• No results found

Performance analysis of Dutch listed equity funds

N/A
N/A
Protected

Academic year: 2021

Share "Performance analysis of Dutch listed equity funds"

Copied!
34
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Performance Analysis of Dutch Listed Equity Funds

Bachelor Thesis

Abstract

This research provides empirical tests for abnormal performance of Dutch listed equity funds between 2002-2014. Two periods were constructed; before and after the start of the financial crisis in 2008. A dataset of 29 equity funds, divided in three investment groups, was evaluated using five performance measures. Funds investing in European equity were most severely affected by the financial crisis as appears from significantly deteriorating abnormal returns. In addition, the Fama & French Three Factor model seems to be the best model for performance evaluation of the funds included in this research.

Student: Martijn Teunissen Student number: 10191054

Supervisor: L. Zou

Specialization: Finance & Organization Research Field: Asset Pricing

(2)

2

Contents

1.

Introduction ………..……… 3

2.

Literature ……….. 4

2.1 Efficient Market Hypothesis ……… 4

2.2 Capital Asset Pricing Model………... 5

2.3 Performance measures ... 6

2.3.1 Sharpe Ratio ... 6

2.3.2 Treynor Measure ... 7

2.3.3 Jensen’s Alpha... 7

2.3.4 Fama & French Three-Factor Model ... 8

2.3.5 Carhart Four-Factor Model ... 9

2.4 Previous empirical research ... 10

3. Methodology and data ... 12

3.1 Methodology ... 12

3.2 Data ... 13

4.

Empirical Results ... 16

4.1 Sharpe Ratio ... 16

4.2 Treynor Measure ... 17

4.3 Jensen’s Alpha ... 18

4.4 Fama & French Three-Factor Model ... 19

4.5 Carhart Four-Factor Model ... 20

5.

Discussion ... 21

5.1 Results ... 21

5.2 Limitations and further research ... 22

6.

Conclusion ... 24

7.

References ... 25

(3)

3

1. Introduction

The attractiveness of investing in Dutch mutual funds seems to rise which is reflected in a 6.6% growth rate of investment size up to €739 billion in the third quarter of 2014 according to figures of the Dutch Central Bank. Actively managed mutual funds investing in equity represent a large category. These funds typically take on higher risks and therefore should yield a higher rate of return compared to the market in theory.

The performance of equity funds is an extensively researched subject in financial literature. Most of these researches have focused on funds investing in the US stock market. On the one hand, studies suggest that these managers are able to reach outperformance due to superior skills. Others oppose this statement and assert that, especially after deducting the related costs, the funds perform equal to the markets. The main purpose of this research will be to provide new empirical evidence from Dutch listed equity funds. The selected funds will be examined on their ability to reach abnormal returns, both positive and negative, compared to the market. In particular, the most recent available data will be used. Because the period included the financial crisis starting in 2008, the sample was split in a period prior to this crisis and a period after.

Another aspect that will be considered in this research is the evaluation of the models used. Five widely used performance measures will be examined on their explanatory power. It should be noted that this examination employs positive testing as opposed to normative testing (examining the assumptions). Afterwards, one can conclude which model appeared to be the most empirically valuable in explaining the returns of this particular group of interest.

(4)

4

2. Literature

In this section the theory underlying the performance measures will be considered. In addition, five commonly used performance measures in academic research will be discussed. The section ends after an extensive review of the empirical results regarding these performance measures.

2.1 Efficient Market Hypothesis

The Efficient Market Hypothesis (EMH) can be regarded as a very influential economic theory concerning research on asset pricing and performance evaluation. Based on earlier research, Fama (1965) was the first to define an “efficient” market as a liquid market in which securities reflect all available information. The assertion behind this is that heavy competition among well-backed business analysts on security information would eliminate positive Net Present Value opportunities ensuring that asset prices are at their proper levels. A second implication of this theory is that active traders should not be able to outperform the market. Therefore, security prices can be seen as reliable guides in allocating real capital where investors should obtain an equilibrium rate of return given the riskiness of their portfolios (Bodie et al., 2011).

Three versions of the EMH are distinguished: the weak, strong and semi-strong forms. The fundamental differences between these forms lies in their notions on how the assumption “all available information” is interpreted. The weak form poses that all data derived from historical data (i.e. technical analysis) is reflected in the prices of securities (Fama, 1970). This view does not exclude that one can achieve outperformance by fundamental analysis. The strong form is more extreme stating that all relevant information, including insider information, is reflected in prices which implies that no one can achieve consistent outperformance. Finally, the last form states that both technical and fundamental analysis are already discounted in prices. The only way to achieve outperformance in this form is by possessing foreknowledge.

Despite its dominant role in the theory of Finance, it has been widely criticized, mainly on its underlying assumptions. Grossman and Stiglitz (1976) oppose the assertion that all relevant information determining security prices is available free of charge. They argue that if markets were efficient, investors would have no incentive to uncover the information which is assumed to be reflected in the prices. Consequently, investors would not trade on the basis of information implying that financial markets cannot be perfectly efficient. This is

(5)

5

what came to be known as the Grossman-Stiglitz paradox. In their model, financial markets are close to efficient where well-informed investors are seeking to exploit mispricing in the short term. The Arbitrage Pricing Theory (APT) model of Ross (1976) extents on this idea that arbitrageurs and speculators play an influential role by gathering costly information in order to correct for mispriced securities.

Another major criticism regards the assumption that investors have rational expectations and always try to maximize their utility. Proponents of the behavioral finance discipline assert that this assumption not realistic and markets are not perfectly efficient as a result of psychological biases in the trading behavior of investors (Bodie et al., 2011). Various researches including De Bondt and Thaler (1990) have investigated these biases and concluded that irrational behavior by investors provide opportunities to achieve outperformance.

Fama (1991) acknowledged that financial markets are not perfectly efficient, but close to efficient. In his EMH model he made adjustments that allowed more space to investors that actively exploit short-term mispricing in order to regain efficiency. The last decade, Lo (2004) proposed the Adaptive Market Hypothesis (AMH) which reconciles the EMH with the critics of the behavioral finance discipline. In this theory, evolutionary concepts as competition, innovation and natural selection are incorporated. These are used to explain why investors have different views on the market developments and adjust their expectations continuously to secure their ‘survival’ in the market. This contradicts the EMH assuming that investors have identical future expectations.

The degree to which extent markets should be seen as efficient is still subject of a fierce debate. At the one extreme, in the literature perfectly efficient markets is not seen as very realistic. On the other spectrum, completely inefficient markets would not be realistic either. The main issue in testing the validity of the EMH is that it is very complicated to construct a valuation model of securities that takes into account all possible information. Leaving undecided whether the EMH is correct or not, it is still the dominant theory for generating models that test outperformance by active management.

2.2 Capital Asset Pricing Model

One widely used model for analyzing the returns on risky assets is the Capital Asset Pricing Model (CAPM). Based on earlier work of Markowitz (1952), the model was developed in

(6)

6

articles by Sharpe (1964), Lintner (1965) and Mossin (1966). The model is constructed as follows:

Rit = Rft + βi * (Rmt - Rft) + εt (1)

Where Rit is the return of asset iat time t, Rft is the return on the riskfree asset at time t, βi is

thesystematic risk of asset i, Rmt is the return on the market portfolio at time t and εt is the

residual at time t.

