Profitable Portfolio Research Articles

The Wikipedia effect: Analyzing investor attention for strategic investment decisions

The industries with the highest increase in Wikipedia page views are the telecommunications, consumer durables, and high-technology sectors. On average, firms with increasing views have higher returns, lower earnings per share, and higher price-to-earnings ratios. Moreover, long-short investment strategies using changes in Wikipedia page views yield profitable portfolio returns, outperforming the market. Factor model regressions demonstrate that these portfolios exhibit positive abnormal returns relative to various benchmark models. In particular, the high-minus-low portfolio shows statistically significant alphas that common risk factors cannot explain. The findings suggest that changes in firms’ Wikipedia page views are a valuable indicator of stock performance.

Portfolio Optimization with Prediction-Based Return Using Long Short-Term Memory Neural Networks: Testing on Upward and Downward European Markets

AbstractIn recent years, artificial intelligence has helped to improve processes and performance in many different areas: in the field of portfolio optimization, the inputs play a crucial role, and the use of machine learning algorithms can improve the estimation of the inputs to create robust portfolios able to generate returns consistently. This paper combines classical mean–variance optimization and machine learning techniques, concretely long short-term memory neural networks to provide more accurate predicted returns and generate profitable portfolios for 10 holding periods that present different financial contexts. The proposed algorithm is trained and tested with historical EURO STOXX 50® Index data from January 2015 to December 2020, and from January 2021 to June 2022, respectively. Empirical results show that our LSTM neural networks are able to achieve minor predictive errors since the average of the MSE of the 10 holding periods is 0.00047, the average of the MAE is 0.01634, and predict the direction of returns with an average accuracy over the 10 investment periods of 95.8%. Our prediction-based portfolios consistently beat the EURO STOXX 50® Index, achieving superior positive results even during bear markets.

Constructing Portfolio Optimization: Analysis in Indonesia Non-Cyclical Industry (Markowitz Approach and Skewness and Kurtosis)

Purpose: The purpose of this research is to examine the application of portfolio optimization in the context of real-world financial issues, particularly in light of the challenges posed by the COVID-19 pandemic. Traditional portfolio optimization strategies, such as those proposed by Markowitz, often rely on the assumption of normally distributed returns, which may not accurately capture the risks associated with extreme events like the COVID-19 crisis. This study aims to shed new light on portfolio optimization methods by exploring various approaches and considering the implications of non-normally distributed returns on portfolio construction. Methods: This research employs a quantitative approach to analyze portfolio optimization techniques in the face of non-normally distributed returns. Using data from financial markets impacted by the COVID-19 pandemic, the study investigates different portfolio construction methods, including risk-free rate for equally weighted portfolios, optimal risk portfolios, minimum variance weights, and maximum expected returns. Various risk metrics such as variance, Sharpe ratio, and standard deviation are considered to evaluate portfolio performance under different constraints. Results and discussion: The empirical findings highlight the limitations of traditional portfolio optimization techniques, particularly in accurately assessing and managing risk in the presence of non-normally distributed returns. Assets exhibit heavily tailed returns, leading to an underestimation of risk when using standard approaches. The study identifies that certain assets offer high returns but also entail significant risks, necessitating a nuanced approach to portfolio construction. By considering stable distribution models and optimizing for both maximum expected return and minimum variance weight, investors can build more profitable and diversified portfolios while managing risk effectively. Implications of the research: The research findings have important implications for investors and financial practitioners, particularly in navigating uncertain market conditions such as those brought about by the COVID-19 pandemic. By recognizing the limitations of traditional portfolio optimization methods and embracing more sophisticated approaches that account for non-normally distributed returns, investors can make more informed decisions and better manage portfolio risk. These insights can inform the development of robust investment strategies tailored to mitigate the impact of extreme events on portfolio performance. Originality/value: This research contributes to the literature by offering a fresh perspective on portfolio optimization under the backdrop of the COVID-19 pandemic. By systematically evaluating different portfolio construction methods and considering the implications of non-normally distributed returns, the study advances understanding in the field of financial risk management. The identification of stable distribution models and the emphasis on balancing maximum expected return with minimum variance weight provide practical guidance for investors seeking to build resilient and profitable portfolios in turbulent market environments. Overall, this research underscores the importance of adapting portfolio optimization strategies to address the realities of contemporary financial markets.

