Are post-IPO buy recommendations which are made within one year after an IPO for firms that had an IPO in 2013 by affiliated analysts who work for one of the top five analyst firms, compared to post-IPO buy recommendati

1 Are post-IPO buy recommendations which are made within one year after an IPO for firms that had an IPO in 2013 by affiliated analysts who work for one of the top five analyst firms, compared to post-IPO buy recommendations from unaffiliated analysts who work for one of the top five analyst firms, leading to different one-year excess returns?

Name:

John Molenaar

Student Number:

10582428

Programme:

Economics and Business

Track:

Finance and Organization

Supervisor:

Y. Wang

Date:

June 28 2016

This paper examines if buy recommendations from affiliated analysts and buy recommendations from unaffiliated analysts, for firms that recently had an IPO, lead to different one-year excess returns. On the one hand, the excess returns can be different because affiliated analysts make biased recommendations to generate business for the investment banking division and the brokerage division. On the other hand, the excess returns can be different because affiliated analysts have superior information which is gathered by their colleagues, the underwriters, during an initial public offering of a company. The results of this study suggest that the one-year excess return is significantly different. The one-year excess return is significantly higher after a buy recommendation from an affiliated analyst, which can be explained because they have access to superior information.

2 This document is written by Student John Molenaar who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

3

Table of contents

1 Introduction

4

2 Literature review

8

2.1 Related previous research

8

2.2 Underwriters

8

2.3 Analysts

9

2.4 Biased recommendations

10

2.5 Superior information

11

2.6 SRO regulations and Global Analyst Research

12

Settlement

2.7 Fama- French 3-factor model

13

3 Data

14

4 Methodology

16

4.1 Models

16

4.2 Hypothesis

18

5 Results

19

6 Conclusion

23

7 Reference list

26

8 Appendix

31

8.1 Correlation Matrix

31

8.2 Variable list

4

1. Introduction

As an illustration, suppose a private company A, which prefer to raise money by issuing shares. To sell new shares to many investors, Company A should be a public company which can be realized via an IPO. An IPO is the first sale of shares to the public (Investopedia, 2016). Normally, as lacking of expertise in this area, company A would seek help from investment banks, which are called lead underwriters in this role. During the IPO process, the lead underwriter is responsible for pricing and distributing the securities (Ellis et el., 2000). A lead underwriter buys shares from Company A before selling them to investors at a stock exchange. Underwriters are compensated for their work with fees (Michaely & O’hara, 2000), which are lucrative for the investment banks (Michaely & Womack, 1999).

Suppose further that company A chooses Barclays as lead underwriter. Barclays is an investment bank that operates in several market areas (Barclays, 2016). One of Barclays’ divisions is specialized in underwriting, moreover, Barclays has another division that is specialized in creating buy/ sell/ hold recommendations for stock trading. People who work at the latter division are called analysts. The analysts try to produce profitable investment advice for their clients. A buy recommendation for a stock of firm is given if analysts believe the stock is undervalued. Contrary, they will give a sell recommendation for a firm if they believe that a stock is overvalued (Kecskes et al., 2015). Analysts either from lead underwriter, Barclays in this case or from non-lead underwriter firms can give a buy/ sell or hold recommendation for Company A. An analyst is called affiliated if the analyst makes a recommendation for a company and the bank for which the analyst works was the underwriter during the IPO of that company (Dugar & Nathan, 1995). In this example, if an analyst who works for Barclays makes a recommendation for Company A, the analyst is called affiliated. If an analyst from another company makes a recommendation for Company A, the analyst is unaffiliated. (Corwin et al., 2015). Lin and McNichols find that recommendations from affiliated analysts are more favorable than recommendations from unaffiliated analysts (1998). About the cause, there exist different ideas in the previous studies.

Some researchers argue that recommendations from affiliated analysts are biased. The research department where the analysts work, does not contribute to preliminary business for investment banks, meaning this research department usually does not generate significant revenue (Jegadeesh et al., 2004). But, the research

5 department can be used by the investment bank as a tool to generate business for other divisions of the investment bank (Bradshaw et al., 2006). Michaely and Womack claim that affiliated analysts give more favorable recommendations to enhance the likelihood that their company also will be chosen as lead underwriter for the next security offering from Company A. This can lead to pressure from the investment bank to make positive recommendations for clients from the investment bank (1999). In this case, the recommendations that are released are more favorable than what they should be, on the basis of the research of the analyst. As a result of the biased recommendations from the affiliated analysts, the excess return should be lower than after a buy recommendation from an unaffiliated analyst (Michaely and Womack, 1999).

However, during the IPO process Barclays and Company A build up a relationship, which can lead to superior information about Company A (Allen & Faulhaber, 1989). Based on this superior information, Barclays can make a favorable recommendation for Company A. If the affiliated analysts have more accurate information about a company than unaffiliated analysts, the buy recommendations from these analysts should have more predictive power and the excess returns should logically be higher after a buy recommendation from affiliated analysts than after a buy recommendation from unaffiliated analysts (Michaely and Womack, 1999).

Analysts can give buy, hold and sell recommendations, but in general, analysts give more buy recommendations than sell recommendations (Womack, 1996). To narrow down the research question, only the buy recommendations shall be discussed in this thesis. To narrow down the research question further, only the IPOs at the NYSE and the NASDAQ that date from 2013 will be discussed. Moreover, only the buy recommendations from the top 5 analyst firms, that were made within one year after an IPO will be examined.

To see if the buy recommendations from affiliated analysts are biased or if the affiliated analysts have superior information, the following research question will be answered:

Are post-IPO buy recommendations which are made within one year after an IPO for firms that had an IPO in 2013 by affiliated analysts who work for one of the top five analyst firms, compared to post-IPO buy recommendations from unaffiliated analysts who work for one of the top five analyst firms, leading to different one-year excess returns?

6 The structure of an investment bank is graphically shown in figure 1. The affiliated analysts work in the Research Department, which is a part of the Brokerage Division. This brokerage division is one division of the investment bank. The investment bank also has an investment banking division, the division of the underwriters. Figure 1: Structure investment bank

This research focusses on the buy recommendations that were made in one year after an IPO from the top five analyst firms, for firms that had an IPO in 2013 at the NASDAQ or NYSE. For all buy recommendations the excess returns are calculated and this research shows that excess returns are significantly higher after buy recommendations of affiliated analysts. This result contributes to the literature, because older research seems dated due to the introduction of the SRO regulations and the Global Analyst Research Settlement.

