Talk:Relative purchasing power parity

It needs to be more developed.

Log math is incorrect in derivation
1. The first approximate equal sign should actually be an exact sign, as we just take the log of both sides of equations, because according to the text "lowercase letters denote natural logarithms of the original variables." So, as log(S2) = s2 (and similarly for other variables), the math should be exactly equal: s2 - s1 = (p2 - p1) - (q2 - q1)

2. Then PI = (P2 - P1) / P1 = P2/P1 - 1, with using x-1 ~= log(x), PI ~= log(P2/P1) = p2 - p1, so big PI, not small pi, is approximately equal p2 - p1. Another way to check is: given small inflation, log(P2/P1) should be approximately zero, as P2/P1 is close to 1. But in current incorrect statement, pi = log(PI) = log(P2/P1), so PI is ~100%???

3. Overall, the math could have been written much simple without introduction of lowercase variables: S2/S1 = (P2/P1) / (Q2/Q1) => log(S2/S1) = log(P2/P1) - log(Q2/Q1) => [Using log(x) ~= x - 1 for all 3 fractions] S2/S1 - 1 ~= (P2/P1-1) - (Q2/Q1-1) => (S2-S1)/S1 === relative change in S ~= PI_A - PI_B 98.237.205.116 (talk) 02:20, 28 January 2023 (UTC)