NYSE Stocks Research Articles

Intraday and daily dynamics of cryptocurrency

This paper examines and compares intraday and intraweek patterns in hourly and daily prices, returns, volumes and volatility of native cryptocurrencies, stablecoins and tokens traded on Bitstamp. We show that native cryptocurrencies and tokens share common intraday periodicity determined by the operating times of the NYSE, LSE and Hang Seng stock exchange markets. Periodic patterns are also documented in the returns on cryptocurrency market portfolio approximated by the PCA applied to intraday and intraweek cross-sectional correlation matrices of cryptocurrency returns. Stablecoins have distinct dynamics and their daily and hourly returns are uncorrelated with one another and with the returns on other cryptocurrencies. We introduce a functional CAPM to accommodate the periodic patterns and estimate it by regressing the functions of intraday and intraweek cryptocurrency returns on the market portfolio. We show that the return functions on Bitcoin, Ether, and Link satisfy affine relationships with the return functions of the market portfolio and their functional betas display periodic intraday and intraweek patterns.

Big portfolio selection by graph-based conditional moments method

This paper proposes a new graph-based conditional moments (GRACE) method to do portfolio selection based on thousands of stocks or even more. The GRACE method first learns the conditional quantiles and mean of stock returns via a factor-augmented temporal graph convolutional network, which is guided by the set of stock-to-stock relations as well as the set of factor-to-stock relations. Next, the GRACE method learns the conditional variance, skewness, and kurtosis of stock returns from the learned conditional quantiles via the quantiled conditional moment method. Finally, the GRACE method uses the learned conditional mean, variance, skewness, and kurtosis to construct several performance measures, which are criteria to sort the stocks to proceed the portfolio selection in the well-known 10-decile framework. An application to NASDAQ and NYSE stock markets shows that the GRACE method performs much better than its competitors, particularly when the performance measures are comprised of conditional variance, skewness, and kurtosis.

Cross-border investment, international consolidation and reversal of US stock exchanges: the gift and the curse of being too big, too liquid and too visible

PurposeThis paper investigates the distinctive role of the US stock exchanges in the process of international consolidation. Besides the USA's leading role in financial markets, the focus on the country is motivated by its uniqueness within the stock exchange consolidation landscape, since, on the one hand, it has been involved in two different stock exchange mergers – with Nasdaq and NYSE – and, on the other hand, it has experienced a “reversal”, having joined and then left the Euronext-NYSE platform.Design/methodology/approachTo investigate the effect of the NYSE-Euronext split on cross-border holdings and the role of the US as a member of the consolidated platform, we adopt a feasible Generalized Least Squares specification correcting for both heteroskedasticity and general correlation of observations across destination-countries, with standard errors adjusted for two-way clustering at the investing-country and year levels.FindingsDifferently from other mergers, we find a weak sensitivity of US inward and outward cross-border investments to stock exchange consolidation, and, consequently, to its reversal. The data suggest that the larger, the more liquid and the more visible the involved stock exchanges are, the less sensitive cross-border investment is to consolidation. Drawing on the cross-listing and cross-delisting literature, we formulate the conjecture that this evidence can be explained by decreasing returns of foreign investment to consolidation: the extraordinary large size, liquidity and visibility of the US stock exchanges diminishes the value of the role played by stock exchange consolidation in reducing cross-border barriers among member countries, so that it makes also the effects of its retreat non-significant.Originality/valueThis paper is the first, to best of our knowledge, to investigate the mirror phenomenon, that is, the “consolidation reversal” process of the NYSE stock exchange, the purpose being to understand its consequences for cross-border holdings. In the first part of this paper, we document no significant effect of the 2014 reversal on cross-border investments. The apparent absence of this effect could be due either to a level of cross-border investments remaining equally high (denoting persistence in investors' behavior) or to an equally non-significant effect of consolidation and reversal of the US stock exchanges on cross-border equity investments. The evidence supports the latter hypothesis and reveals an overall weak sensitivity of US cross-border investments (inward or outward) to stock exchange consolidation and, consequently, to its reversal. We formulate the conjecture, tested in the second part of the paper, that this evidence is due to the presence of diminishing returns of exchange consolidation's scale for foreign investors: the extraordinary large size, liquidity and visibility of the US stock exchanges makes the role of stock exchange consolidation less valuable in dampening cross-border barriers; consequently, also the reversal phenomenon presents no sizeable effects.

