factor
structure driving returns

* Oertmann, Peter (1997); documented in Global Asset Allocation, Chapter 5.

© 2018-2022 Peter Oertmann 25

Identifying common factors for stocks and bonds:
Predetermined global factors

§ Seven global factors (“factor candidates”)

§ Aggregate information on potential global sources of systematic risk
§ Swiss franc perspective
§ The set of predetermined global risk factors includes
– Change in G-7 inflation rates (ING7C)
– Change in G-7 industrial production indices (IPG7C)
– Change in G-7 long-term interest rates (ILG7C)
– Change in G-7 short-term interest rates (ISG7C)
– Change in the price of the G-7 currencies measured in Swiss francs
(CHG7C)
– Dow Jones commodity price index change (DJCIC)
– Excess return on the world stock market (WDSTR)

© 2018-2022 Peter Oertmann 26

Identifying common factors for stocks and bonds:
Empirical design

Step 1 The association between monthly changes of global economic

variables and contemporaneous excess returns on stock and bond
markets is investigated in the setting of factor model regressions.
Step 2 Analysis attends to the question whether some of the global factors
that have been identified to be related to the variance of
international asset returns do also affect the expected returns on
these assets.
Investigation concentrates on those factors that show up with an
influence on both stock and bond markets in the preliminary
regression analysis.
Both an empirical version of the international CAPM and a
multifactor pricing model are estimated.

© 2018-2022 Peter Oertmann 27

Step 1 – Return generating process and factor model

The following linear k-factor model is assumed to explain the relation between returns,
expected returns, and the factors inherently affecting the returns on international stock
and bond markets:

(1a) rit = E[ rit ] + b i1 × d1t + b i 2 × d 2 t + ... + b ik × d kt + e it

Empirical version:

(1b) rit = a i + b i1 × d1t + b i 2 × d 2 t + ... + b i 7 × d 7t + e it

i = 1, ..., 17 (stock markets), i = 1, ..., 8 (bond markets)
t = 1, 2, ..., T

© 2018-2022 Peter Oertmann 28

Step 1 – Factor model regressions

© 2018-2022 Peter Oertmann 29

Step 1 – Factor model regressions (cont.)

© 2018-2022 Peter Oertmann 30

Step 1 – Factor model regressions (cont.)

© 2018-2022 Peter Oertmann 31

Step 1 – Performing Wald tests

Hypothesis 1 “The factor betas are equal to zero for all the markets”

H 0 : b ij = 0
i = 1, ..., 17 (stock markets); i = 1, ..., 8 (bond markets)

Hypothesis 2 “The factor betas are jointly equal across the markets”

H 0 : b ij = b j
i = 1, ..., 17 (stock markets); i = 1, ..., 8 (bond markets)

© 2018-2022 Peter Oertmann 32

Step 1 – Wald tests impose cross-sectional restrictions on betas

bij = 0 bij = 0 ………….. bij = 0

© 2018-2022 Peter Oertmann 33

Step 1 – Wald test results

© 2018-2022 Peter Oertmann 34

Step 1 – Major conclusions

§ International stock market returns seem to be influenced by a broader variety of

global risk factors than international bond market returns.
§ For stock markets there may be four to five persistent driving forces, whereas the
returns in bond markets are predominantly affected by interest rate factors and
shifts in global exchange rates.
§ Three of the seven predetermined global risk factors systematically affect the
returns of both stock and bond investments. These include the excess return on the
world market portfolio, the change in global long-term interest rates, and the
change in the price of the G-7 exchange rate basket. Hence, these variables
represent inherent forces, or at least some of the forces, causing the frequently
observable comovement between international stock and bond markets.

© 2018-2022 Peter Oertmann 35

Step 1 – Major conclusions (cont.)

© 2018-2022 Peter Oertmann 36

Step 2 – Unconditional beta pricing tests

Assuming that the factor model (1) sufficiently characterizes the process generating
international stock and bond market returns, it is postulated that the cross-section of
long-term expected returns of the markets is described by the subsequent beta
pricing relationship:

(2)
E[ ri ] = l 1 × b i1 + l 2 × b i 2 + ... + l k × b ik

Combining equations (1) and (2) yields the following empirically testable non-linear
relationship between market returns, factors changes, and factor rewards:

k k
(3)
rit = å l j × b ij + å b ij × d jt + e it
j=1 j=1

© 2018-2022 Peter Oertmann 37

Step 2 – Unconditional beta pricing tests (cont.)

