SciELO - Scientific Electronic Library Online

 
vol.17 número1La relación entre los precios de las acciones y los componentes del modelo DUPONT: evidencia del mercado de valores mexicanoAnálisis de Eficiencia de Empresas Sustentantes de la Bolsa Mexicana de Valores índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

  • No hay artículos similaresSimilares en SciELO

Compartir


Revista mexicana de economía y finanzas

versión On-line ISSN 2448-6795versión impresa ISSN 1665-5346

Rev. mex. econ. finanz vol.17 no.1 Ciudad de México ene./mar. 2022  Epub 28-Oct-2022

https://doi.org/10.21919/remef.v17i1.570 

Artículos de investigación y revisión

Neuro-wavelet Model for price prediction in high-frequency data in the Mexican Stock market

Modelo de neuro-onda para predicción de precios en datos de alta frecuencia en el Mercado Bursátil Mexicano

Ricardo Massa Roldán1  * 
http://orcid.org/0000-0002-8499-5775

Montserrat Reyna Miranda2 
http://orcid.org/0000-0002-2793-6365

Vicente Gómez Salcido3

1Centro de Investigación y Docencia Económicas, México

2Universidad Anáhuac México, México


Abstract

With the availability of high frequency data and new techniques for the management of noise in signals, we revisit the question, can we predict financial asset prices? The present work proposes an algorithm for next-step log-return prediction. Data in frequencies from 1 to 15 minutes, for 25 high capitalization assets in the Mexican market were used. The model applied consists on a wavelet followed by a Long Short-Term Memory neural network (LSTM). Application of either wavelets or neural networks in finance are common, the novelty comes from the application of the particular architecture proposed. The results show that, on average, the proposed LSTM neuro-wavelet model outperforms both an ARIMA model and a benchmark dense neural network model. We conclude that, although further research (in other stock markets, at higher frequencies, etc.) is in order, given the ever increasing technical capacity of market participants, the inclusion of the LSTM neuro-wavelet model is a valuable addition to the market participant toolkit, and might pose an advantage to traditional predictive tools.

JEL Classification: C45; C53; C88; G14; G17

Keywords: market efficiency; high frequency data; LSTM neural network; wavelet

Resumen

Con la disponibilidad de datos de alta frecuencia y nuevas técnicas para la filtración de señales, es pertinente preguntarse una vez más ¿podemos predecir los precios de los activos financieros? El presente trabajo propone un algoritmo para la predicción de retorno logarítmico del siguiente periodo. Se usan datos en frecuencias de 1 a 15 minutos, para 25 activos de alta capitalización en el mercado accionario mexicano. El modelo consiste en la aplicación de una wavelet seguida de una red neuronal de tipo Long Short-Term Memory (LSTM). En la literatura comúnmente se encuentra el uso de wavelets o de redes neuronales en aplicaciones financieras, la novedad de nuestro trabajo radica en la arquitectura particular que proponemos. Los resultados muestran que, en promedio, el modelo de neuro-wavelet propuesto supera tanto a un modelo ARIMA como a un modelo de red neuronal densa de referencia. Podemos concluir que, aunque más investigación es necesaria, dada la creciente capacidad técnica actual de los participantes del mercado, la inclusión del modelo LSTM neuro - wavelet al abanico de herramientas disponibles es de mucho valor, pues podría representar una ventaja sobre las herramientas predictivas tradicionales.

Clasificación JEL: C45; C53; C88; G14; G17

Palabras clave: eficiencia de mercados; datos de alta frecuencia; redes neuronales LSTM; ondeletas

1. Introduction

Price predictability of a particular security is at the core of the activities of thousands of workers in the financial industry. From an academic standpoint, whether asset prices can be predicted using historical information relates to the evolution of the Efficient Market Hypothesis (EMH). In his seminal work, Fama (1970) describes an efficient market as one in which prices incorporate new information quickly and rationally. The EMH implies that if a market were to be efficient, prices will reflect all past and present information and any newly acquired information would be incorporated into the price instantaneously. If this hypothesis holds, then prices would follow a random walk and it would not be possible to generate and earn excess profits systematically (Malkiel, 2003). Simply put, as all the information is already incorporated into the price, the EMH implies that due to the stochastic behavior of prices, they cannot be predicted using past or present data.

Building over this hypothesis, the process of price forecasting typically assumes, not without an observational support, that markets are not efficient, therefore a price structure to be predicted can be expected. As stated by Hayek (1945), a price system plays a crucial role in incorporating information, the more relevance of such information, the greater the benefit that one agent can obtain in comparison to others. Under this perspective, it is common to find studies exploring the predictive possibility using daily, weekly or monthly returns. However, with the increasing availability of tools that allow the management and analysis of higher frequency data, new contributions can be made to the EMH’s framework.

Violations to the EMH are known as anomalies, a term introduced by Ball (1978), and have been thoroughly discussed, both with high and low frequency data. Hendershott and Riordan (2011) argue that financial instrument prices have non-stationary and non-linear behavior due to the variety of agents operating in a given market, each with a different investment horizon and objective and covering the range from long term investors to intraday speculators to algorithms operating from a server in a datacenter. All agents affect the market not only individually but from their collective actions. The result of this interaction is encoded in a single price signal. Cohen et al. (2002) state that large investors contribute to the efficiency of markets by, first, accelerating the incorporation of information in prices and, second, by offsetting the irrational behavior of individual investors.

However, there are studies that argue that a diversity in participants in the market, with different positions, strategies, points of view and behaviors, creates a signal with noise, resulting in a price signal with very complex behavior (Black, 1986; Jefferies et al., 2001; Giardina and Bouchard, 2003). One argument for this complex behavior is that the EMH does not account for network latency. Latency is the delay between a signal and a response, which is crucial for price formation, and when market participants have different latencies they will receive the same information at a slight time offset, and even if they all shared the same strategy, they would commit transactions asynchronously (Kirilenko and Lamacie, 2015).

It becomes clear that there is a need to develop modeling tools capable of incorporating non-linear elements that, by construction, can capture the complexity of the price generating process.

Some of the most common stock price forecasting methods include ARIMA, GARCH, Ordinary Least Squares, and Maximum Likelihood, all of which have been widely studied (Ariyo et al, 2014). However, building on the price forecasting literature that deals with complex signals, large amounts of high frequency data, and non-linear relationships in data, in the present work we propose the use of tools from the fields of Digital Signal Processing and Machine Learning as an attempt to outperform the traditional models when dealing with non-linear behavior of asset prices.

The proposed methodology consists of 2 steps. The first one involves the decomposition of the log-returns signal into simpler components based on methods in the field of Digital Signal Processing (DSP). These methods are commonly used for audio, image, and video processing, as well as data compression in order to decompose a complex signal into more manageable and simpler signals. In the context of financial price time series, the Wavelet transform has been used to split the single price signal with all its inherent complexity into several time series, each representing the behavior of the original signal across different frequency bands (Reboredo and Rivera-Castro, 2013; Chang and Fang, 2008; Caetano and Yoneyama, 2007). The expected result of this step is that by using wavelets to break up one time series into several time series (divided by frequency) and feeding them as inputs to the neural network model, the neural network will more easily learn how to use the short- and long-term information within the original signal. Due to the properties of the wavelet transform, all information in the original signal is retained in its wavelet representation.