Reducing transaction costs using intraday forecasts of limit order book slopes

AbstractMarket participants who need to trade a significant number of securities within a given period can face high transaction costs. In this paper, we document how improvements in intraday liquidity forecasts can help reduce total transaction costs. We compare various approaches for forecasting intraday transaction costs, including autoregressive and machine learning models, using comprehensive ultra‐high‐frequency limit order book data for a sample of NYSE stocks from 2002 to 2012. Our results indicate that improved liquidity forecasts can significantly decrease total transaction costs. Simple models capturing seasonality in market liquidity tend to outperform alternative models.

A Regularized High-Dimensional Positive Definite Covariance Estimator with High-Frequency Data

This paper proposes a novel large-dimensional positive definite covariance estimator for high-frequency data under a general factor model framework. We demonstrate an appealing connection between the proposed estimator and a weighted group least absolute shrinkage and selection operator (LASSO) penalized least-squares estimator. The proposed estimator improves on traditional principal component analysis by allowing for weak factors, whose signal strengths are weak relative to idiosyncratic components. Despite the presence of microstructure noises and asynchronous trading, the proposed estimator achieves guarded positive definiteness without sacrificing the convergence rate. To make our method fully operational, we provide an extended simultaneous alternating direction method of multipliers algorithm to solve the resultant constrained convex minimization problem efficiently. Empirically, we study the monthly high-frequency covariance structure of the stock constituents of the S&P 500 index from 2008 to 2016, using all traded stocks from the NYSE, AMEX, and NASDAQ stock markets to construct the high-frequency Fama-French four and extended eleven economic factors. We further examine the out-of-sample performance of the proposed method through vast portfolio allocations, which deliver significantly reduced out-of-sample portfolio risk and enhanced Sharpe ratios. The success of our method supports the usefulness of machine learning techniques in finance. This paper was accepted by Agostino Capponi, finance. Funding: This work was supported by the Research Grants Council, University Grants Committee [Grants 11500119, 11505522, 11505721, and 21504818] and the National Natural Science Foundation of China (NSFC) Basic Scientific Center Project [Grant 71988101], entitled as “Econometric Modelling and Economic Policy Studies”, as well as NSFC [Grants 71803166 and 72173104]. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.04138 .

Applications of high-frequency data in finance: A bibliometric literature review

This study aims to provide a bibliometric literature review (BLR) on the applications of high-frequency data in finance. To the best of our knowledge, this is the first BLR on this topic. It aims to map the evolution of the literature, identifying the leading sources of knowledge in terms of the most influential journals, articles, and authors. It also provides a chronological development of the conceptual and intellectual structures of the networks in this topical research area. Using the Scopus database, the study identifies 2920 articles on the application of high-frequency intraday data in finance. These had been published in 393 journals during the period from 1977 to 2019. A thorough content analysis of the 100 most influential papers (ranked based on average citations per year) is also provided concerning research attributes in terms of datasets, asset classes, country of analysis and the major themes and sub-themes of these papers. The Journal of Banking and Finance is the leading journal in terms of the number of publications, whereas the Journal of Finance is the leading journal in terms of citations received on this topic. Tim Bollerslev is the leading author in this area in terms of the total number of publications (36), total citations (7241) and h-index (30). The most cited article in terms of total citations and average citations per year is Andersen, Bollerslev, Diebold, and Labys (2003) titled “Modeling and forecasting realized volatility”, which has appeared in Econometrica. The majority of the top 100 surveyed papers are empirical (66%). Volatility modeling as a major theme is the front runner with 29% of the surveyed papers. The theme “Volatility modeling” has most often been studied with Realized Volatility. The Trade and Quote (TAQ) database and 5-minute interval data appear to be the most favored choices in terms of data usage in high-frequency finance research. 56% of the surveyed papers have used the data on stocks, with NYSE stocks being the most popular, while US financial markets are the most commonly studied markets (65%).

