ArticlePDF Available

Survivorship bias in performance studies

Authors:

Abstract

Recent evidence suggests that past mutual fund performance predicts future performance. We analyze the relationship between volatility and returns in a sample that is truncated by survivorship and show that this relationship gives rise to the appearance of predictability. We present some numerical examples to show that this effect can be strong enough to account for the strength of the evidence favoring return predictability.
A preview of the PDF is not available
... Elton, et al. (1996) find that previous mutual fund studies suffered from survivorship bias as funds that merge or die have worse performance than funds that do not and failing to account for survivorship bias will lead to higher risk-adjusted returns for mutual funds. Brown, et al. (1992) also find that survivorship bias can give a false impression about persistence in mutual fund performance. To avoid this problem, we include in our analysis all alternative mutual funds that ever existed as found in the Morningstar Direct data. ...
... 9.6. Persistence Grinblatt and Titman (1992), Brown, et al. (1992), Hendricks, et al. (1993, Brown and Goetzmann (1995), Goetzmann and Ibbotson (1994), Kahn and Rudd (1995), Malkiel (1995), Elton, et al. (1996), andCarhart (1997) have tested the persistence of mutual fund total returns in time. Grinblatt and Titman (1992) find evidence that differences in performances between funds persists over time and this persistence is consistent with the ability of fund managers to earn abnormal returns. ...
... Studying only surviving funds will overstate performance. Brown, et al. (1992) show that early studies exaggerate the extent of persistence by relying on survivorship-biased data sets. Because survivorship bias has been controlled, there will be no such problems. ...
Article
Full-text available
Alternative Mutual Funds (AMFs) provide the individual investor with the opportunity to invest in funds that follow strategies similar to those of hedge funds and seek returns uncorrelated with the market. Financial planners, advisors, and investors need to be aware of how well AMFs deliver absolute or positive returns regardless of market conditions and their relatively high expense ratios. In this article we analyze the performance of AMFs for the period January 1998 through December 2011 using the Carhart four-factor model and the Fung-Hsieh seven-factor model. Our results indicate that most AMFs have not been able to create any value for their investors over the period of our study. Furthermore, the performance of these funds was even worse during the recent financial crisis. © 2014 Academy of Financial Services. All rights reserved.
... In comparing Llama2-7B with Llama2-70B, an increase in model size generally led to a reduction in most cognitive biases. Yet, for certain biases, such as the Curse of Knowledge [8] and Survivorship Bias [7], the opposite was true. A similar trend was observed in the Vicuna series. ...
Preprint
Full-text available
Detecting cognitive biases in large language models (LLMs) is a fascinating task that aims to probe the existing cognitive biases within these models. Current methods for detecting cognitive biases in language models generally suffer from incomplete detection capabilities and a restricted range of detectable bias types. To address this issue, we introduced the 'MindScope' dataset, which distinctively integrates static and dynamic elements. The static component comprises 5,170 open-ended questions spanning 72 cognitive bias categories. The dynamic component leverages a rule-based, multi-agent communication framework to facilitate the generation of multi-round dialogues. This framework is flexible and readily adaptable for various psychological experiments involving LLMs. In addition, we introduce a multi-agent detection method applicable to a wide range of detection tasks, which integrates Retrieval-Augmented Generation (RAG), competitive debate, and a reinforcement learning-based decision module. Demonstrating substantial effectiveness, this method has shown to improve detection accuracy by as much as 35.10% compared to GPT-4. Codes and appendix are available at https://github.com/2279072142/MindScope.
... Firstly, several researchers identified survivorship bias in sampling as the reason behind the apparent persistence (Brown et al., 1992;Brown & Goetzmann, 1995;Malkiel, 1995). Conversely, other studies identified the presence of momentum and the utilization of momentum strategies not considered by the utilized risk model, offering plausible explanations for performance persistence (Carhart, 1997;Daniel et al., 1997;Wermers, 1997). ...
Article
Full-text available
This study evaluates the persistence of performance of actively managed, equity-based mutual funds in Nigeria using monthly net asset values (NAVs) of 30 funds obtained from the Securities and Exchange Commission over 10 years from 2012 to 2021. We employed a non-parametric technique based on the Contingency Table to test for performance persistence, using the Cross-Product Ratio (CPR) and the Rank Correlation statistics. Evidence shows that mutual funds do not exhibit performance persistence, and the performance of loser funds does not repeat, hence past performance does not predict future performance. Therefore, we recommend that retail investors (and their advisers) should not rely on historical performance to select mutual funds as an investment vehicle. This study provides valuable insight into the performance of actively managed funds in Nigeria and contributes to the ongoing debate about the efficiency of the financial markets and the role of active fund management. It suggests that investors might be better off with passive investment management strategies, given the lack of persistence in the performance of actively managed funds.
... FinRL-Meta's data curation pipeline holds data cleaning and feature engineering, which reduce noise and extract useful signals from the raw data. • Survivorship bias of historical market data: Survivorship bias is caused by a tendency to focus on existing stocks and funds without consideration of those that are delisted (Brown et al., 1992). It could lead to an overestimation of stocks and funds, which will mislead the agent. ...
Article
Full-text available