Investment Bank (Barclays)

Investment Banking Division Who: Underwriters

Goal: Complete transactions

Brokerage Division

Company A Goal: Raise money

Unaffiliated analysts

Goal: Create recommendations with predictive power

Research Department Who: Affiliated analysts Goal: Create

recommendations with predictive power

7 In the literature review will be described which investigations are related to this research. Secondly, it will be discussed what analysts and underwriters primarily do. Thirdly, it will be explained in more detail how analysts can generate business for the brokerage division and the investment banking division and how this generation of business can lead to a conflict of interest between analysts and underwriters. Fourthly, it will be discussed how the cooperation between analysts and underwriters can lead to biased recommendations, but also to recommendations with more predictive power due to superior information. Next, the effect of the SRO regulations and the Global Analyst Research Settlement will be highlighted. Finally, the Fama-French 3-factor model will be discussed, after which the hypothesis will be elaborated on.

In subsequent chapters is explained where the data comes from and how the research is conducted. Thereafter, the results of the regressions are presented and discussed. Finally, the results are summarized and suggestions for future research are given in the conclusion section.

8

2. Literature review

Analysts can work for an investment bank or for an independent research firm. When the analysts work for an investment bank, a conflict of interest can arise (Barber et al., 2007).

Most investment banks make their money in different ways. Although these banks offer different services, their main income is generated by the investment banking division and the brokerage division (Jegadeesh et al., 2004). Another division of the investment bank is the research division. This research division is a part of the brokerage division. This research division usually does not generate significant revenue by itself, but it can be used to bring in work for the other divisions (Dugar & Nathan, 1995). Between the investment banking division and the research division a conflict of interest can arise. This conflict of interest will be discussed in this paper.

2.1 Related previous research

Michaely and Womack did similar research and according to their results, the excess returns are different after buy recommendations from affiliated analysts. In their research, the one-year excess returns are significant lower after a buy recommendation from an affiliated analyst (1999). The paper of Barber, Lehavy and Trueman (2007) is also related to this paper. However, their research is focused on the difference between investment banker analysts and independent analysts. The independent analysts are analysts who work for a firm that does not have an investment banking division. Nevertheless, they found evidence which suggest that buy recommendations from independent analysts are leading to more excess returns than buy recommendations from investment banker analysts. In addition, Womack found that buy recommendations in general, lead to significant excess returns in the 3-day event, the 1-month event and the 6-month event (1996).

2.2 Underwriters

On the one hand, the underwriters work at the investment banking division. The goal of the investment banking division is to complete transactions, for example an Initial Public Offering (IPO) (Michaely & Womack, 1999). After an IPO, the status of a firm changes from private to public, which means that the stocks can be traded on the

9 stock exchange and all investors can buy and sell the shares (Ritter & Welch, 2002). Before the company goes public, one or more lead-underwriters are chosen by the company that goes public. These lead- underwriters are investment banks. The underwriters are compensated for this work with fees (Michaely & Womack, 1999). These fees are lucrative for the investment banks (Michaely & Womack, 1999).

2.3 Analysts

On the other hand work the analysts at the research division. These analysts try to help other investors with profitable information. Analysts try to find mispriced stocks and when they have found one, they make a recommendation for that stock (Jegadeesh et al., 2004). Analysts produce primary information about earnings forecasts of other companies (Hong & Kubik, 2003). These analysts usually have a particular industry in which they specialize. These industry specialists collect industry-specific information (Michaely & Womack, 1999). Based on these earnings forecasts, firm-specific information and industry-specific information, the analysts make buy/ sell/ hold recommendations for other firms (Womack, 1996).

An analyst gives a buy recommendation for a security if he or she believes that the security is undervalued (Womack, 1996). In this case, it is believed by the analyst that the price for a security should be higher than the current market price, which means that you can make a profit if you buy a security relatively cheap now and sell it later for more. If an analyst gives a sell recommendation for a security, the analyst believes that the security is overvalued.

However, information provided by analysts itself does not lead to significant revenues for most investment banks (Jegadeesh et al., 2004). Nevertheless, analysts are useful for investment banks. According Krigman, Shaw and Womack analysts are used by the investment banks to generate business for other divisions of the investment bank (2001).

10

2.4 Biased recommendations

Some researchers have already studied the conflict of interest, which arises if an investment bank is responsible for both an Initial Public Offering and a recommendation. Lin and McNichols have found evidence which suggests that recommendations from affiliated analysts are more favorable than recommendations from unaffiliated analysts (1998). An explanation is that analysts try to generate business for the investment bank.

It could be that these investment banks will give more favorable recommendations for firms which they have helped to go public, to improve the relationship between the investment bank and the other company (Dugar & Nathan, 1995). These favorable recommendations can be given to enhance the likelihood that their investment bank will also be chosen as lead underwriter for the next equity offering. If the bank is chosen again as a lead underwriter, the bank earns lucrative fees (Michaely & Womack, 1999). Ljunqvist, Marston and Wilhelm found evidence which suggests that this tactic works (2006).

Analysts can also create new jobs for the investment bank via the brokerage division of the investment bank. This is because investors cannot buy or sell shares directly, which they have to do via a broker. The broker is an intermediate between the buyer and the seller of a share. For this service, a broker receives trading commission. To ensure that more people buy or sell the shares, in order to earn more trading commissions, analysts can give buy or sell recommendations (Dugar & Nathan, 1995). Analysts can give sell recommendations to ensure that more people sell their shares, but this only works if investors have already held the shares or if they sell the shares short. When an investor does not have the ownership of a stock and he or she expects a price decrease of that stock, he or she can sell the stock short. The idea of short-selling is that an investor can lend a stock from a broker. The investor can sell these borrowed stocks for a relative high price and when the investor has to give the stock back to the broker, he or she can buy the share later for a relative low price (Investopedia, 2016). However, this short selling is expensive (Mehran & Stulz, 2007). The result is that analysts give buy recommendations for firms, to ensure that more people buy the shares in order to earn more trading commissions (Dugar & Nathan, 1995). These trading commissions are a main income of an investment bank (Jegadeesh et al., 2004).

11 In addition, it is possible that the analysts may not give negative recommendations for clients of the investment bank. An internal memo from JP Morgan showed that analysts were not allowed to give negative recommendations for clients of the investment bank (Michaely & Womack, 1999). Also the pay-structure of analysts could lead to biased recommendations, since analysts could be paid on the basis of their help to complete transactions. (Michaely & Womack, 1999).