Model (3) serves as the starting point for setting up the unconditional beta pricing
test. This model is implemented via the following system of restricted seemingly
unrelated regression equations:

æ r1t ö æ b11 ö æ b1k ö æ b11#b1k ö æ d1t ö æ e1t ö

ç ÷ ç ÷ ç ÷ ç ÷ ç ÷ ç ÷
ç ! ÷ = l 1 × ç ! ÷ ×" × l k × ç! ÷ + ç! ÷ × ç! ÷ + ç! ÷
çr ÷ ç ÷ ç b ÷ ç b #b ÷ ç d ÷ ç e ÷
è nt ø è b n1 ø è nk ø è n1 nk ø è kt ø è nt ø

© 2018-2022 Peter Oertmann 38

Step 2 – Risk premium estimates

© 2018-2022 Peter Oertmann 39

Step 2 – Major conclusions

§ The single-factor CAPM identifies a significantly positive premium for the exposure
to world market risk in the long-term returns of both stock and bond markets.
However, for some markets the model leaves relatively large pricing errors.
§ The 3-factor model seems to provide a more distinctive picture of the risk premiums
in both stock and bond market returns. The estimation results indicate that average
returns of both asset classes include rewards for interest rate risk in addition to the
market premium. In addition, there is weak evidence that also exchange rate risk is
priced across international asset.
§ The fit of the 3-factor model to the cross-section of international asset returns is
clearly superior to the fit of the single-factor CAPM.

© 2018-2022 Peter Oertmann 40

Conclusions of the study

§ Multiple sources of global economic risk affect both the variability and the
expected values of returns on international stock and bond markets.
§ Hence, to control the variance and measure the performance of an internationally
diversified portfolio including both stock and bond positions a framework including
multiple global risk factors must be preferred to a single-factor model specified in
the spirit of the international CAPM.
§ When return benchmarks are required for such a balanced portfolio it appears
essential including a premium for the risk of global interest changes, in particular.

© 2018-2022 Peter Oertmann 41

Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 42

Setup of the study *

§ Specification of a multifactor model to explore risks in emerging markets,

following the approach of Harvey (1995)
§ Sample
– 25 emerging markets (MSCI EM plus Argentina, Jordan, Pakistan)
– 23 developed markets (MSCI world)
– Period from December 30, 1992 to March 31, 2010
– Analyses in EUR and local currencies
§ Factors
– World stock market return (WDRET)
– Change in long-term interest rates (INTR)
– Return on S&P Goldman Sachs Commodity Index Industrial Metals (CMDTY)
– Change in the external value of the euro (CRNCY)
– Emerging market return, orthogonal to WDRET (EMF)

* Kirner, Franziska (2010), Sources of Risk in Emerging Markets, diploma thesis at TUM Business School

