Fisher: 'Bubbling stock markets? Price/earnings ratio is not an infallible compass'

Many investors regard the predictive power of price/earnings (P/E) ratios and other valuation parameters as gold. This is why the current high levels of global P/Es are frightening, presumably also threatening 'undervalued' Italian equities.

Valuations, however, are not indicative of the direction of the lists. They never have been! History shows that a high P/E, in itself, means nothing.

These are not my predictions for 2026, which I will publish soon, but the ratings will certainly not influence my opinions. Does this sound crazy? Observers often extol parameters such as P/E and Price-to-Sales Ratio (PSR, which I devised more than 40 years ago) as timing tools. This conception stems from the belief that low P/E ratios are indicative of 'undervalued' stocks, thus inviting 'buy low', while high P/E ratios indicate effervescence and weak returns in the future, thus 'sell high' is advisable. These are the considerations behind the current fears.

However, this logic seems unfounded. Let us take the MSCI Italy index and a very good statistical indicator called 'R-squared' (R-squared or R2), which indicates the extent to which one recurring phenomenon can explain another.

The annual starting P/E of Italian stocks (taking the previous 12 months' earnings as a reference) and future one-year returns from 2000 onwards have an R2 coefficient of 0.03. To clarify, a figure of zero indicates no potential causality, while 1.00 indicates complete causality - in other words, this means that as little as 3% of Italian returns could be derived from P/Es. A derisory percentage. What is the R2 of three- and five-year yields? At 0.01 in both cases. P/Es are responsible for just 1% of these returns. Irrelevant.