Daily Cryptocurrency Returns Forecasting and Trading via Machine Learning

We execute a comparative analysis of machine learning models for the time-series forecasting of the sign of next-day cryptocurrency returns. We begin by compiling a proprietary dataset that encompasses a wide array of potential cryptocurrency valuation factors (price trends, liquidity, volatility, network, production, investor attention), subsequently identifying and evaluating the most significant factors. We apply eight machine learning models to the dataset, utilizing them as classifiers to predict the sign of next day price returns for the three largest cryptocurrencies by market capitalization: bitcoin, ethereum, and ripple. We show that the most significant valuation factors for cryptocurrency returns are price trend variables, seven and thirty-day reversal, to be specific. We conclude that support vector machines result in the most accurate classifications for all three cryptocurrencies. Additionally, we find that boosted models like AdaBoost and XGBoost have the poorest classification accuracy. At length, we construct a probability-based trading strategy that secures either a daily long or short position on one of the three examined cryptocurrencies. Ultimately, the strategy yields a Sharpe of 2.8 and a cumulative log return of 3.72. On average, the strategy’s log returns outperformed standalone investments in all three cryptocurrencies by a factor of 5.64, and Sharpe ratios more than threefold.

Markov Regime-Switching Autoregressive Model of Stock Market Returns in Nigeria

This study is designed to model and forecast Nigeria’s stock market using the All-Share Index (ASI) as a proxy. By employing the Markov regime-switching autoregressive (MS-AR) model with data from April 2005 to September 2019, the study analyzes the stock market volatility in three distinct regimes (accumulation or distribution – regime 1; big-move – regime 2; and excess or panic phases – regime 3) of the bull and bear periods. Six MS-AR candidate models are estimated and based on the minimum AIC value, MS(3)-AR(2) is returned as the optimal model among the six candidate models. The MS(3)-AR(2) analysis provides evidence of regime-switching behaviour in the stock market. The study also shows that only extreme events can switch the ASI returns from regime 1 to regime 2 and to regime 3, or vice versa. It further specifies an average duration period of 9, 3 and 4 weeks for the accumulation/distribution, big-move and excess/panic regimes respectively which is an evidence of favorable market for investors to trade. Based on Root Mean Square Error and Mean Absolute Error, the fitted MS-AR model is adjudged the most appropriate ASI returns forecasting model. The study recommends investments in stock across the regimes that are switching between accumulation/distribution and big-move phases for promising returns.

Estimation of Direction of Exchange Rate, Gold Price and Stock Market Returns with High Order Markov Chain Models

This study uses a hybrid high order Markov Chains Model to predict direction of exchange rate, gold price and stock market returns with the Artificial Neural Network Algorithm as an estimator of transition probability matrix. Many forecasting techniques are used to examine the direction of returns forecasting in the literature such as Markov Chains Model and Artificial Neural Network Algorithm. In this study, it is aimed to combine these two techniques and to utilize the predict values of the Artificial Neural Network Algorithm for calculate transition probabilities matrix. Calculations show that the hybrid model gives high correct classification probabilities besides of well approximated transition probabilities. Returns series of USD/TRY exchange rate, closing price of Borsa Istanbul Stock Exchange and gold prices cover the period of 01/01/2003 and 31/01/2016. All series are obtained from database of Central Bank of Turkey. As a result, although the transition probabilities almost equal to 0.5 and so estimation of these series are not easy, the transition probabilities and correct classification probabilities gained from the hybrid model provide substantial information related to direction of returns forecasting. Besides, estimated model provide valuable information to individual investors and companies, and could help them to take position against to risks.

Dynamic Return-Volume Relations in the Saudi Stock Market: Evidence from Quantiles Regressions

This paper examines the causality patterns between the lagged trading volume and returns of the Saudi stock market (TASI) for the period from2003:01 to April 2013:05, along with two consecutive sub-periods to account for pre- and post- market collapse of 2006. Using the quantile regression approach, the study finds that the return-volume relations are heterogeneous across quantiles with symmetric tendency across the mean for the full sample period. On the contrary, the study could not support the heterogeneous and symmetric effects for the first sub-sample period. The second sub-sample period is characterized by homogenous across quantiles with statistical evidence of symmetry. Thus, the study concludes that the dependence structure between the lagged volume and subsequent market returns seems to be randomly relying on the chosen period which makes volume unsuitable to be used as explanatory power for returns forecasting.