© 2018-2022 Peter Oertmann 43

Return statistics – Emerging markets
Country Starting Mean Volatility Monthly Returns Auto Skewness (Excess)
Date return p.a. p.a. min max median correlation Kurtosis
Argentina 1992.12 6.08% 39.91% -43.69% 43.04% 1.65% 0.05 -0.60 2.26
Brazil 1992.12 18.44% 41.45% -48.08% 35.08% 2.75% 0.07 -0.70 1.94
Chile 1992.12 10.14% 25.66% -35.20% 18.41% 0.50% 0.11 -0.60 2.32
China 1992.12 -1.99% 39.29% -54.38% 39.58% 1.01% 0.05 -0.48 2.84
Colombia 1992.12 16.31% 33.62% -27.76% 29.05% 1.90% 0.20 -0.20 0.55
Czech Republic 2000.12 20.91% 24.73% -29.19% 18.61% 2.14% 0.20 -0.74 2.42
Egypt 2000.12 20.38% 32.91% -30.47% 32.79% 1.33% 0.29 -0.14 1.05
Hungary 2000.12 11.85% 33.27% -46.42% 27.85% 2.74% 0.29 -1.08 4.69
India 1992.12 9.77% 31.78% -27.09% 24.65% 0.92% 0.13 -0.22 -0.06
Indonesia 1992.12 6.39% 47.80% -54.11% 45.74% 1.18% 0.18 -0.54 2.45
Israel 1992.12 7.26% 26.65% -23.52% 25.85% 2.04% 0.12 -0.39 0.80
Jordan 1992.12 4.72% 20.36% -26.42% 19.81% 0.11% 0.23 -0.11 2.38
Korea 1992.12 7.22% 39.70% -39.76% 54.71% 0.43% 0.08 0.13 2.84
Malaysia 1992.12 5.01% 33.69% -37.97% 39.85% 1.06% 0.09 -0.16 3.41
Mexico 1992.12 9.41% 34.09% -41.08% 19.70% 1.86% 0.15 -1.21 3.16
Morocco 2000.12 10.41% 19.75% -18.88% 18.81% 0.82% -0.02 -0.15 1.45
Pakistan 1992.12 4.62% 43.43% -78.33% 33.05% 0.34% 0.05 -1.37 8.20
Peru 1992.12 17.17% 34.04% -41.23% 31.85% 1.83% -0.03 -0.63 2.42
Philippines 1992.12 0.63% 33.56% -36.57% 35.13% 0.51% 0.15 0.04 1.79
Poland 1992.12 14.32% 45.98% -43.66% 77.38% 2.05% 0.11 0.56 5.55
Russia 2000.12 15.05% 37.30% -38.00% 22.16% 2.98% 0.30 -0.68 0.76
South Africa 1992.12 8.78% 27.31% -37.26% 16.86% 1.28% 0.06 -1.00 2.49
Taiwan 1992.12 4.42% 33.00% -27.09% 39.63% -0.12% 0.05 0.43 1.44
Thailand 1992.12 0.86% 41.74% -44.15% 37.96% 1.04% -0.04 -0.37 2.02
Turkey 1992.12 13.55% 57.19% -52.07% 54.85% 3.01% 0.06 -0.33 0.22
Sample Average 9.67% 35.13%

MSCI EM 1992.12 8.43% 25.68% -34.98% 16.48% 1.68% 0.19 -0.93 2.29
MSCI World 1992.12 6.53% 16.45% -15.14% 10.92% 0.84% 0.14 -0.60 0.19

© 2018-2022 Peter Oertmann 44

Return statistics – Developed markets
Country Starting Mean Volatility Monthly Returns Auto Skewness (Excess)
Date return p.a. p.a. min max median correlation Kurtosis
Australia 1992.12 11.17% 20.50% -19.22% 14.29% 1.41% 0.03 -0.50 0.41
Austria 1992.12 4.53% 23.18% -36.04% 17.59% 1.04% 0.25 -1.59 6.61
Belgium 1992.12 6.30% 20.80% -35.27% 14.47% 1.35% 0.27 -2.08 8.11
Canada 1992.12 11.05% 22.07% -25.40% 14.00% 1.22% 0.12 -0.71 1.13
Denmark 1992.12 11.93% 19.33% -19.44% 17.04% 1.44% 0.05 -0.63 1.37
Finland 1992.12 15.73% 34.49% -37.09% 28.43% 1.58% 0.19 -0.39 1.40
France 1992.12 7.64% 19.34% -17.41% 12.71% 1.48% 0.11 -0.54 0.48
Germany 1992.12 7.95% 22.42% -28.67% 19.02% 1.50% 0.05 -0.84 2.40
Greece 1992.12 8.29% 29.95% -35.50% 36.54% 0.89% 0.15 -0.20 2.65
Hong Kong 1992.12 8.31% 28.18% -36.64% 27.78% 1.10% 0.08 -0.15 2.47
Ireland 1992.12 2.48% 22.29% -25.47% 14.76% 1.33% 0.23 -1.04 1.74
Italy 1992.12 7.36% 23.22% -16.69% 19.85% 0.92% 0.00 0.01 0.28
Japan 1992.12 0.78% 20.64% -15.94% 15.43% 0.03% 0.18 0.05 -0.07
Netherlands 1992.12 8.75% 19.99% -20.37% 13.16% 1.55% 0.09 -1.05 1.88
New Zealand 1992.12 7.21% 22.50% -23.21% 16.81% 1.40% -0.06 -0.58 1.06
Norway 1992.12 10.71% 26.24% -33.34% 14.67% 1.80% 0.14 -1.34 4.06
Portugal 1992.12 7.99% 20.92% -22.28% 15.78% 0.95% 0.12 -0.51 1.58
Singapore 1992.12 6.59% 27.30% -24.01% 23.35% 1.12% 0.08 -0.42 1.64
Spain 1992.12 12.03% 22.19% -25.20% 16.07% 1.41% 0.05 -0.59 1.46
Sweden 1992.12 12.53% 26.76% -21.18% 22.90% 1.32% 0.08 -0.39 0.80
Switzerland 1992.12 10.33% 15.92% -17.85% 12.68% 1.24% 0.19 -0.73 1.63
UK 1992.12 6.61% 15.82% -12.04% 12.07% 1.10% 0.20 -0.55 0.09
US 1992.12 7.11% 17.72% -15.82% 12.89% 0.67% 0.10 -0.44 0.25
Sample Average 8.41% 22.69%

