ACTIVE PORTFOLIO MANAGEMENT AND PORTFOLIO CONSTRUCTION- IMPLEMENTING AN INVESTMENT STRATEGY

(1)

AUTHOR: JOHAN CHRISTIAN HILSTED DATE OF SUBMISSION: DECEMBER 14^TH 2012

ACTIVE PORTFOLIO MANAGEMENT AND PORTFOLIO CONSTRUCTION

- IMPLEMENTING AN INVESTMENT STRATEGY

SUPERVISOR: JEPPE SCHOENFELDT NUMBER OF PAGES: 77

NUMBER OF CHARACTERS: 145.260 SIGNATURE:

(2)

1

Abstract

This thesis aims at creating an investment strategy for active portfolio management to outperform the MSCI Denmark from 1992 to 2011. The index development of the Danish stock market has been quite impressive as it has performed remarkably better than other national indices. It is therefore interesting to investigate whether active portfolio management constitutes a winning strategy superior to investing in the MSCI Denmark.

There is no generally accepted approach to conduct active portfolio management. This thesis approaches the subject by comparing two internationally diversified portfolios to the MSCI Denmark as benchmark - one portfolio submitted to a 20% maximum asset representation restriction, the other portfolio left unrestricted.

From investment strategy we conclude that combining strategic and tactical asset allocation constitutes an appropriate investment strategy for active portfolio management, as it limits the long-term portfolio investment opportunities and allows for short-term portfolio repositioning. The information ratio constitutes the performance measure of active portfolio management, as it optimizes portfolio construction by comparing expected returns of portfolio and benchmark – the residual return. The Capital Assets Pricing Model (CAPM) was utilized for return estimations for both investment opportunities and benchmark. Mean-variance portfolio construction was conducted based upon investment opportunities expected to outperform the benchmark.

The MSCI Denmark provides realized average monthly return of 0,65%, while the actively managed portfolios produce average realized monthly return of 0,34% and 0,37%, respectively. In that regard, active portfolio management has not outperformed the benchmark, and statistical findings cannot suggest portfolio timing skill. However, considering the systematic risk adjusted return, both portfolios yield significant alpha, or value added, with the unrestricted portfolio being the best performing portfolio.

In conclusion, active portfolio management cannot produce higher return than the MSCI Denmark, but has proven to benefit the investor, as the market risk exposure justifies both inferior and superior portfolio return to the benchmark.

(3)

2

1. Introduction

A common objective of the portfolio investor is to achieve a higher portfolio risk adjusted return as opposed to investment in a single asset. Combining assets into a portfolio carries the opportunity of risk reduction and at the same time acquiring a higher return compared to single asset investment.

As financial markets experience different phases, different regimes reside in the markets and many investment portfolios incur both losses and gains if it is not managed in accordance with the investor’s expectations to future market developments. For long-term investments such as pension investments, incurring losses in the short term is of little concern as the investment time frame allows for the opportunity to reduce such losses by gaining future positive returns. However, for some investors, in practice, this is not a feasible strategy since they are constrained by consumptions and liabilities.

Accordingly, investors need to liquidate some of their investments in order to fulfill financial needs and obligations. In other words, the investor buys and sells stocks, and the basis of this decision is the conviction that abnormal investment returns can be gained. However, the efficient market hypothesis, which can be traced back to Samuelson (1965) and Farma (1970), states that market prices incorporate all information rationally and instantaneously, eliminating the possibility for the investor to achieve abnormal returns and should this hypothesis hold in practice, the only optimal portfolio strategy would be to conduct portfolio investment and hold the portfolio throughout a predetermined time frame¹².

However, assuming stock markets are not efficient, in terms of market paradigm we turn to the adaptive market hypothesis by Lo (2004) who acknowledges the problematic issue of the assumption of market efficiency³. This paradigm carries some implications that necessitate portfolios to be actively managed.

First, a relationship between asset risk and return exists, but is unlikely to be stable over time. Second, arbitrage opportunities arise over time. A third implication is that investment strategies might not perform equally well in different economic environments. A fourth implication is survival, which is enabled by evolving markets and financial technology. This thesis will not seek to investigate the extent of these implications but yields an important conclusion: a portfolio has to be actively managed. The most

1 Samuelson (1965): p. 43

2 Farma (1970): p. 383

3 Lo (2004): p. 18

(6)

5 important reasons are the changing market behavior, and the advances in market research which will lead to improved tools in portfolio management.

Active portfolio management is a widely used concept where investors compare their investment performance to the market or a benchmark portfolio in order to determine whether their investment decision has yielded a higher return than either of these. Commonly applied benchmarks in active portfolio management are large and highly liquid indices such as the S&P 500 or the Dow Jones Index. In addition, the investment opportunities are usually limited to the underlying stocks of that benchmark categorized into sectorial based indices. The advantage of this approach is that the benchmarks underlying indices are likely to follow a somewhat similar return pattern as the overall market, making it less difficult to allocate portfolio assets. We will in this thesis deviate from this approach, as we apply the MSCI Denmark as benchmark and ten globally based sectorial indices as investment opportunities subjected to active portfolio management.

In order to assemble optimal portfolios, Harry Markowitz (1952)⁴ introduced the concept of efficient portfolio, which either optimizes the return of an asset or minimizes the risk of the asset for a given level of return. The concept is realized by diversifying assets in a portfolio, which is achieved by investing in a variety of different stocks that change differently in relation to each other – stocks with low covariance.

Therefore, as the results of this thesis are of a theoretical nature, the aim is to apply financial modeling of Markowitz’ modern portfolio theory, in order to solve the optimal portfolio construction problems. With regards to portfolio optimization another important topic is considered: Portfolio repositioning – the process of over- and underweighting portfolio assets on a periodic basis.

4 Markowitz (1952): p. 82

(7)

6

1.1 Research Objectives

The following research questions will be answered in this thesis with regards to active portfolio management.

