International direct real estate asset allocation a fuzzy decision making approach

Summary Although the classical Markowitz mean-variance asset allocation framework can be used to enable decision-making in international and direct real estate investing, and that many i

Acknowledgments

I am greatly indebted to my supervisor, Associate Professor Ho Kim Hin, David He has been a tremendous source of support, ideas, and useful advice during the three years we worked together His energy and keen interest in challenging problems have been an example and a constant inspiration for me

I would like to acknowledge gratefully the financial support of the National University of Singapore (NUS) for granting me the prestigious NUS Research Scholarship I am also thankful to Professor Ong Seow Eng, Associate Professor Fu Yuming, Associate Professor Sing Tien Foo and Associate Professor Tu Yong, for their concern, willingness to listen, and their constant efforts to make some suggestions

My friends at NUS have helped me in numerous ways To Li Yun, Chen Zhiwei – I thank you

The love of my family has been my greatest blessing, and it has helped me to go on in

my most difficult moments I fondly thank my parents, who have always been an example of ambition and hard work but also of devotion, caring, and humanity I am grateful to my sister for her efforts to support me and take care of my parents Your encouragement gave me the strength to pursue my professional goals, and your very existence reminded me that there is so much more to life I thank you with all my heart

Contents Acknowledgments I Contents II Summary V List of Tables VI List of Figures VII

Chapter 1 Introduction 1

1.1 Background and Motivation 1

1.1.1 Background 1

1.1.2 Motivation 2

1.2 Research Question and Issues 6

1.3 Objectives 7

1.4 Data and Methodology 8

1.5 Scope of the Research 9

1.6 Potential Results and Contributions 9

1.7 Structure of This Dissertation 10

Chapter 2 Literature Review 12

2.1 Real Estate 12

2.1.1 Definition of Real estate 12

2.1.2 Characteristics of real estate 13

2.1.3 The real estate investing process 13

2.1.4 The benefits of real estate investing 15

2.1.5 Why choose international and direct real estate investing, and not the real estate investment trust (REITs)? 16

2.2 Direct Real Estate Asset Allocation 18

2.2.1 Asset allocation 18

2.2.2 Direct real estate asset allocation 20

2.2.3 Asset Allocation Models 37

2.2.4 Shortcomings of traditional asset allocation approaches 43

2.3 Fuzzy Set Theory and Fuzzy Decision Making 45

2.3.1 Fuzzy set theory 45

2.3.2 The fuzzy logic benefit 46

2.3.3 Why must computer software programs enable decision to be made? 46

2.3.4 Why fuzzy logic? 47

2.3.5 Fuzzy logic’s underlying principle 47

2.3.6 Why is there a need for an “Inference Engine”? 48

2.3.7 The Extension Principle 49

2.3.8 Probability theory and fuzzy set theory 50

2.3.9 Popularity of Fuzzy Set Theory Application Research 50

2.3.10 Fuzzy decision making 50

Chapter 3 The Fuzzy Strategic Asset Allocation (FSAA) Model 52

3.1 Introduction 52

3.2 FSAA 53

Chapter 4 Fuzzy Tactical Asset Allocation (FTAA) Decision Making Models 62

4.1 Expert judgement and fuzziness 63

4.2 Optimization in the fuzzy environment 65

4.3 Definition of the fuzzy decision (Bellman and Zadeh) 65

4.4 Flexible Model: Zimmerman’s Symmetric Fuzzy Linear Programming Model 67 4.4.1 Zimmermann’s Symmetric Fuzzy Linear Programming model 67

4.4.2 Fuzzy optimizer 73

4.5 FTAA Robust Model: Ramik and Rimanek’s Robust Programming Model 74

Chapter 5 Validation of the Fuzzy Asset Allocation Models 77

5.1 The Data 77

5.1.1 JLL Data Source 77

5.1.2 The Required Data 78

5.1.3 Some Key definitions used in by JLL REIS-Asia 80

5.2 Calculation of Total Return Correlation Coefficient 84

5.3 The Fuzzy Asset Allocation Models 86

5.4 Comparisons with the Traditional and Fuzzy Asset Allocation Models 88

Chapter 6 Conclusion and Future Studies 90

6.1 Review of Research Objectives 90

6.2 Summary of Key Findings 90

6.3 Conclusion and Implications 92

6.4 Theoretical Contributions 93

6.5 Practical Contributions and Policy Implication 93

6.6 Limitations of This Research 93

6.7 Recommendations for Future Studies 96

References 97

Appendix A fuzzyTECH 5.12e (08-Mar-2000) 103

AppendixB MPT and Efficient Frontier 109

Appendix C The Fuzzy Tactical Asset Allocation Models 114

Summary

Although the classical Markowitz mean-variance asset allocation framework can be used

to enable decision-making in international and direct real estate investing, and that many

institutional investors have used it to support their decision-making, such a

mean-variance framework still needs to be enhanced in order to capture the multi-causal factors

influencing international and direct real estate investing A fuzzy decision-making

approach can be a more intuitive and a rigorous alternative in this regard The aim of this

dissertation is to enhance the classical Markowitz mean-variance asset allocation

framework through making it more appropriate for decision-making in international and

direct real estate investing This dissertation is concerned with the model formation and

estimation of the institutional investors’ fuzzy strategic asset allocation (FSAA) and

fuzzy tactical asset allocation (FTAA) The dissertation utilizes the Jones Lang Lasalle

Real Estate Intelligence-Asia office-sector dataset in order to integrate the fuzzy

decision-making approach with the classical Markowitz’s asset allocation mean-variance

framework to provide institutional investors, engaged in international and direct real

estate investing, with a more intuitive way of uniquely and rigorously capturing their

expert judgement in optimal asset allocation The findings indicate that the new and

estimated FSAA model and FTAA models can be the appropriate ones to enhance

decision-making in international and direct real estate investing

Keywords: Direct real estate, fuzzy set theory, fuzzy strategic asset allocation, fuzzy tactical asset allocation, Zimmermann’s fuzzy linear programming, Ramik & Rimanek’s fuzzy optimization

List of Tables

Table 1.1 Portfolio size of institutional investors across the globe (in billions of US$) 2

Table 1.2 Estimates of the size of the investible institutional real estate portfolio in 2000 3 Table 3 1 Project Statistics 53

