In which I introduce the mathematics involved in a Pairs Strategy; using Python, I will demonstrate how to implement it — The fundamental idea of pairs trading is taking a market neutral position and using it to exploit an equilibrium between two assets. INTRODUCTION A pairs trade strategy is based on the historical correlation of two securities. The securities in a pairs trade must have a high positive correlation. This is the…

I will be using a new Data Driven Exponentially Weighted Moving Average along with the ARCH & GARCH models to predict the forward volatility of Bitcoin, Ethereum and Solana. — By understanding the forecasted volatility we are able to better limit risk and improve our trading strategies. Volatility is a statistical measure of the variance of returns for a given security. I will be using the absolute value of the log(daily return) to model the volatility of Bitcoin.

The Gaussian Mixture model (GMM) will be used to detect different moods in the stock market. The GMM model will be applied twice: first using price data, and subsequently with macro-economical data. Introduction Over the last decade, the stock market has become increasingly accessible to the average layman. …

I will be using a HMM to classify the market into three volatility regimes. The Data The ETF ‘SPY’ will be used to represent the the S&P500. The data will be daily from Yahoo.

I will be predicting the number of monthly Sunspots using time series analysis. The models that will be explored in-depth are the ARMA and LSTM models. — Introduction Firstly: what is a sunspot? Sunspots are a temporary phenomena on the Sun’s photosphere that appear darker than the surrounding areas. The reason why I have selected the sunspots dataset for time series analysis is sunspots appear on an 11-year solar cycle, meaning we should expect to see a seasonality…

I will model the volatility of the S&P500 to classify the market into three different segments to enhance algorithmic trading strategies. Using market segmentation to create more profitable algorithmic trading strategies The primary idea will be to classify the S&P500 into three segments based on the modelled volatility and using…

Using the Tensorflow package in python and testing using a LSTM model — Here I would like to introduce an innovative new loss function. I am defining this new loss function as the MSE-MAD. The loss function is constructed using the exponential weighted moving average framework and using MSE and MAD in combination. …

Statistically Modelling the early days of the COVID-19 pandemic using Rstudio. Introduction The purpose of this article will be to demonstrate how statistical models can be used for forecasting disease spread. The disease that will be investigated will be COVID-19 during the initial 3 months. …

Introduction I will be introducing a new volatility forecasting approach, that can be applied to any stationary time series. This new method was first introduced in “A Novel Algorithmic Trading Strategy Using Data-Driven Innovation Volatility” , it is denoted by DD-EWMA. In this example we will be looking at the volatility…