1.1.1 Superior Research Objective

The research objective of this thesis is to devise an investment strategy by assembling a diversified portfolio with the aim of outperforming the MSCI Denmark by altering positions of portfolio assets on a short-term basis over a 20 year timeframe. The purpose of this investigation is to determine whether return generated from such strategy is warranted by its systematic market risk. In that regard, the aim is to determine whether active portfolio management is more attractive than high performing benchmark investing on a long-term basis.

1.1.2 Subordinate Research Objectives

In order to accommodate the superior research objectives, a portfolio will be constructed, subjected to active portfolio management and compared to the portfolio’s benchmark – the MSCI Denmark. In order to provide reliable results to support the conclusion, the following subordinate research questions must be answered.

1. How does the mean-variance portfolio model conduct asset allocation in the context of active portfolio management?

2. How does active portfolio management perform compared to MSCI Denmark between 1992 and 2011?

3. Does active portfolio management performance indicate investment skill on the part of the investor?

4. With regards to portfolio and benchmark systematic risk, does active portfolio management add value to the investor?

(8)

7

1.2 Structure of Thesis

The approach to this thesis is based upon the structure illustrated in figure 1.1.

Figure 1.1: Structure of the Thesis

Figure 1.1 divides the content of the thesis into four sections, each containing underlying chapters.

References made will be with regards to the underlying chapters. This section will continue by explaining the thesis methodological approach, the limitations imposed and applied data.

Section 2 will start by discussing of the concepts of strategic asset allocation and tactical asset allocation, in order to frame the investment strategy necessary for the purpose of active portfolio management. It will continue by providing theoretical considerations of determining and limiting the investment opportunities available for portfolio construction. Such limitation is deemed necessary in order for the mean-variance model to construct stable portfolios. Ten globally assembled sector Indices are considered for active portfolio construction in addition to the MSCI Denmark. The active portfolio management strategy attempts to outperform the MSCI Denmark by altering portfolio positions of the investment opportunities. Based on empirical and practical findings with regards to Strategic and Tactical Asset Allocation, a definition of active portfolio management is given and the importance of the definition of the benchmark is highlighted. Additionally, appropriate performance measurements will be presented.

• Chapter 1: Introduction

Section 1: Introduction

• Chapter 2: Investment Strategy

• Chapter 3: Active Portfolio Management

• Chapter 4: Return and Risk Management

• Chapter 5: Portfolio Construction

Section 2: Areas of Focus and Theoretical Framework

• Chapter 6: Portfolio Performance Evaluation

Section 3: Analysis and Evaluation

• Chapter 7: Conclusion

Section 4: Conclusion

(9)

8 The section continues by providing return estimates as input variables for portfolio construction. Based on theoretical and practical findings, issues regarding the risk active portfolio management are exposed to will be analyzed and evaluated. The section then concludes by applying the Markowitz mean-variance portfolio model for portfolio construction. Implementation possibilities with regards to portfolio repositioning are additionally presented and briefly discussed. The performance of the mean-variance model in the context of active portfolio management will then be examined. First in terms of asset allocation, i.e. whether the model has produced stable portfolios during portfolio repositioning. Second, return estimates will indicate whether outperformance is present on a long-term basis.

Section 3 will present the results of the active portfolio management process. From the portfolio strategy, the portfolio model presented in section 2 will have assembled portfolios based on the provided risk and return estimates. This section therefore answers the questions of whether active portfolio management has added value to the investment and whether these results are due to skill rather than luck.

Section 4 concludes by summarizing the answers to the research questions.

1.3 Methodology

1.3.1 Financial Assets and Risk

An enduring element regarding portfolio performance concerns the relationship between risk and return on investments. A financial investment, in contrast to a real investment, which involves tangible assets such as land and production facilities, is an allocation of money whose value is supposed to increase over time. Therefore a security is a contract to receive prospective benefits under stated conditions like stocks and bonds.

The two main attributes that distinguish securities are time and risk. Usually the interest rate or rate of return (depending on whether the security is a bond or stock) is defined as the gain or loss of the investment relative to the initial value of the investment. An investment always contains some sort of risk, categorized into two types – systematic and unsystematic risk, and the higher such risks the higher return is demanded by investors.

(10)

9 Financial assets are divided into two categories; traditional and alternative investments. Figure 1.2 summarizes the considered assets.

Figure 1.2 Financial Assets

The main traditional assets are cash, fixed income, equities real estate and foreign exchange. Cash is assumed to be stored in a bank account, yielding an interest rate often referred to as the risk free rate.

Fixed income securities, are government or company issued securities with a year to maturity, issued with the purpose of managing short-term cash needs. Two important money market interest rates are the London Interbank Offered Rate (LIBOR), which is the interest rate at which large banks in London lend money to each other. The other interest rate of importance is the Treasury Bill.

The long-term borrowing needs of corporations are met by issuing bonds. A bond contract provides periodic coupon payments and the principal value at maturity of the bondholder⁵.

Stocks are issued by corporations which convey rights to the owners, as they can elect the board of directors and have a claim of the earnings of the company. The shareholders are compensated with cash dividends, whose amount is determined by the company’s Management and Board of Directors. When referring to stocks, publically traded stocks are considered, which are regulated by governments. The process of arranging the sale of private stocks to the public market is called an IPO (Initial Public Offering).

In contrast, private equity refers to stocks held by private individuals or organizations.

5 http://www.investopedia.com/terms/b/bond.asp#axzz2CCmQC8CG

Traditional Investments Alternative Investments

Cash Hedge Funds

Fixed Income Managed Funds

Equity (stocks) Private Equity Funds

Real Estate Physical Assets (comodities, art etc) Foreign Exchange Securitized Products (debt obligations etc) Source: Own Creation

(11)

10 Real estate investments and foreign exchange derivatives are also found in portfolios, the latter for hedging against currency risks. Alternative investments emphasize the widening spectrum of investment.