Model and upper–lower bound estimation scheme for portfolio optimisation considering uncertain investment time horizon

In this paper, we study the portfolio selection problem considering the uncertain investment time horizon. We employ the Period Value at Risk (PVaR) to characterise the risk in the special cases and establish a fundamental mixed integer linear programming (MILP) model to obtain the optimal portfolio solutions. Identifying that the symmetric property of PVaR can significantly reduce the computation burden of the CPLEX solver, an enhanced MILP model is proposed. To verify the quality of obtained solutions, we also develop fast lower and upper bound estimation schemes respectively. Using the real-world data sets from NYSE and NASDAQ stock markets, numerical results are provided to show the efficiency of the enhanced model.

The origin of LYONs: A case study in financial innovation

The origin of LYONs: A case study in financial innovation

Multi-Timeframe Algorithmic Trading Bots Using Thick Data Heuristics with Deep Reinforcement Learning

This article presents an augmented Artificial Intelligence (AI) algorithmic trading approach that combines Thick Data Heuristic (TDH), with Deep Reinforcement Learning (DRL), to successfully learn trading execution timing policies. Combining the augmented AI human trader's intuition and heuristics with DRL techniques to provide more focused drivers for trading order execution timing is explored in this study. In this research, the goal is to solve the sequential decision-making problem of AI for profitable day and swing trading order timing executions. Enabling trading bots with cognitive intelligence and common-sense heuristics will offer traders including automatic traders an insight to understand the day-to-day swing trading timeframes indicators and arrive at mature trading decision-making. This article examines the performance of bots with Nasdaq and NYSE stocks that have a strong catalyst (info. which increases directional momentum) to find that they outperform benchmark algorithmic trading approaches. The research illustrates to the reader how to combine TDH and Deep Q-networks (DQN) into a TDH-DQN augmented AI trading bot. The bot learns through test data to predict the optimal timing of order executions autonomously on idealized trading time series data.

Predictability of stock returns using neural networks: Elusive in the long term

The rapidly growing neural network literature continually reports successful stock price forecasting results. Many of these studies use relatively short evaluation periods, spanning only a couple of years. In this paper, sustainability of the neural network forecast quality over the long term is analysed. Feedforward and recurrent networks are used to predict the direction of monthly stock price movements, with past price data as predictors. The analysis is conducted on the NYSE stocks over the 1971–2015 period and all the evaluations are performed out-of-sample. Statistically significant directional predictability for selected assets is found. However, the trading simulations reveal that directional predictability does not guarantee trading performance better than the benchmark buy-and-hold strategy. The opportunities for investors to use the tested models for profit appear to be episodic and periodically enhanced e.g. in periods of recession.

Homo-Hetero Pairing Regression Model: An Econometric Predictive Model of Homo Paired Data

The study aimed to examine the technical and fundamental hypotheses in NYSE, NASDAQ and S&P 500 stock exchange markets. The main determinants (variables) that were examined were stock trading volumes, closing stock prices and stock information available in the stock exchange market. The 240 days, 197 days and 253 days data of closing stock prices and trading volumes at NYSE, S&P500 and NASDAQ stock exchange markets were systematically collected from June 2021 to June 2022. The data was analysed by using the Homo-Hetero Pairing (HHP) Regression Model. This model was developed to detect the linear and non-linear behaviour of data. The study evidenced that both the technical and fundamental hypotheses in NYSE, S&P500 and NASDAQ stock exchange markets are defined by the inverse and S-curved models in two distinctive pairing classes called the positive-positive pairing (PPP) class and the negative-positive pairing (NPP) class. The study concluded that the optimal prediction of the stock price or return is achieved by the fundamentalists in the stock exchange markets. The study recommends that stock investors should priorities the use of the fundamental hypothesis to make their portfolio investment decision. Moreover, the study recommends the application of the HHP regression model in financial markets, economics, psychology, sociology, and medicine studies. In addition, the HHP regression model is recommended for the prediction of water waves in the investigation of hydrodynamic and erosion-accretion processes