The fundamental idea behind this model is that total risk facing investors is divided into systematic risk and idiosyncratic risk. Idiosyncratic risks are security specific risks that can be mitigated through diversification. Systematic risks are the market risks that cannot be diversified away. After diversification, the latter will be the only risk incurred by the investor and is captured by β in the model.

2.3 Performance measures for mutual funds

The need to measuring an asset’s performance has led to the development of sophisticated models, of which the most widely used originate from the basic CAPM model. This section presents five widely used performance measures.

2.3.1 Sharpe Ratio

A fairly basic performance measure is the Sharpe ratio. Introduced by Sharpe (1966) it was defined as the “reward-to-volatility ratio” measuring a portfolio’s average excess return per unit of total risk. The Sharpe ratio is defined as:

S =

𝑅𝑝σ-𝑅𝑓

𝑝 (2)

Where Rp is the return on a portfolio, Rf is the risk-free rate and σp is the standard deviation

on the portfolio.

As it adjust for risk, the ratio is suitable to make comparisons between portfolio’s. According to this model, the higher the ratio, the better the performance. It is assumed that the portfolio is well diversified. This assumption could pose a problem since not all mutual funds can satisfy it. Another weakness lies in the presumed normality in the distribution of the returns. Failing to meet this assumptions might weaken the reliability of the outcomes.

(7)

7

2.3.2 Treynor Measure

A similar risk-adjusted performance measure is the Treynor Measure. This measure is a ratio of return generated by the portfolio over and above the risk-free rate of return, during a given period adjusted for its systematic risk. Again, the portfolios are assumed to be well diversified. As a result, portfolio’s will move in relation with the market and the risk is reflected by β. The Treynor measure is defined as:

S =

𝑅𝑝𝛽-𝑅𝑓

𝑝

(3)

Where Rp is the return on a portfolio, Rf is the risk-free rate and βp is the beta of a portfolio.

When a sample of well diversified portfolios is used, the Treynor Measure and the Sharpe ratio should produce similar results. If not, the Treynor Measure cannot accurately measure the risk underlying relatively undiversified portfolios.

2.3.3 Jensen’s Alpha

Developed by Jensen in 1967, this model is built upon the CAPM. For that reason, the same assumptions apply for Jensen’s alpha as for the CAPM. The Jensen’s alpha can be defined as:

α

p,t

= (R

p,t

– R

f,t

) – β

p

(R

M,t

– R

f,t

)+ ε

p,t (4)

Where αp,t denotes the performance measure for abnormal return at time t, Rp,t denotes the

return on a portfolio in time t, Rf,t denotes the return on the risk-free rate in time t, RM,t

denotes the return on the market at time t, βp denotes the systematic risk of a portfolio in time

t and ε denotes the error term at time t. The alpha is the return over and above that predicted by the CAPM and can easily be used for statistical testing.

As with the CAPM, Jensen’s alpha has received the same criticism that the underlying assumptions are hard to fulfil in reality. Roll (1977, 1995) particularly stresses that the choice for a certain market benchmark is of crucial importance for the results of this performance measure. He argues that different market benchmarks produce different betas which can result in different results. This would imply that it is not an unambiguous performance measure.

An example of the importance of this notion concerns the research of Ippolito (1989). In this research he found evidence that mutual funds had been able to reach outperformance

(8)

8

compared to the market. Elton et al. (1993) reviewed the research and found that the mutual funds had not been able to outperform the market.

2.3.4 The Three-Factor Model by Fama & French

Fama and French (1966) added two more risk factors to the CAPM model. These are the difference in returns between a portfolio of small and large stocks (SMB) and the difference in returns between a portfolio with high book-to-market stocks and low book-to-market stocks (HML). These factors seemed to predict average returns well and therefore could to some extent explain risk premiums. The model can be defined as:

α

p,t

= (R

p,t

– R

f,t

) - β

1

(R

M,t

– R

f,t

)- β

2

(SMB

t

) – β

3

(HML

t

) +

ε

p,t (5)

Where αp,t denotes the performance measure for abnormal return, Rp,t denotes the

return on a portfolio in time t, Rf,t denotes the return on the risk-free rate in time t, the

coefficients β1, β2, β3 are slopes of the risk factors in a time –series regressions in time t and ε

denotes the error term.

Although SMB and HML do not immediately appear to be ligitimate risk factors, Fama and French (1966) argue that these factors serve as proxies for exposures of systematic risk which are not captured in the traditional CAPM. They point out that the model is justified on empirical grounds: historical average returns on stocks of relatively small companies as well as on stocks with high book-to-market ratios appear to be higher than predicted by the traditional CAPM.

Liew (2000) researched whether the two added variables are related to future growth in the economy and therefore may be used as proxies for systematic risk. They found that the returns driven by SMB and HML are positively related to GDP growth and therefore capture part of systematic risk.

Zhang (2005) in addition tried to find an intuitive explanation the SMB factor and focused on the irreversibility of investments. Firms with high book-to-market ratio’s (so called value firms) appeared to be much riskier in bad economic times because of the irreversibility of their relatively large amounts of tangible investments and much less risky in good times compared to growth firms (low book-to-market values). This asymmetric risk of value firms combined with risk averse investors could explain the positive relation between the value premium and the market risk premium. Petkova (2005) also found that part of the risk premium is explained by this value premium.

(9)

9

On the other hand, Black (1993) emphasizes the possibility of data snooping biasedness. This is the idea that given enough variables, it will always be possible to find variables that by pure chance happen to be correlated with the estimation error of the certain returns. Black points out that return premiums to factors such as size have proven to be inconsistent.

Cremers et al. (2010) argue that both the three-factor and four-factor model (to be explained in the next section) suffer from biases. They assert that both the SMB and HML factors assign disproportionate weight to respectively value stocks and small cap stocks. Together, this contributes to an upward bias in alpha estimates.

Although the added factors seemed to be valuable in the study on the US market (Fama, 1993), Titman et al. (1966) found that there was no relation between expected return and the Fama- French risk factors. Also in other markets the extent to which the added factors provide a better explanation vary across researches.

2.3.5 The Four-Factor Model by Carhart

Jedageesh and Titman (1993) uncovered the tendency for particular good and bad performing stocks to persist over time. This is what is called the Momentum Effect. Carhart (1997) added this effect as a fourth factor to the Fama & French Three-Factor model in order to evaluate mutual fund performance. In his research it appeared that the alpha of many mutual funds could to some extent be explained by their sensitivity to momentum in the market. This model has become a widely used model for abnormal performance evaluation concerning a stock portfolio (Bodie et al., 2011). The model can be defined as:

α

p,t

=(R

p,t

– R

f,t

) -β

1

(R

M,t

– R

f,t

)- β

2

(SMB

t

) – β

3

(HML

t

) – β

4

(MOM

t

) + ε

p,t (6)

Where αp,t denotes the performance measure for abnormal return at time t, Rp,t denotes

the return on a portfolio at time t, Rf,t denotes the return on the risk-free rate at time t, the

coefficients β1, β2, β3, β4 are the slopes of the risk factors in a time –series regression in time t

and ε denotes the error term.