Table 3 2 Linguistic Variables 55

Table 3 3 Base Variables 55

Table 3 4 Interfaces 56

Table 3 5 Definition Points of MBF "EconGthProsp" 57

Table 3 6 Definition Points of MBF "MktLiquidity" 57

Table 3 7 Definition Points of MBF "MktTransparency" 58

Table 3 8 Definition Points of MBF "MktVacancy" 58

Table 3 9 Definition Points of MBF "MktPerformance" 59

Table 3 10 Rules of the Rule Block "RB1" 60

Table 3 11The fuzzy strategic asset allocation results 61

Table 5 1 Historical TRs 2000-2005 79

Table 5 2 Historical Correlations among Asian Country TRs (2000-2005) 79

Table 5 3 Forecast (Ex ante) TRs, 2006-2010 80

Table 5 4 Ex-ante Correlations among Asian Country TRs (2006-2010) 80

Table 5 5 General information of ten Asian real estate market 85

Table 5 6 Correlation Coefficients 86

Table 5 7 Covariance Matrix 86

Table 5 8 Zimmerman's FTAA Flexible Programming Model Coefficients 87

Table 5 9 Zimmerman's FTAA Flexible Programming Model Results 87

Table 5 10 The Ramik & Rimanek FTAA Robust Programming Model Coefficients 87

Table 5 11 Ramik & Rimanek FTAA Robust Programming Model Results 87

Table 5 12 Comparision of the different asset allocation models 88

Table 5.13 Portfolio Risk and Return Comparisons……… ….…….………… 89

List of Figures

Fig 1 1 Structure of this Dissertation 10

Fig 2 1 The Real Estate Investing Process 14

Fig 2 2 Participants in the real estate investment 15

Fig 2 3 Opportunity set of portfolios 40

Fig 2 4 Risk indifference curves for investors A and B 41

Fig 2 5 Risk indifference curves for investors A and B with changes in the rate of interest 41

Fig 2 6 Risk indifference curves and the opportunity set 42

Fig 3 1 Structure of the Fuzzy Logic System 54

Fig 3 2 Membership Function of 'temperature' 55

Fig3 3 MBF of "EconGthProsp" 56

Fig 3 4: MBF of "MktLiquidity" 57

Fig 3 5 MBF of "MktTransparency" 57

Fig 3 6: MBF of "MktVacancy" 58

Fig 3 7: MBF of "MktPerformance" 59

Fig 4 1 Decision-making in fuzzy environment 66

Fig 5 1 Efficient frontier of ten Asian real estate market 85

Chapter 1 Introduction

“Precision is not truth.”

Henri Matisse, Impressionist Painter

1.1 Background and Motivation

1.1.1 Background

Real estate investing is a complex human cognitive process involving decision-making

regarding possible uncertain future returns In an ill-defined and complex environment,

human cognition is often overloaded with many interdependent facets of that

environment, resulting in many instances, a sub-optimal judgment Investment analysis

comprises several key analytical techniques, namely the discounted cash flow (DCF)

model, portfolio theory and risk analysis that are essentially structured frameworks,

which enable a more precise and certain evaluation of an investment However, the

success of investment analysis still relies greatly on the reliability and quality of the

inputs to the analytical techniques

In investment analysis, the precise and crisp result of any of its techniques (models) is

derived on the assumption that the variables in the analysis are deterministic or

probabilistic in nature This assumption is pseudo accurate and it fails to take into account

unexpected shocks or perturbations that are possible in the real world Therefore,

investors who rely on sophisticated analytical techniques are not placed in a better

position but are in fact subject to substantial risk Expert judgment offers an acceptable

alternative to non-nạve models as that judgment, which itself is limited by uncertainty, is

attributable to the vagueness and imprecision inherent to the associated expert’s ex ante

information Such a limitation is known as cognitive uncertainty or fuzziness As a result,

‘Fuzzy Set Theory’ is incepted to allow a natural and intuitive way of representing

cognitive uncertainty Fuzzy set theory relaxes the crispness and precision to enable a

robust summary of expert knowledge The incorporation of fuzzy set theory has made

significant inroads relating to the generalization of traditional investment analysis and its

techniques, thereby opening up a new frontier in structured frameworks for evaluating the

investment market

1.1.2 Motivation

Institutional investors like the insurance companies, banks, corporations and pension

funds, are the primary capital players in today’s investment environment The

corresponding and teeming volume of funds interested in international real estate

investing is highlighted in Table 1.1

Table 1.1 Portfolio size of institutional investors across the globe (in billions of US$)

Pension funds (1998 data) Insurance companies(1999 data)

Table 1.2 Estimates of the size of the investible institutional real estate portfolio in

Source: Henderson Investors (2000)

As observed in Table 1.1, the estimated value of the global investment market (i.e

insurancecompanies and pension funds around the world) in year 2000 is about US$23

trillion (Henderson Investors, 2000) The estimated total invested institutional direct real

estate market is much smaller at about US$1.3 trillion, which includes the direct real

estate holdings of insurance companies, pension funds and real estate companies in the

major economies

Recently, industry studies by major investment advisors, including Henderson Investors

(2000), Prudential (1988 and 1990), Jones Lang LaSalle, Lend Lease and AIG, have all

advocated international real estate investing to be the next frontier for institutional

investors, and international real estate to be an alternative and viable investment asset

class

Given the potentially immense volume of funds that is interested in international real

estate investing, it is not surprising that there has been a significant amount of research

focused on the potential benefits of an international real estate investment strategy

Investment management and advisory firms clearly support and promote the

diversification benefits of international real estate investing The diversification benefits

data to improve the quality of international real estate investing research Many studies

on such research adopt the mean-variance, modern-portfolio-theory (MPT) framework

but the adoption of MPT for international real estate investing has been questioned by

researchers

As some of the earliest critiques, Lizieri and Finley (1995) reiterated that a fund adopting

the mean-variance MPT framework in the context of international real estate investing

would have had a disastrous performance They suggested several reasons:

Technical problems with the data and the corner solution often results from the mathematics of modern portfolio theory (MPT)

With unstable returns, the historical mean returns, standard deviations and correlation coefficients between countries may not be the best way to analyze the (direct) real

estate asset class

The majority of the(direct) real estate return series used in the most past studies are short (about 15 years) but the direct real estate investments themselves are

characterized by their long holding period and are not single-period investments

They lack the liquidity of more traditional asset classes (like common stocks and

bonds)