These types of investment are beyond the scope of this thesis.

Risk and return obviously depends upon the type of investment. In order to measure investment return on a frequent basis we turn to equities, or stocks, as return can be realized at any point in time, preferable for an investor who favors frequent trading. The risk of such investment is presented when positive or negative return is realized upon trading. Hence, emphasis on risk and return is important to consider, particularly in the context of active portfolio management, as the investor takes on the risk of frequently realizing positive and negative returns.

The emphasis of the relationship between risk and return is reflected in the applied theoretical models.

These models are described below. As the research objective is based upon a practical exercise it important to also include real-life methods and applications with regards to investment strategies.

Therefore, the methods and information provided by selected theoretical concepts applied is supplemented by the input of two investment professionals. Their contribution to this thesis is described subsequently to the applied theoretical approach.

1.3.2 Applied Theoretical Approach

Throughout the analysis the following theories will assist in performing active portfolio management.

The framework of the investment strategy will be established by the concepts of strategic and tactical portfolio management. The investment opportunity set will be limited and the relevance of the benchmark discussed.

In order to apply the mean-variance portfolio model, expected return estimates and risk measures for each of the sector Indices suitable for the purpose of active portfolio management will be outlined. The former will be calculated by use of the Capital Asset Pricing Model (CAPM) model. The advantage of applying the CAPM model is that we can separate the market risk, beta and expected market return, of each sector index, and in that manner compare risk and return of sectors and benchmark.

(12)

11 Portfolio construction will be conducted by Markowitz’ mean-variance portfolio model as it optimizes asset allocation in Excel based on performance ratios appropriate for the purpose of the investment strategy⁶. Thus, optimizing with respect to the information ratio, which indicates relative benchmark performance with regards to its residual risk, alpha, and residual return, beta, will lead to asset allocations expected to provide the portfolio with a return superior to the benchmark. The asset allocation will be conducted as a repetitive process.

1.3.3 Interviews

In order to limit the amount of assumptions made and to provide a practical point of view when processing empirical data and applying theoretical models, interviews were conducted in order to uncover relevant areas of study. Also, practical reviews were applied, as some exercised concepts of investment

theory carry universal definitions or conditions for execution.

Empirical and practical data have been obtained through interviews and consultation with two investment professionals, who provided relevant information, relating to the research objective.

 Peter Sjøntoft, Vice President, Global Banking, Citigroup Global Markets Limited, London United Kingdom

Peter Sjøntoft has provided his personal viewpoints upon issues where elements of investment theory seem difficult to apply in real-life investment strategies. Furthermore, he has provided suggestions of where to challenge the application of these theories. He has assisted in adding practical viewpoints in establishing the framework for active portfolio management, and discussions regarding the use of investment strategies.

 Claus Vorm, Senior Portfolio Manager, Nordea Investment Management, Copenhagen, Denmark Claus Vorm has been responsible for establishing a team of investment specialists in charge of managing Nordea’s tactical asset allocation, as well as the banks quantitative products.

Furthermore, he has been responsible for managing various balanced portfolios.

Claus Vorm has contributed with insights into the use of the strategic and tactical asset allocation processes in Nordea’s investment strategies, which will be included when forming the investment

6 Benninga (2008): p. 338

(13)

12 strategy for active portfolio management. His contribution also extends to views upon risk management.

Peter Sjøntoft and Claus Vorm were chosen with the purpose of providing a nuanced representation of opinions concerning investment decisions, as well as contributing with informational groundwork for the research objective. Their experiences and information has enabled continuous tightening of the research objective. The interview with Claus Vorm was recorded and stored on the CD attached. Due to geographic differences, interviews with Peter Sjøntoft were conducted by phone, so no conversation was recorded, and therefore no specific citations made. For interview guide, see Appendix 3.

1.4 Assumptions and limitations

In order to maintain a clear focus throughout the thesis it is necessary to define some assumptions and limitations.

1.4.1 Assumptions

The theoretical discussion and the practical calculations with regards to the portfolio construction, will take the point view of a Danish investor who has funds available for investments. In that regard differences between investments by pension funds and private funds, including e.g. tax considerations and consequences will not be discussed.

All calculations are conducted on the basis of monthly data, stated in US dollars from January 1992 to December 2011. Data have been extracted from Datastream, MSCI Barra and Statistikbanken. Monthly observations are opted for as opposed to e.g. daily observations as the former provides clear and adequate information with regards to the development of the index prices and for sufficient data management. The time period is considered appropriate as it provides a sufficient amount of data and it covers significant economic events, affecting the financial markets.

With regards to the issue of transaction costs, we will assume that all transactions have equal expense over the entire period. In addition, in order to limit the constraints to the mean-variance portfolio model,

(14)

13 the transaction costs will be deducted from the return on investment subsequent to portfolio repositioning.

Finally, I will assume that markets are not completely efficient⁷. However, I will assume that the investor is rational suggesting that he prefers more to less. Although market efficiency rests on the assumption that investors are rational, its validity is not undermined by the investor not being rational, given he does not trade randomly. In addition, all hedging of currency is completed by future contracts, meaning that only index movement excluding currency behavior is considered. Moreover, the analysis will not incorporate any tax effects. This is mainly because of the complexities of the Danish tax system. Furthermore, the introduction of such a system is in conflict with the conditions of portfolio theory that leads to saying that the investor should buy the market portfolio.

1.4.2 Limitations

1.4.2.1 Theoretical framework

In portfolio theory there are several models and applications appropriate for portfolio construction. Both Markowitz (1952) and Black Litterman (1992) propose portfolio models applicable for portfolio construction. The thesis will apply Markowitz’ mean-variance portfolio model, but argumentation and justification for the choice of model will be provided. The model will be explained superficially, and derivations will not be included when explaining the model, as I intend to present the model in the most comprehensive manner possible.

