Maths Research Associates 2021 Brunel
Lead Research Organisation:
Brunel University London
Department Name: Mathematics
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Organisations
Publications
| Description | The research funded by this grant proposed a new method for forecasting of correlated exchange rates, using state space models along with the Kalman filter. The research was motivated by the idea that the dollar exchange rates for currencies of developing countries within the same geographic region are affected by the same fundamentals and we should be able to predict each of them out of sample more accurately if we model their evolution jointly rather than individually. This modelling (in terms of both model calibration and exchange rate forecasting) is made tractable through the use of state space formalism and the linear Kalman filter. Through extensive numerical experiments over seven different currency pairs, we successfully established superior performance of the new method in comparison with the existing methods in the literature. We also tested the use of 30 day forward rates in terms of their use in short term exchange rate prediction using the extended Kalman filter and found that the forwards mostly worsen the day-ahead prediction of exchange rates, in keeping with what is reported in some of the academic literature. |
| Exploitation Route | In the academic literature, correlated exchange rate prediction in the format proposed by us has not been reported before. Our method of predicting rates using the linear Kalman filter beats existing benchmark methods by a significant margin in terms of accuracy, on all the data sets used. Accurate prediction of exchange rates have a wide variety of applications within investment banking, currency trading and monetary policy. We hope to build on this work and disseminate it across end user community once we have an accepted publication. |
| Sectors | Financial Services, and Management Consultancy |
| Description | Conference presentation |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Policymakers/politicians |
| Results and Impact | The funded postdoctoral researcher Dr Maunthrooa presented her research at OR64 academic conference in September 2022. Her presentation, titled 'Modelling and forecasting of exchange rate pairs using the Kalman filter' in the financial modelling stream was attended by around 10 delegates, including one policymaker from the UK and one economist from the Federal Reserve. The delegates found the results extremely impressive. One delegate asked Dr Maunthrooa why she wasn't already extremely rich, given that our method can predict exchange rate movements well. |
| Year(s) Of Engagement Activity | 2022 |