Multiple Lineair Regression - Interaction Terms

Hi all,

At the moment I have a multiple regression with 2 variables and and interaction term.

y = exchange rate changes
x1 = macroeconomic news surprises
x2 = speculative net positions
x3 = macroeconomic news surprises*speculative net positions

My question is: Do net positions affect the way exchange rates react to macroeconomic news ??

The regression to run will be:

y = B0 + B1*x1 + B2*x2 + B3*x1*x2

My question now is, what will be the right hypothesis to answer my question.

I was thinking about testing: Ha: B3 = 0 and H0: B3 =/ 0

I'm not sure about this..

I hope somebody can help me with my model and how to test for my question.