Forecasting the Market with Machine Learning Algorithms: An Application of NMC-BERT-LSTM-DQN-X Algorithm in Quantitative Trading

Although machine learning (ML) algorithms have been widely used in forecasting the trend of stock market indices, they failed to consider the following crucial aspects for market forecasting: (1) that investors’ emotions and attitudes toward future market trends have material impacts on market trend forecasting (2) the length of past market data should be dynamically adjusted according to the market status and (3) the transition of market statutes should be considered when forecasting market trends. In this study, we proposed an innovative ML method to forecast China's stock market trends by addressing the three issues above. Specifically, sentimental factors (see Appendix [1] for full trans) were first collected to measure investors’ emotions and attitudes. Then, a non-stationary Markov chain (NMC) model was used to capture dynamic transitions of market statutes. We choose the state-of-the-art (SOTA) method, namely, Bidirectional Encoder Representations from Transformers ( BERT ), to predict the state of the market at time t , and a long short-term memory ( LSTM ) model was used to estimate the varying length of past market data in market trend prediction, where the input of LSTM (the state of the market at time t ) was the output of BERT and probabilities for opening and closing of the gates in the LSTM model were based on outputs of the NMC model. Finally, the optimum parameters of the proposed algorithm were calculated using a reinforced learning-based deep Q-Network. Compared to existing forecasting methods, the proposed algorithm achieves better results with a forecasting accuracy of 61.77%, annualized return of 29.25%, and maximum losses of −8.29%. Furthermore, the proposed model achieved the lowest forecasting error: mean square error (0.095), root mean square error (0.0739), mean absolute error (0.104), and mean absolute percent error (15.1%). As a result, the proposed market forecasting model can help investors obtain more accurate market forecast information.

Improved Dividend Estimation from Intraday Quotes

Liquid financial markets, such as the options market of the S&P 500 index, create vast amounts of data every day, i.e., so-called intraday data. However, this highly granular data is often reduced to single-time when used to estimate financial quantities. This under-utilization of the data may reduce the quality of the estimates. In this paper, we study the impacts on estimation quality when using intraday data to estimate dividends. The methodology is based on earlier linear regression (ordinary least squares) estimates, which have been adapted to intraday data. Further, the method is also generalized in two aspects. First, the dividends are expressed as present values of future dividends rather than dividend yields. Second, to account for heteroscedasticity, the estimation methodology was formulated as a weighted least squares, where the weights are determined from the market data. This method is compared with a traditional method on out-of-sample S&P 500 European options market data. The results show that estimations based on intraday data have, with statistical significance, a higher quality than the corresponding single-times estimates. Additionally, the two generalizations of the methodology are shown to improve the estimation quality further.

Comprehensive Dutch market data analysis shows that e-liquids with nicotine salts have both higher nicotine and flavour concentrations than those with free-base nicotine

ObjectivesRecent years have seen an increase in e-liquids containing nicotine salts. Nicotine salts are less harsh and bitter than free-base nicotine and therefore can facilitate inhalation. Because inhalation-facilitating ingredients are banned in the European Union, we assessed the occurrence and characteristics of nicotine salt-containing e-liquids notified for the Netherlands.MethodsWe analysed data for 39 030 products, submitted by manufacturers in the European Union Common Entry Gate system, as extracted on 30 June 2020.ResultsNicotine salts were present in 13% of e-liquids, especially in pod-related e-liquids (73%) and e-liquids registered from 2018 onwards (over 25%). We found six nicotine salt ingredients (NSIs): nicotine lactate, salicylate, benzoate, levulinate, ditartrate and malate. Nicotine salts also occurred as nicotine–organic acid ingredient combination (NAIC), like nicotine and benzoic acid. Nicotine concentrations were twofold higher in e-liquids with NSI (median 14 mg/mL) and NAIC (11 mg/mL) than for free-base nicotine (6 mg/mL). E-liquids with NSI contained a fourfold higher number (median n=17) and concentration (median 31.0 mg/mL) of flavour ingredients than e-liquids with free-base nicotine (n=4, 7.4 mg/mL). In NAIC-containing e-liquids, these were threefold higher (n=12, 21.5 mg/mL). E-liquids with nicotine salts were less often tobacco flavoured but more often had fruity or sweet flavours.ConclusionsA substantial and increasing share of e-liquids in the Netherlands contains nicotine salts. Their characteristics can make such e-liquids more addictive and more attractive, especially to young and beginning users. Policymakers are advised to consider regulating products containing nicotine salts.

An Ensemble of Random Forest Gradient Boosting Machine and Deep Learning Methods for Stock Price Prediction

Stochastic time series analysis of high-frequency stock market data is a very challenging task for the analysts due to the lack availability of efficient tool and techniques for big data analytics. This has opened the door of opportunities for the developer and researcher to develop intelligent and machine learning based tools and techniques for data analytics. This paper proposed an ensemble for stock market data prediction using three most prominent machine learning based techniques. The stock market dataset with raw data size of 39364 KB with all attributes and processed data size of 11826 KB having 872435 instances. The proposed work implements an ensemble model comprises of Deep Learning, Gradient Boosting Machine (GBM) and distributed Random Forest techniques of data analytics. The performance results of the ensemble model are compared with each of the individual methods i.e. deep learning, Gradient Boosting Machine (GBM) and Random Forest. The ensemble model performs better and achieves the highest accuracy of 0.99 and lowest error (RMSE) of 0.1.

