Past Prices Research Articles

Proposition and modeling analysis of two asset trading strategies of gold and bitcoin under market trading

With the progress of economic globalization, market transactions and investments occur frequently. For a long time, two assets, gold and bitcoin, have become the research hotspot due to payment means, circulation means, storage means, value scale, and world currency. To this end, we built two models. The model I: Diversified BP neural network model. Model II: Portfolio strategy based on mean-standard deviation model. For Model I, we have to predict the future prices of gold and Bitcoin based on past data. We can regard the past price data as a time series. There are relatively mature methods for predicting future values based on time series. For Model II, with future prices predicted, we can invest. It brings up three more questions: should I buy or sell or do nothing now? Buy or sell only gold and bitcoin or both gold and bitcoin? When buying, how should the cash for gold and bitcoin be distributed? Finally, the specific investment strategy is determined by linear programming. Then, we analyzed the effect of commissions on final earnings and derived the relationship between total assets held last and commissions.

Free Hold Price Predictor Using Machine Learning

People who want to buy a new home tend to save more on their budgets and market strategies. The current system includes real estate calculations without the necessary forecasts for future market trends and inflation. The housing market is one of the most competitive in terms of pricing and the same has varied greatly in terms of many factors. Asset pricing is an important factor in decision- making for both buyers and investors in supporting budget allocation, acquisition strategies and deciding on the best plans as a result, it is one of the most important areas in which machine learning ideas can be used to maximize and accurately anticipate prices. As a result, in this paper, we present the different significant factors that we employ to accurately anticipate property values. To reduce residual errors, we can utilize regression models with a range of characteristics. Some engineering aspects are required when employing features in the regression model for improved prediction. To improve model fit, a set of multi-regression elements or a polynomial regression (with a set of varying strengths in the elements) is frequently utilized. In these models, it is expected to be significantly affected by the slope of the spine used to reduce it. Therefore, it directs the best use of regression models over other strategies to maximize the effect. This paper's goal is to predict free hold prices for free hold consumers based on their budgets and goals. Prospective prices can be forecast by evaluating past market trends and price levels, as well as future developments.

Research on the Three-Stage Trading Strategy of Gold and Bitcoin Using Technical Indicators and A LSTM Model

In recent years, in pursuit of higher returns, people began to try the combination of bitcoin and gold. This study explored how to use the current and past prices of gold and bitcoin to get the best investment strategies for the next day to get the maximum return. The obtained historical price data of gold and bitcoin from 2016 to 2021 are divided into three stages according to the time axis, namely, stage Ⅰ (1st day-300th day), stage Ⅱ (300th day-400th day) and stage III (400th day-1825th day). Different methods are used for each stage to determine the trading strategy. In stage I, a comprehensive index is constructed by using four technical indexes closely related to the formulation of daily strategies in the financial market, such as moving average convergence and divergence, deviation rate, stochastic index and detrended price oscillator, and the trading strategy is determined by using this comprehensive index. In stage II, the four technical indicators and product prices in stage I are used as training sets, and the daily trading strategy is given by using the decision tree classification model. In the third stage, the long-term and short-term memory neural network is used to develop the future price prediction model of gold and bitcoin, and the trading strategies of gold, bitcoin and cash asset portfolios are formulated according to the dynamic optimization results of different types of traders' risk tolerance. The results show that during the five years from September 11, 2016 to September 10, 2021, the whole strategy created a net profit of $170,381 for aggressive traders and $7,694 for rational traders. After verification, our strategy is effective and robust, and its volatility is relatively small. The research provides theoretical guidance for investors on how to choose portfolios and trade to get the maximum return.

Testing Weak Form Information Memory: A Study of Indian Futures Market

The greatness and simplicity of EMH (Efficient Market Hypothesis) has surprised investors from time to time regarding the correct pricing and efficient strategies for investment. Applying the same logic to Indian Derivative market, the present paper makes an effort to examine the market efficiency of index futures. The study has undertaken tests of market efficiency of Nifty Index Futures contract from 1st January, 2018 to 1st January, 2021 Random walk theory is tried to be tested in this context and Ljung- Box test is applied to check the fluctuations in future contract prices. The background thought is kept like if market is inefficient at its weak-form that leads to the proof that market does have memory and future prices do not stick to random walk. Findings of the study reveals that past prices of future contracts cannot be used as base to forecast the terminal prices and abnormal profit can be made as they are random in nature.

