Navid Parvini’s research while affiliated with University of Kent and other places

This chapter is concerned with textual and sentiment analysis in agriculture commodities market using the natural language processing (NLP) methods. There are extensive research on textual and sentiment analysis in financial markets however, most of them are focusing on equity market and a minority on other commodities like energy commodities. Therefore, this chapter first reviews research works on textual and sentiment analysis in agriculture market in general. Then, presents textual analysis methods that can be carried out to study the effect of textual data and sentiment in agriculture market. Finally, it presents an example of implementing a topic modelling task and textual regression for forecasting realized volatility of corn returns. To the best of the author’s knowledge, there is no study focusing on textual regression in agriculture market. Additionally, the studies conducting textual sentiment analysis are very limited. In this spirit, this study tries to fill this gap by introducing both well established and new textual and sentiment analysis methods to the agricultural researchers community. The limited experiment carried out with these methods in the present research testifies the superiority of the text-based models in explaining future movements of corn’s volatility. More specifically, the results of one-month-ahead realized volatility regression indicates statistically significant superior performance of both direct textual regression and sentiment regression compared to traditional methods like HAR and ARIMA. In addition, as the most accurate method, textual regression’s accuracy stands higher above that of the sentiment regression model.

In the present study, a machine learning model based on support vector regression (SVR) is proposed for forecasting the closing price of the three most capitalized cryptocurrencies, i.e., Bitcoin, Ethereum, and Ripple. The optimal hyperparameters are obtained by applying the particle swarm optimization (PSO) algorithm, and four historical price characteristics are used as predictors, namely the opening, the highest, and the lowest prices, and the cryptocurrency trading volume. Based on a sample of daily cryptocurrency prices spanning from August 8, 2015 to May 10, 2019, the proposed PSO-SVR approach is tested and compared with a class of neural network algorithms such as the multi-layer perceptron, the long short-term memory and the bi-directional long short-term memory, all of which are optimized by PSO. The results obtained indicate that the novel PSO-SVR model significantly outperforms all the neural network rivals in forecasting.

Despite the growing literature in cryptocurrency forecasting and their price drivers, the relationship between their price and other financial time series is an ongoing matter of debate. This study proposes a three‐step methodology to cover these arguments. First, we conduct an ad hoc analysis using transfer entropy (TE) to study the causal relationship between Bitcoin (BTC) returns and a vast array of financial time series. Then, we utilize variables with a significant amount of information flow toward BTC returns to forecast multi‐step‐ahead BTC returns. Finally, we use explainable artificial intelligence post hoc analysis methods to discover the contribution of each input feature to the overall forecasting. The results indicate a significant change in the information flow pattern in the first days of the COVID‐19 pandemic outbreak. Additionally, our proposed TE‐based feature‐selection method outperforms both benchmarks, a nonfeature‐selection model, and backward stepwise regression.

Investigating Bitcoin price forecasting has attracted academic attention recently. However, despite some studies on potential economic determinants of Bitcoin price, a consensus on the best predictors is not reached yet. This paper investigates different predictors from various markets including Gold, Oil, S&P500, VIX, USDI, Ether and Ripple as well as Bitcoin historical price in predicting one-step-ahead Bitcoin returns. We propose a two-stage forecasting that comprises discrete wavelet transform as the decomposition method and a deep long short-term memory network as the forecasting algorithm. Beside analyzing forecasting for both univariate and multivariate regression, we design a simulated trading system to put the forecasts into practice and analyze economic profitability of the predictors. In addition, we shed light on the black box method by implementing sensitivity analysis. To investigate the effectiveness of the predictors efficacy through time and consider the effects of early 2018 price spike, the dataset is split into two periods: 1) prior and including the spike and 2) after the spike. According to the experiments, it is hard to choose one predictor over the other in the first period. However, in the second period, Gold and Oil show the highest statistical accuracy, while S&P500 is the most profit-making predictor.