scholarly journals The Influence of Cognitive Biases and Financial Factors on Forecast Accuracy of Analysts

2022 ◽  
Vol 12 ◽  
Author(s):  
Paula Carolina Ciampaglia Nardi ◽  
Evandro Marcos Saidel Ribeiro ◽  
José Lino Oliveira Bueno ◽  
Ishani Aggarwal
Keyword(s):  
Machine Learning ◽  
Text Analysis ◽  
Positive Relationship ◽  
Cognitive Biases ◽  
Explanatory Power ◽  
Negative Relationship ◽  
Fair Value ◽  
Statistical Regression ◽  
Learning Methods ◽  
Financial Factors

The objective of this study was to jointly analyze the importance of cognitive and financial factors in the accuracy of profit forecasting by analysts. Data from publicly traded Brazilian companies in 2019 were obtained. We used text analysis to assess the cognitive biases from the qualitative reports of analysts. Further, we analyzed the data using statistical regression learning methods and statistical classification learning methods, such as Multiple Linear Regression (MRL), k-dependence Bayesian (k-DB), and Random Forest (RF). The Bayesian inference and classification methods allow an expansion of the research line, especially in the area of machine learning, which can benefit from the examples of factors addressed in this research. The results indicated that, among cognitive biases, optimism had a negative relationship with forecasting accuracy while anchoring bias had a positive relationship. Commonality, to a lesser extent, also had a positive relationship with the analyst’s accuracy. Among financial factors, the most important aspects in the accuracy of analysts were volatility, indebtedness, and profitability. Age of the company, fair value, American Depositary Receipts (ADRs), performance, and loss were still important but on a smaller scale. The results of the RF models showed a greater explanatory power. This research sheds light on the cognitive as well as financial aspects that influence the analyst’s accuracy, jointly using text analysis and machine learning methods, capable of improving the explanatory power of predictive models, together with the use of training models followed by testing.

scholarly journals The Effect of Daily Stock Price Limits on the Investment Risk: Evidence from the Egyptian Stock Market

2017 ◽  
Vol 6 (4) ◽  
pp. 1 ◽  
Author(s):  
Mai Ahmed Abdelzaher ◽  
Khairy Elgiziry
Keyword(s):  
Stock Prices ◽  
Stock Price ◽  
Trading Volume ◽  
Positive Relationship ◽  
Descriptive Analysis ◽  
Negative Relationship ◽  
Fair Value ◽  
Price Limits ◽  
Stock Volatility ◽  
Closing Price

The study aims to investigate the relationship between daily price limits and stock volatility, trading volume, delayed adjustment of stock prices, and its fair value. To achieve this goal, we used the data of the listed firms in EGX30. We analyzed the data using descriptive analysis then we applied General linear model, ARCH and GARCH models. Based on our analysis results show a positive relationship between upper daily limit and stock volatility, a positive relationship between daily price limits (upper limit- lower limit) and trading volume, a positive relationship between upper daily limit and the return between the closing price and the opening price on the same day, a positive relationship between lower daily limit and the return between the closing price and the opening price in the next day, a negative relationship between upper daily limit and the return between the closing price and the opening price in the next day, and a positive relationship between daily stock price limits and the fair value.

Does the Fed Impact Stock Prices?

2019 ◽  
Keyword(s):  
Interest Rate ◽  
Interest Rates ◽  
Stock Prices ◽  
Positive Relationship ◽  
Negative Impact ◽  
Negative Relationship ◽  
Fair Value ◽  
Equity Returns ◽  
Equity Prices ◽  
The Us

The influence of the Fed’s actions on equity prices has been a source of significant speculation in recent years. This article uses a well-regarded measure for the “fair” value of interest rates to measure the degree to which the Fed is influencing interest rate and then relates that level of interference to equity returns. We find that Fed’s actions are correlated with a modest negative impact on US equity prices—that is Fed interference has a slight negative relationship with broader equity returns. In contrast, outside the US, Central Bank interference generally has a stronger positive relationship to equity returns.

scholarly journals Cultural complexity and demography: the case of folktales

2016 ◽  
Author(s):  
Alberto Acerbi
Keyword(s):  
Population Size ◽  
Cultural Evolution ◽  
Positive Relationship ◽  
Cognitive Biases ◽  
Negative Relationship ◽  
Small Populations ◽  
Cultural Complexity ◽  
Cultural Domain ◽  
Folk Narratives

We investigate the relation between cultural complexity and population size in a non-technological cultural domain for which we have suitable quantitative records: folktales. We define three levels of complexity for folk narratives: the number of tale types, the number of narrative motifs, and, finally, the number of traits in variants of the same type, for two well known tales for which we have data from previous studies. We found a positive relationship between number of tale types and population size, a negative relationship for the number of narrative motifs, and no relationship for the number of traits. The absence of a consistent relation between population size and complexity in folktales provides a novel perspective on the current debates in cultural evolution. We propose that the link between cultural complexity and demography could be domain dependent: in some domains (e.g. technology) this link is important, whereas in others, such as folktales, complex traditions can be easily maintained in small populations as well as large ones, as they may appeal to universal cognitive biases.