Stock Market Prediction using Machine Learning Algorithms

Machine learning is effectively implemented in forecasting the stock prices. The objective is to predict the stock prices in order to make more informed and accurate investment decisions. The essence of the issue lies in exactly forecasting the future stock price of the reputed firm, based on historical or past prices, using Recurrent Neural Network algorithm, LSTM-LongShort Term Memory, ANN-Artificial Neural Network, CNN - Convolutional Neural Network. We also used regression as a trial and error method and found the limitations in the case of a non-continuous dataset. The findings of the study would lend a hand to a common investor averting huge losses and optimizing the stockholder investing in beneficial stocks. While predicting the actual prices of a stock is an uphill climb , we can build a model that will predict whether the price will go up or down.

Combined trading strategy of bitcoin and gold

Market traders buy and sell volatile assets frequently, with a goal to maximize their total return. We have been asked to develop a model that uses only the past stream of daily prices to date to determine each day if the trader should buy, hold, or sell their assets in their portfolio. The assets that can be traded are Bitcoin and gold. We will start with $1000 on 9/11/2016 and try to maximize the total return until 9/10/2021. We will start from forecasting prices and developing trading strategies. In terms of price prediction, we use MSE and Trend_Acc as indicators, and use XGBoost, a representative strong learning algorithm in traditional machine learning, and LSTM, which is good at time series prediction in deep learning, to fit and forecast the data respectively. At first glance, the curve fitting MSE are satisfactory , but, a closer look reveals that the model either firmly remembers the data of the training set, resulting in a lack of generalization ability for unknown data (XGBoost), or tends to take the previous day's results as the predicted results, resulting in a significant lag in the prediction curve (LSTM), all of which are reflected in the models' poor performance in predicting whether prices will rise or fall in the future. In our view, since a large number and complexity of factors affecting prices, it is unrealistic to predict future prices accurately from past prices alone, unless we can get rid of the limitation of the problem, use additional data to assist the prediction, or use all the data as a training set for fitting, we cannot achieve good results in the prediction, but such behavior is inconsistent with our original intention. We established restricted trading model based on composite index judgment. The model not only applies the traditional economic Relative Strength Index and Stochastics Oscillator Index, but also introduces the K-Lipschitz limitation in deep learning into the model. The model dynamically adjusts each transaction strategy according to the changes of working capital and total assets, purchase cost, selling profit and other factors, and the total income of the model is $132433.1. In the horizontal comparison, the profit of our model is more than 39.6%-283.6% than that of the traditional moving average strategy and RSI-STC strategy, and 51.3% higher than that of the random walk model using Montmarlowe algorithm. In addition, we collected the transaction data of bitcoin and gold from 2012-01-01 to 2022-02-21, and applied the model to the historical data, and received good returs. For example, from 2012-1-1 to 2022-2-21, the return was $8418074.9. It is proved that the model has high generalization performance and strong stability. To test the sensitivity of the model to transaction costs, we also analyze the changes in trading strategies that should occur when fees rise. The analysis shows that the traders' single trading volume decreases first and then increases with the increase of the commission fee. The results of sensitivity analysis show that the return change is less than 3%, which proves the robustness of the model. Finally, we made a memo to summarize our work.

Community Impact on a Cryptocurrency: Twitter Comparison Example Between Dogecoin and Litecoin

Context: The third generation of cryptocurrencies gathers cryptocurrencies that are as diverse as the market is big (e.g., Dogecoin or Litecoin). While Dogecoin is seen as a memecoin, the other gathers a very different category of investors. To our knowledge, no study has independently assessed the crypto community’s economical impact on these cryptocurrencies. Furthermore, various methodological possibilities exist to forecast cryptocurrency price—mainly coming from online communities.Method: Our study has retrospectively studied (from 01/01/2015 to 03/11/2021)—using open access data—the association strength (using normalized mutual information) and the linear correlation (using Pearson’s correlation) between Twitter activity and cryptocurrency economical attributes. In addition, we have computed different models (ADF, ARIMA, and Interpretable MultiVvariable Long Short-Term Memory recurrent neural network) that forecast past price values and assessed their precision.Findings and conclusions: While the average Dogecoin transaction value is impacted by tweets, tweets are impacted by Litecoin transactions number and average Litecoin transaction value. Tweet number is impacted by Dogecoin whale behavior, but no significant relationship was found between Litecoin whales and tweets. The forecasting error resulting from our ARIMA (0,0,0) models was 0.08% (with Litecoin) and 0.22% (with Dogecoin). Therefore, those are just the beginning of scientific findings that may lead to building a trading robot based on these results. However, in itself, this study is only for academic discussion, and conclusions need to be drawn by further research. The authors cannot be liable if any financial investment is made based on its conclusions.