A Profitable Portfolio Allocation Strategy Based on Money Net-Flow Adjusted Deep Reinforcement Learning

A Profitable Portfolio Allocation Strategy Based on Money Net-Flow Adjusted Deep Reinforcement Learning

A continuous heterogeneous agent model for multi-asset pricing and portfolio construction under market matching friction

We propose a novel continuous Heterogeneous Agent Model (HAM) that applies to the general n>1 risky assets. In this model, every heterogeneous investor is identified with the continuous variables reflecting their personal characteristics. The heterogeneity of the market is captured by the distribution of these variables. By introducing market friction via a matching-based pricing mechanism, we derive a pricing equation from market clearing. The equation extends the classical CAPM and specifies how the bounded rationality and heterogeneity of investors can drive the actual price of risky assets away from their CAPM price. A system of ordinary (stochastic) differential equations are derived to govern the evolution of the total wealth and market value of risky assets held by the entire market (individual investors). From the equation system, a rolling-window-based maximum likelihood algorithm is developed to calibrate the model, which turn our HAM to a forecast tool for stock returns and an adaptive adjustment strategy for portfolio construction. This bridges the theoretical study on HAMs with the empirical studies of asset pricing and portfolio theory. We apply this tool to the Chinese A-share market and make a comparative analysis on its performance. The results demonstrate that compared to the classical forecast approaches in literature, our HAM has the advantage in identifying the tail risk in real market, which results in a more profitable portfolio in a variety of different investment horizons.

Can machine learning identify sector-level financial ratios that predict sector returns?

Academics and practitioners use fundamental ratios to evaluate an industry's financial and operating performance. WRDS has large firm-level and sector/industry-level databases that provide over 70 financial ratios that are applied in finance classes to compare and assess relative financial performance. However, there has been a lack of sophisticated econometric methods assessing these ratios' importance in predicting sector-level stock return performance. Using Elastic Net methods, we identify financial ratios that significantly forecast out-of-sample sector stock returns and find that these predictive ratios vary across sectors. We form long and long-short portfolios that consistently outperform the market over-time. Long portfolios generate significant alpha and large utility gains, boost the Sharpe and Sortino ratios, and a cumulative investment portfolio exceeds the market benchmark by five times. Long-short portfolios generate Fama-French 4-factor and 6-factor alphas between 4–9% and cumulative investment gains from six to fourteen times. Our research establishes that machine learning can identify financial ratios that significantly predict sector returns and generate profitable portfolio allocation.

IMPACT OF FINANCIAL LITERACY AND INVESTMENT SKILLS ON INVESTORS’ BEHAVIOUR IN COLOMBO STOCK EXCHANGE – MEDIATING ROLE OF DIGITAL LITERACY

"Rationality among the investors in decision making is essence for the performance of Colombo stock exchange (CSE) and also to development of country’s economy. Investor’s behaviour is unpredictable by any economist and even by equity market experts. Investor in the CSE can exhibit rational or irrational behaviour due to the multiple factors affecting investment. Therefore, study of investors behaviour is vital for improve the rationality among investors. As reviewed from literatures of many article, researchers found that, the increasing level of financial literacy and high level of practice in investment skills can improve the rationality of investment decision. The Covid-19 pandemic has driven in the explosive growth of ""Digital"" Financial Services (DFS) and online trading in CSE. Therefore, it is vital for the investor to familiar with digital advancement. Accordingly, Present study amid to analyse the impact of financial literacy and investment skills on investors behaviour and further carried the mediation analyse of digital literacy. Researchers point of view is “financial literacy teach investor to use money efficiently while, investment skills provide the ability to track, collect, analyse and research data to make decision in order to select the most profitable portfolio in CSE. However digital literacy is vital to apply the knowledge of Finance and investment skills in making investment decision in CSE”. This study conducted to identify the rationality of investor’s behaviour from assessing the impact of financial literacy and investment skills level of investors in CSE, and find the mediating role of Digital literacy. A self-administrated questionnaire was issued to the investors in CSE and 227 responses were collected via online. Present study adopted descriptive and inferential statistics to measure the independent- dependent relationship and Mediation relationship of the study. The findings revealed that, financial literacy and investment skills have impact on investor’s behaviour whereas digital literacy does not impact on investor’s behaviour. Moreover, findings revealed that, primal mediation of the study was insignificant accordingly there is no mediation predicted in present study.