After the wavelet transform, the second step consists of stacking a neural network. It is a Machine Learning technique that is able to capture the non-linear behavior contained in the prices. This Neurowavelet model has been used on traffic forecasting (Dunne and Ghosh, 2013), in streamflow forecasting (Mehr et al., 2013; Kisi, 2008), solar wind prediction (Nappoli et al., 2010). In financial time series analysis, the Neurowavelet model has been found to have superior performance to ARIMA and ARIMAX (Ortega and Khashanah, 2014). Building on the work of Ortega and Khashanah (2014), the novelty of this work lies in the use of a Maximal Overlap Discrete Wavelet Transform (MODWT) decomposition and a Long-Short Term Memory neural network and to predict the next period-return for frequencies of 1, 5 and 15 minutes.

The empirical strategy consists of using the information of the 25 most liquid assets in the Mexican Stock Exchange to perform their next-step prediction and compare the results with two alternative tools. The benchmark for this Wavelet-LSTM model is a Wavelet-Dense model, which uses a typical dense hidden layer instead of the LSTM hidden layer. Additionally, a conventional ARIMA (Box and Jenkins, 1970) model is also considered.

Our findings show that the Wavelet-LSTM outperforms, on average, both models. The remainder of the paper is structured as follows. The next section presents the literature review on wavelets, neural networks, and neuro-wavelets in finance. An explanation of the methodological aspects of the proposed model is found in the third section, followed by the description of data and the results of the empirical strategy. Concluding remarks are found in the last section.

2. Literature Review

2.1 Wavelets

Just as Ramsey (1999) stated, the contribution of wavelets to the finance field had great potential. By that time, several contributions had already been made, which can be reviewed in his work. Other noteworthy contributions are the work of Dremin and Leonidov (2008) where they filtered stock prices with a wavelet transform and found that the filtered series is characterized by volatility autocorrelation with large amplitude.

Caetano and Yoneyama (2007) use a wavelet decomposition to explain abrupt changes in stock prices while Aktan et al. (2009) propose the use of wavelets for the estimation of systematic risk in the Istanbul stock exchange, and Bruzda (2019) proposed their use to measure macroeconomic risk.

Building on their findings, we use a particular type of wavelet, the MODWT. This wavelet has been used to study asset prices because, as described by Zhu et al. (2014), it is useful to transform non-stationary and long-range dependent data into stationary and short-range dependent. On that end, Ardila and Sornette (2016) use the MODWT to analyze financial cycles. Ismail et al. (2016) compare a MODWT-EGARCH and a MODWT-GARCH model for prediction in African stocks, while Gupta et al. (2018) use a MODWT-VAR model to review the relationship between returns and volume in stock markets in India and China. Applications of MODWT for the Mexican or Latin-American case are, at most, scarce.

2.2 Neural networks

Neural networks are a Machine Learning tool used primarily for regression or classification tasks to approximate a function. They have also been used extensively in the study of financial data. However, some of the most notable works related to their performance in price forecasting, to our knowledge, are the ones by Maciel and Ballini (2008), Khashei and Bijari (2010), Adnan et al. (2011), Maknickiene and Maknickas (2012), and Ariyo et al. (2014), who found that a simple neural network outperforms SARIMA and GARCH models as well as other simple forecasting methods.

The type of neural network that seems to be a better fit for financial time series, particularly when modeling prices, is the recurrent neural network (RNN) due to its theoretical ability to retain information across time steps, Oancea and Ciucu (2014).

A problem that arises in practice is the difficulty for the RNN to reliably remember information from past time steps due to a phenomenon known as the Vanishing Gradient (Pascanu et al, 2013). Long-Short-Term-Memory (LSTM) is a type of RNN with gates that prevent the gradient from vanishing or exploding. LSTM introduces the concept of self-loops with context-conditioned weights. With a gated self-loop the time scale can be changed dynamically based on the input sequence. LSTM networks have been very successful in handwriting recognition, speech recognition, machine translation, and image captioning.

To that end, Fisher and Krauss (2018) state that LSTM is a suitable model for application in finance, although not much has been done. They find the LSTM outperforms a random forest, a deep neural net, and a logistic regression classifier when predicting the direction of change in the S&P500 index.

In a similar vein, Siami-Namini and Namini (2018) compare a LSTM neural network to ARIMA for the forecasting of financial data and find that former outperforms the latter. Choi and Lee (2018) use a series of stacked LSTM networks to predict financial time series and conclude that the model can capture non-linear behavior, while Hanson (2017) uses a LSTM neural network to forecast stock indices and shows that the outputs of the LSTM networks are similar to those obtained by ARMA and GARCH, but the neural network outperforms when predicting the direction of change.

2.3 Neurowavelets

Aussem and Murtag (1997) made one of the first works applying a wavelet decomposition and a neural network, they used a dynamic recurrent neural network trained on each resolution scale with the temporal-recurrent backpropagation algorithm and tested it on sunspots time series data.

The present work does not follow the neuro-wavelet concept introduced by Murtagh et al (2004), who define a wavelet network as a neural network in which activation functions have been replaced by wavelet functions. Instead, what we take a stance by using the concept of a neurowavelet as a layered model using first a wavelet transform, and then a neural network.

Such a model has been used in the prediction of financial time series by Minu et al. (2010), who find that a neurowavelet model outperforms GARCH and simple feedforward neural network models (FFNN models), while Zhang et al. (2001) used neurowavelet and FFNN models to forecast futures contracts and passed the forecast to a money management system to generate trades. They found that the neurowavelet model doubled the profit per trade.

More recently, Jamazi and Aloui (2012) use a wavelet and a multilayer backpropagation neural network to predict oil prices, they find the model performs better than the backpropagation neural network. Bao et al. (2017) use a wavelet transform followed by stack autoencoders and then a LSTM neural network for the forecast of the next day closing price for six market indices and their futures. They find their model outperforms a recurring neural network.

On the realm of high frequency data, Ortega and Khashanah (2014) apply a neurowavelet approach to 1-minute frequency data of Apple AAPL stock prices. They apply a non-decimated Haar wavelet decomposition followed by the Jordan (1997) and Elman (1990) networks, due to their ability to capture temporal patterns. They compare one, three and five-step ahead forecasting to ARIMA and ARIMAX and find that the neurowavelet model outperforms the other models. Arevalo et al. (2018) propose a discrete wavelet transform (DWT) and a deep neural network (DNN) to forecast one-minute and 3-minute log-returns of 19 stocks in the Dow Jones Index, and they find that the model has an accuracy rate between 64% and 76%. Our work is directly inspired by both studies and builds on them by proposing a neurowavelet model that uses Long Short-Term Memory (LSTM) units, which have internal logical gates that provide the model with an increased ability to remember data from previous time steps and to model even more complex non-linear behavior.

To the best of our knowledge, there is no existing literature that combines the Wavelet transform with an LSTM neural network in the context of high frequency price data.

3. Methodology

3.1 Wavelets and Neurowavelets

The proposed methodology consists of 2 steps. The first one involves the decomposition of the log-returns signal into simpler components using a MODWT transform, while the second one consists in stacking a neural network, commonly known as Neurowavelet.

Wavelets are an extension of Fourier analysis used to analyze nonstationary signals. According to Otazu (2008), the most important difference is that the Wavelet transform is defined in both the spatial frequency and spatial location, while the Fourier transform is only defined in spatial frequency. The Maximal Overlap Discrete Wavelet Transform (MODWT) is appropriate in the price forecasting process as it is time invariant (Percival and Walden, 2000).

