A leading global investment bank’s macro strategy research team, renowned for its top-down approach to forecasting equity indices, was facing a challenge—limited ability to scale coverage across European indices. The reason? Its dependence on large and resource-intensive traditional forecasting models, which also constrained the team’s capacity to provide comprehensive research to its clients
It engaged us to expand the coverage to include global indices and major stocks, and enhance the value proposition of its research services
Crisil Integral IQ solution
We employed a Principal Component Analysis-based framework on Python that distilled a wide universe of equity indices into a few key components
Each component was rigorously back-tested against macroeconomic drivers, identifying patterns with relevant statistics and enabling forecasts with high confidence levels, based on house views for the underlying drivers
We used these components in reconstructing the respective indices to complete with forward-looking predictions
This approach improved model efficiency and helped identify new macroeconomic drivers not used in the legacy models. This, in turn, increased overall model quality
Efficient python coding ensured forecasts for the entire suite of coverage indices were generated in minutes, significantly streamlining the research process
This modelling approach was immediately adopted by the bank and the research team integrated the underlying analysis and research into its publications
Client impact
The efficient model framework helped enhance coverage of equity indices by 20%
The strategy team could initiate top-down forecasts for all stocks under coverage enabling new research avenues
Time spent on model maintenance reduced by 30%, allowing the team to sharpen focus on client communication and marketing
The high-quality forecasts sparked increased interest from investor clients of the Bank