There are additional risks associated with an international diversification strategy that includes asset-specific, the domestic/international sector, the domestic/ international

market and currency markets It is difficult to incorporate these risks into the

traditional mean-variance MPT framework for a mixed-asset portfolio and for the

most part they are ignored in most past research

Published research on the diversification benefits of international real estate investing has

been mixed Most research seem to believe that international real estate investing would

add diversification benefits to a direct real estate or a mixed asset portfolio, but some

disagree, or believe that other sources of international diversification (such as for

common stock) are superior Worzala (1992) reiterates that “diversification benefits can

be found from… overseas property” while Chenh, Ziobrowski, Caines, and Ziobrowski

(1999) pose the question of whether international real estate investing is as effective at

providing international diversification as other international asset classes Regardless, the

one thing they all might agree on is the lack of high quality data on direct and

international real estate markets and past performance Given the broad nature of the

subject matter, it is understandably hard to find reliable data sources for major markets

that are appropriate for enabling meaningful comparison with one another As a result,

many studies have focused on the best data that is available

What proportion of capital should institutional investors decide to devote to direct and

international real estate is subject to uncertainty? More recently, Brounen and Eichholtz

(2003) unsmoothed autocorrelations in domestic private and public real estate markets

adopting methods developed by Geltner (1989a, 1989b) and Giliberto (1990) They use

the resulting returns and confirm that the Sharpe maximizing mixed-asset portfolio

should contain approximately a 10% portfolio weight to real estate Pension Consulting

Alliance (PCA) remain conservative in relation to the inclusion of direct and international

real estate for strategic investment, preferring instead the tactical use of higher-return

investments in certain markets

In contrast, Prudential Real Estate suggests that costs aside, there are excellent

opportunities for direct real estate abroad if investors shift their thinking somewhat and

consider alternative frameworks The allocation that Lowrey (2002) recommends

investors should make in international real estate investing is a staggering 20%-30%

portfolio weight, an allocation that he feels is large enough to optimize the diversification

benefits and to maintain “a wide range of potential portfolio strategies” According to an

MIT survey, average international real estate allocations (within the real estate portfolio)

across pension funds had grow from a 2.1% portfolio weight five years ago to a 7.2%

portfolio weight today (Mullins, 2004)

As a result, there are still puzzles in the asset allocation problem of international real

estate investing Such puzzles cannot be easily resolved without better data for the direct

and international real estate markets Then and only then can more rigorous analysis be

conducted to enable the diversification benefits of geographical diversification, real estate

sector diversification and time diversification (time being one of the key sources of risk

or uncertainty) For the latter, it would be imperative to take advantage of different stages

of the direct real estate market cycles through superior real estate market analysis

1.2 Research Question and Issues

A main and relevant research question to pose is whether we can incorporate fuzzy set

theory into the classical asset allocation models, which optimizes systematic market risk

or uncertainty at the portfolio level, in order to enable direct real estate investment

decision-making to be efficient in international real estate investing?

Perhaps, a corresponding investigative research to the above research question can be

undertaken in two parts that are somewhat akin but different from the classical

mean-variance MPT framework: a unique and rigorous fuzzy strategic asset allocation (FSAA)

and a unique and rigorous fuzzy tactical asset allocation (FTAA) Both the FSAA and

FTAA should be practical to adopt

1.3 Objectives

The objectives for the investigative research of this dissertation comprise the following:

1 To improve the traditional mean-variance modern portfolio theory (MPT) framework

for direct real estate asset allocation in international real estate investing

2 To enable efficient decision-making in international real estate investing for

institutional investors, who are interested in direct real estate investments, as the benefits

of such investments are well documented (Goetzmann and Ibbotson 1990) Unfortunately,

individuals and smaller institutional investors have traditionally had difficulty in

obtaining these benefits since real estate investments prior to the 1990s were dominated

by illiquid units of commingled funds, and by direct investments that were primarily

directed at large institutional investors (Martin 1997, p 145; Bogle 1994, p 52)

However, the recent trend towards securitized real estate investments, particularly the real

estate investment trusts (REITs), has increased the availability of liquid and indirect

real estate security investments for individual and small institutional investors (National

Association of Real Estate Investment Trusts, 1998) Therefore, the study in this

dissertation is only concerned with the institutional investors’ decision-making through

direct real estate asset allocation in international real estate investing

1.4 Data and Methodology

Prime office annual total returns, comprising annualized rental yields (real estate

capitalization rates) and capital value (CV) appreciation on a pre-tax and pre-leveraged

basis, are obtained for ten Asian real estate sectors, namely those for the Beijing Central

Business District (CBD), Shanghai CBD, Seoul CBD, Tokyo CBD, Hong Kong Central

& major business districts, Manila’s Makati CBD, Jakarta CBD, Singapore’s Raffles

Place CBD, Kuala Lumpur CBD, and Bangkok CBD The direct real estate total return

data set is essentially ex post, spanning the period from 2000 to 2005 in US$ terms As

such, there is no conversion carried out for the total return data set in hedged US$ terms

in contrast to a forecast dataset, where the hedged US$ conversion would be appropriate

In Asia, the Jones Lang Lasalle Real Estate Intelligence Service

(JLL REIS-Asia), which is based in Singapore, is the sole service provider that maintains

a reliable transaction-based set of indicators for the market performance of the prime

office sectors in ten countries of the Asian region JLL REIS-Asia also produces 5-year

total return forecasts in local currency terms for each of the markets and several key real

estate market indicators (i.e market demand growth, completions, vacancy, rental change,

yield and capital value growth) In effect, the JLL REIS-Asia data set is a research asset

class index type of database, as opposed to a peer universe type

In general and in terms of research methodology, the following step-wise procedure can

be adopted:

1 Examine the behavioral uncertainty behind international real estate investing in direct

real estate investments

2 Incorporate fuzzy set theory into the strategic asset allocation process

3 Incorporate fuzzy set theory into the tactical asset allocation process

4 Validate the above behavioral uncertainty and the corresponding process models

1.5 Scope of the Research

The research study in this dissertation is scoped along the following confines:

1 International and direct real estate investments (excluding corporate real estate)

2 Fuzzy strategic asset allocation (FSAA) and fuzzy tactical asset allocation (FTAA)

formulation and estimation

3 Inter-sector portfolio diversification but within a direct real estate portfolio because

direct real estate is found to be an effective portfolio diversifier, even more so when both

domestic and international real estate sectors (assets) are considered

1.6 Potential Results and Contributions

Fuzzy set theory can be incorporated into traditional mean-variance MPT asset allocation

models and can therefore improve the efficiency of asset allocation decision-making in

international real estate investing for direct real estate investments Thus, the

contributions of this dissertation are achievable in the following ways:

• Modern and new ideas with regard to fuzzy set theory and fuzzy logic can be

introduced into the international real estate investing process for direct real estate

investments

• The study in this dissertation can incorporate human and intuitive thinking into

the direct real estate asset allocation process within the international context

• New and direct real estate asset allocation models are developed in international

real estate investing

1.7 Structure of This Dissertation

This dissertation consists of six chapters and Fig 1.1 shows the relationships among the

chapters The chapters are outlined below

Chapter 1 Intruduction

Chapter 2 Direct Real Estate Asset Allocation & Fuzzy

Optimization Application to Finance: A Literature Review

Chapter 3 Fuzzy Strategic Asset

Allocation (FSAA) Decision

Making Models

Chapter 4 Fuzzy Tactical Asset Allocation (FTAA) Decision Making Models

Chapter 5 Results Comparison and Interpretation &

Chapter 2 reviews the related literature of direct real estate asset allocation, within

international real estate investing, and the extension of fuzzy set theory, fuzzy logic and

fuzzy optimization in this regard

Introduction

Chapter 3 discusses the fuzzy strategic asset allocation (FSAA) models, where four

macro market indicators are examined in such models

Chapter 4 discusses the development of fuzzy tactical asset allocation (FTAA) models In

addition, the flexible programming and the robust programming models are examined in

depth

Chapter 5 discusses the estimation, results and makes a comparison between different

asset allocation models Furthermore, these models are validated This chapter also

demonstrates the practical use of fuzzy asset allocation in international real estate

investing for direct real estate investments

Chapter 6 concludes this dissertation

Chapter 2 Literature Review

“It is the mark of an instructed mind to rest satisfied with that degree of precision which the nature of the subject admits, and not to seek exactness where only an approximation of the truth is possible.”

Aristotle, Ancient Greek Philosopher

Chapter 2 is concerned with reviewing the related literature on direct real estate asset

allocation within international real estate investing This chapter also extends the

conceptions of fuzzy set theory, fuzzy logic and fuzzy optimization to international and

direct real estate asset allocation

2.1 Real Estate

2.1.1 Definition of Real estate

It is first imperative to properly define real estate itself and according to Graskamp, “Real

estate is Space and Money over Time.”

The space dimension- covering fundamental policy analysis, housing markets (especially user cost and sub-markets)

The cash flow dimension- covering securitized real estate (i.e uncoupling cash flows from physical real estate), the real estate investment trust (UBS DoRE public talk Oct

2004: real estate itself comprises 12% of global investment assets, tremendous

growth in REIT in North American, Europe and Asia), and asset-backed

securitization

Space & Money dimensions- covering the pricing of attributes, technology and cash flow

2.1.2 Characteristics of real estate

Next, the characteristics of real estate can be defined to include the following:

Immobility: fixed in location

Heterogeneity: unique, scarce (particular use at a particular time and location)

Durability: services, good loan collateral

Sizeable investment outlay: financing

Indivisibility of investment

Illiquid: long transaction process

Studies relating to the characteristics of real estate have been considerably hampered by

the lack of reliable data In comparison with the equities market, our knowledge of even

the distributional characteristics of real estate returns and performance are relatively

limited This has been the result of data and information confidentiality and data

inadequacy This situation is gradually changing, and good quality data is now more

readily available and our understanding of real estate market analysis this sector is

improving

2.1.3 The real estate investing process

When one investor makes a real estate investing decision, he should systematically

analyze the factors and contingencies that impact the value of a real estate investment It

is widely accepted by the real estate investment management community that investment

in real estate is about sacrificing present consumption for future benefits; and that the real

estate value is the present worth of rights to the future benefits arising from ownership

At the same time, the real estate investor must identify his objectives, goals and

constraints The basic objective for the investors would be wealth maximization while the

investor makes his investing decision, based on the mean-variance (return-risk) i.e the

risk pricing principle This typical real estate investing process is depicted in Fig 2.1

STEP1: Identify investors’

objectives, goals, and constraits

STEP2: Analyze investment climate and market conditions

STEP3: Develop financial

analysis Financing decision

STEP4: Apply making criteria

decision-Discounted cash flow techniques

STEP5: Investment decision

Fig 2 1 The Real Estate Investing Process

Source: Jaffe and Sirman 1995; Author, 2007

In this thesis, we investigate how to undertake that real estate asset allocation

decision-making when the institutional investor invests in the international and direct real estate

market, to optimize the systematic market risk for maximizing return at the direct real

estate portfolio level and given that the specific risks of the real estate assets (or sectors)

cancel out one another

As illustrated in Fig 2.2, there are many participants that invest in real estate investment

mainly comprise the equity investor, the mortgage investor, the tenant-occupier and the

government that regulates their behavior and space-user restriction

Tenant 1.Residential 2.Commercial 3.Industrial 4.Special purpose 5.Others

REAL ESTATE INVESMENT

Types 1.Short- term 2.Long-term

Mortgage Lender Sources

1.S&L 2.Insurance 3.Bank 4.Individual 5.REIT

RRRReeeesssstttt rrrriiiiccccttttiiiioooonnnnssss oooonnnn

LLLLeeeennnnddddeeeerrrrssssPPPPrrrroooommmmiiiissss

ssssoooorrrryyyy NNNNoooo

MMMMoooo rrrrtttt ggggaaaa ggggeeee DDDDoooo ccccuuuu mmmmeeee nnnntttt

LLLLeeeeaaaasssseeee DDDDooooccccuuuummmm

eeeennnntttt

RRRReeeessssttttrrrriiiiccccttttiiiioooonnnn oooonnnn UUUUsssseeeerrrr

DDDDeeee bbbbtttt SSSSeeee rrrrvvvv iiiicccc eeee

tttthhhh TTTTaaaa xxxxaaaa ttttiiii oooonnnn

Restrictions on use:

taxation, eminent domain, police power

Mortgage Laws

Property Rights Cash Flow

Landlord-Tenant Laws

Fig 2 2 Participants in the real estate investment

Source: NUS lecture

So, although there are many participants in the real estate investing process, this

dissertation is concerned with institutional investors that are interested and are actively

engaged in the international and direct real estate asset allocation process

2.1.4 The benefits of real estate investing

Several key investment and financial benefits result from real estate investing and they

include the following:

Correlation coefficient is low among the direct real estate sectors or assets (the direct real estate sector diversification and the geographical diversification)

Pride of direct real estate ownership, locally and abroad

Reduce / control operational costs (a financial benefit at the direct real estate project level)

Steady rental yield and long-term capital growth (an investment benefits at the direct real estate portfolio level)

Inflation hedge and tax shelter (an investment benefits at the direct real estate portfolio level)

Risk diversification (an investment benefits at the direct real estate portfolio level)

Financial leverage (a financial benefit at the direct real estate project level, consistent with the capital asset pricing model and theory)

2.1.5 Why choose international and direct real estate investing, and not the real estate investment trust (REITs)?

It is worthwhile noting the adage that as property “lives forever”, and no one knows

exactly what the future cash flows from any property will be for all years into the future,

commercial real estate is certainly a “risky asset” (Geltner, 1989) In the 1984 special

issue on valuation published by the Journal of the American Real Estate and Urban

Economics Association (AREUEA), its editor Lusht surveyed 71 professional appraisers,

asking them what valuation topics they considered most important for research The

second most frequently cited response was “Estimating risk and determining the proper

discount rate”

In short, there is no clear consensus on how “risky” commercial real estate properties are,

even on a relative basis compared to common stocks, and there seems to be considerable

confusion even regarding how to measure or think about this issue We seem to know

much less about risk in direct real estate returns than we do about risk in stock market

returns, even though direct real estate is of comparable magnitude to the stock market in

capitalized value, and in some ways the direct real estate assets are much simpler and

easier to understand than modern industrial corporations

The basis, obtainable upon empirically and observable data like cash flows and the

appraisal returns, is thus famed more for the purpose of developing our intuition about the

nature of direct real estate return risk, rather than the required empirical analysis

So, why not simply do an empirical study of the returns to indirect securitized real estate

portfolios, like the REITs, which trade on the stock exchanges Securitized real estate and

their portfolios offer regular and frequently true returns data, based on stock prices and

dividends, and so offer a more direct and theoretically accurate data source of returns A

short answer to the above question is that this dissertation’s investigation is not only

primarily empirical but it is also seeking to understand the nature and determinants of

direct real estate return risk rather than to describe what that risk has historically been ex

post However, several studies have analyzed the indirect securitized REIT returns, and

the REIT data does provide an important source of empirical information about indirect

real estate risk and returns (Smith & Shulman, and Burns & Epley.) In general, such

studies have found that REIT returns behave much like the typical common stock returns,

similar in particular to the common stocks of public utility companies REITs generally

have higher than average yields and lower than average volatility but have their REIT

betas to be smaller than average and significantly positive, all with respect to the stock

market REIT returns are found to be highly correlated with the overall stock market

return At the same time, there are not very many equity REITs and many of them have

small and changing portfolios, and/or have not existed nor have they been publicly traded

for very long

Therefore, it is difficult enough to obtain very clear and specific information about real

estate return risk through studying the REITs alone Perhaps, more serious is the widely

held perception among real estate academicians and practitioners that “REITs are not

Real Estate”, in the sense that the REIT risk and return characteristics are perceived to

differ significantly from those of direct real estate and even from those of the direct real

estate portfolios, which underpin the REITs’ values Various explanations offered to

account for the widely held perception, ranging from arguments that the “stock market is

inefficient” to arguments that the REIT return risk reflects the intangible REIT’s

management risk more than the risk of the tangible direct real estate assets themselves

which the REIT owns Another possible explanation is that the direct real estate assets in

turn are very heterogeneous, such that even seemingly large and diversified direct real

estate portfolios can differ significantly in their risk and return profile

2.2 Direct Real Estate Asset Allocation

2.2.1 Asset allocation

In general, asset allocation in finance theory refers to the process of securing the most

favorable return and risk trade off involving competing interests that are concerned with

risk reduction and return enhancement at the portfolio level, subject to various constraints

(Kritzman, 1992) The importance of asset allocation in arriving at optimal investing

decision-making cannot be over-stated (Brinson, Hood and Beebower, 1986; Grieger,

1987; and Ankrim, 1992) Traditionally, the asset allocation decision is based on the

expected mean-variance theory (EMV) wherein the return and risk profile at the portfolio

level has to be balanced and subject to constraints (Markowitz, 1952) However, recent

studies have increasingly criticized the effectiveness and appropriateness of the EMV

theoretical approach A frequent criticism is the sensitivity of the results to estimation

errors (Chopra and Ziemba, 1993) Other commentators have highlighted its sensitivity to

time factors (Kritzman, 1990, pp.129-144) and to fundamental factors (Benari, 1990)

An alternative and very meaningful way to overcome these above problems is to rely on

expert judgment (Lee et.al., 1990) The use of expert judgment, though well established

(Fischhoff, 1989; Keeney and Von Winterfeldt, 1989), primarily relies on less than

precise descriptors of investment asset returns, risks and the correlations among the

results As Fischhoff (1989) noted, the difficulty with the expression of expert knowledge

utilizing numerical parameters arises from the observation that experts do not think in

terms that are the same as the language of the models Most experts express the

investment asset’s expected returns, risks and the return correlations in vague terms rather

than in precise and crisp figures In other words, expert judgments tend to be fuzzy Until

only recently has there growing interest in the field of mathematics to represent the vague

quantification of human expression, and to manipulate the resulting human linguistic

quantification The study in this dissertation seeks to investigate the adoption of fuzzy set

theory and fuzzy logic to incorporate expert knowledge, in the form of corresponding

fuzzy parameters and within the classical quadratic programming approach, to enable

portfolio asset optimization Two fuzzy mathematical programming models are uniquely

specified and estimated in this dissertation, namely, the ‘Flexible’ and ‘Robust’ models