Finally, the biased recommendations can be explained by a cognitive bias. This bias means that affiliated analysts believe that firms that are underwritten by their bank are special. For this reason, they believe that these firms are better than firms that are not underwritten by their bank (Kahneman & Lovallo, 1993). As a result of this cognitive bias, the affiliated analysts have a distorted view and make more positive recommendations for firms which are underwritten by their bank (Michaely & Womack, 1999).

In this case, if the recommendations are not a good reflection of the value of a company, the excess return should be lower after a buy recommendation from an affiliated analyst (Michaely & Womack, 1999).

2.5 Superior information

A reason that affiliated analysts from investment banks give more accurate recommendations, could be because they have more information than the market about a company which they have helped with an IPO (Allen and Faulhaber, 1989). Just after an IPO, the information asymmetry between a company and unaffiliated analysts is the greatest (Michaely & Womack, 1999). Affiliated analysts can profit from this situation, because their bank has built up a relation with the company during the IPO process and has access to firm specific information. Affiliated analysts can further exploit this advantage if they also get access to accurate company-specific information after the IPO of the company to maintain a good relationship (Michaely & Womack, 1999).

This advantage of affiliated analysts is enhanced, because analysts of investment banks follow significant more firms that have an investment banking relationship with their investment bank (Clarke et al., 2007). The data from this study

12 also shows that the analysts from the 4 investment banks that employ underwriters, Barclays, JP Morgan, Morgan Stanley and Bank of America Merrill Lynch, issue more buy recommendations for firms with an investment banking relationship, details of which can be found in Table 1.

Table 1: Number of recommendations from affiliated analysts and number of recommendations from unaffiliated analysts

Affiliated Unaffiliated Total

Barclays 24 10 34

JP Morgan 48 3 51

Morgan Stanley 32 6 38

Evercore ISI 0 11 11

Bank of America Merrill Lynch 63 17 80

Total 167 47 214

Moreover, investment banks have the resources to attract the best analysts, which can lead to recommendations with more predictive power compared to recommendations from non-investment bank analysts (Boni & Womack, 2003). If the affiliated analysts have more accurate information than unaffiliated analysts have, the predictive power of their recommendations should be higher and the excess returns should be higher from a buy recommendation from an affiliated analyst compared to a buy recommendation from an unaffiliated analyst (Michaely & Womack, 1999).

2.6 SRO regulations and Global Analyst Research Settlement

To reduce the dependency of the research department from the investment banking department, the Self-Regulatory Organization (SRO) regulations were implemented in the summer of 2002 and the Global Analyst Research Settlement was introduced in December 2002.

After the implementation of the SRO regulations it was not allowed anymore to compensate analysts on the basis of ‘’helpfulness’’ for completing transactions. Another rule is the extension of the quiet period of 25 days to 40 days (Kadan et al., 2009). During these 40 days after an IPO are analysts not allowed to issue

13 recommendations for a company that went public. The extension of the period should lead to less biased recommendations from affiliated analysts, because the pressure on affiliated analysts to make favorable recommendations is greatest right after the IPO (Bradley et al., 2003). Moreover, the subject firm may not alter the recommendation before it is published. The subject firm is the firm for which a rapport is written by analysts. In addition, analysts should give more information about themselves and their company when they issue a recommendation. Furthermore, analysts should mention if they are an officer of director at the firm for which they make the recommendation. Moreover, analysts should mention if their salary depends on the revenue of the investment bank. In addition, whether subject firm has an investment banking relationship with the firm of the analyst should be mentioned (Kadan et al., 2009).

The purpose of the Global Settlement was to separate the research department and the investment banking department further. This settlement is an agreement between different parties, including the NASD, NYSE and the top 10 investment banks which include J.P. Morgan, Merrill Lynch and Morgan Stanley (Corwin et al., 2015). After this agreement went into effect, the research department and the investment banking department must be physically separated (Kadan et al., 2009). Moreover, the 10 investment banks had to pay $1.2 billion penalties (Barber et al., 2007). According to several researchers the new regulations led to less dependent analysts (C.Y. Chen & P.F. Chen, 2009) (Kadan et al., 2009).

2.7 Fama-French 3-factor model

The Fama- French 3-factor model is used to calculate the excess returns after buy recommendations (Womack, 1996). This model is based on the Capital Asset Pricing Model (CAPM). The formula for the CAPM is:

𝑅𝑒 = 𝑅𝑓 + 𝛽(𝑅𝑚 − 𝑅𝑓) + 𝜀

In the CAPM, the expected return of a stock 𝑅_𝑒 is explained by the risk-free rate 𝑅_𝑓, plus a risk premium (𝛽(𝑅_𝑚− 𝑅_𝑓))This risk premium is the β of a firm, multiplied by the market risk premium. β is a parameter that measures how sensitive an asset is for changes in market return and the market risk premium is the difference between the expected return of the market and the risk free rate (Gaunt, 2004). However, the CAPM

14 has some limitations. The β in the CAPM ignores several returns for large and small companies. Moreover, the β ignores different returns for firms with different book-to-market ratios (Fama & French, 1996). However, small firms perform better than large firms overall (Banz, 1981). Furthermore, firms with high book-to-market ratios generate in general higher returns (Lakonishok et al., 1994). Therefore, Fama and French introduced a new model, which included factors for these firm size and book-to-market ratios. This model is better able to determine the cross-section of returns than CAPM (Fama & French, 1993). The formula for the Fama- French 3 factor model is:

𝑅𝑖− 𝑅𝑓 = 𝛼𝑖+ 𝑏𝑖(𝑅𝑚− 𝑅𝑓) + 𝑠𝑖𝑆𝑀𝐵 + ℎ𝑖𝐻𝑀𝐿 + 𝜀𝑖

In this model, the return of a stock is explained by the excess return (α), the market premium (𝑅_𝑚− 𝑅_𝑓), the spread between returns of big and small firms (𝑆𝑀𝐵) and the spread between returns of firms with a high book-to-market ratio and firms with a low book-to-market ratio (𝐻𝑀𝐿). To calculate the market premium, the returns of a value-weighted portfolio, including stocks of the NYSE, AMEX and NASDAQ, are calculated and subsequently the risk free rate is subtracted. To calculate the 𝑆𝑀𝐵 factor, six portfolios are created of which three contain several small stocks and three contain several big stocks. Next, the returns of the big firms are subtracted from the returns of the small firms (French, 2016). The formula is given below:

𝑆𝑀𝐵 =1

3(𝑆𝑚𝑎𝑙𝑙 𝑉𝑎𝑙𝑢𝑒 + 𝑆𝑚𝑎𝑙𝑙 𝑁𝑒𝑢𝑡𝑟𝑎𝑙 + 𝑆𝑚𝑎𝑙𝑙 𝐺𝑟𝑜𝑤𝑡ℎ) − 1

3 (𝐵𝑖𝑔 𝑉𝑎𝑙𝑢𝑒 + 𝐵𝑖𝑔 𝑁𝑒𝑢𝑡𝑟𝑎𝑙 + 𝐵𝑖𝑔 𝐺𝑟𝑜𝑤𝑡ℎ)

Finally, to calculate the factor 𝐻𝑀𝐿, four portfolios are created of which two contain several value stocks and the other two contain several growth stocks. Thereafter, the returns of the growth stocks are subtracted from the returns of the value stocks. The formula (French, 2016) is given below:

𝐻𝑀𝐿 =1 2(𝑆𝑚𝑎𝑙𝑙 𝑉𝑎𝑙𝑢𝑒 + 𝐵𝑖𝑔 𝑉𝑎𝑙𝑢𝑒) − 1 2(𝑆𝑚𝑎𝑙𝑙 𝐺𝑟𝑜𝑤𝑡ℎ + 𝐵𝑖𝑔 𝐺𝑟𝑜𝑤𝑡ℎ)

3. Data

This thesis only studies buy recommendations made in 2013 and 2014 for firms that had an IPO in 2013. All the data that has been used is accessible on public websites.

15 The data of the firms that had an IPO at the NASDAQ and NYSE during 2013 are available at the site of the NASDAQ (NASDAQ, 2016). Further, information about the lead underwriters for these IPOs, which are Barclays, JP Morgan, Morgan Stanley and Bank of America Merrill Lynch, come from Zephyr (Zephyr, 2016). Additionally, the list of the top 5 Research Firms includes Barclays, JP Morgan, Morgan Stanley, Bank of America Merrill Lynch and Evercore ISI sources from Institutional Investor (Institutional Investor, 2016). All recommendations are available at Institutional Brokers’ Estimate System (I/B/E/S) (Kadan et al., 2009). I/BE/S is a part of Wharton Research Data Services (WRDS). Moreover, daily stock prices, cash dividends, number of shares outstanding and SIC Codes are sourced from The Center for Research in Security Prices (CRSP) (CRSP, 2016). CRSP data is available via WRDS. Finally, the daily data of 𝑆𝑀𝐵, 𝐻𝑀𝐿, market returns and the risk-free rates for the Fama-French model are available on Kenneth French’ website (French, 2016).

In table 2, the minima, maxima, means, standard deviations and number of observations of the dependent and independent variables are presented.

Table 2: Descriptive statistics

Variable Minimum Maximum Mean Std. Deviation Observations

ER -0,60 0,69 0,0269 0,149 214 UR 0 1 0,7804 0,415 214 Size 0,00 0,02 0,0031 0,005 214 DFirst 0 1 0,5935 0,492 214 DMining 0 1 0,0373 0,190 214 DConstruction 0 1 0,0093 0,096 214 DManufacturing 0 1 0,0981 0,298 214 DTransportation 0 1 0,0981 0,298 214 DWholesale 0 1 0,0281 0,298 214 DRetail 0 1 0,0467 0,211 214 DFinaneInsuranceRealEstate 0 1 0,1215 0,327 214 DServices 0 1 0,2617 0,441 214 DPublicAdministration 0 1 0,2991 0,459 214

16 the explanations for variables is in the 8.2 variable list.

4. Methodology

4.1 Models

To narrow down the research question, the researched firms that make buy recommendations and the researched firms that receive buy recommendations are limited. The researched firms that make buy recommendations are the top five analysts firms, which are selected by Institutional Investor. Institutional Investor makes annual lists of the best Research Teams, based on opinions of more than 3800 money managers and more than 1000 firms (Womack, 1996). The top five Research Teams are Barclays, Bank of America Merrill Lynch, JP Morgan, Morgan Stanley and Evercore ISI (Institutional Investor, 2016). In this sample, Barclays, Bank of America Merrill Lynch, JP Morgan and Morgan Stanley were at least once a lead underwriter for a firm that had an IPO at the NASDAQ or NYSE during 2013. However, Evercore ISI was not a lead underwriter for a firm that had an IPO at the NASDAQ or NYSE during 2013.

First, all IPOs at the NASDAQ and NYSE in 2013 are collected. Thereafter, the investment banks that were lead underwriters are hand-matched to the firms that had an IPO. The co-lead managers and lead managers are referred to as lead underwriters (Barber et al., 2007). Afterwards, the relevant recommendations are collected. The relevant recommendations are all buy recommendations that were made in one year after an IPO, by analysts from the top five analysts firms, for firms that had an IPO in 2013 at the NASDAQ or the NYSE (Michaely & Womack, 1999). Next, for each individual relevant recommendation is examined whether the recommendation has been made by an affiliated or by an unaffiliated analyst.

The relevant buy recommendations are divided into two groups. The first group consists of the buy recommendations from unaffiliated analysts for firms that had an IPO during 2013. The other group comprises the buy recommendations from affiliated analysts.

Subsequently, for each individual buy recommendation the one-year excess return is calculated, started at the day before the buy recommendation (Michaely & Womack, 1999). This one-year excess return is calculated by using a Fama-French model (Barber et al., 2007).

17 In this regression, α is the variable of interest. The α is the excess return, 𝑅_𝑖 is the raw return of a stock, inclusive dividends. 𝑅𝑓 is the risk-free rate, 𝑅𝑚 is the return

of a broad market portfolio, 𝑆𝑀𝐵 (small minus big) is a factor for the difference between returns of big stocks and small stocks and 𝐻𝑀𝐿 (high minus low) is a factor for the difference between returns of stocks with a high book-to-market ratio and stocks with a low book-to-market ratio (Fama & French, 1995).

