How do I know if a stock is overvalued?

There is no single number that settles it, but there is a fast first pass. Take the P/E ratio, price divided by earnings per share, and compare it to two things: the company’s own P/E over the last several years, and the P/E of a few similar companies. If it is trading at 45 times earnings while its peers trade at 20 and its own history is 22, the market is pricing in something extraordinary.

Then ask whether the extraordinary thing is plausible. A high P/E is justified if earnings are growing fast; a company doubling profits every three years can reasonably trade at 40 times today’s earnings. The PEG ratio, P/E divided by growth rate, is a rough way to fold that in: near 1 is fair, well above 2 is expensive.

For companies with no profits, P/E does not work, and people use price-to-sales instead. The same logic applies: compare to peers and history, and ask what has to go right.

The honest caveat is that "overvalued" stocks can stay overvalued for years, and "cheap" ones can get cheaper. Valuation tells you the odds, not the timing. It is most useful as a reason not to buy something at a silly price, and less useful as a reason to bet against it.