MSCI EM 1992.12 8.43% 25.68% -34.98% 16.48% 1.68% 0.19 -0.93 2.29
MSCI World 1992.12 6.53% 16.45% -15.14% 10.92% 0.84% 0.15 -0.60 0.19

© 2018-2022 Peter Oertmann 45

Five-factor model results – Emerging markets

Start Date Const. WDRET INTR CMDTY CRNCY EMF R² adj. R² DW

Argentina 1992.12 -0,0042 1,2295 *** -5,5831 * 0,2012 * -0,3759 0,9571 *** 0,4295 0,4153 2,16
Brazil 1992.12 0,0078 1,6665 *** 2,4265 0,0031 -0,3190 1,1631 *** 0,6175 0,6079 2,26
Chile 1992.12 0,0020 0,8374 *** 0,6689 0,0921 0,3533 ** 0,7623 *** 0,6003 0,5903 2,03
China 1992.12 -0,0092 1,0326 *** -3,8404 0,0001 -0,0128 1,3267 *** 0,4406 0,4267 2,02
Colombia 1992.12 0,0079 0,7049 *** 2,9625 0,1192 0,1677 0,6512 *** 0,2683 0,2501 1,89
Czech Republic 2000.12 0,0096 1,0910 *** -7,3720 *** 0,0095 -0,5034 *** 0,6691 *** 0,5950 0,5757 1,99
Egypt 2000.12 0,0083 1,1127 *** -9,7124 ** 0,0595 -0,0064 0,7878 *** 0,3557 0,3250 1,78
Hungary 2000.12 0,0033 1,3493 *** -3,0746 0,0788 -1,0956 *** 0,5255 *** 0,5515 0,5301 1,88
India 1992.12 0,0025 0,9167 *** 2,0946 0,0005 0,0558 0,8834 *** 0,4206 0,4062 2,20
Indonesia 1992.12 -0,0034 1,3723 *** -4,6440 0,0471 -0,2452 1,3230 *** 0,3956 0,3806 1,88
Israel 1992.12 0,0009 0,9824 *** 1,6568 -0,0130 -0,0310 0,2006 ** 0,3825 0,3671 1,93
Jordan 1992.12 -0,0009 0,2580 *** -0,3170 0,1074 0,5090 *** 0,1189 0,1699 0,1493 1,62
Korea 1992.12 -0,0006 1,1970 *** 5,6802 * 0,1326 0,1259 0,7811 *** 0,4240 0,4097 2,14
Malaysia 1992.12 -0,0016 0,8574 *** -1,6169 -0,1254 0,5418 ** 1,0807 *** 0,4075 0,3928 2,10
Mexico 1992.12 -0,0001 1,4639 *** -3,1905 -0,0448 0,2649 0,9363 *** 0,6667 0,6584 1,93
Morocco 2000.12 0,0031 0,0722 0,1820 0,1970 ** 0,0376 0,1680 0,0862 0,0426 2,23
Pakistan 1992.12 -0,0042 0,2058 -1,9902 0,3693 ** 1,3019 *** 0,8603 *** 0,2251 0,2058 2,21
Peru 1992.12 0,0062 0,9635 *** -4,1699 0,1786 * -0,4377 * 0,9773 *** 0,4546 0,4411 2,13
Philippines 1992.12 -0,0065 0,9506 *** -2,2626 0,0312 0,4696 ** 0,8705 *** 0,4168 0,4023 2,02
Poland 1992.12 0,0051 1,7144 *** -4,1520 -0,2494 * -0,4877 0,9445 *** 0,3848 0,3695 1,96
Russia 2000.12 -0,0011 1,3645 *** -2,2289 0,0093 -0,5256 * 1,3156 *** 0,5599 0,5389 1,80
South Africa 1992.12 0,0004 1,0348 *** -0,5535 0,0836 -0,1714 0,8041 *** 0,6206 0,6112 1,99
Taiwan 1992.12 -0,0034 1,0537 *** 0,8610 0,0842 0,3768 * 0,9338 *** 0,5403 0,5289 2,14
Thailand 1992.12 -0,0093 1,2630 *** -8,2838 ** 0,1392 0,1038 1,2394 *** 0,4861 0,4733 2,13
Turkey 1992.12 0,0026 1,9131 *** 0,0846 0,0075 -0,0996 1,0086 *** 0,3733 0,3577 2,12

