>David,
>I appreciate your reply but please note that the decimal value in my example was a numerical and not a currency value. What I thougth would happen is that the currency value would be multiplied by the numerical and the result converted to a currency. What is happening is the numerical is being converted to a currency and then multiplied. Is this planned???

I would expect it to be; the currency data type sets the precision limit of the computation. I'd expect that I'd have to force a currency value into a numeric variable to force the computation to be done with floating point precision and then explcitly round after the computation to get the effect you describe.