The financial market is a particularly challenging playground for deep reinforcement learning due to its unique feature of dynamic datasets. Building high-quality market environments for training financial reinforcement learning (FinRL) agents is difficult due to major factors such as the low signal-to-noise ratio of financial data, survivorship bias of historical data, and model overfitting. In this paper, we present an updated version of FinRL-Meta, a data-centric and openly accessible library that processes dynamic datasets from real-world markets into gym-style market environments and has been actively maintained by the AI4Finance community. First, following a DataOps paradigm, we provide hundreds of market environments through an automatic data curation pipeline. Second, we provide homegrown examples and reproduce popular research papers as stepping stones for users to design new trading strategies. We also deploy the library on cloud platforms so that users can visualize their own results and assess the relative performance via community-wise competitions. Third, we provide dozens of Jupyter/Python demos organized into a curriculum and a documentation website to serve the rapidly growing community. The codes are available at https://github.com/AI4Finance-Foundation/FinRL-Meta
... The study used firms listed in the DAX 30 index between 2010 and 2019 to avoid potential survivorship bias (Brown, Goetzmann, Ibbotson, & Ross, 1992). This time frame was chosen to exclude the Global Financial Crisis and the COVID-19 pandemic (Issah, Anwar, Clauss, & Kraus, 2023;Kraus et al., 2020). ...
Article
Full-text available
In an era of hypercompetition, research and development (R&D) investments are vital for organizations to stay competitive. This microlevel study draws on dynamic managerial capability (DMC) theory to explore the mechanisms contributing to competitive advantages. It posits that DMCs enhance firm performance by increasing R&D spending, and explores the moderating role of slack resources due to their effect on resource availability. Employing hierarchical regression analysis and bootstrapping methods on a longitudinal sample comprising 31 German DAX firms, the findings robustly demonstrate that DMCs facilitate firm performance by fostering R&D expenditures and confirm the moderating effect of specific slack resources. However, only internal but not external slack resources amplify the relationship between DMCs and R&D intensity. Overall, this study emphasizes the critical role of managers’ microlevel capabilities in determining firm performance and sheds light on how different slack resources influence the relationships between DMCs, R&D intensity, and firm performance.
... Também não podemos deixar de mencionar o viés de sobrevivência da amostra, previsto por Sanvicente e Sanches (2002), visto que 312 fundos de ações elegíveis deixaram de existir entre jan/2011 e dez/2020, período utilizado para o estudo, e foram desconsiderados. Isso pode implicar em um alfa superestimado, tal como descrito por Brown et al. (1992), onde os fundos mais longevos tendem a ser aqueles com melhor desempenho histórico. ...
Article
Full-text available
Este trabalho procura mostrar se o desempenho histórico dos gestores dos 141 maiores fundos de investimentos de ações do Brasil, de gestão ativa, supera o retorno previsto pelo modelo de 3 fatores de Fama e French (1993) ao longo de um período de 10 anos. Foi verificado que, na média, os gestores não conseguem gerar retorno excessivo positivo. Adicionalmente, o trabalho realiza uma análise para verificar se o desempenho desses mesmos fundos sofre influência de altas e baixas nos preços das ações. Os resultados indicam que a direção do mercado não possui influência no retorno excessivo. Por fim, o estudo verifica se o tamanho do patrimônio dos fundos tem impacto em seu desempenho. Os dados indicam que existe uma relação negativa explicada por ineficiências de escala. A amostra exclui fundos restritos e exclusivos, assim como aqueles cuja carteira seja composta predominantemente por investimentos no exterior e outros fundos de investimento.
... The reason behind the variety of models used is that these models are standard in the existing literature ( Blitz & Fabozzi, 2017) compare the results to other papers. In addition, firms that existed in 2000 but ceased to exist before 2020 will be included in the regression to avoid survivorship bias, i.e., the tendency to overestimate past performance by considering only surviving companies' stocks (Brown et al., 1992). ...
Article
Full-text available
This study examines the relationship between the wealth of European societies and their investment decisions in «sinful» industries, including tobacco, alcohol, and gambling. The study aims to challenge the widely held belief that wealthier countries are more socially responsible in their investment choices and to investigate the impact of familiarity bias on investment decisions in these industries. An experimental research design with panel data compares the returns from a portfolio of sin stocks from Northern Europe with a portfolio of sin stocks from Southern and Eastern Europe. The study utilises multiple models, including the CAPM single-factor, the Fama-French three-factor, and the Fama-French five-factor, to measure the risk-adjusted returns of sin stocks across various European countries. Findings reveal that sin stocks from wealthier countries tend to have higher risk-adjusted returns compared to those from less wealthy countries. Sin stocks have a significant relation with the market, but their volatility is consistently lower. Countries that drink more alcohol are more willing to invest in alcohol stocks than countries that drink less, as these stocks outperform the market during economic downturns. Sin stocks impact financial performance, investor behaviour, social responsibility, market efficiency, and regulations. The study uncovers the influence of familiarity bias, indicating that investors from countries more accustomed to «sinful» activities are less reluctant to invest in such industries than countries with lower familiarity. This finding highlights the importance of cultural and social factors in shaping investment decisions and challenges traditional concepts of market efficiency.
Article
Full-text available
Mutual funds represent one of the fastest growing type of financial intermediary in the American economy. The question remains as to why mutual funds and in particular actively managed mutual funds have grown so fast, when their performance on average has been inferior to that of index funds. One possible explanation of why investors buy actively managed open end funds lies in the fact that they are bought and sold at net asset value, and thus management ability may not be priced. If management ability exists and it is not included in the price of open end funds, then performance should be predictable. If performance is predictable and at least some investors are aware of this, then cash flows into and out of funds should be predictable by the very same metrics that predict performance. Finally, if predictors exist and at least some investors act on these predictors in investing in mutual funds, the return on new cash flows should be better than the average return for all investors in these funds. This article presents empirical evidence on all of these issues and shows that investors in actively managed mutual funds may have been more rational than we have assumed.
ResearchGate has not been able to resolve any references for this publication.