Options
Dubois, Michel
Nom
Dubois, Michel
Affiliation principale
Fonction
Professeur ordinaire
Email
michel.dubois@unine.ch
Identifiants
Résultat de la recherche
Voici les éléments 1 - 3 sur 3
- PublicationAccès libreThree essays on financial analysts' performance(2016)
; This dissertation is composed of three chapters. The first chapter explores the importance of previously identified factors in explaining the variation in analysts’ earnings forecast error. As earnings forecasts are the main input in determining price targets and consequently stock recommendations, much of the process through which analysts process their input remains in a so-called “black box”. This study attempts to shed light on these inputs. First, it reveals that forecast errors are stable over time, and analysts do not efficiently integrate past information in their forecasts. Second, analysts do not factor in expectations related to the macroeconomic conditions for the underlying forecast horizon. Analysts overreact (underreact) to positive (negative) macroeconomic expectations on both GDP and consumer sentiment index. Third, this study decomposes analysts’ forecast errors variance by observable characteristics and fixed effects. Importantly, the analysis shows that there is an unobserved, time-invariant component related to the firm-analyst dimension that explains much of the variance in the forecast errors. This component is not yet captured by the existing observable characteristics which, at date, have a trifling effect on their own in explaining the variation in analysts’ forecast error.
In the second chapter, I investigate the role of financial reporting frequency in analysts’ earnings forecasts. I addresses two questions. First, does mandatory quarterly reporting benefit financial analysts in decreasing their earnings forecast error and dispersion? Second, to what extent common accounting standards increase the convergence of analysts’ information set for firms with different reporting frequencies? I find little support to the claim that regulation forcing firms to issue more frequent financial information benefits financial analysts. Compared to a control sample of semiannual reporting firms in the European market, analysts issuing earnings forecasts for firms with mandatory quarterly frequency experience higher forecast error and dispersion. When firms are mandated to report not only on a quarterly frequency, but also under International Financial Reporting Standards (IFRS), analysts’ both forecast error and dispersion decrease. However, while IFRS does benefit analysts by increasing the quality of their information set in absolute terms, they do not wipe out the relative noise associated with mandatory quarterly statements.
The third chapter focuses on how financial analysts adapt to the passage of regulations aiming at limiting conflicts of interest in the investment banking industry. This last chapter investigates analysts’ price targets and recommendations, and unravels a new form of conflicts of interest. Specifically, it investigates whether affiliated brokers issue unfavourable ratings on their clients’ competitors in the product market (rivals). The findings document an important gap between ratings for affiliated and rival firms. Specifically, brokers issue persistently higher ratings on firms with which they are affiliated compared to their rivals. Importantly, the Sarbanes-Oxley Act and the related financial regulations aiming at curbing the conflicts of interests had no significant impact in reducing this gap. As such, affiliated brokers continue to indirectly favour their clients. This form of conflict was devoid of adequate attention in prior research. Furthermore, investors are unaware of the existence of such conflict in the short-run. - PublicationAccès libreThree essays on portfolio management(2015)
; This dissertation is constituted of three distinct chapters. The first two study the information role of sell-side analysts from two specific angles of attack. The first chapter focuses on the investment value of target prices. Based on a sample of more than 590’000 expected return revisions over the 1999-2011 period, I construct tercile portfolios that buy (sell) stocks with the highest (lowest) expected return revisions. The strategy initiated at the end of announcement day and held for a month that is long the highest tercile and short the lowest tercile yields a risk-adjusted performance of 0.48% per month. Similar results are obtained when the expected return revisions are industry or market adjusted. The risk-adjusted return remains significant if the position is initiated five days after the announcement (0.29% per month). Given the high number of target price revisions, I identify ex-ante likely valuable target prices. The risk-adjusted performance of the portfolio based on this subset increases to 0.81%. The downside exposures to SMB and MOM factors are negative and statistically significant at the 1% and 5% level, respectively. I demonstrate that more weight is given to pro cyclical (neutral) stocks when the expected probability of recession is low (high). Finally, we show that the results are not driven by firm specific events, post earnings announcement drift (PEAD), limited investors’ attention or illiquid stocks.
In the second chapter I analyze the information conveyed by analysts’ research and how it is perceived by investors. I introduce a methodology that disentangles the information conveyed through analysts’ target prices according to its availability and scope. The purpose is to investigate if investors correctly interpret analysts’ research by analyzing whether there is correspondence between investors’ reaction and the type of information conveyed through analysts. The empirical results provide evidence that investors duly process analysts’ research and appropriately incorporate this information into prices. Indeed, public information is not associated with any abnormal return, whereas private information is. Moreover, the reaction to firm-specific private information is confined to the firm analyzed, but industry-wide private information is associated with a reaction that spreads to the whole industry. The decomposition also shows that target prices are based on an equal amount of private and public information, and that private information is mostly firm-specific.
In the last chapter I turn my attention to hedge fund managers, a category of sophisticated users of analysts’ research. More specifically, I analyze how the remuneration structure of hedge funds affects the performance to investors to rationalize the persistent abnormal performance of hedge funds. I show that when managers expect to receive a performance fee payment, the commitment to deliver an absolute return, the decreasing returns to scale to which hedge fund strategies are subject, and the performance-linked remuneration combine with the income-maximizing behavior of managers to effectively align the interests of investors and managers. In consequence of the coexistence of these elements, managers have an incentive to control the size of the funds. Therefore, performance-diluting flows do not occur and abnormal performance persists. The model quantitatively reproduces many empirical facts about hedge funds. - PublicationAccès libreCountry versus sector influences and financial analysts' specialization(2007)