Financial transaction taxes and the informational efficiency of financial markets: A structural estimation

We develop a new methodology to estimate the impact of a financial transaction tax (FTT) on financial market outcomes. In our sequential trading model, there are price-elastic noise and informed traders. We estimate the model through maximum likelihood for a sample of 60 NYSE stocks in 2017. We quantify the effect of introducing an FTT given the parameter estimates. An FTT increases the proportion of informed trading, improves information aggregation, but lowers trading volume and welfare. For some less liquid stocks, however, an FTT blocks private information aggregation.

An ensemble of a boosted hybrid of deep learning models and technical analysis for forecasting stock prices

For several years the modeling as well as forecasting of the prices of stocks have been extremely challenging for the business community and researchers as a result of the existence of noise in samples and also the non-stationary behaviour of information samples. Notwithstanding these drawbacks with improved deep learning, it is now possible to design schemes that will efficiently perform the feature learning task. For this work, we proposed a brand-new end to end algorithm labeled EHTS toward solving the stock price forecasting problem. The AB-CNN and CB-LSTM modules extract features from the stock price dataset and soon after amalgamating the results. Thus, the output of the concatenation stage was feed into the concluding stage which is a stand-alone MLP module. The inclusion of the LSTM and Attention Mechanism in our architecture is to extract long-range and exceptionally long-term stock price information. We experiment the proposed algorithm on two popular stocks both from the NYSE stock market namely “Johnson & Johnson” code-named, “JNJ” and the Bank of America (BAC). In terms of the rMSE, MAE and MAPE error metrics, our proposed scheme gives the lowest error value in all for all datasets. Also, five percentage training window sizes are experimented and EHTS outperforms all the baseline schemes for the different window sizes in all the two datasets with the 70% window size having the highest performance. In terms of number of epochs, EHTS uses the lowest number of epochs for training than the other schemes in all the datasets. Finally, we as well study our stock’s information to point out short-range trading opportunities by performing simulations on our stock price data. The metrics considered in the simulation are as follows: Moving Average (MA), Moving Average Convergence Divergence (MACD) curve, MACD histogram, Signal line, Relative Strength Index (RSI), Returns (R), Annual Returns (AR), Sharpe Ratio (SR), Annual Volatility (V), Maximum DrawDown (MDD) and Daily WinningRate (DWR). For all the aforementioned metrics, EHTS performs better than the baselines. Experimental results revealed that our proposed scheme outperformed the stand-alone deep learning schemes, statistical algorithms, and machine learning models from the beginning to the end.

A Practical Guide to harnessing the HAR volatility model

• Considers forecasting benefits from the HAR model looking at alternatives to OLS on raw RV. • Moving to WLS is useful with Robust Regression often the best. • Transforming RV is useful with log RV the best. • There are smaller marginal benefits from using WLS or RR with transformed RV. • Easy practical approaches to get better performance from the standard HAR. The standard heterogeneous autoregressive (HAR) model is perhaps the most popular benchmark model for forecasting return volatility. It is often estimated using raw realized variance (RV) and ordinary least squares (OLS). However, given the stylized facts of RV and well-known properties of OLS, this combination should be far from ideal. The aim of this paper is to investigate how the predictive accuracy of the HAR model depends on the choice of estimator, transformation, or combination scheme made by the market practitioner. In an out-of-sample study, covering the S&P 500 index and 26 frequently traded NYSE stocks, it is found that simple remedies systematically outperform not only standard HAR but also state of the art HARQ forecasts.