Short sales will not be introduced throughout this thesis. The reason for this is that the mean-variance model tends to incorporate extreme values in the asset positions when short sales are included providing portfolios of poor applicability. Another reason is that major stock exchanges have unique short sales regulations⁸. Furthermore, financial gearing is prohibited. Both assumptions contribute to portfolio robustness, meaning altering investment positions are comparable with changes in return and covariance estimates.

7 Shleifer (2000): p. 3

8 http://www.sec.gov/spotlight/keyregshoissues.htm

(15)

14 1.4.1.2 Data

In order to construct a globally representative portfolio ten sector indices have been chosen, as they represent a significant proportion of the world stock market. The indices are presented in chapter 1.5. The MSCI Denmark and MSCI World Market indices are obtained from MSCI Barra⁹.

Before selecting the sector Indices I examined their historical returns, variances and covariance along with the correlation and size of their market capitalization (see Appendix 2). The examination showed a moderate pattern of a positive risk-return relationship among the sectors. Some of the sectors with the highest risk-return payoff even had some of the lowest correlations towards other markets. The general picture, however, showed high correlations among sectors leaving only a few inter-correlations below 0,5 suggesting high integration among sectors (only Technology showed correlation below 0,5). High market integration is not an attractive property from perspective of investment theory as it constructs portfolios based upon high return estimations and low covariance, hence low correlations. However, this scenario highlights whether one major advantage in portfolio management, diversification, can provide the portfolio with a higher risk adjusted return compared to the benchmark.

1.5 Data

In this chapter the definition and source of the equity sector return applied in the thesis are presented.

1.5.1 Equity Sector Return Data

There is a large industry of providing investors with benchmark data of sectors. They use Global Industry Classification Standards (GICS) which operates with a ten sector classification (MSCI Barra, 2010): Energy, Materials, Industrials, Consumer Discretionary, Consumer Staples, Healthcare, Financials, Information Technology, Telecommunication Services, Utilities. These sector groups can again be classified into 24 industry groups, 68 industries, and 154 sub-industries¹⁰.

Equity return data in this thesis is obtained from Datastream and Morgan Stanley Capital International (MSCI), since they offer data on world sector return covering the needed time periods. With regards to

9 http://www.msci.com/products/indices/country_and_regional/all_country/performance.html

10 MSCI Barra (2010): p. 82

(16)

15 Datastream, within each market, stocks are allocated to industrial sectors using the Industry Classification Benchmark jointly created by FTSE and Dow Jones. The groups are formed from stocks registered on the largest equity market in 53 countries and all sector Indices consist of 6 sub-indices. The data set divides the market into 10 sector classifications (somewhat similar to the CICS): Basic Materials (BMATR), Consumer Goods (CNSMG), Consumer Services (CNSMS), Financials (FINAN), Healthcare (HLTHC), Industrials (INDUS), Oil & Gas (OILGS), Technology (TECHN), Telecommunications (TELCM) and Utilities (UTILS)¹¹. The basis for the selection of these sector Indices are based on a preference for global representation, which is discussed further in chapter 2. Table 1.1 provides summary statistics of the ten sector Indices along with the MSCI Denmark, based on monthly observations over the time period 1992- 2011.

Table 1.1: Summary Statistics – Monthly Return

Table 1.1 indicates that Oil & Gas and Technology has provided the highest average monthly return over the period. Technology and Basic Materials have been the most volatile sectors with a standard deviation

11 Datastream (2008): p. 3

Mean Std.dev. Max. Min. Obs

Panel A. Sector Indices

Basic Materials 0,43% 6,22% 18,97% -19,65% 240

Consumer Goods 0,47% 4,63% 12,62% -19,11% 240

Consumer Services 0,35% 4,36% 10,54% -30,06% 240

Finance 0,21% 5,64% 17,85% -12,11% 240

Healthcare 0,48% 3,44% 8,95% -27,41% 240

Industrials 0,47% 5,45% 15,70% -26,09% 240

Oil & Gas 0,67% 5,66% 15,53% -31,10% 240

Technology 0,69% 7,58% 19,98% -17,25% 240

Telecommunications 0,33% 5,31% 16,18% -16,89% 240

Utilities 0,26% 3,84% 8,74% -29,67% 240

Panel B. Benchmark Index

MSCI Denmark 0,65% 5,92% 16,79% -29,67% 240

Panel C. Market Index

MSCI World 0,34% 4,47% 10,35% -21,13% 240

Summary Statistics

Source: Own creation, Datastream, MSCI Barra

(17)

16 above 7% and 6%, respectively. Note, that market and benchmark indices are two different indices. When referring to the benchmark, it is the MSCI Denmark targeted for outperformance, while referring to the market relates to the MSCI World index. The latter will only be applied for return calculations and systematic risk estimations. From Appendix 4 all indices are concluded to be stationary indicating that positive monthly returns are likely to be followed by negative return and vice versa, which in fact complicates active portfolio management, as it makes the decision of market timing rather difficult. This problem will be addressed with regards to portfolio construction in chapter 5.

Within the literature of stock returns there is no general convention regarding the use of simple and log- returns. For example Campbell and Thomson (2008) use the simple mean approach while Benninga (2008) argues log-returns, which is marginally more precise when modeling assets in Excel¹²¹³. Therefore, the monthly return data is calculated as log-returns, and these returns are applied throughout this thesis. The simple and log-return methods were tested on a few sector indices and only small differences were found.

Thus, we do not expect the conclusions to be significantly different if simple returns had been used instead.

Equation 1.1 shows the total return of sector i held from time t-1 to time t. The total return constitutes the dividends and capital gains. Di,t is the dividend from sector i received by the investor during period t.

Dividends for sector indices were incorporated in the return data upon extraction from Datastream.