2.2.2 Direct real estate asset allocation

Published research on the benefits of international real estate investing has been mixed

Several research studies believe that international real estate investment would enhance

diversification to a direct real estate portfolio or to a mixed asset portfolio but some

disagree or believe that other sources of international diversification (such as equities) are

superior Regardless, the one thing they all might agree on is the lacking of high quality

data on direct and indirect international real estate markets and their past performance It

is understandably hard to find reliable and authoritative data sources for major markets of

direct and indirect international real estate that are appropriate for comparative

examination As a result, many studies have focused on the best data that is available In

particular, the indirect securitized real estate returns have been commonly utilized, along

with private direct real estate returns from the more developed countries or markets, like

the US, Japan, Great Britain and Hong Kong

Worzala (1994) had conducted a study of overseas investment by pension funds and other

institutional investors Her study makes an interesting comparison for this dissertation, as

the subject matter is nearly identical What varies is of course the time of her study (10

years earlier) and the nature of the investors Worzala (1994) focused on a pool of

investors, most of which were British or from the European continent: 60% was British,

23% was Dutch and only one respondent was American The focus was intentionally on

the British investors while the investor fund size was quite large at about US$2.1 billion

Worzala highlighted the debate over the benefits of international and direct real estate

investing In her article she referenced several authors whom she claimed did present

evidence both for and against international and direct real estate investing She cited

Webb and Rubens (1989) as arguing against international and direct real estate investing,

as institutional investors would not have benefited from it during the past twenty years,

while Worzala maintained that other authors at the time supported such a case for

international and direct real estate inclusion For example, she claims Sweeney (1989a

and 1989b), Asabere et al “(1991), and Worzala (1992) reiterated that “diversification

benefits can be found from… overseas property.” Rather than entering further into the

foregoing debate, her 1994 study did shed light on the behavior and decision-making

thought processes of institutional investors, who were actively involved in international

and direct real estate investing The study took the form of a mailed investor survey with

several components It investigated the background information on direct real estate

investments abroad, investment techniques inclusive of asset allocation and their

usefulness, the corresponding motivating factors behind such investing decisions, the

impact of currency fluctuations and the perceived trends in international and direct real

estate investing

Worzala concluded that some institutional investors were in fact active internationally

and that these investors were primarily seeking portfolio diversifications, even though

empirical findings on the benefits of portfolio diversification had been inconclusive She

inferred that the investors surveyed were not very sophisticated in dealing with currency

risk, and were not particularly interested in the “… potential negative impact of currency

fluctuations and large transaction costs.” She highlighted international and direct real

estate investing was likely to increase in future

Additional research has been published in favor of international and direct real estate

investing as an imperative source of diversification Eichholtz, Huisman, Koedijk and

Schuin (1998) found that north American returns were being driven by a strong

continental factor, suggesting that US investors needed to go abroad to achieve optimal

portfolio diversification This study found that Asia does not have a continental factor

while Europe does It also found some correlation between the US and Asia direct real

estate returns The implication is that Asia can provide crucial direct real estate portfolio

diversification for Europe, and for the Asian countries themselves North America should

even benefit from direct real estate portfolio diversification from Europe and vice-versa

At the same time, it suggests that direct real estate investors can only attain optimal

portfolio diversification by investing across continents, as opposed to simply investing in

neighboring counties Furthermore, the authors, while being sensitive to the costs and

difficulties involved in investing internationally, suggest that most of the direct real estate

portfolio diversification benefit can be attained by investing in one country or market in a

given continent, through taking advantage of the “continental factor” The continental

factor has a twofold definition First, country returns should depend significantly on the

aggregate market returns in the home continent, and second, this relationship should be

stronger than the dependence on other continents’ markets There are two important

reasons why continental factors are relevant for portfolio managers First, if continental

factors are important, real estate investors can only achieve optimal international

diversification by investing inter-continentally Secondly, the existence of continental

factors implies that international investors can acquire a near-optimal country allocation

by just selecting a country from each continent This can save transaction and monitoring

costs, which can be substantial for real estate investments

Nevertheless, Chenh et al (1999) pose the challenging question of whether or not

international and direct real estate investing is effective in providing portfolio

diversification as other international investment asset classes The co-authors suggest

international and direct real estate investing can still be an effective diversifier They

admit that the studies by Giliberto and Testa (1990), Gordon (1991) and Eichholtz (1996)

did support the international diversification of a direct real estate portfolio but they

qualify that such a conclusion may well be limited in that they do not consider the

resulting effects on a mixed asset portfolio They cited Ziobrowski and Curcio (1991)

who found that international and direct real estate investing offers no “unique” portfolio

diversification benefits to a mixed asset portfolio while asserting that international

common stocks, bonds, and treasury bills can provide the same benefits It is implicit that

the more liquid stocks, bonds and bill are easier (cheaper) to invest in, and they should be

preferred as a means of achieving full portfolio diversification The authors claim that

further research work by Ziobrowski and others that incorporate financial engineering are

incapable of improving the efficiency of portfolio diversification through adopting

international and direct real estate investing

In a 1999 study, the authors also find that by adopting the ‘bootstrap’ method to model

uncertainty, it can be established that primarily because of the powerful influence of

currency fluctuation, all foreign investment assets effectively behave as one and the same

asset (that is, their individual return differences to one another are drowned in the

massive effects of currency movements) They claim this effect weakens the argument

that any particular investment asset class is best for international portfolio diversification

Running their model, the authors show that only a small benefit arisen from international

and direct real estate investing, and they suggest an appropriate allocation may well be

1-3% most of the time For a small percent of the time, say about 5% of the time, however,

the situation would occur when the appropriate allocation of 18% or more would

accorded to international and direct real estate investing They then suggest that a rang of

5-10% would be the appropriate long run allocation to international and direct real estate

investing while maintaining that other investment assets can be used most of the time An

interesting observation about this study, however, is that it only used data sources from

Japan and Great Britain to model the diversification benefits of a US-based portfolio