Equity-Market-Neutral Strategy Portfolio Construction Using LSTM-Based Stock Prediction and Selection: An Application to S&P500 Consumer Staples Stocks

In recent years, a great deal of attention has been devoted to the use of neural networks in portfolio management, particularly in the prediction of stock prices. Building a more profitable portfolio with less risk has always been a challenging task. In this study, we propose a model to build a portfolio according to an equity-market-neutral (EMN) investment strategy. In this portfolio, the selection of stocks comprises two steps: a prediction of the individual returns of stocks using LSTM neural network, followed by a ranking of these stocks according to their predicted returns. The stocks with the best predicted returns and those with the worst predicted returns constitute, respectively, the long side and the short side of the portfolio to be built. The proposed model has two key benefits. First, data from historical quotes and technical and fundamental indicators are used in the LSTM network to provide good predictions. Second, the EMN strategy allows for the funding of long-position stocks by short-sell-position stocks, thus hedging the market risk. The results show that the built portfolios performed better compared to the benchmarks. Nonetheless, performance slowed down during the COVID-19 pandemic.

Incorporating a choice-based diffusion model into a bi-objective multi-generation product optimization problem under consignment stock policy

Rapid changes and evolution in technological innovation have spilled over many industries and compelled firms to release new products periodically. The advent of new products intensifies the importance of analyzing sales data trends and finding the most profitable portfolio that can no longer be ignored in production planning. The topic of diffusion processes has evoked considerable interest from academics and practitioners to forecast sales patterns of two or three competitive generations. However, there is still a lack of in-depth discussion on the effects of core determinants on customers' preferences and choices. This paper bridges the gaps in the literature by conceptualizing the interaction effects among current and new products on diffusion processes that could affect supply chain operations. Specifically, the Bass diffusion model (BDM) is modified to estimate market volume, advertisement effects, and sales data trends, then combined with a choice model to examine how consumers' propensities affect the choice probability of alternatives. The proposed choice-based diffusion model is incorporated into a supply chain problem for joint planning on a production-location-inventory problem concerning the economic and social aspects of designing multi-generation production lines. The model formulation is a bi-objective mixed-integer nonlinear problem with a computation burden. Thus an outer approximation-based algorithm is proposed to reduce computational complexity and combined with an augmented ε-constraint to generate efficient Pareto sets. The model is validated by conducting a case study in Iran and compared to other established models. The computational results demonstrate the proposed model outperforms the expert forecasts, and there are no significant differences between the total sales estimated by the proposed model and the actual sales. The manager is advised to adopt a conscious policy for adoption engagement under supply constraints. It should price the products in a way be more tightly aligned with customers’ desirability and overcome the supply constraints by increasing production or storage capacities.

From deterministic to stochastic: an interpretable stochastic model-free reinforcement learning framework for portfolio optimization.

As a fundamental problem in algorithmic trading, portfolio optimization aims to maximize the cumulative return by continuously investing in various financial derivatives within a given time period. Recent years have witnessed the transformation from traditional machine learning trading algorithms to reinforcement learning algorithms due to their superior nature of sequential decision making. However, the exponential growth of the imperfect and noisy financial data that is supposedly leveraged by the deterministic strategy in reinforcement learning, makes it increasingly challenging for one to continuously obtain a profitable portfolio. Thus, in this work, we first reconstruct several deterministic and stochastic reinforcement algorithms as benchmarks. On this basis, we introduce a risk-aware reward function to balance the risk and return. Importantly, we propose a novel interpretable stochastic reinforcement learning framework which tailors a stochastic policy parameterized by Gaussian Mixtures and a distributional critic realized by quantiles for the problem of portfolio optimization. In our experiment, the proposed algorithm demonstrates its superior performance on U.S. market stocks with a 63.1% annual rate of return while at the same time reducing the market value max drawdown by 10% when back-testing during the stock market crash around March 2020.

Portfolio Construction in China’s Stock Market: Take Automobile Industry as Example

With the development of new energy vehicles, investors focus on the automobile subsidy in rural areas, and automobile section of China’s A share has been active recently. Based on ARMA model and Mean-Variance model, this paper hopes to give advice on portfolio construction. All the closing price of the stocks are download from AkShare database, from the 9th June, 2020 to the 9th June, 2022. After the stationary test and correlation test for time series, make predictions about return rates for the following 20 days based on ARMA model, construct a profitable portfolio based on Mean-Variance model, compare the accumulative returns and Sharp ratios of constructed portfolio and the one with equal weights. The result reveals that the constructed portfolio outperforms the latter portfolio obviously, so it can be considered that the construction is reasonable. This paper follows the current affairs about the policies on car subsidy in rural areas, uses statistical methods to provide suggestions of wealth allocation for investors, expands the application of the two models.

