Thanks, that was my gut instinct, but I couldn’t validate the logic in my head because I always viewed CAGR as a computation of total return — since you’re not earning any interest on the dividends.
Playing around with it, I think it has less to do with time-value, as opposed to where you’re money-weighting the dividends. For example, if I take the sum of the dividends received ($0.90) and applied to any single-year, theoretically, receiving the dividend in the earlier years should be worth more than receiving them in the later years, due to TVM. Using this concept, I would also expect the CAGR formula presented above to yield the same result as back-weighting all dividends to year 5 and calculating the geometric mean (which it does).
If, however, the stock price showed a loss as opposed to a gain, the TVM of the dividends should still be worth more in the earlier years than the later years. I changed the stock price to start at $37.00 in year 0 and fall by $1.00 every year, while keeping the dividend the same. The return of the stock is the same regardless of when you receive the dividend, however, if I weight all $0.90 of the dividend in year 1, I get a geometric mean of -2.38%, whereas if I weight it all in year 5, I get a higher geometric mean of -2.32%. Under TVM, I would expect receiving a positive cash inflow earlier on would raise the rate of return (unless, of course, there’s a negative discount rate).
I think I answered my own question as I was typing this out…