A Practical Guide to Harnessing the HAR Volatility Model

The standard heterogeneous autoregressive (HAR) model is perhaps the most popular benchmark model for forecasting return volatility. It is often estimated using raw realized variance (RV) and ordinary least squares (OLS). However, given the stylized facts of RV and well-known properties of OLS, this combination should be far from ideal. The aim of this paper is to investigate how the predictive accuracy of the HAR model depends on the choice of estimator, transformation, or combination scheme made by the market practitioner. In an out-of-sample study, covering the S&P 500 index and 26 frequently traded NYSE stocks, it is found that simple remedies systematically outperform not only standard HAR but also state of the art HARQ forecasts.

Asset Price Dynamics with Limited Attention

Abstract We identify long-lived pricing errors through a model in which inattentive investors arrive stochastically to trade. The model’s parameters are structurally estimated using daily NYSE market-maker inventories, retail order flows, and prices. The estimated model fits empirical variances, autocorrelations, and cross-autocorrelations among our three data series from daily to monthly frequencies. Pricing errors for the typical NYSE stock have a standard deviation of 3.2 percentage points and a half-life of 6.2 weeks. These pricing errors account for 9.4$\%$, 7.0$\%$, and 4.5$\%$ of the respective daily, monthly, and quarterly idiosyncratic return variances.

Measuring the stock's factor beta and identifying risk factors under market inefficiency

We provide a closed-form measure of the stock's factor-specific beta coefficient that allows for a delay in the reaction of stock prices to systematic information. This measure explicitly relates the stock's factor beta to the investment horizon and enables investors to determine the return measurement interval that is required for the stock's factor beta to fully reflect the stock's factor risk. Exploiting daily data on individual NYSE, Nasdaq, and AMEX-listed common stocks, we use this measure to estimate and investigate the properties of the stock's beta with respect to the market.

Product market competition, reputation, and dividend policy

ABSTRACT This paper examines the effect of product market competition on the dividend policy of firms. Previous studies report different relationships between product market competition and dividend payout policies. This paper aims to clarify the relationship between product market competition and dividend payout policies by dividing dividend-paying firms into two groups, high dividend firms and low dividend firms, based on the level of dividends. We hypothesize that when the product market is more competitive, high dividend firms pay more dividends, but low dividend firms pay less dividends. According to our test using all firms listed on the NYSE, AMEX, and NASDAQ stock exchanges from the CRSP, the results support our hypothesis. This paper also emphasizes the importance of the dividend reputation in the analysis of dividend-related matters. Firms with dividend reputation pay higher dividends than firms with no reputation.

Forecasting large covariance matrix with high-frequency data using factor approach for the correlation matrix

We apply the factor approach to the correlation matrix to forecast large covariance matrix of asset returns using high-frequency data, using the principal component method to model the underlying latent factors of the correlation matrix. The realized variances are separately forecasted using the Heterogeneous Autoregressive model. The forecasted variances and correlations are then combined to forecast large covariance matrix. Our proposed method is found to perform better in reporting smaller forecast errors than some selected competitors. Empirical application to a portfolio of 100 NYSE and NASDAQ stocks shows that our method provides lower out-of-sample realized variance in selecting global minimum variance portfolio.

Grey Wolf optimization-Elman neural network model for stock price prediction

Over the past two decades, assessing future price of stock market has been a very active area of research in financial world. Stock price always fluctuates due to many variables. Thus, an accurate prediction of stock price can be considered as a tough task. This study intends to design an efficient model for predicting future price of stock market using technical indicators derived from historical data and natural inspired algorithm. The model adopts Elman neural network (ENN) because of its ability to memorize the past information, which is suitable for solving stock problems. Trial and error-based method is widely used to determine the parameters of ENN. It is a time-consuming task. To address such an issue, this study employs Grey Wolf optimization (GWO) algorithm to optimize the parameters of ENN. Optimized ENN is utilized to predict the future price of stock data in 1 day advance. To evaluate the prediction efficiency, proposed model is tested on NYSE and NASDAQ stock data. The efficacy of the proposed model is compared with other benchmark models such as FPA-ELM, PSO-MLP, PSOElman, CSO-ARMA and GA-LSTM to prove its superiority. Results demonstrated that the GWO-ENN model provides accurate prediction for 1 day ahead prediction and outperforms the benchmark models taken for comparison.