T2V_TF: An adaptive timing encoding mechanism based Transformer with multi-source heterogeneous information fusion for portfolio management: A case of the Chinese A50 stocks

The Stock prediction has traditionally been an attractive and challenging topic for investors and researchers. Traditionally, people concern more about predicting stock prices, less effort has been made to recommend stocks for constructing a profitable portfolio. Moreover, in existing methods for stock prediction, most of them construct models based on one or two kinds of features like stock prices, news sentiment, or simple technical indicators, and disregard the importance of multi-source information fusion. In response to this concern, we propose a novel model T2V_TF based on deep learning by combining both Time2Vec and Transformer technologies. To introduce more diverse information into the proposed model, we further conduct an in-depth exploration of the extraction and fusion of multi-source heterogeneous information, which includes the trading data, time–frequency features, Alpha 101 and Alpha 191 technical indicators, and sentiment scores. Moreover, to increase the ranking ability of our model, T2V_TF takes the ranking loss as the loss function instead of the widely used regression loss. Finally, all the technological innovations of this paper are verified on the portfolio constructed based on the A50 stocks from the Chinese stock market. The experimental results demonstrate that our proposed T2V_TF can get better portfolio cumulative return, compared with other models including the multi-layer perceptron, the support vector machine, the gradient boosting decision trees, the long short-term memory model, and the attention-based long short-term memory model, and the Transformer.

EVALUATING MEAN TIME TO OPERATION (MTTO) AND OTHER RELIABILITIES METHODS FOR COMPUTING CRITERION RELIABLE RISK OF RETURN (RRR) IN TEHRAN STOCK EXCHANGE

This study evaluates a new criterion with high efficiency in the decision matrix as a criterion with quality in Multi-Criteria Decision Making (MCDM) methods. This research discusses some concepts such as reliability techniques, MCDM, decision matrix (DM) and real return also proposes a criterion Reliable Risk of Return that is called RRR. This criterion can show how an investor can trust to the return of a company that their return is more than the specific amount that the investor defines it. For this purpose, by applying a creative to use reliability techniques that are usually used in other science, especially engineering. This research proposes a new technique MTTO and uses other reliability techniques creatively. Results of this research can be helpful for investors to choose the best company from the list of companies also recommend profitable portfolios. The MTBF, MTTF, FIT and MTTO are famous techniques that are used to calculate the reliability of a system.

Cryptocurrency portfolio optimization with multivariate normal tempered stable processes and Foster-Hart risk

We study portfolio optimization of four major cryptocurrencies. Our time series model is a generalized autoregressive conditional heteroscedasticity (GARCH) model with multivariate normal tempered stable (MNTS) distributed residuals used to capture the non-Gaussian cryptocurrency return dynamics. Based on the time series model, we optimize the portfolio in terms of Foster-Hart risk. Those sophisticated techniques are not yet documented in the context of cryptocurrency. Statistical tests suggest that the MNTS distributed GARCH model fits better with cryptocurrency returns than the competing GARCH-type models. We find that Foster-Hart optimization yields a more profitable portfolio with better risk-return balance than the prevailing approach.

Stock Price Reactions to Publications of Financial Statements: Evidence from the Moscow Stock Exchange

This paper analyses the effects of financial statements on the efficiency of the Russian stock market. Specifically, we analyse the impact of financial reporting on stock prices of the firms listed on the Moscow Stock Exchange. By means of the widely used event study method, which dates back to Ball and Brown [1], we analyse how corporate news publication affects stock prices.
 Our research analyses 1000 samples, each consisting of 30 events, independent of the underlying stocks/firms and analyses the relation between the behaviour of the share prices and the release of the firms’ annual, quarterly, and unscheduled financial statements. We use the daily stock price data of 56 components of the Russia Trading System Index from the years 2014 to 2020 in order to analyse the relation between the behaviour of the shares’ prices and the releases of the firms’ annual, quarterly, and unscheduled financial statements.
 Using an ordinary least squares market model, we estimate the market parameters and especially the so-called normal returns, i.e. benchmark values. With this, we calculate the abnormal returns, i.e. the price changes caused by the events cf. [1; 2]. We perform several statistical tests for non-Gaussian distribution of these abnormal returns and find that there is a significantly non-Gaussian relationship between the publication of financial statements and the prices of the shares, which should not be the case in an efficient market [2].
 Our results indicate that stock price volatility on the publication of financial statements may be caused by some information asymmetry, and demonstrate that the Russian stock market responds significantly to new information. Thus, we discuss recommendations to improve the information content of financial statements in Russia. This means analysts and fund managers can use new information to predict future stock returns and, thus, construct profitable portfolios.