Given the discrete wavelet transform with wavelet filter {hj,i} and scaling filter {gj,i} , the MODWT filters are defined as h-j,i=hj,i/2j/2 and g-j,i=gj,i/2j/2 . Then, if we have a time series {Xt} , the wavelet coefficients at level j are described by equations 1 and 2:

W-j,i=i=0Ij-1h-j,iXt-imodN (1)

V-j,i=i=0Ij-1g-j,iXt-imodN (2)

On the other hand, neural networks are Machine Learning tools used primarily for regression or classification tasks to approximate a function f* . An important property of neural networks is their ability to model nonlinear behavior. Neurons -building blocks for function approximation machines that achieve statistical generalization- can be linear or nonlinear, and a combination of nonlinear neurons will result in a nonlinear network. Thus, the nonlinear behavior is not held at a neuron in particular, but rather spread across the network.

The basic elements of a neuron are: i) the connections between neurons -each carrying a weight that can be positive or negative-, ii) an adder -to sum the weighted input signals-; and iii) an activation function -to limit the amplitude of the output of a neuron in a linear or nonlinear manner-. The neuron can then be defined as shown in equations 3 and 4:

uk=j=1mxjwkj (3)

yk=γbk+uk (4)

Where, x1,, xm are the input signals; wk1,, wkm are the weights of neuron k ; uk is the combined linear output of the weighted input signals; bk is the bias weight, γ(.) is the activation function. yk is the output signal of the neuron. The bias weight bk applies an affine transformation to the output uk .

Unsupervised learning algorithms attempt to learn a structure within a dataset without explicitly linking an input state with an output state. The learning algorithm used here falls strictly under supervised learning. It is necessary to have a training set with a number of inputs that correspond to known outputs -target response-. The network is presented with a training example and the network weights are adjusted in order to minimize the error or the distance between the predicted response and the expected response.

In a neural network, neurons can be stacked layer upon layer, where every layer feeds its output as the input to the next layer. The output is directed, so there are no feedback connections, thus there is no recurrence. A feedforward neural network (FFNN) is defined as described by equation 5.

y=fx;θ (5)

Where θ represents the parameters that best approximate f* . Each sample x has a corresponding label y f*(x) , which specifies what the output layer must be at every point x . The learning algorithm iterates to modify the parameters θ , the weights of every neuron, to reach the desired output. The training data provides no indication to the learning algorithm about what it should do with the parameters or layers in the model, so it must decide how to best use these layers to approximate f* .

The layers between the input layer and the output layer are known as hidden layers. By adding more hidden layers, the network can extract higher order statistics. Hidden layers are usually vectors, and their dimensionality determines the width of the model. Each element of the vector represents a neuron. This representation allows neurons to work as parallel units.

Neural networks are referred to in the form m-h1-h2-q where m is the number of source nodes, h1 and h2 the number of neurons in the hidden layers, and q the number of neurons in the output layer. Thus a 10-5-1 network has 10 inputs, 1 output, and 5 neurons in a single hidden layer.

Recurrent neural networks (RNN) are best suited to process sequential data x(1),,x(τ) (Roman and Jameel, 1996) which is the case for financial market price data. A FFNN is modified with a connection that provides feedback, so it becomes a recurrent neural network.

Recurrent nueral networks can scale to longer sequences than other types of networks. In that sense, the RNN updates the equations 6, 7 and 8 for every timestep between t=1 and t=τ :

at=b+Wht-1+Uxt (6)

ht=tanhat (7)

o(t)=c+Vh(t) (8)

Where b and c are bias vectors, and matrices U (input to hidden), V (hidden to output) and W (hidden to hidden).

A problem that arises in RNN is that long term dependencies are not properly captured by the model because the gradients over timesteps either vanish or explode, resulting in instability during optimization (Pascanu et al., 2013). The exponentially smaller weights of long-term interactions eventually become so small that they lose effect. A widely used optimization algorithm to minimize the loss function during neural network training is stochastic gradient descent (Sirigiano and Spiliopoulos, 2017). This is the algorithm used in the empirical strategy of this work.

Because the computational cost of additive cost functions is O(m) , as the training set grows, the training time required increases significantly. To solve this problem, stochastic gradient descent approximately estimates the expectation of the gradient using a minibatch of samples. A minibatch Β=x(1),,x(m') is drawn uniformly from the training set, where m' has a fixed size. Fitting a training set with billions of samples using updates from hundreds of samples is now possible, yet it is still a rather time-consuming task. The estimate of the gradient is given by equation 9.

g=1m'θi=1m'Lxi,yi,θ (9)

And it follows the estimated gradient in equation 10:

θθ-ϵg (10)

Where ϵ is the learning rate.

However, in practice the learning rate is not fixed, it decreases gradually over time, so that the learning rate at iteration k is ϵk . For that, the conditions shown in equations 11 and 12 have to be met in order to guarantee convergence:

k=1ϵk= (11)

k=1ϵk2< (12)

As pointed out by Goodfellow et al., (2016), linear decay of the learning rate is often used in practice, which has the structure specified in equation 13:

ϵk=1-αϵ0+αϵτ (13)

Where τ is the iteration and α=kτ , and ϵ . Selecting the learning rate is a balance between learning speed and avoiding instability. If ϵ0 is too large then the learning curve oscillates and the cost function increases, if set too low learning becomes slow and may become stuck.

3.2 Long Short-Term Memory (LSTM) neurowavelets

Finally, the LSTM is a type of RNN with gates that prevent the gradient from vanishing or exploding (Hochreiter and Schmidhuber, 1997). LSTM introduces the concept of self-loops with context-conditioned weights.

With a gated self-loop, the time scale can be changed dynamically based on the input sequence. Each unit is composed of LSTM cells, and each cell has the same inputs and outputs as a common recurrent network, with more parameters and a system of gates for flow control. Let fi(t) represent a gate unit, for timestep t and cell i , then it can be defined as shown in equation 14:

fi(t)=σbif+jUi,jfxj(t)+jWi,jfhj(t-1) (14)

Where bf is the bias, Uf the input weights, and Wf the recurrent weights. The external input gate unit can then be expressed as seen in equation 15:

gi(t)=σbig+jUi,jgxj(t)+jWi,jghj(t-1) (15)

And the output gate qi(t) is described by equation 16 and 17:

qi(t)=σbi0+jUi,j0xj(t)+jWi,j0hj(t-1) (16)

hi(t)=tanhsi(t) qi(t) (17)

Where the internal state si(t) is updated as shown in equation 18, leading to the internal logical gates in LSTM solving the vanishing gradient problem.

si(t)=fi(t)si(t-1)+gi(t) σbi+jUi,jxj(t)+jWi,jhj(t-1) (18)

3.3 Activation Function and Architecture

To get the output from one layer to the next in the neural network, an activation function is used. There is a wide variety of them and Gomes et al (2011) present a comprehensive review of activation functions in finance.

For the present work, the selected activation function is the sigmoid, which is described in equation 19, and which has an S shaped graph. We use this activation function because it is non-linear, therefore it can capture non-linear behavior, and it is differentiable at every point.

γv=1e-t+1 (19)

The structure of the network determines how many neurons it has, the connections between them, and how the layers are stacked. The stacking of layers is what gives depth to the model.