) 1 . 1 ( log

1 ,

, ,

, 









 



 t i

t i t i t

i P

D R P

The numerical difference between simple and log-returns are usually small for high frequency data. Both concepts have their advantages and disadvantages in terms of portfolio and time aggregation. The reader is referred to Tsay (2001) for more details¹⁴.

12 Campbell, Thomson (2008): p. 1517

13 Benninga (2008): p. 503

14 Tsay (2001): p. 3

(18)

17 The excess return is defined by the return in excess of the risk free asset.

) 2 . 1

, (

,

, it ft

e t

i R R

R  

Here, R_i^e_,_t is the monthly excess return of investment i at time t, R_i_,_t is the return of investment i at time t, and R_f_,_t is the monthly return of the risk free asset at time t.

Active return is defined by the excess return of an asset or portfolio in excess of the benchmark return at time t.

) 3 . 1

, (

,

,t it Bt

i R R

R  



Here, R_i_,_tis the active return and R_B_,_tthe benchmark return at time t. Substituting the right hand side of equation 1.2 into 1.3 and subtracting the risk free rate from the benchmark return yields the following calculation for active excess return:

 

t B t i e t

t f t B t f t i e t

R R R

R R R R R

, ,

, , ,

, (1.4)













From equation 1.4, active return is identical to active excess return.

(19)

18

1. Investment Strategy

In order to frame the concept of active portfolio management a specified investment strategy is required. Investors buy and sell stocks based upon individual incentives such as a desire for abnormal investment returns or because they are constrained by consumption and liabilities. Hence, there is no single approach defined to actively manage a portfolio as the allocation of assets differs depending upon the investment objective. Thus, an investment strategy appropriate for answering the research objective is called for. We will apply the concepts of strategic and tactical asset allocation, two debated methodologies to portfolio management, in order to establish a framework for such strategy¹⁵.

2.1 Investment Strategy as the Source of Portfolio Performance

The accepted advice about spreading investments on several asset classes in order to minimize risk, and not let emotions or gut feelings control the continuous portfolio allocation process, has become common knowledge. Nevertheless, the opinions and expectations of the individual investor still dominates the decision making process when they assemble portfolios¹⁶.

According to Peter Sjontoft, identifying an investor who has managed to outperform a benchmark is not necessarily rare or difficult, when selecting among investors who base their investment decisions on gut feelings, emotions or even randomly select stocks to invest in. On average, these investors may in fact outperform a given benchmark, not because they are smart investors with extensive market insights, but because the stocks they invest in simply happen to perform better than the market over the given timeframe. Berk (2005) supports this statement, claiming that little evidence of performance persistence exists – investors that perform well in one year are no more likely to perform as well next year¹⁷. One possible explanation for this is that markets are, as in this case stationary, which opens the possibility for investor lucky as they happen to buy stocks when markets are down and sell with a positive return. Such scenario promotes the assumption that investor performance can be due to luck rather than skill and that active portfolio management could be just as successful as random portfolio management – in an

15 Schneeweis et.al. (2010): p. 91

16 http://www.proinvestor.com/finansnyhed/10294074/Fodboldaktier-koebes-som-legetøj

17 Berk (2005): p. 30

(20)

19 individual case, not on average. The investor who chose e.g. Novo Nordisk or Apple five years ago would likely have outperformed most indices¹⁸. The question is really if someone who over-performed was lucky or skilled, i.e. is his chances of repeating the performance next year higher than average? On that basis the investor’s management skills is put into context as this constitutes the performance difference between passive and active portfolio management. In that regard active portfolio return stem from the investment strategy. The investment strategy constitute the process of asset allocation and security selection, as these determine what to invest in at which point in time in order to generate positive active return.

2.2 Strategic Asset Allocation

Strategic asset allocation is a static approach to asset allocation where the portfolio assets are allocated based on a long-term return estimate. Asset positions are retained throughout the period, meaning that as asset returns change their portfolio weights change correspondingly, and the investor then rebalance the portfolio by adjusting asset weights back to their initial positions¹⁹. Once the assets have been selected for portfolio construction, no other asset will be introduced at any time.

Markowitz’ extensive work on the modern portfolio theory has yielded a dimension to the asset management theory, i.e. diversification. This reward emerges when assets, which correlate negatively or independently with each other, are assembled into a portfolio²⁰. Strategic asset allocation sets forth long- term allocations for assets possessing such characteristics.

The allocation process relies on the assumption that markets are efficient, which makes it impossible to obtain abnormal return on investments, and therefore makes market timing irrelevant. On the contrary active trading will increase transaction costs, which reduces the realized return more than expected. In order to illustrate the argument against market timing, Ibbotson Associates made the following analysis²¹:

18 http://www.euroinvestor.dk/boerser/nasdaq-omx-copenhagen/novo-nordisk-b/205365 and http://www.euroinvestor.dk/boerser/nasdaq/apple-inc/38687

19 Schneeweis (2010): p. 99

20 Markowitz (1959): p.102

21 Spar Invest (2007): p. 17

(21)

20 Figure 2.1 indicates that an investors who invested 1 dollar in the S&P500 in 1926 index and then did nothing, has achieved a remarkably higher return as opposed to the investor attempting to time the market and therefore unfortunately has failed to invest during the 35 best months of the period. As a result, attempting to time the market yielded a return on approximately the same level as T-bills.

Brinson et.al. (1986) conducted an empirical analysis which investigated the extent to which the investment policy, the market timing and the selection of specific assets, affect the deviations in the total portfolio return. Based on historic data from US pension funds they investigated which investment decision had the greatest influence on the observed return and its volatility.