Stock market investment and different behavioural patterns: an exploratory study

PurposeThe purpose of the research is to gain an understanding about how stock market investors impact various behavioural personality traits in various consumer groups with differing levels of motivation and capacity to absorb emerging stock market data.Design/methodology/approachThe research has used structural equation modelling (SEM) to test the validity of the theoretical model.FindingsThe current paper is the first study that uses stock market data from an emerging economy to examine the relationship between stock market investment and different behavioural patterns such as stock market attachment, trust, satisfaction and loyalty. The authors observe the presence of direct positive relationships between stock market investment and different behavioural personality traits. Moreover, the authors also observe that stock market attachment can be seen as an intermediary variable between stock investment involvement and satisfaction. The empirical findings also suggest the presence of indirect relationships between stock investment involvement and satisfaction and between stock market attachment and loyalty. The authors find that the indirect relationship between stock market attachment and loyalty occurs when the level of satisfaction is higher. Therefore, satisfaction appears to facilitate the relationship between stock market attachment and loyalty.Research limitations/implicationsOne major limitation of the study is data availability. More specifically, the study was conducted with customers of eight different banks in the province of Eskisehir, Turkey. From the 250 questionnaires distributed, 173 were returned, yielding a response rate of 69.2%.Practical implicationsBy identifying the trait characteristics of segments of stock market participants relative to their propensity to invest in stocks, it is possible to tailor messages that influence people to invest for the long term.Originality/valueThe paper deploys stock market data from an emerging economy to investigate the relationship between stock market investment and different surface traits such as stock market attachment, trust, satisfaction and loyalty. To the best of the authors' knowledge the current paper is the first such study.

Pragmatic pharmacoeconomic analyses by using post-market adverse drug reaction reports: an illustration using infliximab, adalimumab, and the Canada vigilance adverse reaction database

Background The prediction of the real-world cost of adverse drug reactions (ADRs) has historically relied on the data from randomized controlled trials (RCT). However, trial conditions do not always reflect the real-world applications of pharmaceutical products; hence, they may not accurately portray the actual risks of ADRs associated with them. The objective of this study is two-fold: (a) demonstrate whether and how post-market and RCT ADR data could lead to different conclusions for a set of drugs of interest, and (b) evaluate the potential economic impact of the post-market ADRs associated with those drugs. Methods We selected two TNF-α inhibitor biologics, infliximab and adalimumab, and used the Canada Vigilance Adverse Reaction (CVAR) online database as a source of post-market ADR data. Adverse reaction data from RCTs were obtained from ClinicalTrials.gov. Direct healthcare costs associated with adverse reactions were obtained from Canadian Institute for Health Information (CIHI) or Interactive Health Data Application, Alberta. We calculated post-market ADR rates and compared them with those found in the randomized controlled trials of these two drugs. Using the post-market data, we estimated the costs associated with serious ADRs from three perspectives: patient, health system, and societal. Results For both drugs, the post-market and RCT data exhibited significantly different adverse reaction rates for several different clinical outcomes. As a general trend, more serious adverse reactions, such as death, appeared to have a higher rate in post-market applications compared to the clinical trials. The estimated average annual economic burden of the severe adverse reaction outcomes ranged from $10 million to $20 million for infliximab and $6 million to $19 million for adalimumab. Conclusions The frequency and severity of post-market adverse reactions associated with pharmaceutical products may significantly differ from those detected in the clinical trials. Despite possible methodological differences, this is due to the fact that post-market data reflect the externalities of the real-world that are absent in RCTs. The economic burden of adverse reactions can be substantial, and the cost calculated using post-market data is better reflective of the cost of ADRs in the real-world.

Fuzzy Transfer Learning in Time Series Forecasting for Stock Market Prices

Transfer learning involves transferring prior knowledge of solving similar problems in order to achieve quick and efficient solution. The aim of fuzzy transfer learning is to transfer prior knowledge in an imprecise environment. Time series like stock market data are non-linear in nature and movement of stock is uncertain, so it is quite difficult following the stock market and in decision making. In this study, we propose a method to forecast stock market time series in the situation when we can use prior experience to make decisions. Fuzzy transfer learning (FuzzyTL) is based on knowledge transfer in that and adapting rules obtained domain. Three different stock market time series data sets are used for comparative study. It is observed that the effect of knowledge transferring works well together with smoothing of dependent attributes as the stock market data fluctuate with time. Finally, we give an empirical application in Shenzhen stock market with larger data sets to demonstrate the performance of the model. We have explored FuzzyTL in time series prediction to unerstand the essence of FuzzyTL. We were working on the question of the capability of FuzzyTL in improving prediction accuracy. From the comparisons, it can be said fuzzy transfer learning with smoothing improves prediction accuracy efficiently.

Wann ist ein Unternehmen profitabel?

This paper analyses how the profitability of public companies in the Federal Republic of Germany has been measured since the 1950s and under which conditions corporations were considered successful. For this purpose, textbooks and arguments of leading business economists, speeches and publications of managing directors and companies’ annual reports are surveyed, in order to identify trends and policy changes. The paper demonstrates that the introduction of shareholder value approaches, based on financial market data, in the 1990s led to a fundamental change in management practices, connected to innovative financial accounting techniques. Since then, companies’ profitability has been assessed in relation to benchmarks derived from financial market data. Financial markets thus became increasingly relevant for decisionmaking processes in the real economy.