This thesis is an attempt to advance our understanding of investor behaviour in one of the world’s largest markets, i.e. the mutual fund industry. It consists of three essays that answer the following questions: Does investor fund-selection ability explain the impressive growth of the U.K. mutual fund industry? Does the behaviour of U.S. mutual fund investors vary across the business cycle? And, how do investors react to U.S. mutual fund name changes? The first essay explores the role of investor fund-selection ability in explaining the growth of the mutual fund industry given that previous studies find that mutual funds underperform their benchmarks on average. I examine such ability in the context of the remarkable growth experienced by U.K. mutual funds during the decade of 2000-2010. Using three alternative measures of selection ability and two for performance measurement, I find that fund-selection ability is explained away by the momentum factor due to investors naively chasing recent winners. In addition, this essay is the first to examine the impact of fund visibility on selection ability. I find that fund visibility is an important factor in the investment decision-making process, and one that fund managers can potentially manipulate to their advantage. The second essay is motivated by recent findings that benchmark-adjusted returns to the fund industry are positive in periods of economic contractions. Previous literature is silent on investor behaviour in the face of superior average returns. This essay fills the gap in literature by examining investor’s fund-selection ability across the business cycle. I examine U.S. fund data from 1970-2011 and find that while genuine selection ability does not exist in any period, investors do behave differently across the business cycle. Specifically, investors no longer chase recent winners during contractions, despite no change in fund performance consistency. Instead, I find that investors are more concerned about controlling their risk exposure, especially to the market, during periods of economic downturn. The third essay examines investor reactions to U.S. mutual fund name changes, following the adoption of a new SEC ruling in 2001 to curtail misleading names. We uncover striking evidence that funds continue to undertake cosmetic name changes, and that such changes appear to mislead investors. I find that investors react more positively to cosmetic name changes than non-cosmetic ones. This result is not driven by marketing efforts. Instead, further examination reveals that this arises because cosmetic name changes frequently include industry ‘buzzwords’ in the new name, a tactic that is rewarded with higher flows to such funds. I also find that additional name changes by a fund continue to attract significant flows, although the magnitude of the flows decreases over each successive event. This essay provides compelling evidence in favour of investor irrationality and has implications for both practitioners and academics.

Magistro baigiamajame darbe išanalizuoti ir įvertinti Lietuvos II pakopos pensijų fondų investicijų grąžos pokyčiai verslo cikluose, iškelta fondo pasirinkimo problema bei pateikti siūlymai kaip šią problemą spręsti pensijų fondų dalyviams. Pirmoje darbo dalyje teoriniu aspektu analizuojama Lietuvos pensijų sistema, pateikiami teigiami ir neigiami kaupiamųjų fondų aspektai ir pateikiama verslo ciklų samprata. Antroje dalyje atliekama II pakopos pensijų fondų metinėse ataskaitose skelbiamų rodiklių analizė, nagrinėjami dažniausiai mokslinėje literatūroje sutinkami pensijų fondų vertinimo metodai bei pateikiamas darbo tyrimo modelis ir apibrėžiama darbo eiga. Trečioje dalyje pateikiama trumpa 2013 metų pabaigoje veiklą vykdžiusių II pakopos pensijų fondų apžvalga ir panaudojant Šarpo metodiką bei kitus pagrindinius fondų vertinimo kriterijus yra atrenkami efektyviausiai valdomi skirtingų strategijų pensijų fondai. Identifikavus verslo ciklus Lietuvoje, atliekama efektyviausiai valdomų skirtingų strategijų pensijų fondų investicijų grąžos analizė kiekvienoje verslo ciklo fazėje. Atliekama techninė analizė ir sudaromos tiesinės daugianarės regresijos lygtys, naudojamos prognozuoti investicijų grąžos pokyčius remiantis faktiniais fondų apskaitos vienetų vertės pokyčiais ir makroekonominiais rodikliais. / Master's Work analyzed and evaluated Lithuanian pillar II pension funds return on investment changes in business cycles, a series of suggestions is given for pension funds participants how to solve the problem of pension fund selection. The first part examines theoretical aspect of Lithuanian pension system, an overview of its positive and negative aspects and defined concept of business cycles. In second section analyzed indicators provided in the annual reports of pillar II pension funds, mostly encountered pension fund valuation methods in the scientific literature and workflow is defined. The third part present short review of Lithuanian pillar II pension funds which operated in 2013 and using Sharpe methodology and other most important valuation methods are selected efficiently managed by different strategies of pension funds. After identification of the business cycles in Lithuania, carried out in most effectively managed, by different strategies of pension funds, the return on investment analysis for each phase of the business cycle Technical analysis and the straight multiple regression equations used to predict changes in investment return based on the actual value of the fund units of accounting changes and macroeconomic indicators.

