We examine Sentix sentiment indices for use in tactical asset allocation. In particular, we construct monthly relative sentiment factors for the U.S., Europe, Japan, and Asia ex-Japan by taking the difference in 6-month economic expectations between each region's institutional and individual investors. These factors (along with one-month forward equity returns) then serve as inputs to a wide array of machine learning algorithms. Employing combinatorial cross-validation and adjusting for data snooping, we find relative sentiment factors have robust and significant predictive power in all four regions; that they surpass both standalone sentiment and time-series momentum in terms of informational content; and that they demonstrate the ability to identify the subsequent best- and worst-performing global equity markets from along a cross-section. The results are consistent with previous findings on relative sentiment, discovered using unrelated datasets.
We show, using machine learning, that fund characteristics can consistently differentiate high from low-performing mutual funds, as well as identify funds with net-of-fees abnormal returns. Fund momentum and fund flow are the most important predictors of future risk-adjusted fund performance, while characteristics of the stocks that funds hold are not predictive. Returns of predictive long-short portfolios are higher following a period of high sentiment or a good state of the macro-economy. Our estimation with neural networks enables us to uncover novel and substantial interaction effects between sentiment and both fund flow and fund momentum.
Of the $69.1 trillion global financial assets under management across mutual funds, hedge funds, real estate, and private equity, fewer than 1.3% are managed by women and people of color. Why is this powerful, elite industry so racially homogenous? We conducted an online experiment with actual asset allocators to determine whether there are biases in their evaluations of funds led by people of color, and, if so, how these biases manifest. We asked asset allocators to rate venture capital funds based on their evaluation of a 1-page summary of the fund’s performance history, in which we manipulated the race of the managing partner (White or Black) and the strength of the fund’s credentials (stronger or weaker). Asset allocators favored the White-led, racially homogenous team when credentials were stronger, but the Black-led, racially diverse team when credentials were weaker. Moreover, asset allocators’ judgments of the team’s competence were more strongly correlated with predictions about future performance (e.g., money raised) for racially homogenous teams than for racially diverse teams. Despite the apparent preference for racially diverse teams at weaker performance levels, asset allocators did not express a high likelihood of investing in these teams. These results suggest first that underrepresentation of people of color in the realm of investing is not only a pipeline problem, and second, that funds led by people of color might paradoxically face the most barriers to advancement after they have established themselves as strong performers.
We examine gender differences in the language of CFOs who participate in quarterly earnings calls. Female executives are more concise and less optimistic, are clearer, use fewer idioms or clichés, and provide more numbers in their speech. These differences are particularly strong in the more spontaneous Questions and Answers (QA) section of the calls and are reflected in stronger market and analyst reactions. Gender differences seem to be associated with CFO overconfidence.
How investors understand and use central bank communications, aka FEDSPEAK, is oftentimes cryptic and difficult to analyze. This study attempts to provide some clarity to this issue by applying textual analysis to both high-frequency price and communication data, to focus on episodes whereby stock price movements are identifiable and on investors’ reactions to specific sentences communicated by the Fed.
“Employees are our greatest asset” is a phrase often heard from companies. However, due to accounting rules requiring that most expenditures related to employees be treated as costs and expensed as incurred, the value of employees is an intangible asset that does not appear on any balance sheet. That leaves the interesting question of whether employee satisfaction provides information on future returns.
The primary idea behind this research is that a more sophisticated statistical technology (in the sense of reducing predictive mean squared error) produces predictions with greater variance than a more primitive technology. These technologies range from a simple logistic regression of default outcomes based on borrowers and default variables to random forest machine learning models. Said differently, improvements in predictive technology act as mean-preserving spreads for predicted outcomes—in this case, predicted default propensities on loans, which also means that there will always be some borrowers considered less risky by the new technology, or “winners”, while other borrowers will be deemed riskier “losers”, relative to their position under the pre-existing technology.
Firm characteristics are often missing, which forces both researchers and practitioners to come up with workarounds when handling missing data. Previous approaches resorted to either dropping observations with missing entries or simply imputing the cross-sectional mean of a given characteristic. As both procedures accompany serious drawbacks (see below), there is a need for more advanced methods. The authors set up an attention-based machine learning model, motivated by recent advances in natural language to find some answers
The question of whether or not the FED considers or responds to the stock market in its policy decisions has been studied fairly extensively, the subject of the existence of the "FED put" continues to pop up in the literature. In this particular revival of the issue, the authors are among the first to study FOMC minutes, transcripts, and other sources of information using textual analysis in order to provide an answer to the question: Does the FED respond to stock market events and if it does, what is the nature of the response?
Option Return Predictability with Machine Learning and Big Data Bali, Beckmeyer, Moerke, WeigertA version of this paper can be found hereWant to read our summaries of academic finance papers? Check out our Academic Research Insight category [...]
Financial Media, Price Discovery, and Merger Arbitrage Buehlmaier and ZechnerReview of Finance, forthcomingA version of this paper can be found hereWant to read our summaries of academic finance papers? Check out our Academic Research Insight category What are [...]
Part 1: The End of Accounting This is the first part of a series of guest posts by Kai Wu, the CIO & Founder of Sparkline Capital. The Factor Zoo As readers of Alpha Architect’s [...]
Responsible Investing: The ESG Efficient Frontier Pedersen, Fitzgibbons, and PomorskiJournal of Financial Economics, 2020A version of this paper can be found hereWant to read our summaries of academic finance papers? Check out our Academic Research Insight category What [...]
CFO Gender and Financial Statement Irregularities V.K.Gupta, S. Mortal, B. Chakrabarty, X. Guo, D. B. TurbanAcademy of Management Journal, 2019A version of this paper can be found hereWant to read our summaries of academic finance papers? [...]
Zero-Revelation RegTech: Detecting Risk through Linguistic Analysis of Corporate Emails and News S.R. Das, S. Kim, B. KothariJournal of Financial Data Science, Spring 2019A version of this paper can be found hereWant to read our summaries [...]
How news and its context drive risk and returns around the world Charles Calomiris and Harry MamayskyJournal of Financial Economics, August 2019A version of this paper can be found here.Want to read our summaries of academic [...]
We at ENJINE are big believers in the potential of machine learning (or as some call, “artificial intelligence”) to transform asset management. However, it’s fair to say that machine learning hasn’t received mass adoption in [...]
Predicting Bond Returns: 70 Years of International Evidence Guido Baltussen, Martin Martens, Olaf PenningaWorking PaperA version of this paper can be found hereWant to read our summaries of academic finance papers? Check out our Academic Research Insight category [...]
When More or Less is Less: Managers' Clichès J. Klevak, J. Livnat, and K. SuslavaJournal of Financial Data Science, Summer 2019A version of this paper can be found hereWant to read our summaries of academic finance [...]
Quantitative factor portfolios generally use historical company fundamental data in portfolio construction. The key assumption behind this approach is that past fundamentals proxy for elements of risk and/or systematic mispricing. However, what if we could [...]