That's right. These are all lumped together as regular assets and, in economics mode, they are assigned the rate of return you give them in Assumptions. The categories there are just designed to trigger your memory. In Assumptions (Taxes tab) you can indicate what percentage of this regular asset income is taxed at the capital gains rate.

For example, a user might have a brokerage account with positions in intermediate term bonds, s&p index fund, small caps, and some global growth funds or something. Anyway, the user presumably has a target or benchmark for nominal return and should enter that nominal return at 5% or 3% or 8% or whatever he or she expects to receive on that asset allocation.

My regular assets are primarily in intermediate term bond funds with a bit of S&P index fund, and the expectation for nominal return is 2% or 3%.

You can use the Monte Carlo method and enter your actual positions (your asset classes or you can create a custom asset class with your actual fund or stock position) and the program will pull from that to show the return. But this approach has the downside of calculating returns based on a somewhat opaque history that may not be anything like the future. Consider a person that has just one asset class, say intermediate term bonds. The compound annual growth rate on that asset class since 1991 has been 6.08% nominal. If I enter that return history in the Monte Carlo mode or create a custom asset class with that return and variance history it will assume a return something like that depending on how long of a history it uses. Given that the yield on intermediate term bond is 2.2% right now, it seems misguided to me to let the program assume 6% return. I'd rather control it myself and assign 2% or 3% return.

Since 1989, the S&P index has an annual return of 9.6%. That seems quite optimistic going forward and I'd prefer to adjust these down using my own conservative bias. I know that I'm going beyond your question here, sorry, but I thought I'd just throw this out as something to think about. I much prefer entering my own assumed nominal rate of return and running several models where I assume the worst and assume the best to see the impact on the discretionary spending.