If a valuation ratio such as the earnings yield (Earnings / Price) is non-stationary tested with ADF, can the inverted price-earnings ratio (Price / Earnings) be stationary?
I use monthly data on valuation ratios of the S&P 500, and found that the E/P is non-stationary, while the P/E is stationary, and wonder how this can be the case, I couldn't find anything wrong with the data, and the ratios use the same data to be calculated. I find it strange for the same data to have different time-series properties, but is this due to the relationship being unequal, or some other factor I am unaware of?
Thank you.
I use monthly data on valuation ratios of the S&P 500, and found that the E/P is non-stationary, while the P/E is stationary, and wonder how this can be the case, I couldn't find anything wrong with the data, and the ratios use the same data to be calculated. I find it strange for the same data to have different time-series properties, but is this due to the relationship being unequal, or some other factor I am unaware of?
Thank you.