At first sight, similar to the factors added in the Three-factor model, it does not seem very intuitive to reflect the momentum factor as a source of risk influencing return. However, recent work resulted in a growing appreciation of the idea that part of the momentum effect might be related to market liquidity in determining asset pricing. Based on the work of Amihud and Mendelson(1986) and increasingly after the financial crisis of 2008-2009, liquidity risk is more and more viewed as another important factor in determining asset

(10)

10

pricing and returns for which investors want to be compensated (Bodie et al., 2011). In their research, Pastór and Stambaugh (2003) test the effect of high liquidity on alphas computed by the four-factor model and suggest that liquidity risk to some extent may account for the momentum effect. More recent research by Asness et al. (2013) find a similar relationship. Again, as the Four-Factor model is an extension of both the Jensen’s Alpha model and the Three-Factor model, the critics on the previous models apply to this model as well. Although this model is widely used in academic literature, much debate exists about whether it is a significant improvement over the CAPM (Cooper, 2005). One particular area where the Four-Factor model does appear to provide a better specification of risk is on actively managed mutual funds. Using the basic CAPM, researchers found that fund with high historical returns have future positive alphas (Grinblatt et al., 1992 & Hendricks et al., 1993). When Carhart (1997) repeated this tests using his own model, he did not find evidence of positive alphas for that particular group.

2.4 Previous empirical research

As mentioned before, in accordance with the EMH theorem, the more efficient the market the harder it will be for investors to outperform the market. The semi-strong form allows sophisticated investors, such as mutual fund managers, to exploit mispriced assets in order to outperform the market without taking on additional risk, i.e. alpha return. Presumably, the competitive advantage of a sophisticated investor is that he is able to access information that is not widely distributed. However, competition among these sophisticated investors will eliminate these opportunities quickly, pushing prices back to their zero-alpha level (Berk, 2011). For this reason, the semi-strong form is regarded as closest to reality and therefore is widely used as starting point in analyzing mutual fund performance.

In his early study for the period 1945-1964 Jensen (1968) found little evidence that US mutual funds on average were able to outperform the market index. Moreover, after deducting the costs of active management the funds as a group appeared to reach underperformance compared to the market. Grinblatt and Titman (1989) investigated the period 1975–1984 and came to the same conclusion as a group but did find significant outperformance by some funds individually. Similar studies performed by Azar and Al Hourani (2010) and Ippolito (1989), also found evidence for certain individual outperformance. Again, most alpha’s were not large enough to overcome the management fees.

(11)

11

Studies by Brown (1992) and Malkiel (1995) stressed that the survivorship bias appears to be more important than previously assumed. This methodological issue exists when certain mutual funds disappear from the research sample during the timespan of the research. To avoid this problem, he used a dataset containing the returns of all equity funds for each year in the period 1971 to 1991. Nonetheless, the study concluded that mutual funds tended to underperform the market returns.

Shukla & van Inwegen (1995) found evidence that mutual funds investing in foreign equity underperform the funds investing in domestic equity. Coval & Moskowiz (1999) also underline the existence of this home bias. The main explanation of this phenomenal is concerned with information disadvantages that mutual funds face when investing abroad. Although not as extensively as on the US market, researches have also focused on other markets. Rouwenhorst (1998) analyzed 12 European countries from and did not find significant differences compared to the results on the US market. Otten and Bams (2002) also analyzed the European funds, including Dutch mutual funds, using the Four-factor model in the period 1991 to 1998. Before the deduction of costs, four out of five countries reached significant positive alpha returns, including the Netherlands. However, only funds in the UK provide significant results after deduction. The findings on the UK market contradict the results from Blake and Titman (1998) who found evidence for underperformance.

Horst et al. (1998) studied the performance of Dutch mutual funds and concluded that funds investing in Dutch equity were able to outperform the market benchmark. Analyzing mutual funds that invested worldwide using a sample of 27 countries in the period 1997-2007, Ferreira et al. (2011) draw the overall conclusion that the funds were underperforming.

As pointed out in this brief literature review, the results regarding potential outperformance of mutual funds is far from unanimous. In most studies encountering abnormal returns, most were found at the individual level. Berk and Green (2004) respond that these findings are not per se inconsistent with efficient markets. They argue that few skilled equity fund managers can produce positive alpha’s on a short time horizon. These managers will attract new funds until the additional costs and complexity of these extra funds cannot produce sufficient abnormal returns anymore. This way alpha’s will be driven down to zero. Based on the academic literature, most studies were not able to produce overwhelming evidence supporting outperformance relative to the various market indices.

(12)

12

3. Methodology and data

3.1 Methodology

This thesis will evaluate the performance equity funds listed in the Netherlands in a period before and after the recent financial crisis. Of these funds three main groups are established; Dutch equity funds investing in:

I. Dutch equity II. European equity III. Worldwide equity

The central research question is:

Are there significant differences in abnormal performance of Dutch listed equity funds relative to the market in the period preceding the start of the financial crisis compared to the period after the start of the crisis?

In order to be able to formulate an answer to this question two sub questions are constructed:

1. Did Dutch listed equity funds outperform, underperform or perform equally to their market benchmarks in the years prior to the start of the financial crisis in 2008? 2. Did Dutch listed equity funds outperform, underperform or perform equally to their

market benchmarks in the years after the start of the financial crisis in 2008?

Performance of a security is initially shown by its return. However, this has to be adjusted for the level of risk incurred since expected return and risk are positively correlated. In academic research several performance measures can be used. The results of performance evaluation, as discussed in section 2 can vary widely between those measures. In this research, the five most commonly used in academic literature are chosen to be utilized. These comprise:

1. The Sharpe ratio 2. The Treynor ratio 3. The Jensen’s Alpha

4. The Three-Factor Model of Fama & French 5. The Four-factor model of Carhart

After application of these measures, the outcomes in the two different time periods can be compared and inferences can be drawn. As discussed in section 2, each of the risk-adjusted performance measures has their own limitations. The Sharpe ratio and Treynor measure are not suitable for evaluation of possible outperformance since there is no alpha parameter

(13)

13

included in the model. These models can be used for performance measurement against the market benchmark. As a support for the main question, these two measures will be employed to test whether the funds are efficient portfolio’s. According to the theory, a portfolio is considered to be efficient when its ratio is equals or exceeds the market benchmark ratio.

For the Jensen’s Alpha, the Three-factor Model and the Four-factor model regression analysis will be used in order to test whether the fund was able to underperform, outperform or perform equally to the benchmark. For each fund within all groups an OLS regression will be applied in both periods of time.

It should be noted that the results will be interpreted from the perspective of investors with a long positions. So a short-position point of view will not be analyzed.