Finally, these excess returns are used in another regression to test if the buy recommendations from affiliated analysts are leading to different excess returns, compared to excess returns after buy recommendations from unaffiliated analysts. Here is used an OLS estimation procedure. The other model, used by Michaely and Womack (1999), is:

𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝛽3𝐷𝐹𝐼𝑅𝑆𝑇 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑒 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀

The dependent variable is the one-year excess return after a buy recommendation. As the excess returns are often consistently low, they are multiplied by 100 to make interpretation more easy. The variable of interest is a dummy variable which has value 1 if an affiliated analyst made the recommendation and has value 0 if an unaffiliated analysts made the buy recommendation. The first control variable (SIZE) is the the market capitalization in billions at the end of the quiet period, which is calculated by the formula (Michaely & Womack, 1999) as below:

𝑆𝑖𝑧𝑒 = (𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠ℎ𝑎𝑟𝑒𝑠 𝑜𝑢𝑡𝑠𝑡𝑎𝑛𝑑𝑖𝑛𝑔 ∗ 𝑆ℎ𝑎𝑟𝑒 𝑃𝑟𝑖𝑐𝑒)/ 1000.000.000

The quiet period ends 40 calendar days after an IPO (Bradley et al., 2003). To calculate the market capitalization, this moment is chosen because analysts are allowed to issue recommendations for a company since that moment. Therefore, this is the first moment investors get access to the opinions of analysts about firm values (Michaely & Womack, 1999). This control variable is added, because previous research has shown that the long-run performance after an IPO is related to size (Ritter, 1991). Moreover, previous research has shown that the post-IPO performance of a small company is higher than the post-IPO performance of a big company (Reinganum, 1981). The second control variable is a dummy variable DFIRST, which has value 1 if the recommendation is the first recommendation after the IPO (Michaely & Womack, 1999).

18 Other variables that are included in the model are Industry dummies, based on SIC codes (Michaely & Womack, 1999). These dummies are included as there may be differences between different industries for the supply and demand of recommendations (Bhushan, 1989). SIC codes divide companies into separate business categories. The ten different businesses are: Agriculture/Forestry/ Fishing, Mining, Construction, Manufacturing, Transportation & Public Utilities, Wholesale Trade, Retail Trade, Finance/Insurance/Real Estate, Services and Public Administration (SICCODES, 2016). No firm with a code that matched with the business Agriculture/Forestry/Fishing is included in the sample. All sample firms are divided into the other 9 business categories.

A t-test is used to see if the variables have a significant effect on the one-year excess return after a buy recommendation.

4.2 Hypothesis

The hypothesis is whether buy recommendations of affiliated analysts are leading to the same excess returns, compared with excess returns after buy recommendation of unaffiliated analysts or are leading to different excess returns. Here the null hypothesis implies that excess returns after buy recommendations of affiliated analysts are not significantly different than excess returns after buy recommendations of unaffiliated analysts. If the null hypothesis is not rejected, it means that excess returns after buy recommendations of affiliated analysts are not significantly different than excess returns after buy recommendations of unaffiliated analysts. Contrary, if the null hypothesis is rejected, it means that excess returns after buy recommendations of affiliated analysts are significantly different than excess returns after buy recommendations of unaffiliated analysts.

Statistical hypothesis is:

𝐻0: 𝛽1 = 0

𝐻₁: 𝛽₁ ≠ 0

This thesis employs four models to explain the research question as follows.

19 𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝜀

𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀

𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝛽3𝐷𝐹𝐼𝑅𝑆𝑇 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀

 ER: One-year excess return after a buy recommendation, multiplied by 100  UR: Dummy variable which has value 1 if a recommendation is made by an

affiliated analyst

 SIZE: Market Capitalization of a firm in billions at the end of the quiet period  Industry dummies: Ten dummy variables that divide companies into ten

different businesses, based on SIC codes

 DFIRST: Dummy variable which has value 1 if the buy recommendation is the first buy recommendation after an IPO

In these models, the variable of interest is the first explanatory variable 𝑈𝑅. The results of these four models will be described in the next chapter.

Based on previous background and research, I expect that the one-year excess returns after buy recommendations from affiliated analysts and unaffiliated analysts are not the same, even under the circumstances of the SRO regulations and the Global Analyst Research Settlement. This implies that I expect that the null hypothesis will be rejected, since previous studies show the excess returns can be different due to biased recommendations from the affiliated analysts (Michaely & Womack, 1999) and the superior information that the affiliated analysts have (Allen & Faulhaber, 1989).

5

Results

In table 3 the regression results are described. In model 1, the one-year excess return is explained by a dummy underwriter with the following regression (model 1):

20 The coefficient of underwriter is 0.064, which is significantly at 1% level.

In model 2, the one-year excess return is explained by a dummy underwriter and the size of a firm with the following regression (model 2):

𝐸𝑅 = 𝛽₀+ 𝛽₁𝑈𝑅 + 𝛽₂𝑆𝐼𝑍𝐸 + 𝜀

The coefficient of underwriter is 0.057, which is significantly at 5% level. Moreover, the coefficient of size is 5.478, which is significantly at 1% level.

In model 3, the one-year excess returns are explained by a dummy underwriter, the size of a firm and dummies for several industries with the following regression (model 3):

𝐸𝑅 = 𝛽₀+ 𝛽₁𝑈𝑅 + 𝛽₂𝑆𝐼𝑍𝐸 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀

The coefficient of underwriter is 0.050, which is significantly at 5% level. Moreover, the coefficient of size is -4.453, which is significantly at 5% level. In addition, the coefficient of DMining is -0.032 and the coefficient of DWholesale is -0,090, which are not significantly different from zero. Moreover, the coefficient of DConstruction is -0.255, the coefficient of DFinanceInsuranceRealEstate is -0.115 and the coefficient of DPublicAdministration is -0.107. These coefficients are significantly at 5% level. Finally, the coefficient of DManufacturing is -0.175, the coefficient of DTransportation is -0.194 and the coefficient of DServices is -0.167, which are significantly at 1% level.

In model 4, the one-year excess returns are explained by a dummy underwriter, the size of a firm, dummies for several industries and a dummy for the first recommendation after an IPO with the following regression (model 4):

𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝛽3𝐷𝐹𝐼𝑅𝑆𝑇 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀

The coefficient of underwriter is 0.050, which is significantly at 5% level. Moreover, the coefficient of size is -4.567, which is significantly at 5% level and the coefficient of DFirst is -0.017, which is not significantly different from zero. In addition, the coefficient of DMining is -0.026 and the coefficient of DWholesale is -0,090, which are not significantly different from zero. Moreover, the coefficient of DConstruction is -0.247, the coefficient of DFinanceInsuranceRealEstate is -0.114 and the coefficient of

21 DPublicAdministration is -0.104. These coefficients are significantly at 5% level. Finally, the coefficient of DManufacturing is -0.173, the coefficient of DTransportation is -0.192 and the coefficient of DServices is -0.104, which are significantly at 1% level.