© 2018-2022 Peter Oertmann 46

Risk exposure of emerging and developed markets

Percentage of emerging and developed markets with significant exposure to the five
risk factors examined
100% 100%

90% 90%

80% 80%

70% 70%

60% WDRET 60% WDRET

INTR INTR
50% 50%
CMDTY CMDTY
40% 40%
CRNCY CRNCY
30% EMF 30% EMF
20% 20%

10% 10%

0% 0%
Emerging Markets Developed Markets

© 2018-2022 Peter Oertmann 47

Five-factor model: summary regression output

Summary regression output for emerging and developed markets

Multi-factor
regression WDRET INTR CMDTY CRNCY EMF adj. R²
EM 1.0643 -1.8550 0.0607 -0.0001 0.8516 41.83%
(t-stat) 7.6068 -0.5415 0.5814 0.0724 6.306 -
min 0.0722 -9.7124 -0.2494 -1.0956 0.1189 4.26%
max 1.9131 5.6802 0.3693 1.3019 1.3267 65.84%
median 1.0537 -1.9902 0.0595 -0.0064 0.8834 40.97%
*** 92% 4% 0% 16% 88%
** 92% 12% 8% 28% 92%
* 92% 20% 20% 40% 92%

DM 1.0840 0.0414 0.0019 -0.3770 0.1499 61.97%

(t-stat) 16.7824 -0.0549 0.1504 -2.9223 1.9913 -
min 0.7828 -5.7539 -0.1411 -0.8337 -0.1079 36.38%
max 1.5973 4.8840 0.1175 0.3933 0.7811 91.53%
median 1.1042 0.1819 -0.0013 -0.4830 0.0423 62.02%
*** 100% 9% 4% 61% 35%
** 100% 17% 17% 83% 39%
* 100% 30% 26% 87% 48%

© 2018-2022 Peter Oertmann 48

Part 4: Rational asset pricing and multifactor models

The idea of rational asset pricing

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 49

Case – the setup

Data (“Case2.xls”)
§ Monthly total return index data over the 1st decade of the 21st century, from
January 2000 to December 2009, denominated in EUR, is provided for
– 10 European stock markets
– 4 global risk factors

Tasks
§ Analyze the relationships between the returns of the stock markets and the
changes of the global factors. Perform the following regressions for each market:
– Market return on the MSCI World index return (single factor model)
– Market return on the 4 global factors (4-factor model)
§ Report regressions coefficients, t statistics, and R squares in a table and in a
chart.
§ Comment your results!