The first layer can then be defined as shown in equation 20,

h(1)=g(1)W(1)Tx+b(1) (20)

The second layer is shown in equation 21,

h(2)=g(2)W(2)Th(1)+b(2) (21)

And layer n is shown in equation 22,

h(n)=g(n)W(n)Th(n-1)+b(n) (22)

The universal approximation theorem (Hornik et al, 1989; Cybenko 1989) states that a feedforward network with a linear output layer and one or more hidden layers with any activation function can approximate any Borel measurable function from one finite-dimensional space to another with any desired non-zero error if the network has enough hidden units.

The derivatives of the feedforward network can also approximate the derivatives of the function. Any continuous function on a closed and bounded subset of Rn is Borel measurable and therefore may be approximated by a neural network. While the universal approximation theorem states that a large enough feedforward network can represent the function, it doesn't guarantee that the training algorithm is able to learn that function. The theorem does not say how large the network must be. Therefore, our strategy consists in comparing two types of architecture, a dense architecture - where every node in a layer is connected to every other node in the forward layer-, and a LTSM architecture.

3.4 Optimization and Backpropagation

Optimization in neural networks consists of finding the parameters θ for a neural network to reduce the cost function J(θ) according to a performance measure P . This way the performance measure P is optimized indirectly. The cost function is defined by equation 23:

Jθ=Ex,y~p^dataLf,x;θ,y (23)

Where L is the per example loss function, fx;θ is the predicted output for input x , p^data is the empirical distribution, and y is the target output.

To minimize the objective function, the expectation is taken from the data generating distribution pdata and not only over the finite training set, as shown in equation 24:

Jθ=Ex,y~pdataLf,x;θ,y (24)

In order to transform a machine learning problem into an optimization problem we minimize the expected loss on the training set by replacing the true distribution p(x,y) with the empirical distribution p^(x,y) . To minimize the empirical risk, this is shown in equation 25:

Ex,y~p^dataLf,x;θ,y=1mi=1mLf,x(i);θ,y(i) (25)

A word of caution must be made regarding this approach as the empirical risk minimization can lead to overfitting - a situation where the model memorizes the training set and cannot be generalized properly for samples outside the training set.

Finally, backpropagation is the method used to allow a neural network to learn by itself. If a neuron j is fed by a layer of neurons before it, the induced local field produced at the input of the activation function associated with neuron j is as described by equation 26:

vjn=i=0mwjinyin (26)

Where m is the total number of inputs applied to neuron j . The function signal yi(n) at the output of a neuron j at iteration n is given by equation 27:

yin=γvin (27)

The backpropagation algorithm applies a correction Δwji(n) to the weight wjin which is proportional to the partial derivative ϵ(n)/wji(n) , which represents a sensitivity factor, determining the direction of search in weight space for the weight wji . This is shown in equation 28

Δwjin=-ηϵnwjin (28)

Where η is the learning rate parameter for the backprogapation algorithm.

4. Data and Results

Figure 1 depicts the proposed methodology. The target output of our forecast is the log-return for one timestep in the future. Therefore, the two steps involved are:

  1. Apply a MODWT to decompose the signal of log-returns.

  2. Fit a LSTM neural network to the filtered data. The optimization target is the Mean squared error (MSE).

While this seems enough to build and train a Wavelet-LSTM neural network, in practice there are many hyperparameters inherent to the network that must be specified and their values can be the difference between the model converging or not converging, the speed of convergence, can result in overfitting, reaching different local minima, and computational cost.

The benchmark against which we compare the wavelet-LSTM model is an ARIMA model fitted to the data. We also compare the results to a wavelet-dense model in order to have a sense of the performance of the neural network itself.

We measure the ability of the model to reduce its mean squared error during training and to correctly forecast the direction -if the predicted return will be positive or negative-.

Figure 1 Wavelet-LSTM neural network. Source: Own elaboration. 

4.1 Data

We use price data from 25 assets of the Mexican Stock market with time intervals for 1, 5, and 15 minutes. The selection of assets is based on the largest market capitalization value of the Mexican Stock Exchange for the period of analysis. The selection of intervals is based on the availability of data from the data provider (Bloomberg), and on previous studies: Bakhach et al (2018) and Ortega and Khashanah (2014) make use of a 1 minute interval, Carrion (2013) uses a 5 minutes interval, Martens (2002) uses 5 and 15 minutes intervals. For internal validation purposes, the period of analysis goes from February 2017 to July 2017, considering the Open, Maximum, Minimum and Close prices, as well as Volume. Data were obtained from Bloomberg. The list of assets can be found in Table 1.

Table 1 List of assets from the Mexican Stock Exchange 

AMXL MM Equity ELEKTRA* MMM Equity
WALMEX* MM Equity PE&ONES* MM Equity
FEMSAUBD MM Equity GCARSOA1 MM Equity
GFMEXICOB MM Equity IENOVA* MM Equity
GFNORTEO MM Equity ALFAA MM Equity
KOFL MM Equity CUERVO* MM Equity
TLEVICPO MM Equity KIMBERA MM Equity
SANMEXB MM Equity GAPB MM Equity
CEMEXCPO MM Equity ASURB MM Equity
AC*MM Equity GRUMAB MM Equity
BIMBOA MM Equity MEXCHEM* MM Equity
GFINBURO MM Equity PINFRA* MM Equity

Source: Own elaboration with tickers from Bloomberg (2017).

Both the Wavelet-LSTM and Wavelet-Dense models were initialized with a diverse combination of hyperparameters to determine convergence, stability, and performance. Then a training and calibration of 1926 experiments, each with a different combination of hyperparameters, was performed.

This requires considerable computing power, and for this we have assembled a small 8-node cluster with 12 Graphic Processing Unit, GPU, accelerator cards. To program and train the neural networks we use the Keras Deep Learning Library (version 1.0.4, Jun 06 2016) -an open source project started by Google researcher: Francois Chollet-.

The closing prices, C, are transformed to log returns and used as inputs for the benchmark ARIMA model. These log returns are also the output values and training target in the Wavelet-LSTM and Wavelet-Dense models.

The time series for the 25 assets are separated in eight frequency levels using a MODWT transform with J=8. This decomposition generates 9 time series for each one of the 25 stocks analyzed, 8 detail series and 1 wavelet approximation. These time series are then divided into a moving window of size 12 and fed as input to a layer of 512 LSTM units, followed by a Dense layer fully connected with 512 neurons. It is finally followed by an additional Dense layer with 512 inputs and 1 output with a linear activation function, which corresponds to the forecast of the log returns. The Wavelet-LSTM and Wavelet-Dense models are illustrated in Figure 2 and Figure 3, respectively.

Figure 2 Wavelet-FSTM model. Source: own elaboration. 

Figure 3 Wavelet-Dense model. Source: own elaboration. 

The time series are split 90% for training the network (training set) and 10% for out of sample validation (validation set). We trained the model with 500 epoch iterations (an epoch is one complete pass over the training data set), checking for congruence between the training error and validation error to avoid overfitting and detecting instability.

The optimization target used is the Mean Squared Error (MSE) and the scale is adjusted in the interval [-1,1] in order to match the original center to 0 so the sign of the return can be preserved while pairing the interval to the one of the activation functions.

Finally, to compare the performance of the Wavelet-LSTM model with the selected benchmarks, we count the number of times the sign of the forecast corresponds with the actual sign of the return, that is, the times the direction of the forecast is correct.