Another 1999 study by Eichholtz, Veld, and Vestbirk (1999) considers the use of indirect

securitized real estate for portfolio diversification They looked at three issues related to

investing in securitized real estate abroad They confirmed that while securitized real

estate represented only an estimated 3% of the global stock of commercial real estate,

pricing was found to be reasonable, that the securities were liquid, and that the markets

were significant in size, thereby making the largest public real estate markets viable for

institutional investing The implication was that such securitized real estate may well be a

substitute for private real estate on a global basis

Building on their previous study, Eichholtz, Koedijk, and Schweitzer (2001) highlighted

the trade-off between the benefits of diversification from international and direct real

estate investing and the high costs of undertakes such investing They suggested that

those costs can be greatly reduced by investing in securitized international real estate

investing in the form of local real estate companies Their study compares the

performance of ‘international’ real estate investment trusts (REITs) and real estate

company shares to ‘local’ REITs and ‘local’ real estate company shares They found that

the international had underperformed their local counterparts and that this

underperformance was not due to factors like leverage, transaction costs, and currency

but rather that the local real estate companies seemed to have an overall performance

advantage, although as the international counterparts grew larger then their performance

improved, and thus narrowed the performance gap

More recently, Brounen and Eichholtz (2003) did unsmooth the autocorrelations among

the domestic private and public property markets by adopting the methods developed by

Geltner (1989a, 1989b) and Giliberto (1990)

The authors found that returns for securitized and private real estate investments are

increasingly similar when such methods are applied They posited that the true

‘unobserved’ property index was somewhere between the lagged, the smoothed private

index and the public market, whose volatility is increased by the stock market on the

whole In the end, they used the resulting returns and confirmed that the Sharpe

maximizing mixed-asset portfolio should contain approximately a 10% allocation to real

estate

The articles by Eichholtz and others helps to close the gaps in the understanding of

private and public real estate returns, both domestic and international They confirm the

potential role of direct and indirect real estate in the mixed asset portfolio, affirm the

value of securitized real estate for diversification and help to project the true, unobserved

returns of real estate They do not, however, provide conclusive evidence of the absolute

diversification benefits of international real estate investing or its superiority in providing

optimal international portfolio diversification of the mixed portfolio Such answers may

not be forthcoming until the global international direct and indirect real estate markets

provide better information about their private market returns

At the real estate investment advisory level, Pension Consulting Alliance (PCA),

provides investment advice and consulting services to pension plan sponsors, typically

working directly for a fund’s investment board In some cases, PCA manages

discretionary accounts and will make investment decisions for its clients In all, PCA’s

clients have invested over US$25 billion in direct and indirect real estate, making the firm

one of the most influential voice in the industry

PCA is on the conservative side of the international and direct real estate investing and it

issue recommends a tactical approach, rather than a strategic one In a January 1998

white paper entitled, “US Pension Fund Investment in International Real Estate,” Nori

Lietz and Claudia Faust of PCA address the role of international and direct real estate

investing in an institutional investor’s mixed asset portfolio This paper begins with a

brief history of such investing and the authors cite the common reasons for institutional

investors to so invest because of portfolio diversification benefits and superior

risk-adjusted returns They consider the decision to invest in international and direct real

estate investing simply because it makes up a significant portion of the world’s investible

real estate but still deemed to be insufficient They also consider the diversification

argument to be incomplete and so they advise investors to seek superior risk-adjusted

returns as the sole criterion for making investingt decisions

The PCA paper points to the important differences between public and private markets

They agree that the initial [research] data supports the argument of portfolio

diversification with public international and direct real estate investing but stress that the

time period of the data was rather short (suggesting the long term benefits are unclear),

and that much of the data occurred during a period marked by a strong global market for

real estate (direct and indirect), while many countries of interest has few public real estate

companies These constraints limited the usefulness of these indirect real estate

investments While they conceded the US stock market is not highly correlated with

domestic direct real estate (implying it was not correlated with international and direct

real estate, either), they suggested that strategic commitments of significant size would be

necessary to achieve portfolio diversification benefits, and they point to the following

literature as that evidence such an approach would not be reasonable:

The PCA paper cites several studies in favor of portfolio diversification via international

and direct real estate investing It cites Eichholtz (1996) as international and direct real

estate investing offers portfolio diversification and it suggests that private international

and direct real estate investing would offer more portfolio diversification because the

securitized and international and direct real estate investing linked to the public markets

Hartzell, Watkins, and Laposa (1996) reiterated the common stocks of international real

estate investment companies to be more effective at portfolio diversification than

international equities Also, they assert that Gordon and Canter (1997) found that US and

foreign real estate markets to be less correlated than US and foreign equity markets

Asabere, Kleiman and McGowam (1991) as well as Steinfeld and Zisler (1996) showed a

weak correlation between the US and international real estate stocks Gordon (1992)

showed that in terms of the private market, US investors can improve portfolio

diversification by buying direct real estate in the UK Moore, Freeman, Allen and Ingram

(1995) demonstrated that public and private real estate returns track each other within a

country (suggesting that public real estate securities can be considered a proxy for the

private markets) Gordon and Canter (1997) suggested that international real estate

securities can successfully diversify a US mixed-asset portfolio while Quan and Titman

(1996) found that private direct real estate capital returns and public indirect real estate

equities were not correlated within the US but that these were highly correlated in the UK

and Japan as well as across Asia in general These findings suggest that a significant

correlation may well exist between public indirect real estate equities and certain private

international and direct real estate markets, but the authors point out that the results were

not statistically significant Lastly, they recount the conclusion by Eichholtz et al (1997)

that a ‘continental factor’ exists in Europe and North America but not in Asia

While Lietz and Faust (1998) agreed there was evidence that international and direct real

estate investing can provide portfolio diversification, their conclusion is that certain

markets offer more portfolio diversification than others and that private direct real estate

has not been established strongly enough to be the best source of international portfolio

diversification They do believe investors should make allocations to markets with the

best possibility for portfolio diversification but they assert that international

diversification for a mixed asset portfolio can be best accomplished on the equities side,

and that international and direct real estate investing should always meet or exceed the

risk-adjusted returns for US direct real estate Regardless, PCA thinks that to make an

impact (in terms of international portfolio diversification), then roughly 10% of an

investor’s mixed-asset portfolio should be allocated to international investments,

inclusive of real estate, and that investors should carefully consider such an allocation

percentage to be feasible

Lietz and Faust (1998) also point out that not all data support the higher or even the

equivalent risk-adjusted returns to be attainable in international markets In some cases,

they claim that international and direct real estate investing was found by researchers to

having lower risk-adjusted returns than the domestic and direct real estate

According to these two authors, the study by Eichholtz (1996) “drew no conclusions” as

to whether or not international and direct real estate investing offers different

risk-adjusted return’s compared to the US common stocks and bonds They refer to a study by