The first studies were conducted based on data from 1974-1983 and they concluded that a better return could be achieved by solely focusing on a passive strategic asset allocation approach²². By passive is meant retaining asset position by portfolio rebalancing. Their calculations showed that portfolio managers who tried to time the market or buy specific stock achieved a lower return compared to the passive investment strategy. These conclusions emphasize that asset classes with weights comparable to a benchmark contribute with the largest proportion of the total return of the portfolio. In other words Brinson et.al (1986) argues for buying a market portfolio, with fixed asset positions which in the long-term will outperform any other portfolio combination.

22 Brinson et.al (1986): p. 136

2306,45

18,83 26,9

0 500 1000 1500 2000 2500

S&P500 T-bills S&P500 excl. the best 35 months

Figure 2.1 By how much has 1 dollar increased in value from 1926 to 2006?

Source: Own Creation, Sparinvest (2007)

(22)

21 In addition to providing a better return, 93,6% of the deviations of return can be explained by strategic asset allocation²³. Therefore, as regards to long-term investments, literature has proved that in terms of returns, the best idea is to conduct portfolio investment, and not try to alter investment positions on a periodic basis. In addition, resources committed to the investment decision should be concentrated around the strategic decisions – the determination of the investment opportunity set - since the risk of the investment can be observed here.

In order to substantiate these results, Brinson et.al. (1991) conducted a new analysis on behalf of US pension funds during the period 1977-1987. They reached the same results as just discussed. For this period 91,5% of the standard deviation of return could be explained from strategic asset allocation. Only 1,8% of the return deviations was explained by tactical asset allocation - the process of over- and underweight portfolio sectors²⁴.

Aside from explaining a majority of return deviations, the strategic approach to asset allocation contributes to a better risk adjusted return in contrast to attempting to time the market. Such approach offers investors the possibility of achieving a higher return at a lower risk. This is in accordance with Markowitz (1952) portfolio theory, as he developed the efficient portfolio based on long-term historical data.

2.2.1 Discussing Empirical Findings and Brinson et.al.

The findings supporting the strategic asset allocation as the prevailing investment strategy assumes market efficiency, which makes it impossible to obtain abnormal returns, and thus makes market timing irrelevant. However, to believe investors have the same information available and that the full use of this information is reflected in asset prices is a naïve assumption. Empirical findings within the area of behavioral finance have provided evidence that the assumption of a fully informed rational investor is rather bold. Grossman and Stiglitz (1980) go even as far as stating that if markets were perfectly efficient there would be no profit from gathering information, in which case there would be no reason to trade, leading markets to eventually collapse²⁵. In order to adopt a strategy for active portfolio management,

23 Brinson et.al (1986): p. 137

24 Brinson et.al. (1991): p. 45

25 Grossman and Stiglitz (1980): p. 393

(23)

22 trading is thus required and although markets are stationary, we can therefore not assume full efficiency in the market, but refer to the adaptive market hypothesis instead.

Despite their conclusions, Brinson et al. (1986) have, however, received criticism for their analysis. Jahnke (1997) criticized their use of the variance as risk measurement as opposed to the standard deviation²⁶. If the variance were to be replaced with the standard deviation, the strategic asset allocation would instead explain 79% of the return deviations, which is not nearly as seminal as 93,6%.

In addition, the argument that risk of a passive investment strategy can be more easily identified is rather obsolete. Only if assets expected returns are constant over time, should the asset weights remain constant. The problem with such assumption is that stocks perform differently over different timeframes due to their stages in business cycles and changes in macroeconomic developments. Hence, investors assume changing risk premiums on stocks, leading expectations of asset prices, and as a result, their portfolio weight to change. On this basis, investors should consider asset allocation as a dynamic process which allows asset positions to change as their expected premiums change.

Among the critics towards Brinson et.al are also Statman (2000). He concludes that a portfolio manager who consequently invests in the appropriate asset classes (stocks, bonds and holds cash as well) every year between 1980 and 1997, achieves a return 8,1% per year in excess of passive strategic approach²⁷. In addition, 89,4% of the return deviations could be explained by the investment strategy, and the result is therefore not far from the analysis of Brinson et.al. (1986). His point is that the investor should not only focus on the strategic approach, but adopt a more active approach as well.

Regardless of whether the investor believes in the conclusions of Brinson et.al or joins the group of critics, there is no doubt that the opted investment strategy has major influence upon the portfolio return deviation. Strategic Asset allocation offers an attractive feature in the asset selection. It is fixed, which means determining a fixed opportunity is supported by literature and carries practical advantage for portfolio construction as we don’t have to dedicate resources to identifying new opportunities. With regards to asset allocation we turn to the dynamic approach of tactical asset allocation.

26 Jahnke (1997): p. 2

27 Statman (2000): p. 19

(24)

23

2.3 Tactical Asset Allocation

Contrary to strategic asset allocation strategy, another practice within asset allocation assumes a more active approach to asset management. In order to conduct a thorough investigation of whether active investment management can outperform a benchmark, a different strategy is considered. The main points of tactical asset allocation are outlined in the following.

Following a strategic asset allocation approach, changes in the portfolio will exclusively occur ex-post, meaning that the investor will modify his portfolio as a reaction to events occurred, by rebalancing the asset weight back to their initial target weights. The initial portfolio composition will therefore be altered as a result of general market developments. In periods of negative returns for some assets the portfolio must be rebalanced so the optimal proportion of assets is recreated.

However, asset allocation must be based on expectations of short term future returns in order to continuously ensure positive active return. In 1971 William Fouse launched the first index fund²⁸. His work made it possible to control different asset classes simultaneously, and this technique has later been known as tactical asset allocation. On that basis, compared to strategic asset allocation, tactical asset allocation is on the other hand an ex-ante investment strategy where the investor proactively adjusts his portfolio based on market historic developments and expectations²⁹. Thus, the dynamic nature of tactical asset allocation requires active adjustments to the investment opportunities in response to short-term changes in the economic environment. Its objective is to adjust the allocation in order to take advantages of temporary pockets of market inefficiency³⁰.