4.2 Results

When building a forecasting model, parsimony is desirable. A reduced number of estimated parameters is easier to understand and explain, and every estimated parameter adds estimation variation. In the case of ARIMA each unnecessary parameter causes the variance of the one step ahead forecast error to increase approximately, where n is the number of observations (Ledolter and Abraham, 1981). This is also a matter of noise versus signal. Models that are overly complex can mistake noise for a signal. Occam's razor is a metaphor for parsimony: removing needless complexity and leaving only theories, models, and hypotheses that are as simple as possible without being false.

By finding a balance between goodness of fit and model complexity we can avoid overfitting. In the case of Machine Learning and more specifically the Wavelet-LSTM model proposed, we use a training set and an out-of-training-sample validation set, and we evaluate the loss function for each. If the training loss becomes measurably lower than the validation loss, we have a condition of overfitting or no convergence. One technique to prevent overfitting is Early Stopping, which halts the iterative training algorithm once the training loss compared to the validation loss shows no improvement. Table 2 contains the percentages of the correct direction predicted by ARIMA, Wavelet-LSTM y Wavelet-Dense models for each of the 25 stocks and the time scales selected. Table 3 contains the percentage difference in the performance of Wavelet-LSTM model against the selected benchmarks.

It can be seen that, on average, Wavelet-LSTM improves the forecasting of the one-step ahead direction when compared to the Wavelet-Dense and the ARIMA models.

For the case of the first benchmark, the biggest difference between the Wavelet-LSTM precision and the Wavelet-Dense appears in the 5 minutes window.

For the case of the second benchmark, as the time window becomes larger, the Wavelet-LSTM forecasting improvement increases in magnitude compared to the forecasting of the ARIMA model. Altogether, the results are in line with the ones exposed in the literature review.

Table 2 Percentage of correct prediction of the one-step ahead direction 

Wavelet-LSTM Wavelet-Dense ARIMA
Stock 1m 5m 15m 1m 5m 15m 1m 5m 15m
AMXL MM Equity 25.74% 37.34% 46.80% 25.90% 36.40% 49.97% 25.51% 36.20% 42.55%
WALMEX* MM Equity 36.59% 45.44% 51.89% 36.30% 42.68% 50.88% 35.16% 42.84% 45.77%
FEMSAUBD MM Equity 41.39% 48.69% 54.46% 41.21% 47.91% 50.93% 42.23% 47.46% 49.66%
GFMEXICOB MM Equity 36.62% 46.16% 52.88% 36.18% 44.36% 48.17% 34.84% 43.39% 46.78%
GFNORTEO MM Equity 34.99% 42.87% 53.25% 35.14% 47.08% 51.29% 34.90% 44.45% 47.97%
KOFL MM Equity 40.41% 49.45% 48.15% 40.43% 48.40% 55.49% 40.05% 46.64% 49.33%
TLEVICPO MM Equity 42.46% 54.74% 57.34% 41.31% 48.35% 52.70% 40.97% 46.31% 47.90%
SANMEXB MM Equity 34.15% 43.92% 50.11% 33.78% 44.60% 51.62% 33.81% 43.59% 46.75%
CEMEXCPO MM Equity 28.70% 42.09% 51.24% 28.76% 39.84% 49.51% 28.37% 39.34% 44.49%
AC*MM Equity 38.32% 49.47% 53.94% 38.04% 47.76% 51.75% 37.63% 46.23% 48.33%
BIMBOA MM Equity 32.15% 48.17% 52.95% 31.37% 41.97% 51.34% 31.09% 40.91% 46.43%
GFINBURO MM Equity 33.33% 58.21% 68.86% 30.23% 42.61% 49.31% 29.66% 39.96% 45.64%
ELEKTRA* MMM Equity 42.04% 58.25% 59.85% 39.69% 49.35% 55.99% 39.96% 47.70% 49.38%
PE&ONES* MM Equity 40.79% 52.54% 59.10% 40.04% 47.54% 56.90% 39.28% 46.89% 48.89%
GCARSOA1 MM Equity 35.79% 46.27% 53.19% 34.10% 43.64% 54.87% 34.20% 41.92% 46.63%
IENOVA* MM Equity 34.61% 50.91% 57.03% 34.20% 45.24% 52.93% 32.13% 40.12% 44.43%
ALFAA MM Equity 32.01% 47.82% 53.56% 30.60% 42.35% 50.49% 30.41% 40.05% 44.88%
CUERVO* MM Equity 25.65% 37.59% 44.16% 25.25% 34.63% 45.09% 25.00% 35.41% 41.98%
KIMBERA MM Equity 34.85% 54.41% 62.92% 31.29% 42.17% 45.00% 30.93% 40.10% 45.02%
GAPB MM Equity 39.23% 49.55% 55.15% 39.21% 47.20% 56.40% 38.24% 45.04% 48.28%
FUNO11 MM Equity 28.50% 41.29% 50.26% 28.34% 39.01% 49.87% 27.88% 37.52% 43.67%
ASURB MM Equity 38.78% 49.82% 57.47% 39.31% 48.39% 49.07% 38.79% 46.59% 47.79%
GRUMAB MM Equity 35.29% 46.52% 51.59% 33.44% 45.48% 51.03% 34.98% 44.75% 48.33%
MEXCHEM* MM Equity 33.60% 52.13% 57.27% 32.64% 43.57% 57.34% 32.88% 42.19% 47.41%
PINFRA* MM Equity 35.75% 48.34% 53.14% 35.80% 45.76% 50.87% 34.95% 44.53% 48.15%

Source: Own elaboration with the results of the study.

Table 2 contains the percentages of the correct direction predicted by ARIMA, Wavelet-LSTM y Wavelet-Dense models for each of the 25 stocks and the time scales selected. It can be seen that, in general, the Wavelet-LSTM yields higher percentages of correct prediction than those from the Wavelet-Dense and the ARIMA models. Of the predictions considering a 1-minute window, the Wavelet-LSTM has the highest percentage in 18 out of the 25 forecasts. The average of correct prediction for this window is 35.27% for the Wavelet-LSTM, while for the Wavelet-Dense and the ARIMA models the average are 34.5% and 34.15% respectively. For the 15 minutes window, again, the Wavelet-LSTM has the highest percentage in 18 out of the 25 forecasts. Interestingly, the average of correct prediction increases for the three models: 54.26% for the Wavelet-LSTM, 51.55% for the Wavelet-Dense and 46.66% for the ARIMA model. Finally, for the 5 minutes window the Wavelet-LSTM has the highest percentage in 23 out of the 25 forecasts and the average of correct prediction is 48.08% for the Wavelet-LSTM, 44.25% for the Wavelet-Dense and 42.81% for the ARIMA model. In sum, the Wavelet-LSTM outperforms both benchmark models in 59 of the 75 forecasts and appears to be more precise in the 5 minutes window than the Wavelet-Dense and the ARIMA models.

