-
50%
Cost savings
-
~10
currency pairs covered for fair value
CLIENT CHALLENGES
- A leading European bank wanted to develop a fair value model for major FX pairs to check whether any currency pair was undervalued or overvalued compared to its long-run fair value.
- The bank also wanted to develop a quantitative method to predict how long it would take for the actual FX price to converge to its fair value.
OUR APPROACH
- Preparation
- Data collection from public sources and other data subscriptions
- Data cleaning for model input
- Data manipulation and transformation for model development
- Data maintenance for future iterations and model validation
- Model fitting
- Used the BEER approach
- Identified four significant predictors for FX rates
- Used panel regression for G10 currencies
- Used Vector Error Correction Model (VECM) to know the speed of convergence to fair value
- Used R and Python for model development
- Result interpretation and publishing
- Helped the client publish the findings of the model in research papers and other publications
- Maintained the model to keep it up-to-date and in line with econometric assumptions
- Tested tweaks in the model to improve forecasting accuracy
IMPACT DELIVERED
- Set up a stable and robust model to estimate the fair value of the FX pairs and provide a direction for FX price movements in the medium to long term
- Provided skilled manpower at a fraction of global costs
Thank you for sharing your details
Your file will start downloading automatically
If it does not download within 1 minute,
Share this on
