Finding stocks to buy is a separate skill from analysing them. Most investors start with a screener, which returns companies matching criteria you already chose, so it can only surface what you already knew to look for. The better sources produce candidates a screen would never return.
What a screener can and cannot give you
A screener applies your filters to a database. That makes it excellent at finding companies matching a hypothesis you already hold and incapable of generating the hypothesis. Everyone running the same popular filters gets the same list, which is the opposite of an edge.
It remains useful, and it belongs second rather than first. The five sources below produce candidates worth screening, in roughly ascending order of how much work they take and how well they repay it.
The five sources
Start from what you already understand
The cheapest edge available to an individual is knowing an industry from the inside. If you work in logistics, healthcare or software, you already understand who is winning and why in a way no screener reports. Peter Lynch's version of this is overquoted and still correct: the advantage is not the tip, it is the context you have that the market does not price efficiently.
Screen for quality, not for cheapness
A filter for low P/E reliably surfaces companies whose earnings are about to fall. Screen instead for the characteristics of a good business: ROIC above 15% sustained, stable or expanding gross margin, positive and growing free cash flow, and a share count that is not rising. Apply valuation afterwards, as a test rather than a search.
Read the supply chain of a company you already like
Every business you admire buys from someone and sells to someone. Those suppliers and customers are often less followed, and the reasons the original company is good frequently apply to them. Segment disclosure in a 10-K names the largest customers, which is a free list of candidates nobody screened for.
Watch what changes, not what is
A screener describes the present. Most opportunity sits in transitions: a company crossing into operating leverage, a first year of positive free cash flow, a moat widening. Screening for the change rather than the level surfaces companies before the market has finished repricing them.
Keep a watchlist, and let price come to you
Most good businesses are not attractively priced most of the time. The discipline is to research first, decide what you would pay, write it down, and wait. A list of twenty companies you understand with a price attached to each is worth more than a screen you run when you happen to have cash.
A quality screen that is actually worth running
Four filters, chosen so that passing all of them is difficult and meaningful.
ROIC > 15% for five consecutive years
Gross margin flat or rising over five years
Free cash flow positive in each of the last three years
Diluted share count flat or falling over five years
This returns a short list, and almost everything on it will be a genuinely good business. Notice that none of the four mentions price. Valuation comes after you have decided the company is worth owning, which is the sequence that stops the price from writing your thesis.
Where ideas come from that you should distrust
- Anything trending. By the time a company is widely discussed, the easy information advantage is gone and the price reflects the enthusiasm.
- Institutional filings. 13F disclosures arrive up to four and a half months after the fact, exclude shorts, and tell you nothing about cost basis or conviction. A source of names, not a signal.
- A recommendation you did not derive. The problem is not that it might be wrong. It is that you cannot evaluate it when it stops working, so you will hold too long or sell at the bottom.
What to do with a candidate
Finding is the easy half. Every name from every source above still needs the same work: understand the business, establish its lifecycle phase, test the moat, judge management, then value it. That is the seven-step framework, and the ordering matters more than the sources do.
Turn a candidate into a decision
Finding a name is the easy half. Phase Check reads any US-listed company's financials, places it on the lifecycle and names the metric that decides it, which is the fastest way to tell whether a candidate is worth the full analysis.
Try Phase Check freeNo credit card. One check without an account, unlimited with a free one.
Frequently asked questions
Related
- How to analyze a stock, what to do with every candidate you find
- Best stock screeners, the tools, ranked honestly
- Economic moat, what quality screening is really looking for
- How many stocks should you own, how big the list should get