Replicability of Machine Learning Models in the Social Sciences

2018 ◽  
Vol 226 (4) ◽  
pp. 259-273 ◽  
Author(s):  
Ranjith Vijayakumar ◽  
Mike W.-L. Cheung
Keyword(s):  
Social Sciences ◽  
Machine Learning ◽  
Variable Selection ◽  
Model Building ◽  
Predictive Accuracy ◽  
Explanatory Power ◽  
Multivariate Adaptive Regression Splines ◽  
Support Vector ◽  
Learning Methods ◽  
Machine Learning Methods

Abstract. Machine learning tools are increasingly used in social sciences and policy fields due to their increase in predictive accuracy. However, little research has been done on how well the models of machine learning methods replicate across samples. We compare machine learning methods with regression on the replicability of variable selection, along with predictive accuracy, using an empirical dataset as well as simulated data with additive, interaction, and non-linear squared terms added as predictors. Methods analyzed include support vector machines (SVM), random forests (RF), multivariate adaptive regression splines (MARS), and the regularized regression variants, least absolute shrinkage and selection operator (LASSO), and elastic net. In simulations with additive and linear interactions, machine learning methods performed similarly to regression in replicating predictors; they also performed mostly equal or below regression on measures of predictive accuracy. In simulations with square terms, machine learning methods SVM, RF, and MARS improved predictive accuracy and replicated predictors better than regression. Thus, in simulated datasets, the gap between machine learning methods and regression on predictive measures foreshadowed the gap in variable selection. In replications on the empirical dataset, however, improved prediction by machine learning methods was not accompanied by a visible improvement in replicability in variable selection. This disparity is explained by the overall explanatory power of the models. When predictors have small effects and noise predominates, improved global measures of prediction in a sample by machine learning methods may not lead to the robust selection of predictors; thus, in the presence of weak predictors and noise, regression remains a useful tool for model building and replication.

Disclosure Sentiment: Machine Learning vs. Dictionary Methods

2021 ◽  
Author(s):  
Richard Frankel ◽  
Jared Jennings ◽  
Joshua Lee
Keyword(s):  
Machine Learning ◽  
Explanatory Power ◽  
Regression Tree ◽  
Random Forest Regression ◽  
Conference Call ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Time Period ◽  
Machine Learning Approach ◽  
Tree Method

We compare the ability of dictionary-based and machine-learning methods to capture disclosure sentiment at 10-K filing and conference-call dates. Like Loughran and McDonald [Loughran T, McDonald B (2011) When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks. J. Finance 66(1):35–65.], we use returns to assess sentiment. We find that measures based on machine learning offer a significant improvement in explanatory power over dictionary-based measures. Specifically, machine-learning measures explain returns at 10-K filing dates, whereas measures based on the Loughran and McDonald dictionary only explain returns at 10-K filing dates during the time period of their study. Moreover, at conference-call dates, machine-learning methods offer an improvement over the Loughran and McDonald dictionary method of a greater magnitude than the improvement of the Loughran and McDonald dictionary over the Harvard Psychosociological Dictionary. We further find that the random-forest-regression-tree method better captures disclosure sentiment than alternative algorithms, simplifying the application of the machine-learning approach. Overall, our results suggest that machine-learning methods offer an easily implementable, more powerful, and reliable measure of disclosure sentiment than dictionary-based methods. This paper was accepted by Brian Bushee, management science.

