Screening public equities

A screen is a filter applied to a large universe of companies to produce a short list worth reading about. It saves time and it introduces bias, and both effects should be understood before it is used.

Build the screen around the thesis

Screening on low valuation multiples alone returns companies that are cheap for a reason. A screen is more useful when it encodes the thesis: stable gross margin, returns on capital above the cost of capital, moderate leverage and revenue that has not been bought with acquisitions.

• Return on invested capital over several years

• Gross margin stability rather than level

• Net debt to EBITDA

• Free cash flow conversion

• Share count over time

Survivorship and restatement

Historic screens run on current index members quietly remove the companies that failed. Financial data is also restated, so a backtest can use numbers that were not available at the time. Both make a screen look better than it was.

What comes after

The screen ends where research starts. The output is a list of candidates, and each one still needs the filings read, the competitive position understood and the price compared with what the business can earn.

Screens are reviewed each term. Where a screen produced a company that later failed for reasons the criteria should have caught, the criteria change.