I read an emperical study that:
"echange rate change"= coefficient*(Interest rate country1-interest rate country 2) + coeffecient*orderflow + error term
This formula had 64% explanatory power when they made a regression. Thats awesome. They were looking at customer and interdealer order flow, from reauter 2000 data. And showed remarkable results as well when trying to forecast prices,where forecasting power decreased with time horizons, as fundamentals became more important. Interest rate differentials is of importance, but when taken alone, it had only 0.01 explanatory power.....
So what are the practical steps?
"echange rate change"= coefficient*(Interest rate country1-interest rate country 2) + coeffecient*orderflow + error term
This formula had 64% explanatory power when they made a regression. Thats awesome. They were looking at customer and interdealer order flow, from reauter 2000 data. And showed remarkable results as well when trying to forecast prices,where forecasting power decreased with time horizons, as fundamentals became more important. Interest rate differentials is of importance, but when taken alone, it had only 0.01 explanatory power.....
So what are the practical steps?