A semiparametric approach to estimating reference price effects in sales response models

It is well known that store-level brand sales may not only depend on contemporaneous influencing factors like current own and competitive prices or other marketing activities, but also on past prices representing customer response to price dynamics. On the other hand, non- or semiparametric regression models have been proposed in order to accommodate potential nonlinearities in price response, and related empirical findings for frequently purchased consumer goods indicate that price effects may show complex nonlinearities, which are difficult to capture with parametric models. In this contribution, we combine nonparametric price response modeling and behavioral pricing theory. In particular, we propose a semiparametric approach to flexibly estimating price-change or reference price effects based on store-level sales data. We compare different representations for capturing symmetric vs. asymmetric and proportional vs. disproportionate price-change effects following adaptation-level and prospect theory, and further compare our flexible autoregressive model specifications to parametric benchmark models. Functional flexibility is accommodated via P-splines, and all models are estimated within a fully Bayesian framework. In an empirical study, we demonstrate that our semiparametric dynamic models provide more accurate sales forecasts for most brands considered compared to competing benchmark models that either ignore price dynamics or just include them in a parametric way.

Search intensity, search time and prices: evidence from retail diesel markets in France

We study the effect of price variations over time and across space on search intensity and search time by consumers in retail markets for diesel in France. The main contribution of the paper is that existing work in industrial organisation in this area has already studied the effect of such variability on the first measure of search but not on the second one. Using novel data from daily consultation measures and price information from a government-run website, we find that price dispersion across space increases search activity and the amount of time allocated to search. Furthermore, while contemporaneous price changes do not appear to influence the number of visits and time per visit to the website, several coefficients on past price changes are positive and statistically different from zero. Our results thus suggest that price dispersion and price variability play a role in inducing search by current and potential customers.

The Impact of Historical Price Information on Purchase Deferral

In this article, the authors examine how historical price information influences consumers’ decision to defer a purchase. They focus on two aspects of historical price information: the direction and the frequency of past price changes. The authors advance a theoretical framework postulating that the interaction between these two factors shapes consumers’ decisions to buy now versus later. Controlling for the total magnitude of price changes, the authors propose that consumers are more likely to defer purchase when the price of the product has previously increased compared with when the price has decreased. Importantly, the authors hypothesize that this effect is more pronounced when consumers observe a single large change in price (e.g., a decrease of $100 vs. an increase of $100) compared with when they observe multiple smaller changes that establish a trend (e.g., four decreases of $25 vs. four increases of $25). The authors argue that these effects are driven by differences in consumers’ expectations about future prices. They test their predictions, as well as two moderators of the proposed effects—the monotonicity and the timing of price changes—in six well-powered preregistered experimental studies (N = 5,713) using both hypothetical and actual purchases.

Modelling the mean and volatility spillover between green bond market and renewable energy stock market

<abstract><p>In this paper,we investigate the mean and volatility spillover between the price of green bonds and the price of renewable energy stocks using daily price series from 02/11/2011 to 31/08/2021. The unrestricted trivariate VAR-BEKK-GARCH model is employed to examine potential causality,mean,and volatility spillover effects from the green bond market to the renewable energy stock market and vice-versa. The results from the VAR-BEKK-GARCH model indicate that there exists a uni-directional Granger causality from renewable energy stock prices to green bond prices. While the price of green bonds is positively influenced by its own lagged values and the lagged values of renewable energy stock prices,only the past price value of renewable energy stocks has a positive effect on the current price value. We identified a uni-directional volatility spillover from renewable energy stock prices to green bond prices. However,there was no shock spillover from both sides of the market. This research shows that investors in the green bond market should always consider information from the renewable energy stock market because of the causal link between renewable energy stocks and green bonds.</p></abstract>

Two-Stage Hybrid Machine Learning Model for High-Frequency Intraday Bitcoin Price Prediction Based on Technical Indicators, Variational Mode Decomposition, and Support Vector Regression

