I do not believe that foreign capital investment (at least in the short-run) would directly increase the domestic money supply.
When foreign investors purchase domestic currency, the investors merely transfer that currency from domestic accounts to their own accounts. Then, when the investors purchase capital in the domestic market, the purchased currency is transferred back to the domestic sellers. Thus I do not see how net circulation of domestic currency rises in the event of a capital inflow directly.
There are, of course, more complicated indirect effects such as domestic banks issuing an increased number of loans based off new demand deposits that may increase the money supply, but the direct effect that I believe you're alluding to isn't there. As you say, the domestic money supply cannot be increased without banks playing a role in fueling domestic credit.