Contrary to Strategic Asset Allocation the investor has not determined an optimal asset distribution, but adjusts the portfolio in accordance with his expectations to the market development. Therefore, the essence of tactical asset allocation is to proactively position portfolio assets based upon changes in expected returns. Thus, assets with high expected returns are favored, as opposed to the remaining opportunity set, and assets with low expected returns are less desired. The advantage of the tactical asset

28 Lee (2000): p. 12

29 Picerno (2010): p. 153

30 Schneeweis (2010): p. 101

(25)

24 allocation and the main source of its popularity is that it combines Graham and Dodd’s value investment strategy together with Markowitz modern portfolio theory³¹.

The concept of value investing by Graham and Dodd has been an active investment strategy. Investors seek undervalued stocks with low price-to-earnings ratios. On the other hand, Markowitz considers investments within a determined timeframe, and hence only the market portfolio can be considered a risky portfolio. Tactical asset allocation made it possible to combine these two strategies into one, which enables the possibility of active management and thereby a superior return.

2.4 Professional Views upon Asset Allocation

As described above there is no doubt that strategic asset allocation constitutes a relevant investment strategy and carries the advantage of a predetermined long-term investment universe. However, from the viewpoint of the active investor, tactical asset allocation also seems a suitable investment strategy as it proactively repositions the portfolio on a regular basis in accordance with the investor’s expectations. The extent to which investors follow the strategic or tactical approach or even both when constructing optimal portfolios remains unknown. The following will therefore include a review of Claus Vorm’s assessment upon the process of asset allocation and highlight the way Nordea Investment Management applies portfolio theory.

In relation to balanced portfolios, the initial asset allocation is based on long-term investment decisions consistent with strategic asset allocation. The asset allocation process begins by considering levels of equilibrium for interest rate structures, inflation rates etc. From these estimations Nordea attempts to determine the long-term risk premium on each selected stock. The equilibrium expectations enables expected return calculations over a strategic default period, which is within a business cycle of approximately ten years.

Nordea then starts optimizing, by allocating funds into the stocks which are considered below long-term equilibrium – undervalued stocks. On that basis it is reasonable to form a strategy that considers the state of the economy compared to long-term equilibrium. Claus Vorm considers this a strategic analysis,

31 Value investment involves investing in stocks with low P/E ratio

(26)

25 valuable for creating highly diversified optimal portfolios based on estimated expected return and covariance among assets. Obviously, as other portfolio models, this method carries pitfalls, but Claus Vorm believes that the results of these models combined with the investor’s common sense can create strong diversified portfolios.

On the other hand, Claus Vorm also believes advantages can be utilized in tactical asset allocation. Stocks may be underpriced as considered in the strategic asset allocation, but they may have to be priced even lower before returning to equilibrium, hence Nordea will have to purchase the stock e.g. two months later.

The important conclusion here is that both strategic and tactical asset allocation will not necessarily have to be present in the same investment strategy. We implement a long-term investment strategy with changes in short-term portfolio positions, within the boundaries of predetermined risk parameters. This approach can add value in terms allocating funds on a strategic long-term basis and reposition the portfolio on a tactical basis. Hence, both the strategic and tactical approach is implemented in a balanced portfolio, as the portfolio assets are repositioned on a regular basis. Claus Vorm suggests a useful strategy would be to balance assets within determined boundaries, e.g. portfolio assets can be reweighted ±20% of their initial portfolio weight each month.

Strategic and tactical asset allocation will both be implemented as part of the investment strategy, in balanced capacities. Strategic asset allocation offers advantages in terms of limiting the investment opportunities which combined can be submitted to periodic repositioning in accordance with tactical asset allocation.

2.5 Benchmark and Investment Opportunities

The choice of benchmark is of significant importance, as its performance obviously plays a key role in determining the success of the investment strategy. Additionally, the investment opportunities must be given careful consideration as combining such should constitute a more attractive investment compared to the benchmark.

(27)

26 2.5.1 Benchmark

Rationalizing the choice of benchmark, the Danish index has been considered a “Safe Haven” investment, which it has proven by retained and even increased returns relative to other major indices in market turbulences making it a high performing index difficult to outperform³² ³³. Thus, it remarkable performance relative to other markets makes successful active portfolio management a challenging task.

Figure 2.2 illustrates its performance against other major national economies.

When applying a small index as benchmark such as the MSCI Denmark, the investor must consider the case of corporate domestic market dominance in terms of market capitalization. In writing, Novo Nordisk, constitutes almost half of the basis for the Danish OMXC20 index putting the applicability of MSCI Denmark as a representative benchmark into question³⁴. On the other hand, the same scenario is likely in other small or midsize European markets, where few or a single company possesses major influence upon index development, due to superior market capitalization.

It is evident that MSCI Denmark has performed well relatively to other countries, but for Denmark, high returns have been followed by high losses, by example of the return development between 2006 and

32 http://borsen.dk/nyheder/investor/artikel/1/229201/udlandet_kaster_sig_over_c20-aktier.html

33 http://borsen.dk/nyheder/investor/artikel/1/221690/udenlandske_investorer_vaelter_ind_over_landets_graenser.html

34 http://borsen.dk/nyheder/investor/artikel/1/240054/novo_nordisks_dominans_mere_end_fordoblet.html 0

100 200 300 400 500 600 700

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Index

Figure 2.2 Selected Country Indices (92M1=100)

MSCI Denmark MSCI France MSCI Germany

MSCI UK MSCI USA MSCI World

Source: Own Creation, MSCI Barra

(28)

27 2009. Thus, from its historical performance we can expect the index to carry high market risk – beta. Peter Sjøntoft suggests that investing in high beta stocks is indeed one way to achieve high return, but as these returns comes with high risk, investing in low beta stocks with low returns is not an unwise investment decision. The MSCI Denmark appear as a difficult benchmark to outperform in terms of realized return, but relatively low return on investments therefore constitute an equally successful investment strategy with regards to adding portfolio value, given the condition that low returns are warranted by correspondingly low market risk. From Claus Vorms comments with regards to portfolio construction, value added stem from selecting stocks with different levels of systematic risk, thus different levels of return, and based upon these input data construct portfolios carrying systematic risk at a level justified by the realize a return. The stock selection is determined and limited by the investment opportunities.