3.2 Data

The fall of Lehman brothers in September 2008 can generally be seen as the starting point of the financial crisis. Therefore, in this research that event is chosen as well. In addition, to keep research as relevant as possible, the latest available data was chosen. This is December 2014. So the period after the start of the crisis consists of 75 months. In order to make a correct comparison, the first period needs to consist of 75 months as well, counting backwards from September 2008. This corresponds to May 2002.

Morningstar was used to compile the list of equity funds for each group. The main shortcoming of this international provider of mutual fund information is that only surviving funds at the time it is consulted are included. For that reason, the dataset is not free of the survivorship bias. Although data of new funds that commenced in the period 05/2002-08/2008 was available for the period 09/ 2008-12/2014 via Datastream, these are not taken into consideration in order to make comparison between the two periods as fair as possible. When selecting the specific funds for each group in the analysis, the following criteria had to be met:

1. The mutual fund must predominantly invest in equity, i.e. for more than 95%. 2. The mutual fund must not be an index fund.

3. The fund existed in the entire period 05/2002 – 12/2014.

With respect to the first criterion, the portion not invested in stock may only be invested in money market securities. Equity funds commonly hold 0 to 5% of total assets in these securities to provide liquidity necessary to meet potential redemption of shares (Bodie et al.,

(14)

14

2011). The second criterion is important because the objective of this research is to investigate actively managed funds. Index funds try to match the performance of a broad market index and therefore are presumed to be passively managed.

In total, 29 funds satisfied the criteria for the entire period. Table 3.1 shows the number of funds per group.

Group 05/2002 – 09/2014

Dutch equity 7 European equity 11 Worldwide equity 11 Total 29

Table 3.1 Number of funds per category

The monthly NAV in euro of these funds were retrieved from the Datastream database. Subsequently, the returns were computed as logarithmic returns using these monthly observations. So for each fund existing for both periods, the number of observations amounted to 152 observations.

As can be seen in the regression models, market returns are needed as well. In order to simplify the study and in accordance with other studies, market proxies are used. For the group Dutch equity, the AEX index was chosen. For the European equity and the Worldwide equity group, these are the MSCI Europe Index and the MSCI AC Index respectively. The data was again obtained through Datastream.

Another important variable in the models is the risk free rate of return. The yield on the three month EURIBOR rate provided by the Dutch Central Bank and obtained through Datastream was applied for the Dutch equity and the European equity group. Concerning the Worldwide equity group, the three month LIBOR rate provided by the Swiss National Bank was used and obtained through Datastream.

The SMB, HML and MOM factors were retrieved from the Kenneth R. French Data Library. Since no specific factors were available for the Netherlands within this database, the European factors were used for the both the Dutch and European group. The global factors were applied to the worldwide group. After deducting the risk free rates from the market and fund returns respectively, the excess returns for both were computed. At this point all relevant data is obtained to perform the analysis using the five performance measures, as mentioned in section 3.1.

(15)

15

A market Sharpe ratio and Treynor measure will be computed in order to be able to analyze the results per fund against the market. In addition, the mean Sharpe ratio and Treynor measure will be computed for each group in both time periods.

Concerning the factor models, for each fund and every time period, an OLS regression will be performed. Furthermore, for each group as a whole, a panel regression will be used in order to draw inferences on possible differences before and after the start of the financial crisis. Altogether, for each of the three factor performance measures 58 individual regressions were performed, including both periods. On top of that, for each of these performance measures 6 panel regressions were performed, including both periods.

(16)

16

4. Empirical Results

In this section, the results which have been reached using the different performance measures will be discussed. The descriptive statistics tables can be found in appendix 1.

4.1 Sharpe Ratio

Table 4.1 shows the results of the computed Sharpe ratios per fund. In addition, table 4.2 shows the average Sharpe ratios and market Sharpe ratio for each period. Since most excess returns were negative, most calculated ratios are negative. In the group Dutch Equity, before the start of the crisis, 4 out of 7 funds had a Sharpe ratio higher and 3 had a ratio lower than the market. When dealing with negative values, higher indicates a ratio closer to 0. The average Sharpe ratio for all funds was -0.497 while the market Sharpe ratio was -0.523. Therefore, it appears that the equity funds in this group have performed better than the market in this period; on average they can be considered efficient. This changed after the start of the crisis. On the individual level, only 2 higher ratios appeared. The Sharpe ratio on average was -0.157 compared to a market ratio of -0.142 which might indicate that the return of these funds is lower than should be expected based on the level of risk.

In the European group a more extreme result appeared. All funds were able to reach a higher Sharpe ratio before the financial crisis. Therefore, the average Sharpe ratio was -0.483 for the funds compared to 0.690 for the market. After the start of the crisis, only 4 were able to reach a higher result and 7 performed worse compared to the market. On average, the equity funds in the European group appeared to be performing worse with respect to the market.

In the last group, the same declining trend was shown. Before the start of the crisis 9 out of 11 reached a higher ratio while after the crisis only 4 achieved this result. On average, the equity funds investing worldwide performed better than the market in the first period and worse in the second period.

Prior to crisis After start crisis

Group Higher Lower Higher Lower

Dutch Equity 4 3 2 5

European Equity 11 0 4 7

Wordlwide Equity 9 2 4 7

(17)

17

Prior to crisis After start crisis

Group Average Market Average Market Dutch Equity -0.497 -0.523 -0.157 -0.142 European Equity -0.483 -0.693 -0.176 -0.143 Wordlwide Equity -0.301 -0.343 0.062 0.076

Table 4.2 Average Sharpe Ratio compared to the market

4.2 Treynor Measure

When using the Treynor measure a similar pattern is revealed as can be seen in table 4.3 and 4.4. One remarkable difference compared to the outcomes of the Sharpe Ratio is that the Treynor Measure for Dutch Equity in the period before the crisis is lower than the market Treynor. Although the Sharpe ratio and Treynor measure are quite similar performance measures, differences in outcomes could occur due to different risk used to normalize the performance. For the Sharpe Ratio this is the total risk captured by the standard deviation and for the Treynor Measure this is the systematic risk captured by beta. As explained in section 2.3.2, the Treynor Measure has the specific condition that the underlying portfolio should be well-diversified. If not, changes in outcomes could appear. This might be the case here. Nonetheless, the overall pattern of a worsening trend after the start of the financial crisis can still be found. So these two intuitive measures produce no real contradictory outcomes.

Prior to crisis After start crisis

Group Higher Lower Higher Lower

Dutch Equity 1 6 2 5

European Equity 8 3 3 8

Wordlwide Equity 3 8 6 5

Table 4.3 Separate Treynor Measures compared to the market

Prior to crisis After start crisis

Group Average Market Average Market Dutch Equity -3.804 -3.349 -1.112 -0.867 European Equity -2.778 -3.111 -1.135 -0.706 Worldwide Equity -1.344 -1.406 0.303 0.348

(18)

18

4.3 Jensen’s Alpha

Table 4.5 and 4.6 show the results when using the Jensen’s Alpha as performance measure. In table 4.5 the number of significant alphas are shown and divided between outperformance and underperformance. Table 4.6 shows the alphas of the panel regression per group for each period.

