Search

The aim of this paper is to explore the potential of the class of recurrent neural networks in developing a classification model for the stock market. For this research, the data that replicates the entire Nasdaq stock market limit order book is used. After extracting order book attributes from the raw dataset, feature selection approaches based on conditional entropy and Stochastic Universal Sampling are proposed, in order to highlight potentially informative features for our data classification task. Furthermore, the performances of nine presented recurrent neural network models for classifying the stock market data vector into one of the labels from the set S = {buy, sell, idle} are examined. In this study, the performances of the models based on different networks, namely the gated recurrent unit, the long short-term memory, and the recurrent neural network, are compared. Using a multi-criteria approach it is concluded that the model based on the GRU topology fed with the features selected using the newly proposed feature selection method outperforms other examined models. • The performances of nine different limit order book classification models are examined. • The PROMETHEE method specifies the most suitable model among the observed models. • The proposed feature demonstrates superiority compared to other selection methods. • The models based on different neural network topologies are evaluated. [ABSTRACT FROM AUTHOR]

In developing countries such as Iran, investigating the effect of exchange risks on different companies is a vital issue. Due to the lack of developed financial markets, these countries would be expected to be exposed to financial crises. In this study, the effect of exchange rate risks on Iranian banks was investigated using two market-oriented models. The sample included seven banks listed in Tehran Stock Exchange during 1382 to 1390. The results of estimating both models indicated that there is not a significant relationship between exchange rate fluctuations of Dollar in compare to Rial and stock price in these banks. One of the reasons for this result could be attributed to the lack of investors’ awareness and suitable analysis of the effects of these fluctuations on the companies’ return. Another reason could be the ability of the banks to control the exchange rate fluctuations using varied protecting methods which leads to the decrease in the effect of these fluctuations on companies’ values.

In the aftermath of stock market crash due to COVID-19, not all sectors recovered in the same way. Recently, a stock price model is proposed by Mahata et al. (2021) that describes V- and L-shaped recovery of the stocks and indices, but fails to simulate the U- and Swoosh-shaped recovery that arises due to sharp fall, continuation at the low price and followed by quick recovery, slow recovery for longer period, respectively. We propose a modified model by introducing a new parameter θ = + 1 , 0 , − 1 to quantify investors' positive, neutral and negative sentiments, respectively. The model explains movement of sectoral indices with positive financial anti-fragility (ϕ) showing U- and Swoosh-shaped recovery. Simulation using synthetic fund-flow with different shock lengths, ϕ , negative sentiment period and portion of fund-flow during recovery period show U- and Swoosh-shaped recovery. It shows that recovery of indices with positive ϕ becomes very weak with extended shock and negative sentiment period. Stocks with higher ϕ and fund-flow show quick recovery. Simulation of Nifty Bank, Nifty Financial and Nifty Realty show U-shaped recovery and Nifty IT shows Swoosh-shaped recovery. Simulation results are consistent with stock price movement. The estimated time-scale of shock and recovery of these indices are also consistent with the time duration of change of negative sentiment from the onset of COVID-19. We conclude that investors need to evaluate sentiment along with ϕ before investing in stock markets because negative sentiment can dampen the recovery even in financially anti-fragile stocks. • A sentiment based stock price model is presented for U- and Swoosh-shaped recovery during the COVID-19. • Sentiment of the investors affect the recovery of the stocks and indices. • HHT is applied to identify the time-scale of the shock and recovery during the COVID-19. • The study may help investors plan their investment during different crisis. [ABSTRACT FROM AUTHOR]

We briefly review the statistical properties of the escape times, or hitting times, for stock price returns by using different models which describe the stock market evolution. We compare the probability function (PF) of these escape times with that obtained from real market data. Afterwards we analyze in detail the effect both of noise and different initial conditions on the escape time in a market model with stochastic volatility and a cubic nonlinearity. For this model, we compare the PF of the stock price returns, the PF of the volatility and the return correlation with the same statistical characteristics obtained from real market data. [ABSTRACT FROM AUTHOR]

Abstract: We analyze the hitting time distributions of stock price returns in different time windows, characterized by different levels of noise present in the market. The study has been performed on two sets of data from US markets. The first one is composed by daily price of 1071 stocks trade for the 12-year period 1987–1998, the second one is composed by high frequency data for 100 stocks for the 4-year period 1995–1998. We compare the probability distribution obtained by our empirical analysis with those obtained from different models for stock market evolution. Specifically by focusing on the statistical properties of the hitting times to reach a barrier or a given threshold, we compare the probability density function (PDF) of three models, namely the geometric Brownian motion, the GARCH model and the Heston model with that obtained from real market data. We will present also some results of a generalized Heston model. [Copyright &y& Elsevier]

Abstract: It has been assumed that arbitrage profits are not possible in efficient markets, because future prices are not predictable. Here, we show that predictability alone is not a sufficient measure of market efficiency because of the influence an order has on its dynamics. We instead propose to measure inefficiencies of markets in terms of the maximal profit an ideal trader who can perfectly predict the future behavior of the market can take out from a market. In a stock market model with an evolutionary selection of agents this method reveals that, the mean relative amount of realizable profits P is very limited and we find that it decays with the rising number of agents. Our results show that markets may self-organize their collective dynamics such that it becomes very sensitive to profit attacks, which demonstrates that a high degree of market efficiency can coexist with predictability. [Copyright &y& Elsevier]

We present a novel microscopic stock market model consisting of a large number of random agents modeling traders in a market. Each agent is characterized by a set of parameters that serve to make iterated predictions of two successive returns. The future price is determined according to the offer and the demand of all agents. The system evolves by redistributing the capital among the agents in each trading cycle. Without noise the dynamics of this system is nearly regular and thereby fails to reproduce the stochastic return fluctuations observed in real markets. However, when in each cycle a small amount of noise is introduced we find the typical features of real financial time series like fat-tails of the return distribution and large temporal correlations in the volatility without significant correlations in the price returns. Introducing the noise by an evolutionary process leads to different scalings of the return distributions that depend on the definition of fitness. Because our realistic model has only very few parameters, and the results appear to be robust with respect to the noise level and the number of agents we expect that our framework may serve as new paradigm for modeling self-generated return fluctuations in markets. [Copyright &y& Elsevier]