Great discussion, gentlemen! I'm kinda late to the party (just back in town), but if I might add a couple of pennies' worth...
Using the numbers from J Scott's post, we've got the property purchased for 100K originally, and appreciating at 5% pa for 30 y, to 432K. Loan payments plus the 20K down pmt bring our total cost up to 192,800.
With that fact set, our overall ROI is (432,000 - 192,800) / 192,800 = 124.1%. Equivalently, it's the 224% calculated previously, less 1. The previous calculation didn't first deduct the cost. It's like investing 100 today, and cashing out for 110 in one year. It's true that 110 / 100 = 110%, but clearly the ROI was 10%.
Since most investments assume at least an annual compounding, we should do likewise in "annualizing" our ROI, to give us a decent apples-to-apples. And further, as J Scott points out, only 20K of the cost is paid up front, with the remainder spread linearly over 30 years.
Running those numbers, the annual ROI on this hypothetical comes in at 4.4%.
Instinct immediately says something's wrong...if the property's appreciating at 5% a year, how can the overall ROI be less? The fly in the ointment is the financing. We're assuming a cost of money at 6%, against a property that's appreciating at 5%. Borrowing 80K at 6%, and putting that cash into an investment that grows at 5% has a slight negative return effect on the overall results. Had we simply paid 100K cash up front, no financing, our ROI would have been precisely 5% per year.
But of course, this is relying solely on the appreciation as our return source. Throw in the rent income and it's a whole 'nuther game. And of course, this is all pre-tax stuff.
I'd finally add an agreement that nearly all asset classes are quoted at nominal (pre-inflation), rather than real (post-inflation) returns.
Thanks for listening, guys. Cheers!