Alternative data (local news, social media, blogs) from emerging markets, e.g. China
Investor sentiment, Key Opinion Leader (KOL) sentiment, Public sentiment ...
Environmental, Social, and Governance (ESG) data to understand a company’s operational and product related ESG risks and opportunities.
Estimated open positions of funds and investment banks on certain assets
Sales, P&L, active users, etc. estimated from social media, web traffic and credit card transaction data
Hires, churn, headcounts, etc.
FTSE buy/sell signals
- Derived from multiple data sources: alternative data, trading data, foundamental data, macro data …
- Multiple feature extractors + reinforcement learning, where feature extractors include: Regression, CNN, LSTM, RCNN, Transformer etc.
- Generates “BUY/SELL” signals on a daily basis, ranging from -4 (strong sell) to +4 (strong buy)
Based on simple buy/sell trading strategy
- Multiple predictors have been developed and back-tested using a vanilla trading strategy
- All predictors were shown to be able to generate 3x – 8x annualised return and Sharpe ratio vs FTSE index
- Have continued to outperform FTSE in real trading
We have market signals (e.g. FTSE and CSI) as well as signals for individual stocks and commodities.
At MADE by DATA, we use the latest and the most powerful machine learning technologies to solve the most difficult problem in the investment industry - to deliver alpha. Our natural language processing (NLP) capabilities allow us to digist vast amount of news, blogs, social media information and to extract the most intricate insight. We then use various deep learning and reinforcement learning techniques to infer and to predict.
Our founding team includes PhDs from top universities with 15yrs+ experience in machine learning and bankers from leading investment firms. Our clients include multiple UK banks, hedge funds and one of the biggest sovereign funds in the world.