Very good piece.

This is a good framework to employ when thinking about Buffett / Munger's strategy of compounding "winners" over time. I.e. investing in company's that produce returns that are in excess of their cost of capital over the long run, on the basis of a sustainable competitive advantage or "moat" to enable compounding, and buying such companies at a "fair" price or, when Mr. Market affords the opportunity, at a discount to estimated intrinsic value.

This is a piece of the jigsaw which shows the power of compounding. A potential third analysis above could be a "no dividend / no sell-off" scenario, where all the profits are reinvested at the same rate of ROE (i.e. return on incremental invested capital is equivalent to that of the historical profile - difficult to achieve without a moat), which would show even greater returns (especially if taxes are taken into account in the three scenarios) of compounding at a high ROE.

There are some interesting fundamental multiples which are not widely used, the inputs for which include ROE, ke, % of incremental invested capital which returns excess returns (i.e. "70%" would achieve the 12% ROE in your example, while the remaining 30% would achieve Ke) and the time (in years) of that outperformance (driven by how strong the competitve advanatge is maintained). It's an interesting may of highlighting the power of compounding high ROE/ROIC businesses with moats.

Best regards