3 ms·
You're entirely correct about diminishing returns & difficulty attempting to drive expenses close to zero, but note that the author of the post was saying that
by shoo 6y ago
You're entirely correct about diminishing returns & difficulty attempting to drive expenses close to zero, but note that the author of the post was saying that it isn't necessary for expenses to go to zero. Expenses just need to be sufficiently small (but not arbitrarily small) compared to the percentage of total savings -- I'll interpret "total savings" as the accumulated capital that we are able to invest to generate a return.
As you approach a certain critical ratio of expenses / invested capital your runway of how long you can live off investment returns & capital goes to infinity. You can approach that ratio in two ways: by reducing expenses or increasing the amount of invested capital.
E.g. suppose you have expenses of X / year. Suppose you have accumulated capital worth K that generates a real investment return of K * r every year. If K * r >= X then you can operate indefinitely with investment returns covering all expenses, i.e. the "infinite runway". We haven't talked about the expenses X being zero or frugal. Your expenses X could be lavish, but they just need to be no greater in size than K * r .
A plausible value of r corresponding to capital invested in the stock market with a very small allocation of bonds might be 0.035 -- 0.04 for a 3.5% ... 4% annual real return. So if your annual expenses are X then to cross the threshold into infinite runway you need to have capital invested worth at least K >= K* = X / r , that is K >= X / 0.04 ... X / 0.035 = 25X ... 28.6X .