In order to preserve the maximum amount of accuracy in mathematical calculations, if you're working with decimal values to begin with, I'd always start with floating point numbers.

So, firstly, let's make sure that your variables are type float. Helpfully, these are defined as measure in the variable declaration.

If they're not measure/float, let's convert them to it.
By type:
&myFloatVar := strtofloat(&myStrVar)

&myFloatVar := inttofloat(&myIntVar)

&myFloatVar := curtofloat(&myCurVar)

Then, we can do the calculations just as you would any other:
&myNewFloatVar := &myFloatVar * 7.5

NB: We don't need to divide by 1000 as 7500/1000 is 7.5

Obviously if the number isn't going to be as clean as 1000 each time, then you can do it as another line:
&myNewFloatVar := &myNewFloatVar / 1000

Then if you need to present that value back to the original variable or to some other, just reverse the conversion steps above. So for string - for example:
&myStrVar := floattostr(&myNewFloatVar)

Hope this helps.