2.5.2 Investment Opportunities

Active portfolio management commonly involves altering portfolio asset positions and compares the portfolio performance to the overall market, such as S&P 500 or Dow Jones, which constitutes many international corporations. Thus, the stocks included in the portfolio and benchmark are identical and potential outperformance occurs by altering asset positions in the portfolio. As the Danish market is relatively small in terms of capitalization, it represents few companies dominating their respective industries and in some cases the entire market. To illustrate the point consider the correlation between MSCI Denmark and four of the largest companies residing in the market.

Table 2.1 Correlation between MSCI Denmark and Four Underlying Stocks

The high level of correlation and market capitalization of these companies suggests that if changes in their respective industries or even the market occur, such change is likely attributable to these stocks. Selecting the underlying stock indices of the MSCI Denmark makes active portfolio management a difficult task indeed, as diversification opportunities are narrow, and the investor might be tempted to select only few stocks as those above, as their development have great influence on the benchmark development.

Broader investment opportunities for active portfolio management are therefore required.

A.P.Moller Maersk B Carlsberg B Danske Bank Novo Nordisk B

MSCI Denmark 0,89 0,83 0,79 0,85

Source: Own Creation, Datastream, MSCI Barra

(29)

28 2.5.2.1 Seeking International Opportunities

A variation of different assets will provide the investor with a variability of return in the portfolio and reduces its risk. In order to achieve portfolio optimization the investor must allocate different asset classes to the portfolio. According to Litterman (2003), to find independency among assets, and hence reduce the risk of the portfolio, the investor must consider different markets³⁵. In addition, Markowitz (1959) claims that in order to achieve optimization at an international level, it is important to consider different types of sectors and industries³⁶. Hence, the most beneficial way to allocate assets is to invest globally. Portfolio optimization traditionally requires independency among a diversified selection of assets, thus allocating assets in different types of sectors and industries seems beneficial to achieve optimization.

When considering an investment in common stock, investors tend to divide the vast universe of stocks into categories based on general business lines and by industry within these business lines. One way of dividing this universe of stocks gives classifications for industrial firms, financial institutions and other industry sectors. An alternative classification scheme separates US and foreign common stocks. I have avoided the latter division, because the industry breakdown is more useful when constructing the portfolio of global common stocks. When stocks are sorted by industries rather than geography they constitute a more comprehensive illustration of company-level performance. It is easier to identify which stocks are performing better than others as they not only react to global macroeconomic events, but also individual sectorial events. Therefore, return patterns are assumed to develop somewhat differently, which enables the identification of diversification opportunities. With a global capital market the focus should include all the companies in an industry viewed in a global setting. Extensive market integration advocates for this decision as it has diminished the importance of the geographic location of major companies, while their sectorial location is better characterized.

2.5.2.2 The Rise of Sector Importance

Obviously, the asset allocation process refers to the decision process of determining the amount of funds that should be allocated to each financial asset in the existing opportunity set. It is the investor’s objective to obtain the highest risk adjusted return as possible. Our interpretation from Brinson et. al (1986) showed that the asset allocation decision is by far the most dominant factor of portfolio performance as it explain

35 Litterman (2003): p. 137

36 Markowitz (1959): p. 89

(30)

29 more than 91% of the variation in asset returns. Furthermore, Litterman (2003) suggestions that asset allocation can be divided into two different types of decisions 1) asset allocation between different asset classes, e.g. stocks and bonds and 2) asset allocation within one asset class, e.g. countries and sectors.

Hopkins and Miller (2001) analyzed these dimensions of global equity portfolios (countries, sectors, industries and companies). Their analysis concluded that a significant shift seemed to have occurred in the importance of global sectors and industries at the expense of geography in global investment strategies³⁷. Although this emphasis is likely to shift through time the reward for global sector allocation, as well as organizing stock selection on a sectorial basis, seems to justify allocating resources to sector research.

Their tests further suggested that an industry group orientation, rather than a broader sector-level orientation can add value to asset allocation research. However, in order to maintain track of an affordable amount of data in addition to a global representation among the investment opportunities, further investigation of subgroups below the sectorial level will not be considered in this thesis.

Considering the advice of broad diversification from Claus Vorm, this thesis seeks to allocate assets into a portfolio, which represents a broad global market portfolio, which is solid and easy to measure. Each asset represents a sector index which embodies a range of industries. As introduced in chapter 1 the further analysis considers the Datastream World Indices with ten sub-indices, which have been chosen to represent ten major industry sectors as introduced in chapter 1.5.

2.5.2.3 Market Integration

Only a few decades ago some national markets were considered difficult to gain access to, limiting investment opportunities to only domestic markets. As globalization has integrated national economies, companies, consumers and investors now trade across borders, which results in increasing correlation between markets. Hence, systematic risk arises as economic development in some countries impact that of others, which to some extend align their return patterns, as they are affected by the same developments. However, based on historical performance MSCI Denmark stands out as having performed remarkably better than other major indices. In terms of benchmark comparison, both advantages and disadvantages for applying the MSCI Denmark as benchmark exist. Table 2.2 shows the correlation between each sector index, Denmark and four major national economies.

37 Hopkins and Miller (2001): p. 1

ACTIVE PORTFOLIO MANAGEMENT AND PORTFOLIO CONSTRUCTION- IMPLEMENTING AN INVESTMENT STRATEGY