Table 3 Difference of correct prediction between Wavelet-LSTM model and benchmarks 

Wavelet-LSTM | ARIMA Wavelet-LSTM | Wavelet-Dense
Stock 1m 5m 15m 1m 5m 15m
AMXL MM Equity 0.22% 1.14% 4.25% -0.16% 0.94% -3.17%
WALMEX* MM Equity 1.43% 2.60% 6.12% 0.29% 2.76% 1.01%
FEMSAUBD MM Equity -0.84% 1.23% 4.80% 0.17% 0.78% 3.53%
GFMEXICOB MM Equity 1.78% 2.77% 6.10% 0.44% 1.79% 4.71%
GFNORTEO MM Equity 0.09% -1.58% 5.28% -0.15% -4.22% 1.96%
KOFL MM Equity 0.37% 2.80% -1.18% -0.01% 1.04% -7.34%
TLEVICPO MM Equity 1.49% 8.44% 9.45% 1.14% 6.39% 4.65%
SANMEXB MM Equity 0.34% 0.33% 3.36% 0.37% -0.67% -1.51%
CEMEXCPO MM Equity 0.33% 2.75% 6.75% -0.06% 2.25% 1.73%
AC*MM Equity 0.69% 3.23% 5.61% 0.28% 1.71% 2.19%
BIMBOA MM Equity 1.06% 7.26% 6.51% 0.78% 6.20% 1.60%
GFINBURO MM Equity 3.67% 18.25% 23.22% 3.10% 15.60% 19.55%
ELEKTRA* MMM Equity 2.08% 10.55% 10.47% 2.35% 8.91% 3.86%
PE&ONES* MM Equity 1.51% 5.65% 10.20% 0.75% 5.00% 2.19%
GCARSOA1 MM Equity 1.58% 4.35% 6.56% 1.69% 2.64% -1.68%
IENOVA* MM Equity 2.48% 10.79% 12.60% 0.41% 5.66% 4.10%
ALFAA MM Equity 1.60% 7.77% 8.68% 1.40% 5.47% 3.07%
CUERVO* MM Equity 0.65% 2.18% 2.18% 0.41% 2.97% -0.93%
KIMBERA MM Equity 3.93% 14.32% 17.90% 3.56% 12.25% 17.92%
GAPB MM Equity 1.00% 4.51% 6.87% 0.02% 2.35% -1.24%
FUNO11 MM Equity 0.62% 3.77% 6.59% 0.16% 2.27% 0.39%
ASURB MM Equity -0.02% 3.23% 9.69% -0.54% 1.44% 8.41%
GRUMAB MM Equity 0.31% 1.77% 3.26% 1.84% 1.04% 0.56%
MEXCHEM* MM Equity 0.73% 9.94% 9.86% 0.96% 8.56% -0.07%
PINFRA* MM Equity 0.81% 3.81% 4.99% -0.05% 2.57% 2.27%
Average 1.12% 5.27% 7.61% 0.77% 3.83% 2.71%

Source: Own elaboration with the results of the study.

Table 3 contains the percentage difference in the performance of Wavelet-LSTM against the selected benchmarks. For this analysis, if a positive sign is obtained, the Wavelet-LSTM yielded a higher percentage of correct prediction than the one from the Wavelet-Dense or the ARIMA models. For the case of the difference in percentage of correct prediction between the Wavelet-LSTM and the Wavelet-Dense, 60 out of the 75 forecasts showed a positive sign. This suggests that the Wavelet-LSTM outperformed the Wavelet-Dense in 80% of the presented forecasts. When analyzing the case of the difference in percentage of correct prediction between the Wavelet-LSTM and the ARIMA model, the positive sign is obtained in 71 of the 75 cases, for an outperformance of approximately 95% of the presented forecasts. Altogether, the results are in line with the ones exposed in the literature review: in general, the ARIMA model is outperformed by both Wavelet models, while the Wavelet-Dense is outperformed by the Wavelet-LSTM.

5. Concluding Remarks

The description of price dynamics at various frequencies is of importance both for academics and practitioners in finance. Many are particularly concerned with the possibility of forecasting prices or returns in a specific time horizon.

Previous studies have shown that at high frequencies, returns do not follow the efficient markets hypothesis, hence a “complete” stochastic behavior is not observed (Nath and Dalvi, 2004; Strawińsky and Slepaczuk, 2008; Tapia, 2020). By asset prices not following a random walk a forecasting of their next value is possible and, with the use of novel tools and computational power, one might take advantage of this fact. This directly translates to a potential benefit for the practitioners that have recognized the importance of dealing with high frequency information in their predictive models. Precisely, O’Hara (2014) discussed a new market paradigm: “Trading has become faster, and market structure has fundamentally changed. In today’s market, high-frequency traders (HFTs) act on information revealed by low-frequency traders (LFTs). To survive, LFTs must avoid being detected by predatory algorithms of HFTs. LFTs can thrive by adopting trading strategies appropriate to the high-frequency trading world”.

5.1 Main Results

In the present work, we apply a wavelet followed by LSTM neural network to 25 assets of the Mexican Stoc Market. We compare the performance of the proposed model with the performance of an ARIMA model and a wavelet followed by a Dense neural network model, following the literature.

The measure for comparison is the accuracy in forecasting the direction of the price movement (the sign of the logarithmic return) for time series at three different time windows (1, 5 and 15 minutes).

We find that The LSTM neuro-wavelet outperforms both the ARIMA and the Dense neuro-wavelet for the three studied frequencies. However, the higher accuracy was found for the 5 minutes interval.

5.2 Limitations and Recommendations for Future Research

Results and implementation do not come without limitations. We recognize that the selection of the benchmark models and high market capitalization assets may lead to skewed results. While we selected Wavelet-Dense and ARIMA models based on the consensus when dealing with financial time series, a comparison against other Neurowavelets architectures might be of interest.

On the other hand, we also find value in exploring the proposed methodology in various cases within the Mexican market, such as, considering different market capitalization levels or considering industry indices instead of single assets. We leave that for future research.

Finally, the technical capacity surely represents a constraint for the hyperparameters’ selection and estimation capability may be somewhat hindered by it. In that sense, future research agenda may also include the assessment of whether these results are maintained over long time windows or in markets with a different inherent structure, such as the US stock market.

5.3 Originality

The question of whether prices can be forecasted is not new, the literature on the topic is vast, but most of it focuses on low frequency operations. With the mass availability of computing power, large amounts of data and new filtering tools, we can revisit the question, and particularly, we can bound it to the realm of high frequency trading.

Wavelet filtering has been widely applied in finance. Neural network models have been extensively explored, as well. But recently, the application of both models combined has been proposed (Zhang et al., 2001; Minu et al., 2010; Jamazo and Aloui, 2012; Bao et al. 2017). For high frequency trading, little has been done. Ortega and Khashanah (2014) apply a Haar wavelet followed by a Jordan-Elman neural network to 1 minute data for the Apple (AAPL) stock price. Arevalo et al. (2018) apply a wavelet followed by a Deep Neural Network (DDM) to 1 and 3 minute data on 19 stocks in the Dow Jones Index.

We follow this line, by applying a wavelet transform followed by a Long Short Term Memory neural network, which provides increased ability to remember data from previous time steps, and thus allows to model more complex non-linear behavior.

5.4 Main Conclusions

In this work we illustrate the use of a new tool for forecasting stock prices at high frequencies. The results show that the LSTM neuro - wavelet model proposed outperforms the chosen benchmarks. This leads us to conclude that the addition of this tool to the forecasting toolkit of a market participant would be of great value. Certainly, the application of said tool is computationally intensive, but the possible gains from it, and the ever-greater availability of computing power might make it possible to consider the use of complex algorithms, such as this one.

We believe there is an opportunity to outperform traditional forecasting tools by making clever use of computer power, and such we have tried to show in the present research. This is an important step in showing how ever more sophisticated tools can be applied successfully to gain insight into the structure of the markets and the complexity that interactions of many actors induce.