© 2018-2022 Peter Oertmann 50

Case – the results
Single Factor Model 4 Factor Model
MSCI World Constant R Squared FX USD/EUR EUR 10Y Rate CRB Index MSCI World Constant R Squared
0,692 0,003 0,631 0,206 0,062 -0,114 0,743 0,003 0,673
Switzerland
14,210 1,403 2,696 1,199 -2,366 13,992 1,562
1,260 0,002 0,726 0,609 0,140 -0,165 1,365 0,002 0,808
Germany
17,681 0,631 6,126 2,082 -2,637 19,775 0,636
0,896 0,001 0,826 0,100 0,098 0,033 0,861 0,001 0,839
UK
23,652 0,370 1,638 2,377 0,862 20,402 0,288
1,397 0,003 0,652 0,575 0,035 -0,040 1,492 0,002 0,695
Sweden
14,865 0,659 3,918 0,351 -0,428 14,638 0,414
1,256 0,010 0,555 0,929 0,339 0,423 1,118 0,006 0,749
Norway
12,123 2,069 7,167 3,860 5,173 12,417 1,720
1,047 0,002 0,786 0,489 0,058 -0,053 1,124 0,001 0,854
France
20,834 0,892 7,068 1,231 -1,209 23,402 0,653
0,948 0,001 0,630 0,539 -0,040 0,087 1,019 -0,001 0,710
Italy
14,187 0,446 5,458 -0,604 1,405 14,869 -0,188
1,016 0,007 0,646 0,547 -0,091 -0,114 1,184 0,006 0,741
Spain
14,663 2,034 5,533 -1,369 -1,837 17,257 1,950
1,155 0,002 0,763 0,456 0,054 -0,033 1,220 0,001 0,809
Netherlands
19,504 0,664 5,152 0,909 -0,584 19,836 0,396
0,945 0,006 0,372 1,152 0,155 0,324 0,953 0,002 0,607
Austria
8,360 1,153 7,724 1,533 3,441 9,198 0,451
Source: Own calculations

© 2018-2022 Peter Oertmann 51

Part 4: Rational asset pricing and multifactor models

Motivation of multifactor models

International Arbitrage Pricing Theory (IAPT)
Specification of a global multifactor asset pricing model
Empirical study: Sources of Risk in Emerging Markets
Case: Economic risks of European stock markets
Exercise: Application of a factor model

© 2018-2022 Peter Oertmann 52

Exercise: Application of a factor model

Risk factors Factor 1 Factor 2 Factor 3

Volatility 16,0% 10,0% 12,0%
Risk premium 4,0% -0,2% 0,6%

Stock markets Volatility Factor exposures

USA 20% 0,90 -0,80 0,60
Germany 25% 1,20 -0,60 0,80
Hong Kong 30% 1,40 0,40 0,70

Risk-free rate 1,5%

© 2018-2022 Peter Oertmann 53

Exercise: Application of a factor model

1. How do German stocks react to a 10% increase in factor 2?

2. Is it possible to hedge “factor 2 risk” by a portfolio of stock markets?

© 2018-2022 Peter Oertmann 54

Exercise: Application of a factor model

3. What portion of the German stock market’s variance is captured by factor 1?

4. What portion of the US stock market’s variance is captured by factor 2?

© 2018-2022 Peter Oertmann 55

Exercise: Application of a factor model

5. What portion of the Hong Kong stock market’s variance is explained by the
model?

© 2018-2022 Peter Oertmann 56

Exercise: Application of a factor model

6. What is the expected return for the US stock market?

© 2018-2022 Peter Oertmann 57

Part 4: Selected references

Global Asset Allocation, Chapter 5

Chen, N. F., R. Roll, and S. A. Ross (1986), Economic forces and the stock market, Journal of Business
Fama, E. F., and K. R. French (1993), Common factors in the returns on stocks and bonds, Journal of
Financial Economics
Ferson, W. E., and C. R. Harvey (1994), Sources of risk and expected returns in global equity markets,
Journal of Banking and Finance
Harvey, C.R. (1995), The Risk Exposure of Emerging Equity Markets, The World Bank Economic Review
Ikeda, S. (1991), Arbitrage asset pricing under exchange risk, The Journal of Finance
Merton, R. (1973), An intertemporal capital asset pricing model, Econometrica
Oertmann, P., C. Rendu and H. Zimmermann (2000), Interest rate risk of European financial corporations,
European Financial Management
Oertmann, P. (1997), Global Risk Premia on International Investments, Gabler.
Oertmann, P. (1996), Strands of the Arbitrage Pricing Theory, Working paper, University of St. Gallen
Ross, S. A. (1976), The arbitrage theory of capital asset pricing, Journal of Economic Theory
Solnik, B. (1983), International arbitrage pricing theory, The Journal of Finance

© 2018-2022 Peter Oertmann 58