Why does Purchasing Power Parity hold true for Long Time Horizon Returns but only for Developing Countries?

[Anon9001]

Purchasing Power Parity Theory states that the Long Term Currency Return will be equivalent to the Long Term Inflation Differential between the Two Countries. I am wondering though why this Theory seems to be only valid for Developing Countries. I produced a scatter plot in Excel showing the Correlation between the Long Term Inflation Differential of the Country Relative to USA and the Long Term Currency Return of USD Relative to that Country's Currency for both Developed and Developing Countries and the Correlation is only significant for Developing Countries.

Due note that the R-Squared in the Scatter Plot of Developed Countries its inflated due to presence of outliers in the data. The Spearman Correlation is quite a bit lower than the Pearson Correlation Coefficient of the Developed Countries and is not Statistically Significant as a result (P-Value is >5%). I calculated Spearman Correlations for both the Developed and Developing Countries and only for Developed Countries the Spearman Correlation is much lower than the Pearson Correlation Coefficient whereas for Developing Countries the Spearman Correlation is quite close to the Pearson Correlation Coefficient (Both are Statisically Significant also at the 5% Level for Developing Countries).




My theory is that the difference is due to the fact that many Developing Countries including the one I am living in have a de-facto peg to USD whereas Developed Countries have Market Determined Exchange Rate. Considering Market Exchange Rates have higher volatility than Pegged Exchange Rates this can create bigger deviations from Purchasing Power Parity even over long time horizons but that is just my theory on the matter I cant exactly explain why there is a huge difference here.

Here is link to download the Excel Sheet where I made the scatter plot and calculated the Pearson and Spearman Correlations of the Data.

Thanks.