Due to the inherent chaotic and fractal dynamics in the price series of Bitcoin, this paper proposes a two-stage Bitcoin price prediction model by combining the advantage of variational mode decomposition (VMD) and technical analysis. VMD eliminates the noise signals and stochastic volatility in the price data by decomposing the data into variational mode functions, while technical analysis uses statistical trends obtained from past trading activity and price changes to construct technical indicators. The support vector regression (SVR) accepts input from a hybrid of technical indicators (TI) and reconstructed variational mode functions (rVMF). The model is trained, validated, and tested in a period characterized by unprecedented economic turmoil due to the COVID-19 pandemic, allowing the evaluation of the model in the presence of the pandemic. The constructed hybrid model outperforms the single SVR model that uses only TI and rVMF as features. The ability to predict a minute intraday Bitcoin price has a huge propensity to reduce investors’ exposure to risk and provides better assurances of annualized returns.

Asset price volatility and investment horizons: An experimental investigation

We study the effects of the investment horizon on asset price volatility using a Learning to Forecast laboratory experiment. We find that, for short investment horizons, participants coordinate on self-fulfilling trend-extrapolating predictions. Price deviations are then reinforced and amplified, possibly leading to large bubbles and crashes in asset prices. For longer investment horizons such bubbles do not emerge and price volatility tends to be lower. This is due to the fact that, for longer horizons, there is more dispersion in participants’ forecasts, and participants extrapolate trends in past prices to a lesser extent. We also show that, independent of the investment horizon, if the initial history of asset prices is already relatively stable before participants start their prediction task, price volatility remains small, with prices close to their fundamental values for the duration of the experiment.

Forecasting US Commercial Property Price Indexes Using Dynamic Factor Models

The general purpose of a dynamic factor model (DFM) is to summarize a large number of time series into a few common factors. In this paper we explore several DFMs on 80 granular, non-overlapping commercial property price indexes in the US, quarterly from 2001Q1 to 2017Q2. We examine the nature and the structure of the factors and the index forecasts that can be produced from the DFMs. We consider specifications of one to four common factors. As a major motivation for the use of DFMs is their ability to improve out-of-sample forecasting of systems of numerous related series, we apply the DFM estimated factors in an Autoregressive Distributed Lag (ARDL) model to forecast individual market index returns. We compare for four markets the forecasts to those from a benchmark univariate autoregression. The results show that the DFM & ARDL model predicts the crisis and subsequent recovery really well, whereas the benchmark model typically extrapolates the past price trend.

Second-hand house price index forecasting with neural networks

ABSTRACT The house market in China has been growing rapidly for the past decade and price forecasting has become a significant issue to the people and policymakers. We approach this problem by examining neural networks for second-hand house price index forecasting from 10 major cities for March 2012–May 2020. Our purpose is to build simple and accurate neural networks to contribute to pure technical house price forecasting of the Chinese market. We explore various model settings across the algorithm, delay, hidden neuron, and data spitting ratio, and arrive at a rather simple neural network with three delays and eight hidden neurons, which leads to stable performance of 0.8% average relative root-mean-square error across the 10 cities for the training, validation, and testing phases. Results here can be used on a standalone basis or combined with fundamental forecasting in forming perspectives of house price trends and conducting policy analysis.

ALGORITMA NEURAL NETWORK BACKPROPAGATION UNTUK PREDIKSI HARGA SAHAM PADA TIGA GOLONGAN PERUSAHAAN BERDASARKAN KAPITALISASINYA

Stock is one type of investment where investors can gain profits in the form of capital gains and dividends. Types of shares based on the level of capitalization are divided into 3 types, namely the first layer (blue chips), the second layer, and the third layer. One of the techniques that investors use in order to make a profit is technical analysis, which is using data of past stock prices and volumes based on the assumption that trends can recur following historical data patterns. Based on the assumptions of technical analysis, it is possible to use data mining to predict stock prices. In this study, stock price predictions will be carried out by comparing three types of companies based on their capitalization, for first layer stocks using PT. Bank Central Asia Tbk (BBCA), the second layer using PT. XL Axiata Tbk (EXCL), and third layer using PT Pembangunan Graha Lestari Indah Tbk. The data mining algorithm that will be used is the Neural Network Backpropagation method. The attributes used as predictors are open, high, low, and volume, while the objective attribute is close. This study aims to determine whether daily stock historical data can be used to predict stock prices using the Neural Network Backpropagation method and how to compare the results of predictions between 3 companies with different capitalization levels. The result of RMSE for BBCA by using the most optimal combination of parameters and 3 hidden layer is 123.84. The result of RMSE for EXCL by using the most optimal combination of parameters and hidden layer 2 is 37.36. The result of RMSE for PGLI by using the most optimal combination of parameters and hidden layer 6 is 6.16. So that the backpropagation neural network algorithm is most optimally applied to third layer companies, PT. Pembangunan Graha Lestari Indah Tbk because the RMSE value is the smallest.