Prior to crisis After start crisis

Group Outperformance Underperformance Outperformance Underperformance

Dutch Equity 0 0 0 0

European Equity 2 0 0 1

Worldwide Equity 2 0 0 1

Table 4.5 Number of significant alphas per group for each period

Prior to crisis After start crisis

Group Alpha Alpha

Dutch Equity -0.225 -0.171

European Equity 0.482* -0.372*

Worldwide Equity 0.045 -0.076

Table 4.6 Significant alphas per group as a whole for each period. *= significant at the 5% level.

For the Dutch equity group, no significant alphas revealed in the two periods, both on the individual and overall level. So based on this performance there is insufficient evidence at the significance level of 5% to reject the claim that Dutch listed equity funds performed equally the market benchmark in both periods.

In the European equity group, two significantly positive alphas were shown in the period before the start of the crisis and one significantly negative alpha in the subsequent period. As a whole, a significantly positive alpha of 0.482 was found in the first period and a significantly negative alpha of -0.372 in the second period. Based on a significance level of 5%, there is sufficient evidence to reject the claim that the Dutch funds investing in European equity performed equally to the benchmark in both periods of time.

Similar to the European group, the worldwide equity group showed two significantly positive alphas in the first period and one significantly negative in the second period. As a

(19)

19

whole, no sufficient evidence was found to reject the hypothesis that this group performed equally to the market benchmark in both periods.

4.4

Three-Factor Model by Fama & French

Table 4.7 and 4.8 show the results when using the Fama & French Three-Factor model as performance measure. In table 4.7 the number of significant alphas are shown and divided in outperformance and underperformance. Table 4.8 shows the alphas of the panel regression per group for each period.

Prior to crisis After start crisis

Group Outperformance Underperformance Outperformance Underperformance

Dutch Equity 0 0 0 0

European Equity 2 0 0 1

Worldwide Equity 1 1 0 1

Table 4.7 Number of significant alphas per group for each period

Prior to crisis After start crisis

Group Alpha Alpha

Dutch Equity -0.127 -0.189

European Equity 0.549* -0.383*

Worldwide Equity 0.195* -0.093

Table 4.8 Significant alphas per group as a whole for each period. *significant at the 5% level.

Again, concerning the Dutch equity group, no significant alphas were found on the individual and overall level. The results of the European group show a similar pattern as the Jensen’s Alpha measure. Individually, two funds achieved significant alphas in the prior period and one reached a significant underperformance in the following period. As a whole, significant outperformance (0.549) was found in the prior period and significant underperformance in the following period (-0.383).

The results on the Worldwide equity show a fairly different results on the first period. At the individual level, one significantly positive and one significantly negative alpha appeared. Altogether, this group reached a significant result only in the prior period (0.195). So when using the Three-Factor model, no significant deviation from the results of the Jensen’s alpha is seen for the Dutch and European group. A notable different outcome only appeared in the Worldwide group prior to the crisis where a significant outperformance was reached.

(20)

20

4.5

The Four-Factor Model by Carhart

Table 4.9 and 4.10 show the results when using the Carhart Four-Factor model as performance measure. In table 4.9 the number of significant alphas are shown and divided in outperformance and underperformance. Table 4.10 shows the alphas of the panel regression per group and for each period.

Prior to crisis After start crisis

Group Outperformance Underperformance Outperformance Underperformance

Dutch Equity 0 0 0 0

European Equity 2 1 0 1

Worldwide Equity 2 1 0 1

Table 4.9 Number of significant alphas per group for each period

Prior to crisis After start crisis

Group Alpha Alpha

Dutch Equity -0.132 -0.191

European Equity 0.516* -0.383*

Worldwide Equity 0.200* -0.102

Table 4.8 Significant alphas per group as a whole for each period. *significant at the 5% level.

The results of the Dutch group are again similar compared to the Jensen’s Alpha and Three-Factor models. Regarding the European group, one more significant underperformance is found on the individual level in addition to the results of the former two measures. The panel regression outcomes once more show sufficient evidence to reject the hypothesis that the funds performed equal to the market. A significant pattern of declining performance is shown in the European group as well.

The results of the Worldwide group are similar in comparison with the Three-Factor Model. The only difference is that one more significantly positive alpha on the individual level appears in the first period.

(21)

21

5. Discussion

5.1 Results

The first notable result with respect to the Sharpe ratio measures is the large number of negative outcomes for the mutual funds and markets. This is caused by the negative excess returns. As pointed out by Krimm et al. (2012), one should be cautious when ranking funds on the basis of the Sharpe ratio in bear market conditions. A general admission of the Sharpe ratio is that if two funds with identical excess returns are evaluated, the one with the lowest standard deviation (i.e. a higher Sharpe ratio) is preferred. Krimm et al. argue that this only holds in bull market conditions. In case of bear market conditions, this relationship is reversed: the fund with a higher standard deviation receives a higher (less negative) Sharpe ratio. This seems to undermine the intention that is at the heart of the Sharpe ratio: less risk is preferred, ceteris paribus. For that reason, the results of the Sharpe ratio measure do not seem to be a meaningful assessment of fund performance in these extraordinary times around the start of the financial crisis. The only outcomes that hold is the one for the Worldwide group in the second period since there the excess returns are above zero. However, because the results from the prior period seem to suffer from this so called Market Climate Bias, no reliable comparison can be made between the two periods.

In the same line of reasoning the Treynor ratio might be biased as well. Although first analysis on both measures provides the intuitive result that the performance of these equity funds might have deteriorated compared to the market benchmark after the start of the financial crisis, they do not seem be reliable performance measures in these market conditions.

In appendix 2 the empirical results of the equity funds combined are shown for each group and every period after using panel regressions. When analyzing the Dutch equity group, no significant alphas were found in either period. The explanatory power of the model seems to increase after including the SMB, HML and MOM factors according to the increasing R2 statistics. When testing this formally using partial F-tests with a significance level of 5% in the period before the crisis, all extension models provide significant results meaning that these factors contribute to the variance of the fund’s excess returns and should hence be included. For the second period, the Four-Factor extension produce insignificant results and is therefore no substantial contribution to the Three-factor model in that particular period.

(22)

22

The funds categorized in the European group seemed to be most affected by the crisis since a significant overall outperformance changed in a significantly underperformance compared to the European market. After conducting the partial F-tests on both multifactor models, the null hypotheses that the added factors have no effect on the excess return of the equity funds were rejected in the first period. Therefore, the extended factors are assumed to increase the explanatory power of the model and should therefore be included. In the second period, the partial F-test on the MOM factor does not reject the null hypothesis that the explanatory power of the model remains the same when adding the factor . The F-test on the SMB and HML factors did show up to be a significant addition to the model.

Concerning the funds investing worldwide, on average appeared to be able to achieve outperformance in the period prior to the start of the crisis. After the start of the crisis no sufficient evidence was found against equal performance compared to the market. Only the Three-Factor model provides sufficient evidence to conclude that the inclusion of the SMB and HML factor significantly increase the explanatory power of the model.

These significant alpha results do not object the underlying theory and earlier studies. As described in section 2.1, the semi-strong form of the EMH does not necessarily exclude the possibility of reaching abnormal returns. As described in section 2.4, results of mutual fund performance evaluation among researches vary widely. In most studies few alphas could be found on the individual level which is in line with the results in this research.

