The author then recalls the previous section's interpretation of the capital value of a stream of receipts as the weighted average period of the receipts (weighted by discounted values of receipts), and the comparison of this average period of receipts with the period of a standard stream of receipts to test whether a rise in the interest rate would increase or decrease income. The author asks whether this test can be reinterpreted so as to agree with the common-sense case described above.
The author gives his answer for the case in which prices and interest rates are expected to remain constant. In this case all three approximations of income give the same results, with the standard stream of receipts having the same constant value in all periods. If the average period of the given stream of receipts is greater than that of the standard stream, then the given stream must have lower value initially. But because the two streams must have the same capitalized value, the given stream must catch up later. (In the language of the text, there must a "crescendo.") The author concludes that "The average period turns out to be nothing else but an exact method of measuring the crescendo (or diminuendo) of a stream of values." In the case of a stream of identical quantities, continuing indefinitely, and "discounted throughout by the same rate of interest" the author shows that the average period works out to be the reciprocal of the rate of interest, the calculations being as follows:
Finally, the author gives a formula for the crescendo of a stream of values, with each period's value expanding by the same proportion. The formula for the crescendo c is
c = i – 1 / P
where i is the interest rate, and P is the average period of the stream.