How can we tell if the U.S. stock market is getting too pricey? Well, if you watch CNBC long enough or read enough stories in the financial media, you are likely to learn dozens of ways people will try and answer that question. There is not one right answer. If there was, successful investing would be easy and it is far from it.

I decided to dig into the numbers and present one way we can evaluate the stock market at today's levels relative to prior market peaks, in order to see if we are nearing a point where we should start to get worried. I chose five of the most noteworthy market peaks over the last 25 years or so. After each of these peaks, the S&P 500 index fell at least 20% peak-to-trough. Some of the corrections were relatively normal, mild bear markets (the 1990 recession; -20% and the 1998 Asian financial crisis; -22%). Others were more pronounced (the 1987 crash; -33% and the dot-com and housing bubble bursts of 2000 and 2007; -50% and -58%, respectively).

I have graphed the P/E ratio of the S&P 500 index at each of these five market peaks. At one extreme we have the 1990's bull market led by internet stocks, which saw equity valuations easily reach record-high levels, but at other peaks the results are more uniform, with markets typically topping out with P/E ratios in the high teens or low 20's.

As you can see, today the S&P 500 sits at 16x earnings. While we are approaching levels that should be considered elevated, one can argue that another 10-15% upside in P/E ratios would not be out of line with historical data. That said, making a large bet that valuations will reach the high end of the historical range is not something I would take to the bank. To me, this data says that the market is starting to get pricey, and although we could very well squeeze more upside out of this bull market (largely because with interest rates so low, equity investors are willing to pay more for stocks), I would still be cautious. As a contrarian, ever-higher stock prices only increase my preference to raise more cash and wait for the next correction, even if we don't know exactly when it will come.