Submitted by MFEE Mestrado Profissional em Finanças e Economia Empresarial da EPGE (mfee@fgv.br) on 2015-02-23T22:54:26Z No. of bitstreams: 1 Felipe Cotrim - Versão Final.pdf: 2874262 bytes, checksum: c9adf65f93c0c6c7ad37a5c1dd872036 (MD5) / Approved for entry into archive by Marcia Bacha (marcia.bacha@fgv.br) on 2015-03-04T13:06:22Z (GMT) No. of bitstreams: 1 Felipe Cotrim - Versão Final.pdf: 2874262 bytes, checksum: c9adf65f93c0c6c7ad37a5c1dd872036 (MD5) / Approved for entry into archive by Marcia Bacha (marcia.bacha@fgv.br) on 2015-03-04T13:06:37Z (GMT) No. of bitstreams: 1 Felipe Cotrim - Versão Final.pdf: 2874262 bytes, checksum: c9adf65f93c0c6c7ad37a5c1dd872036 (MD5) / Made available in DSpace on 2015-03-04T13:06:56Z (GMT). No. of bitstreams: 1 Felipe Cotrim - Versão Final.pdf: 2874262 bytes, checksum: c9adf65f93c0c6c7ad37a5c1dd872036 (MD5) Previous issue date: 2012-09-17 / With lhe increase of lhe number of assei managers and an even larger a number of investments alternatives in lhe Brazilian hedge fund industry, Fund of Hedge Funds became an alternative for investors planning to diversify their investments through financiai specialists. The intention of this study is to evaluate lhe capacity of Brazilian Funds of Hedge Funds (FoHF), classified as Multimercados Multigestor, to generate abnormal returns (alpha). For this porpoise we studied a sample of 1421 Fund of Hedg Funds between January of 2005 and December of 2011. The results of multi-factor model regressions, derived from Jensen's model (1968), suggest that only 3.03% of lhe funds in lhe sample can add value. The three main potential sources of alpha generalion in Funds of Hedge Funds come from lhe strategic allocation of lhe portfolios, lhe anticipation of market movements (market timing) and lhe capacity of FoHF managers to select lhe best assei managers in lhe industry to com pose its portfolio (fund selection). To evaluate lhe Brazilian FoHF manager's ability to anticipate market movements we included quadratic terms in lhe multi-factor models, as proposed by Treynor and Mazuy (1966). The results showed lha! managers, on average, could not add value by market timing. To evaluate lhe strategic allocation ability and lhe fund's selection abilities, we created a new variable with lhe information about lhe asseis in lhe porlfolio of each fund in lhe sample in every different month. The tests indicated that FoHF managers, on average, could no! add value by selecting lhe best managers, but lhe strategic allocation ability showed a positive contribution to FoHF's return. We also studied lhe alpha generation capacity before costs. lt raised lhe percentage of funds with positive alpha to 6.39% of lhe funds in lhe sample, but it was no! able to change lhe signal of lhe average alpha, lha! remained nega tive. / Com o aumento do número de gestores especializados em um número cada vez maior de possibilidades de investimentos na indústria de fundos brasileira, os fundos Multigestor se tornaram uma alternativa para os investidores que procuram diversificar seus investimentos e delegam às instituições financeiras o trabalho de alocar os recursos dentro das diferentes estratégias e fundos existentes no mercado. O intuito deste estudo é avaliar a capacidade de gerar retornos anormais (alfa) dos fundos de fundos da indústria brasileira, classificados como Fundos Multimercados Multigestor. Para isso foi estudada uma amostra com 1.421 fundos Multigestor com tributação de Longo Prazo no período de janeiro de 2005 a dezembro de 2011. A análise dos resultados encontrados através de regressões de modelos de vários fatores, derivados do modelo de Jensen (1968), sugere que apenas 3,03% dos fundos estudados conseguem adicionar valor a seus cotistas. Foram estudadas ainda as três principais fontes potenciais de geração de alfa dos fundos de fundos, a escolha das estratégias que compõe a carteira do fundo (alocação estratégica), a antecipação de movimentos de mercado (market timing) e a capacidade selecionar os melhores fundos dentro de cada estratégia (seleção de fundos). A partir da inclusão de termos quadráticos, conforme proposto pelos modelos de Treynor e Mazuy (1966) pode-se verificar que os fundos Multigestor, em média, não conseguem adicionar valor tentando antecipar movimentos de mercado (market timing). Através da construção de uma variável explicativa com a composição estratégica de cada fundo da amostra em cada período de tempo, pode-se verificar que os gestores de fundos de fundos, em média, também fracassam ao tentar selecionar os melhores fundos/gestores da indústria. Já a escolha das estratégias que compõe a carteira do fundo (alocação estratégica) mostrou contribuir positivamente para o retorno dos fundos. Ainda foi avaliada a capacidade de gerar alfa antes dos custos, o que elevou o percentual de fundos com alfa positivo para 6,39% dos fundos estudados, mas foi incapaz de alterar o sinal do alfa médio, que permaneceu negativo.