Price Transmission in Yam Markets in Côte d’Ivoire

Many studies have shown that rising prices are not transmitted in the same way as falling prices and vice versa. There is an asymmetry in price transmission between agricultural commodity markets that is an important issue in household food security. The main objective of this study is to examine the nature of price transmission in yam markets in Côte d’Ivoire. To do so, two regression models were used, namely the Ordinary Least Squares (OLS) method and Vector autoregressive (VAR) model. The results revealed that a variation in prices on Abengourou and Korhogo markets affects the price on the Abidjan market even though trade between these markets is not direct. Similarly, the price one month later on Abidjan market is affected by its past price and that of Korhogo market one month earlier. As recommendation, the study suggests that the government should strengthen and improve road infrastructure as well as communication networks to enhance trade between markets.

Prévisions d’analystes financiers et ordre de grandeur des prix : une approche expérimentale

La littérature récente en comptabilité et en finance montre que le niveau des cours des actions influence les prévisions de prix des analystes (Roger, Roger et Schatt [2018]). Dans le présent article, nous montrons que ce résultat reste valide dans le cadre contrôlé du laboratoire quand des sujets doivent prévoir les prix futurs sur un marché expérimental auquel ils ne participent pas. Chaque sujet fait des prévisions lors de deux marchés successifs : l’un pour lequel la valeur fondamentale est faible et l’autre pour lequel la valeur fondamentale est élevée. Bien que les prix de marché soient plus élevés (par rapport à la valeur fondamentale) sur les marchés à petits prix du fait du biais de « petit prix » des sujets traders, nos résultats indiquent que les prévisions des analystes sont plus optimistes sur les marchés à petits prix que sur les marchés à prix élevés. Les sujets analystes ancrent leurs prévisions sur les prix de marché passés et n’atténuent pas le biais des sujets traders. Nos résultats montrent ainsi que l’existence de ce biais de « petit prix » reflète l’utilisation par les sujets de deux échelles mentales différentes pour traiter petits et grands nombres. Classification JEL : G14.

An Empirical Investigation of the Random Walk Hypothesis in the Nigerian Stock Market

The Random Walk Hypothesis (RWH) states that stock prices move randomly in the stock market without following any regular or particular pattern and as such historical information contained in the past prices of stocks cannot be used to predict current or future stock prices. Hence, stock prices are unpredictable and that investors cannot usurp any available information in the market to manipulate the market and make abnormal profits. This study empirically examines the random walk hypothesis in the Nigerian stock market using the daily quotations of the Nigerian stock exchange from 2nd January, 1998 to 31st December, 2019. The study employs Augmented Dickey-Fuller unit root test, the random walk model, Ljung-Box Q-statistic test for serial dependence, runs test of randomness, and the robust variance ratio test as methods of analyses. The result of the study rejected the null hypotheses of a unit root and random walk in the stock returns. The null hypothesis of no serial correlation in the residuals of stock returns was also rejected indicating the presence of serial correlation/autocorrelation in the residual series. The result of the runs test rejected the null hypothesis of randomness in the Nigerian stock returns. The results of the variance ratio test under homoskedasticity and heteroskedasticity assumptions both strongly rejected the null hypothesis of a random walk for both joint tests and test of individual periods. Based on the results of the four tests applied in this study, it is concluded that the Nigerian daily stock returns under the period of investigation do not follow a random walk and hence the null hypothesis of a random walk is rejected. The results of the study further revealed that the Nigerian stock market is weak-form inefficient indicating that prices in the Nigerian stock market are predictable, dependable, consistently mispriced, inflated, liable to arbitraging and left unprotected to speculations and market manipulations. The study provided some policy recommendations

Joint inventory and pricing control with lagged price responses

This paper studies the single stage, multiple period joint inventory and pricing planning problem with lagged price response. We prove that the optimal joint inventory and pricing policy takes the form of a state dependent list price base stock policy for the additive linear demand function with a first order auto-regressive distributed lag. Both the base stock level and the list price decisions are determined as functions of the effective price, which captures the current and past pricing impacts. Numerical studies are conducted to examine the impact of the lagged price response on the optimal decisions and the total profit.