Table 3: Regression results

In column Model 1, Model 2, Model 3, Model 4, the regression models are 𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝜀, 𝐸𝑅 =

𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝜀, 𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 + 𝛽2𝑆𝐼𝑍𝐸 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀, 𝐸𝑅 = 𝛽0+ 𝛽1𝑈𝑅 +

𝛽2𝑆𝐼𝑍𝐸 + 𝛽3𝐷𝐹𝐼𝑅𝑆𝑇 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀, respectively. The excess return is calculated by

Fama-French three factor model, which is 𝑅𝑖− 𝑅𝑓 = 𝛼𝑖+ 𝑏𝑖(𝑅𝑚− 𝑅𝑓) + 𝑠𝑖𝑆𝑀𝐵 + ℎ𝑖𝐻𝑀𝐿 + 𝜀𝑖. The results

are reported for buy recommendations, made in 2013 and 2014 by the top five analyst firms for firms that had an IPO at the NASDAQ or NYSE during 2013. The t statistics test the null hypothesis that the coefficient is significantly different from zero.

Dependent variable: ER

Model 1 Model 2 Model 3 Model 4

UR 0.064 0.057 0.050 0.050 (2.625)*** (2.369)** (2.156)** (2.147)** SIZE -5.478 -4.453 -4.567 (-2.834)*** (-2.112)** (-2.161)** Dfirst -0.017 (-.869) DMining -0.032 -0,026 (-0.485) (-0.387) DConstruction -0.255 -0.247 (-2.381)** (-2.294)** DManufacturing -0.175 -0.173 (-3.282)*** (-3.245)*** DTransportation -0.194 -0.192 (-3.637)*** (-3.601)*** DWholesale -0.090 -0.090 (-1.260) (-1.261) DFinanceInsurance RealEstate -0.115 (-2.225)** -0.114 (-2.202)** DServices -0.167 -0.167 (-3.491)*** (-3.487)*** DPublic Administration -0.107 (-2.266)** -0.104 (-2.209)** Constant -.023 0.000 0.133 0.142 (-1.060) (-0.018) (2.817)*** (2.937)*** N 214 214 214 214 R2 _{0.031 0.067 0.179 0.182}

22

Adjusted R2 _{0.027 0.058 0.138 0.137}

***: significance at 1% level **: significance at 5% level *: significance at 10% level

Numbers between the brackets are t-values

The detailed explanations for variables is in the 8.2 variable list.

In all regressions the variable Underwriter has a statistically significant positive effect on the one-year excess return after a buy recommendation. This contradicts older research, since other older research showed significant worse performance of stocks that were recommended by affiliated analysts (Michaely & Womack, 1999). However, the results are not economically significant. If an affiliated analyst makes a buy recommendation instead of unaffiliated analysts, this leads to about 0,05% points more excess return (ceteris paribus). The results are only economically significant if they can lead to an investment strategy (Pesaran & Timmermann, 1995).

An explanation for this positive effect of affiliated analysts, is the impact of the SRO regulations and the Global Analyst Research Settlement. After the implementations of these SRO regulations and the Global Analyst Research Settlement, analysts were no longer paid on the basis of their help to bring in new jobs. Moreover, the investment banking department and the research department were physically separated within the investment banks, and the subject firm was no longer allowed to change a recommendation of the analysts before publication (Kadan et al., 2009). As a result of these new rules, affiliated analysts are less dependent on underwriters, while they still have access to firm-specific information which is gathered by underwriters during an IPO-process (C.Y. Chen & P.F. Chen, 2009). The data of the regression indicates that affiliated analysts are better able to determine the value of companies.

Further, the size of a firm influences for the excess return after a buy recommendation. The factor SIZE has a significant negative effect on excess return. If the market capitalization of a firm increases by $1 billion, then the excess return after a buy recommendation decreases by about 5% points (ceteris paribus). The effect of SIZE is explained by the SIZE-factor effect. This SIZE-factor effect implies that small firms, in general, perform better than big firms (Dimson & Marsh, 1986). Another explanation for the size effect is that bad news for small firms reaches investors harder. If there is bad news for a large firm within one year after the buy recommendation, the

23 analysts will notice this and will make it known to investors. If this happens, the share price will decrease and thus, the excess return will decrease as well. Moreover, if there is bad news for a small firm within one year after the buy recommendation, the analysts will probably not notice this, since analysts generally follow more big firms than small firms (Bhushan, 1989). The managers of the small firm prefer high stock prices and therefore they will not reveal the bad news to investors. In contrast, if analysts are not updated on good news of a small company, analysts can deduce the good news by themselves. Under such circumstances only the good news of small companies reaches the investors. In that case, the one-year excess return after a buy recommendation for a small firm will be higher (Hong et al., 2000).

If a buy recommendation is the first buy recommendation after an IPO for a firm, the one-year excess return is 0,017% points lower than the one-year excess return after a buy recommendation that is not the first buy recommendation after an IPO (ceteris paribus). However, this effect is not significant. This negative insignificant effect is in accordance with previous research (Michealy & Womack, 1996).

The data indicates that there are differences in excess returns after buy recommendations in various industries. The retail industry is the control industry in the regression. Buy recommendations for firms in the construction industry, the manufacturing industry, the transportation industry, the Finance/ Insurance/ Real Estate industry, Services industry and Public administration industry are leading to significant less one-year excess return, compared with one-year excess return after a buy recommendation for a firm in the retail industry (Ceteris Paribus). An explanation for the different excess returns after buy recommendations in various industries is that it is harder for some industries to acquire valuable information than for other industries (Bhushan, 1989). Analysts use this industry-specific information to make recommendations (Womack, 1996).

6. Conclusion

In this paper is examined whether buy recommendations from affiliated analysts, who work for one of the top five analyst firms, are leading to different excess returns, compared with buy recommendations from unaffiliated analysts, who work for one of the top five analyst firms. This study shows that the one-year excess returns differ significantly after buy recommendations from affiliated analysts compared to buy

24 recommendations from unaffiliated analysts. In previous research, the excess returns after buy recommendations from affiliated analysts were lower due to biased recommendations (Michaely & Womack, 1996). In this research, the excess returns are significantly higher after buy recommendations from affiliated analysts. An explanation for this change is the introduction of the SRO regulations and the Global Analyst Research Settlement which analysts have become less dependent on underwriters (Kadan et al., 2009).

This study also shows that the one-year excess return after a buy recommendation decreases by about 5% points if the market capitalization of a firm increases by 1 billion (ceteris paribus). This negative relation is in accordance with previous research (Dimson & Marsh, 1986).

