SX5E Dividends Forecasting
I first modelled dividends per share (DPS) for EUROSTOXX 50 companies using a simple ARIMA model that only looks at each firm’s past dividends. Forecast Percentage Error Distribution of ARIMA ...
I first modelled dividends per share (DPS) for EUROSTOXX 50 companies using a simple ARIMA model that only looks at each firm’s past dividends. Forecast Percentage Error Distribution of ARIMA ...
I used a Markov-switching (MS) model to classify the Bitcoin market into three distinct constant-mean regimes from 2012 to 2025. The results are as follows: ...
Coming soon
In 2024, I was tasked with analysing the real estate industry in Hong Kong. Our original database only had around 200 transactions. So I scraped rent prices for around 9000 past transactions on Centaline from May 2024 to August 2024 using Selenium. ...
A fun exercise in Principal Component Analysis (PCA) for dimensionality reduction. ...
While preparing for the Hong Kong public examination, I built an interactive economics question bank from scratch using HTML, CSS, and JavaScript. After analysing 11 years of past papers, I categorised questions into 109 types and added a randomisation function to draw questions unpredictably. I used it for spaced repetition by answering five questions a day, ultimately achieving the top grade, 5**, in the final examination. The interface of the Question Bank Website ...