Data Science for Financial Service: The Essential Financial Data
Ahmad Muzaffar B.
APAC Data Scientist & AI Specialist at IBM. GenAI | LLM | AI Governance | Data Lakehouse | BI Analytics
The major requirement to apply machine learning in any projects is the availability of data. In financial service, there are many types of data available to build model for specific service.
Depending on the nature of data and the service/product to be developed, we need to choose the right, accurate, consistent & unique data to be consumed in modeling.
Generally, there are 4 essential types of financial data (not limited to this list):
Fundamental data
- Assets
- Liabilities
- Sales
- Costs/earnings
- Macro variables
- Transaction
Market data
- Price/yield/implied volatility
- Volume
- Dividend/coupons
- Open interest
- Quotes/cancellations
- Aggressor side
Analytics data
- Analyst
- Recommendations
- Credit ratings
- Earnings expectations
- News sentiment
Alternative Data
- Satellite/CCTV images
- Google searches
- Twitter/chats (sentiment data)
- Metadata
The above data can be linked with each other, to uncover new insights and to drive decision or new products development.
Here are analytics & alternative data resources for finance:
American Economic Association (AEA)
- A good source to find US macroeconomic data.
- AssetMacro is a macroeconomic database that includes 25,000+ indicators for 120+ countries.
- Eurostat Comext includes datasets on trade flows since 1988, organized by commodity.
- The EU Open Data Portal gives access to open data published by EU institutions and agencies about the economy, as well as employment, science, environment, and education.
- Up to date information on financial markets from around the world, including stock price indexes, commodities and foreign exchange.
- Examine and analyze data on internet search activity and trending news stories around the world.
- The International Monetary Fund publishes data on international finances, debt rates, foreign exchange reserves, commodity prices and investments.
- A good source for economic and financial data – useful for building models to predict economic indicators or stock prices.
- Dataset about various aspects of Indian economy, banking and finance with different granularity.
- Datasets covering population demographics and a huge number of economic and development indicators from across the world.
- The CIA World Factbook includes economic stats of countries, as well as other stats on demographics, geography, communications, and military.
In my next post, I will share about real world cases and machine learning methods that can be applied in data science project, specifically for financial service.
Leave your comment below if you have new information to improve this article. Thanks!