Benchmark Study of Supervised Machine Learning Methods for a Ship Speed-Power Prediction at Sea

2021 ◽  
Author(s):  
Xiao Lang ◽  
Da Wu ◽  
Wengang Mao
Keyword(s):  
Machine Learning ◽  
Supervised Machine Learning ◽  
Support Vector ◽  
Statistical Regression ◽  
Learning Methods ◽  
Benchmark Study ◽  
Machine Learning Methods ◽  
Extreme Gradient Boosting ◽  
Ship Performance ◽  
Ship Speed

Abstract The development and evaluation of energy efficiency measures to reduce air emissions from shipping strongly depends on reliable description of a ship’s performance when sailing at sea. Normally, model tests and semi-empirical formulas are used to model a ship’s performance but they are either expensive or lack accuracy. Nowadays, a lot of ship performance-related parameters have been recorded during a ship’s sailing, and different data driven machine learning methods have been applied for the ship speed-power modelling. This paper compares different supervised machine learning algorithms, i.e., eXtreme Gradient Boosting (XGBoost), neural network, support vector machine, and some statistical regression methods, for the ship speed-power modelling. A worldwide sailing chemical tanker with full-scale measurements is employed as the case study vessel. A general data pre-processing method for the machine learning is presented. The machine learning models are trained using measurement data including ship operation profiles and encountered metocean conditions. Through the benchmark study, the pros and cons of different machine learning methods for the ship’s speed-power performance modelling are identified. The accuracy of various algorithms based models for ship performance during individual voyages is also investigated.

scholarly journals Using OpenStreetMap Data and Machine Learning to Generate Socio-Economic Indicators

2020 ◽  
Vol 9 (9) ◽  
pp. 498 ◽  
Author(s):  
Daniel Feldmeyer ◽  
Claude Meisch ◽  
Holger Sauter ◽  
Joern Birkmann
Keyword(s):  
Machine Learning ◽  
Explanatory Power ◽  
Economic Indicators ◽  
Statistical Regression ◽  
Spatiotemporal Resolution ◽  
Global Database ◽  
Governmental Institutions ◽  
Social Environmental ◽  
Complex Phenomena ◽  
The Impact

Socio-economic indicators are key to understanding societal challenges. They disassemble complex phenomena to gain insights and deepen understanding. Specific subsets of indicators have been developed to describe sustainability, human development, vulnerability, risk, resilience and climate change adaptation. Nonetheless, insufficient quality and availability of data often limit their explanatory power. Spatial and temporal resolution are often not at a scale appropriate for monitoring. Socio-economic indicators are mostly provided by governmental institutions and are therefore limited to administrative boundaries. Furthermore, different methodological computation approaches for the same indicator impair comparability between countries and regions. OpenStreetMap (OSM) provides an unparalleled standardized global database with a high spatiotemporal resolution. Surprisingly, the potential of OSM seems largely unexplored in this context. In this study, we used machine learning to predict four exemplary socio-economic indicators for municipalities based on OSM. By comparing the predictive power of neural networks to statistical regression models, we evaluated the unhinged resources of OSM for indicator development. OSM provides prospects for monitoring across administrative boundaries, interdisciplinary topics, and semi-quantitative factors like social cohesion. Further research is still required to, for example, determine the impact of regional and international differences in user contributions on the outputs. Nonetheless, this database can provide meaningful insight into otherwise unknown spatial differences in social, environmental or economic inequalities.

Classification of EEG signals in epilepsy using a novel integrated TSK fuzzy system

2020 ◽  
pp. 1-16
Author(s):  
Yuwen Tao ◽  
Yizhang Jiang ◽  
Kaijian Xia ◽  
Jing Xue ◽  
Leyuan Zhou ◽  
...  
Keyword(s):  
Machine Learning ◽  
Supervised Learning ◽  
Fuzzy System ◽  
Domain Adaptation ◽  
Explanatory Power ◽  
Machine Learning Algorithms ◽  
Eeg Signals ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Training Samples

The use of machine learning technology to recognize electrical signals of the brain is becoming increasingly popular. Compared with doctors’ manual judgment, machine learning methods are faster. However, only when its recognition accuracy reaches a high level can it be used in practice. Due to the difference in the data distributions of the training dataset and the test dataset and the lack of training samples, the classification accuracies of general machine learning algorithms are not satisfactory. In fact, among the many machine learning methods used to process epilepsy electroencephalogram (EEG) signals, most are black box methods; however, in medicine, methods with explanatory power are needed. In response to these three challenges, this paper proposes a novel technique based on domain adaptation learning, semi-supervised learning and a fuzzy system. In detail, we use domain adaptation learning to reduce deviation from the data distribution, semi-supervised learning to compensate for the lack of training samples, and the Takagi-Sugen-Kang (TSK) fuzzy system model to improve interpretability. Our experimental results show that the performance of the new method is better than those of most advanced epilepsy classification methods.

Sign in / Sign up

Export Citation Format

Share Document