In all groups and time periods, the Fama & French Three-factor model appeared to be significantly better in explaining the variation of the excess returns. With respect to the Four-Factor model, the explanatory power only increased within the groups Dutch equity and European equity prior to the crisis. Therefore, the momentum factor did not appear to be a significant contribution to the Three-factor model. The Three-Factor model seemed to be the model with the greatest empirical strength and consequently is preferred in this research.

5.2 Limitations and further research

This research is very specific to a relatively small group of Dutch equity funds. For that reason, one should be cautious when drawing conclusions from these findings and when comparing to other studies. Hence, it is recommended to further test the models across other periods of time and with a larger dataset in order to check their true validity for Dutch listed equity funds.

(23)

23

The results impute that Dutch equity funds investing in European equity were most affected by the economic turmoil encompassing the financial crisis. At first glance a possible explanation could be that this group suffered greatly from the aftermath of the Euro crisis which followed the financial crisis. Yet, this implication might be premature and not fully plausible. Further research could consequently address this implication.

Moreover, it should be noted once more that the funds in this research have existed for the entire period of analysis. Survivorship bias is not controlled for and could possibly pose problems regarding the reliability of the results. Further research could include non-surviving funds in their dataset in order to minimize this bias to some extent.

Another important remark involves the choice of a market benchmark against which the performance is to be compared, also pointed out by Roll (1977, 1995). Firstly, the market proxies do not deliver exactly the same result as the market returns that is intended by studies. This may cause biases. Secondly, it is assumed that investors can directly invest in the market if wanted. However, in reality investors will need a financial intermediary which can provide them with investment products that replicate the market returns. Nevertheless, this transaction involves costs which should be accounted for. In the research, these costs are not considered emphasizing once more that the conclusions need to be treated with a degree of caution.

The main purpose of this study was to examine whether Dutch equity funds were able to reach abnormal returns and which model could explain these return the best. It might be interesting to extent this research in two ways. Firstly, research on the possible determinants causing the abnormal returns of the firms that did not perform equal to the market. Secondly, one could evaluate these funds on performance persistence. This might be helpful in distinguishing between abnormal performance as a result of skill or as a result of good

fortune.

One additional marginal note has to be made. The effect of front-end and back-end loads as well as management fees on the performance of the equity funds are not taken into account. The inclusion of these costs in the model might lead to different outcomes since significant alpha returns could disappear or worsen (when negative). Further research is necessary to draw conclusions on the effect of these costs on Dutch equity fund performance.

(24)

24

6 Conclusion

The literature on performance evaluation is extensive and still a subject of controversy. An important part of the debate is about the usefulness of models derived from the Capital Asset Pricing Model. This research attempts to make a modest contribution to this debate by providing empirical results on the performance of actively managed Dutch equity funds in a time span before and after the recent financial crisis when applying five widely used performance measures based on the CAPM.

The dataset contained three distinct groups of funds investing in Dutch, European and Worldwide equity. For each group as a whole and each fund individually a period before and a period after the start of the financial crisis was investigated. The performance measures used were respectively the Sharpe ratio, Treynor measure, Jensen’s alpha, Fama & French Three-Factor model and the Carhart Four-Factor-model. The intention of using the first two measures was to analyze whether the portfolios were efficient, i.e. able to reach a higher risk-adjusted return compared to the market. The remaining three were utilized to demonstrate whether these funds were able to reach outperformance or underperformance relative to the market benchmark.

The results of the Sharpe ratio and the Treynor measure appeared to be biased due to the considerably bear market conditions. As a consequence, these could not be interpreted accurately.

Empirical results after using the remaining three performance measures found sufficient evidence that the abnormal returns of Dutch equity funds investing in European equity significantly changed after the start of the financial crisis in September 2008. Contrary to the period prior to this crisis, little evidence was found that funds investing worldwide were able to reach outperformance in the subsequent period. Therefore, I can conclude that the abnormal returns within this group were notably impacted by the crisis as well. No evidence on any significant deviation from the market return was found with reference to the Dutch equity group.

Overall, the Fama & French Three factor model seems to be the one with the highest empirical explanatory power. It is therefore the preferred performance measure. Nevertheless, further research is required to gain a more in-depth knowledge on the validity of these findings and what the underlying determinants of abnormal in this specific group of Dutch equity funds compile.

(25)

25

Reference list

Asness, C.S., Moskowitz, T.J. & Pedersen, L.H. (2013) Value and Momentum Everywhere.

Journal of Finance. 68(3), pp. 929 – 985.

Azar, S.A. and Al Hourani, M.A. (2010). The Performance of U.S. Equity Mutual Funds,

Journal of Money, Investment and Banking, 18(1), pp. 13- 28.

Berk J.B. & Green, R.C. (2004). Mutual Fund Flows and performance in Rational Markets. Journal of Political Economy, 112(6), pp. 1269 – 1295.

Berk, J.B. & DeMarzo, P. (2011). Corporate Finance. Pearson Education Limited, 2nd edition.

Black, F. (1993). Beta and Return. The Journal of Portfolio Management, 20(1), pp. 8 – 18. Bodie, Z., Kane, A., Marcus, A.J. (2011). Investments and Portfolio Management.

Mcgraw-Hill, 9th edition.

Brown, S., Goetzmann, W., Ibbotson, R., Ross, S., 1992. Survivorship Bias in Performance Studies. Review of Financial Studies, 5(4), pp. 553-580.

Carhart, M. (1997). On persistence in mutual fund performance, Journal of Finance, 52 (1), pp. 57-82.

Cooper, M., Gutierrez, R & Marcum, B.(2005) On the Predictability of Stock Returns in Real Time. Journal of Business, 78(2), pp. 469 – 500.

Coval, J. & Moskowitz, T. (1999). Home Bias at HomeL Local Equity Preference in Domestic Portfolios. Journal of Finance, 54(6), pp. 2045-2073.

Cremers, K.J.M, Petajisto, A. & Zitzewitz, E. (2010). Should Benchmark Indices Have

Alpha? Revisiting Performance Evaluation. Critical Finance Review, 2(1), pp. 1 – 48. De Bondt, W.F.M., Thaler, R. (1990). Do Security Analysts Overreact? American Economic

Review, 80(2), pp. 52-57.

Elton, E.J., Gruber, M.J., Das & Hlavka, M. (1993). Efficiency with Costly Information: a Reinterpretation of Evidence from Managed Portfolios. Review of Financial Studies, 6(1), pp. 1 – 22.

Fama, E.F. (1965). The Behavior of StockMarket Prices. Journal of Business, 38(1), pp. 34 -105.

Fama, E.F. & French, K.R. (1993). Common Risk Factors in the Returns on Stocks and Bonds, Journal of Financial Economics, 33 (1), pp. 3-56.

Fama, E.F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work.

(26)

26

Grinblatt, M. & Titman, S., (1989). Mutual Fund Performance: An Analysis of Quarterly Portfolio Holding. Journal of Business, 62(3), pp. 393-416.