References

Adnan, R., Hisyam Lee, M. & Nor, E. (2011) Neural Network Forecasting: Error Magnitude and Directional Change Error Evaluation. Recent Advances in Computational Mathematics. pp 92-96. ISBN: 978-1-61804-267-5 [ Links ]

Aktan, B., Mabrouk, A. B., Ozturk, M., & Rhaiem, N. (2009). Wavelet-based systematic risk estimation: an application on Istanbul stock exchange. International Research Journal of Finance and Economics,23(3), 33-45. ISSN 1450-2887 [ Links ]

Ardila, D., & Sornette, D. (2016). Dating the financial cycle with uncertainty estimates: a wavelet proposition.Finance Research Letters,19, 298-304. https://doi.org/10.1016/j.frl.2016.09.004Links ]

Arévalo, A., Nino, J., León, D., Hernandez, G., & Sandoval, J. (2018, June). Deep learning and wavelets for high-frequency price forecasting. InInternational Conference on Computational Science (pp. 385-399). Springer, Cham. https://doi.org/10.1007/978-3-319-93701-4_29Links ]

Ariyo, A. A., Adewumi, A. O., & Ayo, C. K. (2014, March). Stock price prediction using the ARIMA model. In2014 UKSim-AMSS 16th International Conference on Computer Modelling and Simulation(pp. 106-112). IEEE. DOI: 10.1109/UKSim.2014.67Links ]

Aussem, A., & Murtagh, F. (1997). Combining neural network forecasts on wavelet-transformed time series.Connection Science, 9(1), 113-122. https://doi.org/10.1080/095400997116766Links ]

Bakhach, A. M., Tsang, E. P., & Raju Chinthalapati, V. L. (2018). TSFDC: A trading strategy based on forecasting directional change. Intelligent Systems in Accounting, Finance and Management, 25(3), 105-123. https://doi.org/10.1002/isaf.1425Links ]

Ball, R. (1978). Anomalies in relationships between securities' yields and yield-surrogates.Journal of financial economics, 6(2-3), 103-126. https://doi.org/10.1016/0304-405X(78)90026-0Links ]

Bao, W., Yue, J., & Rao, Y. (2017). A deep learning framework for financial time series using stacked autoencoders and long-short term memory.PloS one,12(7). DOI: 10.1371/journal.pone.0180944Links ]

Black, F. (1986). Noise.The journal of finance,41(3), 528-543. https://doi.org/10.1111/j.1540-6261.1986.tb04513.xLinks ]

Bloomberg L.P. (2017) Tickers for stocks in the Mexican Stock Market. Retrieved January 30, 2017 from Bloomberg terminal. [ Links ]

Box, G.E.P. & Jenkins, G.M. (1970). Time Series Analysis Forecasting And Control. XF2006175558. 3. 10.2307/1912100. [ Links ]

Bruzda, J. (2019). Complex analytic wavelets in the measurement of macroeconomic risks.The North American Journal of Economics and Finance,50, 100988. https://doi.org/10.1016/j.najef.2019.100988Links ]

Caetano, M. A. L., & Yoneyama, T. (2007). Characterizing abrupt changes in the stock prices using a wavelet decomposition method.Physica A: Statistical Mechanics and its Applications,383(2), 519-526. https://doi.org/10.1016/j.physa.2007.03.027Links ]

Carrion, A. (2013). Very fast money: High-frequency trading on the NASDAQ. Journal of Financial Markets, 16(4), 680-711. https://doi.org/10.1016/j.finmar.2013.06.005Links ]

Chang, P. C., & Fan, C. Y. (2008). A hybrid system integrating a wavelet and TSK fuzzy rules for stock price forecasting.IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews),38(6), 802-815. DOI: 10.1109/TSMCC.2008.2001694Links ]

Choi, J. Y., & Lee, B. (2018). Combining LSTM network ensemble via adaptive weighting for improved time series forecasting.Mathematical Problems in Engineering,2018. https://doi.org/10.1155/2018/2470171Links ]

Cohen, R. B., Gompers, P. A., & Vuolteenaho, T. (2002). Who underreacts to cash-flow news? Evidence from trading between individuals and institutions.Journal of financial economics ,66(2-3), 409-462. https://doi.org/10.1016/S0304-405X(02)00229-5Links ]

Cybenko, G. (1989). Approximation by superpositions of a sigmoidal function.Mathematics of control, signals and systems, 2(4), 303-314. https://doi.org/10.1007/BF02551274Links ]

Dremin, I. M., & Leonidov, A. V. (2008). Volatility dynamics of wavelet-filtered stock prices.Bulletin of the Lebedev Physics Institute,35(1), 1-5. https://doi.org/10.1007/s11953-008-1001-7Links ]

Dunne, S., & Ghosh, B. (2013). Weather adaptive traffic prediction using neurowavelet models.IEEE Transactions on Intelligent Transportation Systems,14(1), 370-379. DOI: 10.1109/TITS.2012.2225049Links ]

Elman, J. L. (1990). Finding structure in time.Cognitive science,14(2), 179-211. https://doi.org/10.1207/s15516709cog1402_1Links ]

Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work.The journal of finance ,25(2), 383-417. https://doi.org/10.1111/j.1540-6261.1970.tb00518.xLinks ]

Fischer, T., & Krauss, C. (2018). Deep learning with long short-term memory networks for financial market predictions.European Journal of Operational Research,270(2), 654-669. https://doi.org/10.1016/j.ejor.2017.11.054Links ]

Giardina, I., & Bouchaud, J. P. (2003). Bubbles, crashes and intermittency in agent based market models.The European Physical Journal B-Condensed Matter and Complex Systems,31(3), 421-437. https://doi.org/10.1140/epjb/e2003-00050-6Links ]

Gomes, G. S. D. S., Ludermir, T. B., & Lima, L. M. (2011). Comparison of new activation functions in neural network for forecasting financial time series.Neural Computing and Applications,20(3), 417-439. https://doi.org/10.1007/s00521-010-0407-3Links ]

Goodfellow, I., Bengio, Y. & Courville, A. (2016). Deep learning: adaptive computation and machine learning.The MIT Press. [ Links ]

Gupta, S., Das, D., Hasim, H., & Tiwari, A. K. (2018). The dynamic relationship between stock returns and trading volume revisited: A MODWT-VAR approach.Finance Research Letters,27, 91-98. https://doi.org/10.1016/j.frl.2018.02.018Links ]

Hansson, M. (2017). On stock return prediction with LSTM networks. Retrieved February 202, from http://lup.lub.lu.se/student-papers/record/8911069Links ]

Hayek, F. (1945). The Use of Knowledge in Society. The American Economic Review, 35(4), 519-530. Retrieved February, 2021, from http://www.jstor.org/stable/1809376Links ]

Hendershott, T., & Riordan, R. (2011). High frequency trading and price discovery. Manuscript, University of California, Berkeley, 3. Retrieved February 2021. Retrieved February 2021, from: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1928510. [ Links ]

Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory.Neural computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735Links ]

Hornik, K., Stinchcombe, M., & White, H. (1989). Multilayer feedforward networks are universal approximators.Neural networks, 2(5), 359-366. https://doi.org/10.1016/0893-6080(89)90020-8Links ]