Moreover, the effect of various industries on excess returns after buy recommendations is researched in this paper. It appears that buy recommendations for firms in the construction industry, the manufacturing industry, the transportation industry, the Finance/ Insurance/ Real Estate industry, the Services industry and the Public administration industry are leading to significant less one-year excess return, compared to one-year excess return after a buy recommendation for a firm in the retail industry (ceteris paribus). This is because for some industries it is harder to collect industry-specific information than for others (Bhushan, 1989), which is needed to produce recommendations (Womack, 1996).

Finally, if a recommendation is the first recommendations after an IPO, this does not lead to significant different excess return, compared with a recommendation which is not the first recommendation after an IPO. The effect is negative, but not significant. This negative insignificant effect is in accordance with previous research (Michaely & Womack, 1996).

However, this research has some limitations. The conclusions have been drawn on the basis of 214 recommendations. Moreover, only one-year post-IPO buy recommendations from the top 5 research department for firms that had an IPO at the NASDAQ or NYSE in 2013 have been researched. Nevertheless, this research contributes to the literature, because older research appears dated by the introduction of the SRO regulations and the Global Analyst Research Settlement. My suggestion for further research is to research the effect of the SRO regulations and the Global Analyst Research Settlement on biased buy recommendations from affiliated analysts. In my view, it will take several years before the new rules are implemented effectively.

25 Therefore, my suggestion for further research is a research that compares the period from introduction of the new rules to as recent as possible, to the period before the introduction of the new rules with an event study. In this case, the research is based on sufficient observations and it ensures that the results are not dated. Here, the event regards the introduction of the Global Analysts Research Settlement. For this event a dummy variable can be used, which has value 1 if a recommendation is made after the introduction. A recommended research question is: Are post-IPO excess returns after buy recommendations from affiliated analysts different after the introduction of the Global Analyst Research Settlement?

To conclude, as shown by the data, post-IPO buy recommendations from affiliated analysts lead to significant different excess returns, compared to one-year excess returns after post-IPO buy recommendations from unaffiliated analysts. If an affiliated analyst makes a buy recommendation instead of an unaffiliated analyst, this leads to about 0,05% points more excess return (ceteris paribus). Finally, it can be concluded that affiliated analysts make recommendations with more predictive power instead of biased recommendations, based on superior information.

26

7. Reference list

Allen, F., & Faulhaber, G. R. (1989). Signalling by underpricing in the IPO market. Journal of financial Economics, 23(2), 303-323.

Banz, R. W. (1981). The relationship between return and market value of common stocks. Journal of financial economics, 9(1), 3-18.

Barber, B. M., Lehavy, R., & Trueman, B. (2007). Comparing the stock

recommendation performance of investment banks and independent research firms. Journal of Financial Economics, 85(2), 490-517.

Barclays. (n.d.) Retrieved from http://www.investmentbank.barclays.com/

Bhushan, R. (1989). Firm characteristics and analyst following. Journal of Accounting and Economics, 11(2), 255-274.

Boni, L., & Womack, K. L. (2003). Wall street research: will new rules change its usefulness?. Financial Analysts Journal, 59(3), 25-29.

Bradley, D. J., Jordan, B. D., & Ritter, J. R. (2003). The quiet period goes out with a bang. The Journal of Finance, 58(1), 1-36.

Bradshaw, M. T., Richardson, S. A., & Sloan, R. G. (2006). The relation between corporate financing activities, analysts’ forecasts and stock returns. Journal of Accounting and Economics, 42(1), 53-85.]

Chen, C. Y., & Chen, P. F. (2009). NASD Rule 2711 and changes in analysts' independence in making stock recommendations. The Accounting Review, 84(4), 1041-1071.

Clarke, J., Khorana, A., Patel, A., & Rau, P. R. (2007). The impact of all-star analyst job changes on their coverage choices and investment banking deal flow. Journal of Financial Economics, 84(3), 713-737.

27 Corwin, S. A., Larocque, S., & Stegemoller, M. (2015). Investment Banking

Relationships and Analyst Affiliation Bias: The Impact of Global Settlement on Sanctioned and Non-Sanctioned Banks. Available at SSRN.

CRSP. (2016, May 16) CRSP DAILY STOCK. Retrieved from https://wrds-web.wharton.upenn.edu/wrds/ds/crsp/stock_a/dsf.cfm?navId=128

Description of Fama/ French Benchmark Factors. (n.d.) Retrieved from http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/Data_Library/f-f_bench_factor.html

Dimson, E., & Marsh, P. (1986). Event study methodologies and the size effect: The case of UK press recommendations. Journal of financial Economics, 17(1), 113-142.

Dugar, A., & Nathan, S. (1995). The effect of investment banking relationships on financial analysts' earnings forecasts and investment

recommendations*.Contemporary Accounting Research, 12(1), 131-160.

Ellis, K., Michaely, R., & O'hara, M. (2000). When the underwriter is the market maker: An examination of trading in the IPO aftermarket. The Journal of Finance, 55(3), 1039-1074.

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of financial economics, 33(1), 3-56.

Fama, E. F., & French, K. R. (1995). Size and book‐to‐market factors in earnings and returns. The Journal of Finance, 50(1), 131-155.

Fama, E. F., & French, K. R. (1996). The CAPM is wanted, dead or alive. The Journal of Finance, 51(5), 1947-1958.

28 Gaunt, C. (2004). Size and book to market effects and the Fama French three factor asset pricing model: evidence from the Australian stockmarket. Accounting &

Finance, 44(1), 27-44.

Hong, H., & Kubik, J. D. (2003). Analyzing the analysts: Career concerns and biased earnings forecasts. The Journal of Finance, 58(1), 313-351.

Hong, H., Lim, T., & Stein, J. C. (2000). Bad news travels slowly: Size, analyst coverage, and the profitability of momentum strategies. The Journal of Finance, 55(1), 265-295.

IPO. (n.d.) Retrieved from

http://www.investopedia.com/university/ipo/ipo.asp

IPO Activity. (n.d.) Retrieved from

http://www.nasdaq.com/markets/ipos/activity.aspx?tab=pricings&month=2013-01

Jegadeesh, N., Kim, J., Krische, S. D., & Lee, C. (2004). Analyzing the analysts: When do recommendations add value?. The journal of finance, 59(3), 1083-1124.

Kadan, O., Madureira, L., Wang, R., & Zach, T. (2009). Conflicts of interest and stock recommendations: The effects of the global settlement and related regulations.

Review of Financial Studies, 22(10), 4189-4217.