Grinblatt, M., Titman, S. (1992). The Persistence of Mutual Fund Performance. Journal of

Finance, 47(5), pp. 1977 – 1984.

Ferreira et al. (2013). The Determinants of Mutual Fund Performance: A Cross-Country Study. Review of Finance, 17(2), pp. 4863 – 525.

Hendricks, D., Patel, J. & Zeckhauser, R. (1993). Hot Hands in Mutual Funds: Short-Run Persistence of Performance 1974-1988. Journal of Finance, 4(1), pp. 93-130. Horst, J.R. et al. (1998). Style Analysis and Performance Evaluation of Dutch Mutual Funds.

Tilburg University, Center for Economic Research.

Ippolito, R.A. (1989). Efficiency with Costly Information: a Study of Mutual Fund Performance 1965-1984. Quarterly Journal of Economics, 104(1), pp.1-23.

Jensen, M. (1968). The Performance of Mutual Funds in the Period 1945-1964. Journal of

Finance, 23(2), pp. 389-416.

Krimm, S., Scholz, H. & Wilkens, M. (2012). The Sharpe Ratio’s Market Climate Bias: Theoretical and Empirical Evidence from the US Equity Mutual Funds. Journal of

Asset Management, 13(4), pp. 227-242.

Lintner, J. (1965). The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets. Review of Economics and Statistics, 47 (1), pp. 13–37.

Malkiel B. G., 1995. Returns from Investing in Equity Mutual Funds 1971 to 1991. Journal

of Finance, 50(2), pp. 549-572.

Markowitz, H.M. (1952). Portfolio Selection. Journal of Finance. 7(1), pp. 77-91.

Mossin, J. (1966). Equilibrium in a Capital Asset Market. Econometrica, 34 (4), pp. 768– 783 Narasimhan, J. & Titman, S. (1993). Returns to Buying Winners and Selling Losers:

Implications for Stock Market Efficiency. Journal of Finance, 48(1), pp. 65-91. Pastór, L. & Stambaugh, R.F. (2003). Liquidity Risk and Expected Stock Returns. Journal of

Political Economy, 11(3), pp. 642 – 685.

Petkova, R. & Zhang, L. (2005). Is Value Riskier than Growth?. Journal of Financial

Economics, 78(1), pp. 187 – 202.

Roll, R. (1977). A Critique of the Asset Pricing Theory’s Tests Part 1: On Past and Potential Testability of the Theory. Journal of Financial Economics, 4(2), pp. 129-176.

Roll, R. & Ross, S.A. (1994). On the Cross-Sectional Relation between Expected Return and Betas. Journal of Finance, 49(1), pp. 101-121.

(27)

27

Rouwenhorst, K. G., 1998. International Momentum Strategies. Journal of Finance, 53(1), pp. 267-284.

Sharpe, W. (1964). Capital Asset Prices: A Theory of Market Equilibrium. Journal of

Finance, 19(3), pp. 425 – 442.

Shukla, R. & Ingwegen, G. (1995). Do Locals Perform Better than Foreigners?: Analysis of UK and US Mutual Fund Managers. Journal of Economics and Business, 47(3), pp. 241 – 254.

(28)

28

Appendix 1 Descriptive Statistics

Table 1 Dutch equity, 05/2002 – 08/2008

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 -3.35 6.67 -28.32 8.72 SMB 76 0.27 1.91 -6.94 3.62 HML 76 0.71 1.26 -2.01 4.72 MOM 76 1.12 3.87 -13.91 13.80 (Rm - Rf) fund 1 76 -3.19 6.15 -24.37 6.75 (Rm - Rf) fund 2 76 -3.31 6.34 -24.74 9.25 (Rm - Rf) fund 3 76 -2.42 6.36 -20.89 12.21 (Rm - Rf) fund 4 76 -3.34 7.30 -41.19 8.78 (Rm - Rf) fund 5 76 -2.56 5.74 -19.37 7.06 (Rm - Rf) fund 6 76 -2.47 5.71 -19.31 9.50 (Rm - Rf) fund 7 76 -3.30 6.31 -24.74 7.78

Table 2 Dutch equity, 09/2008 – 12/2014

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 -0.87 6.72 -26.22 9.61 SMB 76 0.06 2.02 -4.51 4.85 HML 76 -0.22 2.73 -4.60 7.45 MOM 76 0.52 4.72 -25.96 9.87 (Rm - Rf) fund 1 76 -0.89 6.15 -27.44 9.05 (Rm - Rf) fund 2 76 -1.10 6.51 -27.29 9.71 (Rm - Rf) fund 3 76 -1.26 6.01 -22.18 12.01 (Rm - Rf) fund 4 76 -1.04 6.84 -25.42 11.27 (Rm - Rf) fund 5 76 -0.54 6.17 -25.70 14.97 (Rm - Rf) fund 6 76 -0.53 6.55 -29.35 19.36 (Rm - Rf) fund 7 76 -0.93 6.10 -24.99 9.47

(29)

29

Table 3 European equity, 05/2002 – 08/2008

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 -3.11 4.87 -18.92 5.58 SMB 76 0.27 1.91 -6.94 3.62 HML 76 0.71 1.26 -2.01 4.72 MOM 76 1.12 3.87 -13.91 13.80 (Rm - Rf) fund 1 76 -1.75 7.51 -20.46 16.42 (Rm - Rf) fund 2 76 -3.45 6.27 -22.12 12.16 (Rm - Rf) fund 3 76 -3.24 9.22 -27.17 18.28 (Rm - Rf) fund 4 76 -3.40 7.93 -28.06 14.22 (Rm - Rf) fund 5 76 -3.04 4.84 -18.62 6.90 (Rm - Rf) fund 6 76 -3.18 5.19 -19.11 7.10 (Rm - Rf) fund 7 76 -2.94 6.90 -25.87 12.54 (Rm - Rf) fund 8 76 -1.27 3.65 -9.95 8.47 (Rm - Rf) fund 9 76 -2.74 6.25 -21.25 8.72 (Rm - Rf) fund 10 76 -2.97 5.35 -18.75 7.00 (Rm - Rf) fund 11 76 -2.92 6.82 -25.66 20.05

Table 4 European equity, 09/2008 – 12/2014

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 -0.71 5.51 -18.69 9.78 SMB 76 0.06 2.02 -4.51 4.85 HML 76 -0.22 2.73 -4.60 7.45 MOM 76 0.52 4.72 -25.96 9.87 (Rm - Rf) fund 1 76 -1.47 9.56 -32.14 20.98 (Rm - Rf) fund 2 76 -0.92 5.04 -23.42 7.63 (Rm - Rf) fund 3 76 -0.51 7.43 -24.12 13.44 (Rm - Rf) fund 4 76 -0.77 6.68 -30.68 14.49 (Rm - Rf) fund 5 76 -0.72 5.56 -19.18 10.67 (Rm - Rf) fund 6 76 -0.71 5.99 -19.99 11.01 (Rm - Rf) fund 7 76 -0.93 4.59 -19.19 6.93 (Rm - Rf) fund 8 76 -3.25 7.88 -31.76 12.04 (Rm - Rf) fund 9 76 -0.43 6.78 -26.35 14.58 (Rm - Rf) fund 10 76 -0.83 6.05 -24.88 8.19 (Rm - Rf) fund 11 76 -1.13 6.14 -18.34 11.28

