Base Metal/Stock Price Forecasting from Alternate Data Sources Using AI

Project Reference :


Institution :

National University of Singapore (NUS)

Principal Investigator :

Dr Chua Tat Seng

Technology Readiness :

6 (Technology demonstrated in relevant environment)

Technology Categories :

AI- Machine Learning

Background/Problem Statement

The future trend of base metal prices is affected by a wide range of factors such as supply and demand, and macroeconomic conditions. This information is very important for governments as well as industries.

The spot prices of base metals, which are used for constructing infrastructure and various types of products, can be incredibly difficult to predict. The prices are stochastic and highly non-linear as they are influenced by many factors with complex relationships. They are also affected by macro-economic situations such as currency exchange rates and government policy changes, such as increase in import tax rates.

With greater market volatility and uncertainty, and more data sources available to train machine learning and deep learning models, AI will play a big role in commodity price prediction in the future.


An AI model for prediction of base metal prices that is trained using large-scale alternate data sources such as market data, macroeconomic data, third party estimations, supply and demand data, and relevant social media information.

The machine learning framework consists of three components:

  1. Classification of the direction of price movement using an ensemble of machine learning and deep learning models
  2. Regression of price predictions using a deep learning model
  3. Filter mechanism to align the outputs from 1 and 2 above

The solution also includes new indicators to extract intelligence from alternate data sources relevant to base metal trading such as news and analyst reports to derive short term price movement indicators.

The solution outperforms traditional linear model benchmarks in its accuracy for out-of-sample forecasts for one, tree, five and ten-day horizons.

(Enhancing stock movement prediction with adversarial training)

(Graph Adversarial Training: Dynamically Regularizing Based on Graph Structure)


Monitoring commodity price forecasting and trends play a critical role in the strategic plan of procurement teams and organisations. It empowers them to make data-driven planning and decisions, forecast pricing-related risks, and proactively manage suppliers while minimising disruption in the supply chain as a result of price volatility.

Potential Application(s)

  • Improving the quality of prediction of base metal prices is profitable for the base metal trading market. The solution can be used as a service to traders of base metal commodities and their associated derivatives.
  • Implementation of this technology in the financial world will help financial institutions, business owners as well as government agencies in Singapore manage their risk exposure and strengthen and develop Singapore as a leading fintech hub in the world.


We welcome interest from the industry for collaboration/ co-development / customisation of the technology into a new product or service. If you have any enquiries or are keen to collaborate, please contact us.