The Price/Earnings (P/E) ratio is probably the most overused statistic in financial media. And perhaps one of the least useful.

One reason P/E is so widely used is that it's relatively easy to calculate. It also offers a (seemingly) simple way to compare market valuations across time. That's why it's constantly being pulled out as support in "the market is under/over-valued" arguments.

As we've written before, P/E ratios can be helpful as very broad guides. If the market's P/E is truly at extremes, it can serve as a helpful counter-weight to the prevailing sentiment at the time, which is likely also at an extreme level (good or bad).

The problem is most of the time P/E ratios aren't at extremes. Instead, they typically occupy the very broad middle ground between extreme highs and extreme lows. And when they're sitting in this broad middle ground, they aren't very helpful at all in predicting what the market is likely to do next.

If you look at a chart of the S&P 500's P/E ratio over the last 75 years, you'll find that it has covered quite a range. Looking at such a chart, it would be clear that market P/E ratios over 25 are pretty high and anything under 10 is pretty low. Those are the somewhat helpful extremes I mentioned a moment ago.

Unfortunately, that's about all that's clear. Even within a relatively simple data series like this there's plenty of room for interpretation. For example, the average P/E over the last 75 years is 15.35. But over the last 25 years, it's been 18.90. Is the difference between those levels due to legitimate structural changes like 401(k)s taking over the retirement savings system? Or is it merely that stocks have been relatively overpriced for a while now? It's impossible to say for sure.

The big takeaway here is not to be overly persuaded by arguments based on P/E ratios. For one thing, there are a bunch of different ways to calculate them (based on forward earnings estimates, actual trailing earnings, 10-yr smoothed earnings, etc.). But even leaving that out of the equation, it's relatively easy to get P/E ratios to say whatever message someone wants to convey.

Consider this.

If I want to paint a bullish picture of the market's current potential, I could reasonably claim that if the market simply were to rise back to its average P/E of the past 25 years, based on the latest S&P 500 earnings estimate for June ($111.78), that would imply a rise in the index to 2,112. That's a quick 9% gain. Very reasonable, right?

But if I want to paint a bearish picture, it's just as easy. Simply point out that based on the latest actual quarterly earnings (March's $108.85), the market is currently trading nearly 14% over its long-term average P/E. Scary!

Neither of these even deal with the idea of whether the market is justified in being over or under the long-term averages. I've just used plausible "baseline" figures to create whichever narrative I want. And that's exactly what many in the financial media do in order to tell their stories and either get you excited or fearful (depending on what they want to sell you).

P/E ratios have their uses, but predicting where the broad market is heading next isn't among them. Hopefully this will help you tune out this common, but mostly unhelpful, financial propaganda tool.