(30)

30

Table 5 Worldwide equity, 05/2002 – 08/2008

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 -1.41 4.37 -13.12 6.09 SMB 76 0.24 1.73 -5.12 3.84 HML 76 0.50 1.20 -3.08 3.19 MOM 76 0.71 3.30 -10.4 9.32 (Rm - Rf) fund 1 76 -1.50 4.36 -14.16 7.08 (Rm - Rf) fund 2 76 -1.43 4.31 -13.03 6.05 (Rm - Rf) fund 3 76 -1.47 4.85 -14.45 9.80 (Rm - Rf) fund 4 76 -0.85 6.11 -15.63 13.08 (Rm - Rf) fund 5 76 -1.57 4.81 -17.23 9.36 (Rm - Rf) fund 6 76 -1.47 4.65 -14.49 10.25 (Rm - Rf) fund 7 76 -1.58 4.78 -14.77 8.97 (Rm - Rf) fund 8 76 -1.52 4.51 -14.35 7.71 (Rm - Rf) fund 9 76 -1.45 4.06 -12.80 5.52 (Rm - Rf) fund 10 76 -1.14 6.31 -24.31 12.28 (Rm - Rf) fund 11 76 -1.46 4.46 -15.12 6.61

Table 6 Worldwide equity, 09/2008 – 12/2014

Variable Obs. Mean Std. Dev. Min Max

(Rm – Rf) 76 0.35 4.58 -14.91 9.20 SMB 76 0.03 1.47 -2.75 3.78 HML 76 -0.05 1.83 -4.79 4.34 MOM 76 0.11 4.03 -23.89 7.50 (Rm - Rf) fund 1 76 0.14 4.27 -16.67 7.84 (Rm - Rf) fund 2 76 0.48 4.15 -15.42 8.50 (Rm - Rf) fund 3 76 0.46 3.76 -12.01 7.77 (Rm - Rf) fund 4 76 -0.42 8.24 -34.66 16.40 (Rm - Rf) fund 5 76 0.12 4.39 -15.38 6.92 (Rm - Rf) fund 6 76 0.46 4.41 -14.64 8.61 (Rm - Rf) fund 7 76 0.38 5.04 -15.85 12.71 (Rm - Rf) fund 8 76 0.26 3.98 -16.16 7.48 (Rm - Rf) fund 9 76 0.16 4.49 -15.15 9.51 (Rm - Rf) fund 10 76 0.389 4.74 -17.89 9.71 (Rm - Rf) fund 11 76 0.34 4.64 -16.87 8.25

(31)

31

Appendix 2 Results panel regression

The results are presented in the following two pages

(32)

32

Group

(1) (2) (3)

Independent variables Excess return

funds Excess return funds Excess return funds Dutch equity, 2002-2008

Excess returns market 0.811*

(0.033) 0.855* (0.036) 0.807* (0.040) SMB 0.585* (0.087) 0.683* (0.89) HML -0.149 (0.154) -0.114 (0.153) MOM -0.185* (0.047) Intercept -0.225 (0.154) -0.127 (0.196) -0.132 (0.192) N 532 532 532 F-Statistic 613.95 274.34 236.52 R2 0.7352 0.7625 0.7711 European Equity, 2002-2008

Excess returns market 1.058*

(0.038) 1.087* (0.384) 1.046* (0.039) SMB 0.325* (0.095) 0.390* (0.094) HML -0.090 (0.152) -0.068 (0.152) MOM -0.114* (0.055) Intercept 0.482* (0.163) 0.549* (0.201) 0.516* (0.120) N 836 836 836 F-Statistic 773.91 293.43 232.18 R2 0.6160 0.6239 0.6269 Worldwide Equity, 2002-2008

Excess returns market 1.030*

(0.018) 1.025* (0.017) 1.018* (0.017) SMB -0.012 (0.043) 0.005 (0.047) HML -0.310* (0.067) -0.312* (0.067) MOM -0.025 (0.030) Intercept 0.045 (0.067) 0.195* (0.076) 0.200* (0.077) N 836 836 836 F-Statistic 3162.11 1218.01 967.43 R2 *= 5% significance 0.8475 0.8534 0.8536

(33)

33

Group

(1) (2) (3)

Independent variables Excess return

funds Excess return funds Excess return funds Dutch equity, 2008-2014

Excess returns market 0.838*

(0.027) 0.813* (0.029) 0.814* (0.029) SMB 0.514* (0.063) 0.517* (0.062) HML 0.159* (0.059) 0.166* (0.063) MOM 0.009 (0.033) Intercept -0.171 (0.126) -0.189 (0.119) -0.191 (0.120) N 532 532 532 F-Statistic 996.33 458.38 352.97 R2 0.7886 0.8168 0.8168 European Equity, 2008-2014

Excess returns market 0.975*

(0.035) 0.975* (0.040) 0.975* (0.041) SMB 0.429* (0.081) 0.428* (0.081) HML 0.062 (0.063) 0.061 (0.064) MOM -0.002 (0.042) Intercept -0.372* (0.136) -0.383* (0.133) -0.383* (0.135) N 836 836 836 F-Statistic 766.84 303.59 231.50 R2 0.6428 0.6594 0.6594 Worldwide Equity, 2008-2014

Excess returns market 0.940*

(0.033) 0.962* (0.036) 0.974* (0.036) SMB 0.069 (0.057) 0.091 (0.058) HML -0.155* (0.054) -0.129* (0.057) MOM 0.048 (0.025) Intercept -0.076 (0.085) -0.093 (0.086) -0.102 (0.088) N 836 836 836 F-Statistic 798.62 317.51 240.44 R2 0.7767 0.7804 0.7817 *= 5% significance

(34)

Referenties

GERELATEERDE DOCUMENTEN

Laser texturing of different patterns were conducted in order to improve the surface functionality of metal sheets obtained during the imprinting process.. The contribution

This paper examines the performance of a model based condition monitoring approach by using just operating parameters for fault detection in a two stage gearbox..

2) to incorporate these features in a robust learning frame- work, by using the recently proposed robust Kernel PCA method based on the Euler representation of angles [30], [31].

multimedia, network and user data, environmental inferences  Deriving a mathematical model for context aware routing in mobile phone sensor networks.  Testing performance for

Dit raamplan beschrijft het door alle ULO’s ondersteunde kader waarbinnen voorstellen kunnen worden ingediend voor de opzet, uitvoering, evaluatie en consolidatie van

As AM moved into direct competition with traditional manufacturing for service parts, the need arose to match service properties.. For some systems, post processing

Nederlandse Gasunie wants to invest in the sector funds that are the most efficient, it should invest in the ING Dutch Office Fund, Altera Vastgoed Winkels, Altera Vastgoed

The main goal of this research is to determine whether Dutch fund managers earn abnormal returns compared to what an investor could earn with a passive strategy mimicking a