Ismail, M. T., Audu, B., & Tumala, M. M. (2016). Comparison of forecasting performance between MODWT-GARCH (1, 1) and MODWT-EGARCH (1, 1) models: Evidence from African stock markets.The journal of finance and Data Science, 2(4), 254-264. https://doi.org/10.1016/j.jfds.2017.03.001Links ]

Jammazi, R., & Aloui, C. (2012). Crude oil price forecasting: Experimental evidence from wavelet decomposition and neural network modeling.Energy Economics,34(3), 828-841. https://doi.org/10.1016/j.eneco.2011.07.018Links ]

Jefferies, P., Hart, M. L., Hui, P. M., & Johnson, N. F. (2001). From market games to real-world markets.The European Physical Journal B-Condensed Matter and Complex Systems ,20(4), 493-501. https://doi.org/10.1007/s100510170228Links ]

Jordan, M. I. (1997). Serial order: A parallel distributed processing approach. In Advances in psychology (Vol. 121, pp. 471-495). North-Holland. https://doi.org/10.1016/S0166-4115(97)80111-2Links ]

Khashei, M., & Bijari, M. (2010). An artificial neural network (p, d, q) model for timeseries forecasting.Expert Systems with applications,37(1), 479-489. https://doi.org/10.1016/j.eswa.2009.05.044Links ]

Kirilenko, A. A., & Lamacie, G. (2015). Latency and asset prices.Available at SSRN 2546567. http://dx.doi.org/10.2139/ssrn.2546567Links ]

Kişi, Ö. (2008). Stream flow forecasting using neuro‐wavelet technique.Hydrological Processes: An International Journal,22(20), 4142-4152. https://doi.org/10.1002/hyp.7014Links ]

Ledolter, J., & Abraham, B. (1981). Parsimony and its importance in time series forecasting. Technometrics, 23(4), 411-414. DOI: 10.1080/00401706.1981.10487687Links ]

Maciel, L. S., & Ballini, R. (2008). Design a neural network for time series financial forecasting: Accuracy and robustness analysis.Anales do 9º Encontro Brasileiro de Finanças, Sao Pablo, Brazil. [ Links ]

Maknickienė, N., & Maknickas, A. (2012). Application of neural network for forecasting of exchange rates and forex trading. In The 7th international scientific conference Business and Management (pp. 10-11). DOI:10.3846/bm.2012.017Links ]

Malkiel, B. G. (2003). The efficient market hypothesis and its critics.Journal of economic perspectives,17(1), 59-82. DOI: 10.1257/089533003321164958Links ]

Martens, M. (2002). Measuring and forecasting S&P 500 index‐futures volatility using high‐frequency data. Journal of Futures Markets: Futures, Options, and Other Derivative Products, 22(6), 497-518. https://doi.org/10.1002/fut.10016Links ]

Mehr, A. D., Kahya, E., & Olyaie, E. (2013). Streamflow prediction using linear genetic programming in comparison with a neuro-wavelet technique.Journal of Hydrology,505, 240-249. https://doi.org/10.1016/j.jhydrol.2013.10.003Links ]

Minu, K. K., Lineesh, M. C., & John, C. J. (2010). Wavelet neural networks for nonlinear time series analysis.Applied Mathematical Sciences, 4(50), 2485-2495. [ Links ]

Murtagh, F., Starck, J. L., & Renaud, O. (2004). On neuro-wavelet modeling.Decision Support Systems,37(4), 475-484. https://doi.org/10.1016/S0167-9236(03)00092-7Links ]

Napoli, C., Bonanno, F., & Capizzi, G. (2010). An hybrid neuro-wavelet approach for long-term prediction of solar wind.Proceedings of the International Astronomical Union, 6(S274), 153-155. https://doi.org/10.1017/S174392131100679XLinks ]

Nath, G. C., & Dalvi, M. (2004). Day of the week effect and market efficiency-evidence from Indian equity market using high frequency data of national stock exchange.Available at SSRN 1092765. DOI: 10.2139/ssrn.1092765Links ]

Oancea, B., & Ciucu, Ş. C. (2014). Time series forecasting using neural networks. Proceedings of the CKS 2013 International Conference.arXiv preprint arXiv:1401.1333. [ Links ]

Ortega, L., & Khashanah, K. (2014). A neuro‐wavelet model for the short‐term forecasting of high‐frequency time series of stock returns.Journal of Forecasting,33(2), 134-146. https://doi.org/10.1002/for.2270Links ]

Otazu, X. (2008). Theory and implementation of image fusion methods based on the a trous algorithm.Image Fusion, 139. [ Links ]

O’Hara, M. (2014) High-Frequency Trading and its impact on Markets. Financial Analysts Journal, 70 (3). ISSN: 0015-198X . https://doi.org/10.2469/faj.v70.n3.6Links ]

Pascanu, R., Mikolov, T., & Bengio, Y. (2013, February). On the difficulty of training recurrent neural networks. InInternational conference on machine learning(pp. 1310-1318). [ Links ]

Percival, D. B., & Walden, A. T. (2000).Wavelet methods for time series analysis(Vol. 4). Cambridge university press. [ Links ]

Ramsey, J. B. (1999). The contribution of wavelets to the analysis of economic and financial data.Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences,357(1760), 2593-2606. https://doi.org/10.1098/rsta.1999.0450Links ]

Reboredo, J. C., & Rivera-Castro, M. A. (2013). A wavelet decomposition approach to crude oil price and exchange rate dependence.Economic Modelling,32, 42-57. https://doi.org/10.1016/j.econmod.2012.12.028Links ]

Roman, J., & Jameel, A. (1996, January). Backpropagation and recurrent neural networks in financial analysis of multiple stock market returns. InProceedings of HICSS-29: 29th Hawaii International Conference on System Sciences(Vol. 2, pp. 454-460). IEEE. DOI: 10.1109/HICSS.1996.495431Links ]

Siami-Namini, S., & Namin, A. S. (2018). Forecasting economics and financial time series: ARIMA vs. LSTM.arXiv preprint arXiv:1803.06386. [ Links ]

Sirignano, J., & Spiliopoulos, K. (2017). Stochastic gradient descent in continuous time.SIAM Journal on Financial Mathematics, 8(1), 933-961. https://doi.org/10.1137/17M1126825Links ]

Strawiński, P., & Ślepaczuk, R. (2008). “Analysis of high frequency data on the Warsaw Stock Exchange in the context of efficient market hypothesis”.Journal of Applied Economic Sciences, 3(3), 306-319. [ Links ]

Tapia, A. (2020) Hipótesis de eficiencia en mercados automatizados. Doctoral dissertation. EGADE Business School, Tecnologico de Monterrey. URI: https://hdl.handle.net/11285/636543Links ]

Zhang, B. L., Coggins, R., Jabri, M. A., Dersch, D., & Flower, B. (2001). Multiresolution forecasting for futures trading using wavelet decompositions.IEEE Transactions on Neural Networks,12(4), 765-775. DOI: 10.1109/72.935090Links ]

Zhu, L., Wang, Y., & Fan, Q. (2014). MODWT-ARMA model for time series prediction.Applied Mathematical Modelling,38(5-6), 1859-1865. https://doi.org/10.1016/j.apm.2013.10.002Links ]

*No source of funding for research development

Received: November 05, 2020; Accepted: March 10, 2021

*Corresponding author: montserrat.reyna@anahuac.mx

3

Investigador independiente

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License