Gordon and Canter (1997), which supported Eichholtz’ findings, that public indirect real

estate markets are more volatile than their overall equity markets, some of the public

indirect real estate markets even offered higher risk-adjusted returns than expected, and

that a tactical strategy for international and direct real estate investing would be the best

Quan and Titman (1996) produced similar results but they point out that the time periods

to enable the foregoing studies were limited and that these periods may well be

influenced by market cycles Similarly, they caution against relying too much upon past

performance and they even question whether or not investors should invest strategically,

and so risk periodic underperformance, or whether or not investors should invest

tactically to attain superior performance under market timing

PCA on Direct Investment in International and Direct Real Estate Investing

If an investor does elect to undertake in international and direct real estate investing,

Pension Consulting Alliance (PCA) believes that care should be taken to construct a

resulting portfolio based on the correlations of the investor’s existing direct real estate

investments and its composite portfolios They recommend the construction of

customized portfolios with investments over-weighted in certain markets that minimize

correlations at both levels One particular fear they have is that previous studies on the

performance of private international and direct real estate investing performance lack

sufficient treatment of the effects of foreign taxation on returns Private international and

direct real estate investing may not be tax exempt, and even if leverage or other tax

shields are used, a significant premium may well be needed to compensate for the taxes

Likewise, while currency effects have been factored into some of the previous research,

such effects may not be applicable to private international and direct real estate investing,

either Finally, since previous studies on the diversification benefits of private

international and direct real estate investing, and in the opinion of PCA to be inconclusive,

it is difficult to justify incorporating such investing within the context of modern

portfolio theory (MPT) diversification Their conclusion is that public international and

direct real estate investing has an advantage over private international and direct real

estate investing since the latter’s public securities are less likely to be affected

unexpectedly by the above mentioned factors Thus, in PCA’s opinion, investors should

compare the after-tax private and international and direct real estate investing and the

after-tax public international indirect real estate investing returns to other investment

asset classes, especially for the tax-exempt investors, in order to optimize expected

returns

According to PCA, institutional investors should undertake private international and

direct real estate investing only for higher risk-adjusted returns, since the associated costs

of investing are higher PCA in general thinks that US pension funds should overweight

the domestic market since these additional costs are not present As previously mentioned,

private international and direct real estate investing may not be tax exempt Volatile

exchange rates may further affect returns, and the repatriation of capital may well be a

problem Private international and direct real estate investing can be difficult to monitor

and can be illiquid Finally, more so than with securitized indirect real estate, the political,

social, and economic risks do come into play, inclusive of graft and corruption of the

domestic level In general, PCA believes that strong local knowledge is needed to execute

private market investment, something few US investors can claim to have

Concluding Remarks

Lietz and Faust (1998) believe that international and direct real estate investing should

offer higher risk-adjusted returns than those of domestic and direct real estate investing in

order to be considered by a pension fund Owing to their higher growth, developing

markets may be better off in adopting a strategic approach than a tactical one, and that

strategic investments should be focused on uncorrelated markets Investors seeking

portfolio diversification should (a) invest significantly, (b) make country allocations

systematically, and (c) consider the difficulties of the private markets, including

rebalancing the portfolio over time However, given the conflicting research on portfolio

diversification and the quality of international and direct real estate investing returns,

Lietz and Faust advise against taking a strategic approach They feel that securitized

international and direct real estate investing is better as strategic ‘core’ investments, and

that private investment in emerging markets is better for a tactical ‘opportunity’ approach,

which would provide investors with appropriate returns given the risks and costs involved

Interestingly, Lietz and Faust provide an alternative suggestion for institutional investors

to avoid some of the local costs of investing in private direct real estate They propose

that swapping investments between foreign and domestic tax exempt investors may well

be the most efficient and liquid way to achieve international portfolio diversification

Prudential Real Estate Investors (PREI) is a real estate investment manager for pension

funds and other institutional investors PREI is one of the largest real estate investment

management firms with over US$22 billion in assets under management PREI’s research

department produces articles on real estate investment regularly, many of them on the

subject of international and direct real estate investing The firm manages real estate

investments in multiple markets on behalf of both international and US investors As such,

they advocate international and direct real estate investing for their global clients as an

achievable source of optimal risk-adjusted returns and portfolio diversification, and they

recommend investing across the risk spectrum, from ‘core’ markets (i.e the developed

countries) to ‘emerging’ markets (i.e the developing countries with strong prospects)

PREI on Returns for Private International & Direct Real Estate Investing

Several articles in PREI (Prudential Real Estate Investors)’s library address the issues

pertaining to the associated costs of international and direct real estate investing and its

portfolio diversification benefits Connor, Liang, and McIntosh (1999) attempt to address

what PREI calls “the myth of the foreign risk premium”, a common belief in the

institutional world that investors should automatically earn a higher risk- adjusted return

for going abroad Their study classifies this ‘foreign risk’ into three categories, namely

country risk, market risk, and deal risk to illustrate the argument that not all foreign deals

are the same, in as much as domestic deals are not Therefore, they claim that a very safe,

pre-leased office development in London should not necessarily carry a higher return

than a speculative office development in Houston With country risk signifying long-term

risk, the study next breaks the countries down into three categories, namely Core, Core

Plus, and Emerging These three risk categories allow another dimension of risk to

discriminate, for example, between very similar deals in very different markets, like two

pre-leased class A office developments, one being in Paris and another in Seoul They

assert that in such situations, investors should expect equivalent risk-adjusted returns if

the risks relating to the country, market and deal have been appropriately measured

PREI on International and Direct Real Estate Investing Allocations

Liang and McIntoch (2000) establish a premium for each of the 50 markets of interest

based on PREI’s market-based method The market-based method involves using

international common stock market data from ‘Morgan Stanley’ in conjunction with

country credit ratings by ‘Institutional Investor’ to determine the appropriate risk

premiums This contrasts with the categorization method used by others that assigns risk