Can the probability of extreme returns be the basis for profitable portfolios? Evidence from China

Based on the generalized logit predicting model from Jang and Kang (2019), this paper estimates the ex-ante probability of extreme returns and finds that the significantly negative (positive) influence of the predicted crash (jackpot) probability is robust, whether based on the traditional portfolio construction, orthogonalized portfolio construction and Fame-Macbeth cross-section regression. Further analyses show both the behavioral speculators' trading and rational investors' arbitrage limits could be the sources of mispricing caused by extreme returns. Overall, this paper applies a predicting model to estimate the probabilities of the future extreme returns, and figures out the significant influence and possible sources of the crash and jackpot probabilities in China. Portfolios based on extreme return probabilities can be profitable and steady bases for uninformed investors in the Chinese stock market.

Gold Against the Machine

Despite the increasing significance and the central role of stock markets, investing in gold has remained a popular choice among market participants. The necessity to forecast gold prices has sparked a voluminous literature on the matter, though there is no consensus regarding the variables that drive gold prices evolution or the methodology that adheres to the true data generating mechanism. In this paper, we forecast gold prices comparing econometric and machine learning methodologies in order to produce a model that can better grasps the dynamics of gold prices. To do so, we filter the most prominent variables proposed by the relevant literature exploiting the ability of the Ensemble Empirical Mode Decomposition algorithm to separate noise from the actual evolution of a timeseries. Then, we train Support Vector Regression models coupled with the linear and nonlinear kernels. Our empirical findings suggest that the proposed model adheres closer to gold price evolution than Ordinary Least Square regression and Least Absolute Shrinkage and Selection Operator models used in the literature, while it can be utilized in shaping profitable portfolios.

Boosting Insights in Insurance Tariff Plans with Tree-Based Machine Learning Methods

Pricing actuaries typically operate within the framework of generalized linear models (GLMs). With the upswing of data analytics, our study puts focus on machine learning methods to develop full tariff plans built from both the frequency and severity of claims. We adapt the loss functions used in the algorithms such that the specific characteristics of insurance data are carefully incorporated: highly unbalanced count data with excess zeros and varying exposure on the frequency side combined with scarce but potentially long-tailed data on the severity side. A key requirement is the need for transparent and interpretable pricing models that are easily explainable to all stakeholders. We therefore focus on machine learning with decision trees: Starting from simple regression trees, we work toward more advanced ensembles such as random forests and boosted trees. We show how to choose the optimal tuning parameters for these models in an elaborate cross-validation scheme. In addition, we present visualization tools to obtain insights from the resulting models, and the economic value of these new modeling approaches is evaluated. Boosted trees outperform the classical GLMs, allowing the insurer to form profitable portfolios and to guard against potential adverse risk selection.

Clustering and portfolio selection problems: A unified framework

Given a set of assets and an investment capital, the classical portfolio selection problem consists in determining the amount of capital to be invested in each asset in order to build the most profitable portfolio. The portfolio optimization problem is naturally modeled as a mean-risk bi-criteria optimization problem where the mean rate of return of the portfolio must be maximized whereas a given risk measure must be minimized. Several mathematical programming models and techniques have been presented in the literature in order to efficiently solve the portfolio problem. A relatively recent promising line of research is to exploit clustering information of an assets network in order to develop new portfolio optimization paradigms. In this paper we endow the assets network with a metric based on correlation coefficients between assets’ returns, and show how classical location problems on networks can be used for clustering assets. In particular, by adding a new criterion to the portfolio selection problem based on an objective function of a classical location problem, we are able to measure the effect of clustering on the selected assets with respect to the non-selected ones. Most papers dealing with clustering and portfolio selection models solve these problems in two distinct steps: cluster first and then selection. The innovative contribution of this paper is that we propose a Mixed-Integer Linear Programming formulation for dealing with this problem in a unified phase. The effectiveness of our approach is validated reporting some computational experiments on some real financial datasets.

Alpha Go Everywhere: Machine Learning and International Stock Returns

We apply machine learning techniques and use stock characteristics to predict the cross-section of stock returns in 31 international markets. We conduct a stringent out-of-sample test to examine concerns about overfitting: the models are trained with past U.S. data and used to predict international stock returns. Neural network (NN) models outperform linear models significantly in terms of forming profitable portfolios and modestly in out-of-sample R^2. We achieve stronger results when the models are trained separately for each market, allowing for market-specific return-characteristic relationships. However, regression tree models are prone to overfitting and they underperform linear models when the number of observations is low. We also show that international markets are integrated in the sense that U.S.-based variables can further enhance the return predictability of market-specific NN models.