Kahneman, D., & Lovallo, D. (1993). Timid choices and bold forecasts: A cognitive perspective on risk taking. Management science, 39(1), 17-31.

Kecskes, A., Michaely, R., & Womack, K. L. (2015). Do Earnings Estimates Add Value to Sell-Side Analysts' Investment Recommendations?. Management Science, Forthcoming.

29 Kenneth. R. French- Data library (May 2, 2016). Fama/French 3 Factors [daily]. Retrieved from

http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html

Krigman, L., Shaw, W. H., & Womack, K. L. (2001). Why do firms switch underwriters?. Journal of Financial Economics, 60(2), 245-284.

Lakonishok, J., Shleifer, A., & Vishny, R. W. (1994). Contrarian investment, extrapolation, and risk. The journal of finance, 49(5), 1541-1578.

Lin, H. W., & McNichols, M. F. (1998). Underwriting relationships, analysts' earnings forecasts and investment recommendations. Journal of Accounting and Economics, 25(1), 101-127.

Ljungqvist, A., Marston, F., & Wilhelm, W. J. (2006). Competing for securities underwriting mandates: Banking relationships and analyst recommendations.The Journal of Finance, 61(1), 301-340.

Mehran, H., & Stulz, R. M. (2007). The economics of conflicts of interest in financial institutions. Journal of Financial Economics, 85(2), 267-296.

Michaely, R., & Shaw, W. H. (1994). The pricing of initial public offerings: Tests of adverse-selection and signaling theories. Review of Financial studies, 7(2), 279-319.

Michaely, R., & Womack, K. L. (1999). Conflict of interest and the credibility of underwriter analyst recommendations. Review of financial studies, 12(4), 653-686.

Pesaran, M. H., & Timmermann, A. (1995). Predictability of stock returns:

Robustness and economic significance. The Journal of Finance, 50(4), 1201-1228.

Reinganum, M. R. (1981). Misspecification of capital asset pricing: Empirical anomalies based on earnings' yields and market values. Journal of financial Economics, 9(1), 19-46.

30 Ritter, J. R., & Welch, I. (2002). A review of IPO activity, pricing, and allocations. The Journal of Finance, 57(4), 1795-1828.

Ritter, J. R. (1991). The long‐run performance of initial public offerings. The journal of finance, 46(1), 3-27.

Short selling. (n.d.) Retrieved from

http://www.investopedia.com/university/shortselling/shortselling1.asp

SIC CODES. (n.d.) Retrieved from http://siccode.com/

The 2016 All- Europe Research Team. (n.d.) Retrieved from

http://www.institutionalinvestor.com/Research/6088/Overview.html#.Vw5LCkeDtf4

Womack, K. L. (1996). Do brokerage analysts' recommendations have investment value?. The journal of finance, 51(1), 137-167.

Zephyr. (n.d.) Retrieved from https://zephyr-bvdinfo-

com.proxy.uba.uva.nl:2443/version-201663/Search.EndUser.serv?_CID=11&context=353UCHYUPCHZFBM&UIMode=S treamLined

31

8. Appendix

8.1 Correlation Matrix

ER UR SIZE FIRST MININ

G CONS TRUC TION MAN UFAC TURI NG TRAN SPOR TATI ON WHOL ESAL E RETA IL FINANC E SERVIC ES PUBLIC ADMINI STRATI ON ER 1 ,177** -,205** -,017 ,154* -,067 -,07 -,122 ,067 ,187* * ,067 -,231** ,139* UR ,177** 1 -,101 ,021 ,105 ,052 ,061 -,053 ,09 -,043 ,059 -,121 ,001 SIZE -,205** -,101 1 -,122 -,041 -,045 -,112 -,151* -,04 ,029 -,103 ,474** -,183** FIRST -,017 ,021 -,122 1 ,113 ,08 ,017 ,017 -,032 -,042 -,013 -,135* ,084 MINING ,154* ,105 -,041 ,113 1 -,019 -,065 -,065 -,033 -,044 -,073 -,117 -,129 CONSTRUCTI ON -,067 ,052 -,045 ,08 -,019 1 -,032 -,032 -,016 -,022 -,036 -,058 -,063 MANUFACTU RING -,07 ,061 -,112 ,017 -,065 -,032 1 -,109 -,056 -,073 -,123 -,196** -,215** TRANSPORTA TION -,122 -,053 -,151* ,017 -,065 -,032 -,109 1 -,056 -,073 -,123 -,196** -,215** WHOLESALE ,067 ,09 -,04 -,032 -,033 -,016 -,056 -,056 1 -,038 -,063 -,101 -,111 RETAIL ,187** -,043 -,029 -,042 -,044 -,022 -,073 -,073 -,038 1 -,082 -,132 -,145* FINANCE ,067 ,059 -,103 -,013 -,073 -,036 -,123 -,123 -,063 -,082 1 -,221** -,243** SERVICES -,231** -,121 ,474** -,135* -,117 -,058 -,196** -,196** -,101 -,132 -,221** 1 -,389** PUBLICADMIN ISTRATION ,139* ,001 -0,183** ,084 -,129 -,063 -,215** -,215** -,111 -,145* -,243** -,389** 1 **: significance at 5% level *: significance at 10% level

32

8.2 Variable List

ER One-year excess return after a buy recommendation, multiplied by 100

UR Dummy variable which has value 1 if a recommendation is made by an affiliated analyst

SIZE Market Capitalization of a firm in billions at the end of the quiet period

DFirst Dummy variable which has value 1 if the buy

recommendation is the first buy recommendation after an IPO

DMining Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Mining industry and 0 otherwise

DConstruction Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Construction industry and 0 otherwise

DManufacturing Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Manufacturing industry and 0 otherwise

DTransportation Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Transportation industry and 0 otherwise

DWholesale Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Wholesale industry and 0 otherwise

DRetail Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Retail industry and 0 otherwise

DFinanceInsuranceRealEstate Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Finance industry or the Insurance industry or the Real Estate industry and 0 otherwise

DServices Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Services industry and 0 otherwise

DPublicAdministration Dummy variable which has value 1 if a buy

recommendation is made for a company that is active in the Public Administration industry and 0 otherwise Industry dummies Ten dummy variables that divide companies into ten

different businesses, based on SIC codes

33 Ri The raw return of a stock, inclusive dividends

Rf The risk-free rate

Rm The return of a broad market portfolio

SMB A factor for the difference between returns of big stocks and small stocks

HML A factor for the difference between returns of stocks with a high market ratio and stocks with a low book-to-market ratio