How to find undervalued stocks

It is the most common beginner method and it is close to a screen for distress. The cheapest companies on any multiple are usually cheap because the market has correctly worked out that earnings are about to fall. For a cyclical business the signal inverts entirely: the P/E looks lowest at the top of the cycle.

Start with return on invested capital and how durable it has been. A business earning 20% on capital for a decade while rivals pile in has something structural protecting it. That is the population worth valuing. Everything else is a value trap waiting for a buyer.

In SteadyShares this is the Discover screen: filter by moat score and returns first, add a debt ceiling, and stop at a shortlist you can actually read in an evening.

This is the step almost everyone skips. Rather than asking a model what a company is worth, feed in today's price and solve for the growth rate that would justify it. Now you have one checkable claim: the market thinks this compounds at 14% for a decade.

You can have an informed opinion about that. You cannot really have an opinion about a fair value of 62.40.

Markets are mostly efficient. If you think you have found a bargain, there must be a reason nobody else has taken it: the company is too small for institutions, the good news is two years out, or the market is extrapolating a temporary problem into a permanent one.

If you cannot name the reason, you have probably not found a bargain. You have found something you do not yet understand.

One sentence on why it makes money. One paragraph on why it is mispriced. Three specific things that would prove you wrong. Undated conviction is just a feeling, and feelings sell at the bottom.