The Stock Screening Playbook: Finding Quality Companies Before They Become Consensus Picks

Investor Advantage Playbook Series | Part 5

Here is the uncomfortable pattern behind most retail stock discovery: a name shows up on a “most active” list, a subreddit, a finance app’s push notification, or a colleague’s dinner-table tip, and only then does the research begin. By construction, that process finds stocks after the price has already moved enough to attract attention. You are not discovering an opportunity. You are doing due diligence on a decision the crowd already made for you, three weeks and twenty per cent ago.

A screening system fixes the sequencing problem. It does not predict the future, and it does not guarantee you will find the next great compounder before anyone else does. Nothing does that reliably, and anyone selling you otherwise is selling you something. What it does is replace “I heard about this stock” with “this stock survived a filter I built before I knew its name.” That is a different, more repeatable way to arrive at a shortlist.

The claim of this playbook is narrower than the usual retail version of it. A screen built on multi-year business quality will not systematically surface undiscovered companies before they are consensus picks. Early discovery is partly luck, coverage, and a variant view. What this process can do is find businesses that have already demonstrated durable economics, and put them in front of you before a narrative, a “most active” list, or a wall of buy ratings has done the selecting for you. Some of those names will already be well known. The edge is not obscurity. The edge is that you arrived by a test you wrote down in advance.

This piece is that test: the metrics that separate genuine business quality from things that only look like it, the retail habits that quietly sabotage the process, a five-step system you can run on a schedule, and two teaching cases so the steps are not left as slogans.

What a Screen Is Actually For

A screen is not a verdict. It is a triage tool — a way of taking a universe of several thousand listed companies and mechanically discarding the ones that do not clear a set of quantitative hurdles, so that the qualitative work that actually decides whether you would own something gets spent on twenty names instead of two thousand.

That distinction sets the bar for what a screen needs to do well. It does not need to be clever. It needs to be consistent, and it needs to filter on the handful of things that have actually been shown to correlate with durable business quality — not the things that are easiest to sort a spreadsheet by. Most retail screening tools default to the latter: biggest single-day mover, highest dividend yield, lowest price-to-earnings ratio. Those are the columns that are easiest to click “sort” on. They are not the columns that tell you anything reliable about the business underneath.

A screen also has a selection bias, and naming it is part of using it properly. A filter that demands five consecutive years of high returns on capital will systematically miss turnarounds, cyclicals at the trough, and younger compounders whose economics are still being established. A gate that also demands positive and growing free cash flow, converted at a high ratio, misses a second group: capital-intensive compounders that reinvest heavily and still earn high returns on the capital they employ. Some of the best subsequent returns live in that excluded set. That is an acceptable trade only if you know you are making it. This playbook is optimised for durability, not for early discovery and not for catching last cycle’s winners on the way down. If you want a turnaround screen or a deep-value screen, build a different one. Do not quietly retune this one until it would have caught whatever just worked.

Part One: The Business Quality Filters That Actually Matter

Return on invested capital, held against a disclosed hurdle, sustained across a cycle. If this playbook has one metric to anchor on, it is this one. Return on invested capital — operating profit after tax, divided by the debt and equity capital actually invested in the business — tells you how efficiently a company turns capital into profit.

The textbook comparison is against the company’s weighted average cost of capital. Conceptually that is right: a business earning 8 per cent ROIC against a 9 per cent cost of capital is destroying value every year it operates, no matter how fast revenue is growing. A business earning 25 per cent against an 8 per cent cost of capital has a genuine engine, and the wider that gap, the more durable it tends to be.

The practical problem is that cost of capital is not a known input. Two reasonable estimates of the same company’s WACC can differ by several percentage points, depending on the risk-free rate, the beta, and the capital structure assumed. A 14 per cent ROIC is a clear pass against an 8 per cent hurdle and a borderline case against a 12 per cent one. Treating the denominator as if it were a fact produces false precision.

For a retail screen, use a disclosed proxy and say so. A workable starting hurdle is a fixed 10 per cent after-tax ROIC, held for at least five consecutive years, plus a sector check: the result should also sit clearly above the median of the industry you are comparing it with. A 15 per cent ROIC is exceptional for a regulated utility and ordinary for asset-light software. Absolute thresholds without that relative check will either flood the shortlist with software or empty it of everything else. If you later build a proper cost-of-capital estimate for a name that survives, use it in the underwrite. Do not pretend the first-pass screen already did.

The single-year snapshot matters less than the multi-year pattern. A company posting ROIC comfortably above the hurdle in a downturn as well as a boom is telling you something a company only earning it during the good years is not.

Return on equity is the weaker cousin, and it is worth being specific about why. ROE can be inflated with leverage alone. A company can lift its ROE by taking on more debt, with no improvement in the underlying business. ROIC strips that distortion out by looking at the return on the whole capital base, debt included, rather than just the equity slice. Use ROE as a cross-check. Do not use it as the screen.

Gross margin stability and trend. A gross margin that has sat in a tight band for a decade, or has expanded gradually, is telling you something about pricing power and cost discipline that a single year’s number cannot. A gross margin that swings wildly with the commodity cycle, or that has been quietly eroding for several years while revenue still grows, is a business with less control over its own economics than the top-line growth rate suggests. Read the trend against the industry. Stable-but-ordinary is a different signal from stable-and-wide.

Revenue durability. Not all revenue is the same quality. Recurring, contracted, or subscription revenue is worth more per dollar than one-off transactional revenue, because it is more predictable and cheaper to retain than to replace. A useful screen here is customer concentration: a company where the top five customers account for more than a third of revenue carries a structurally different risk profile from one with a broad customer base, even if today’s growth numbers look identical. Concentration is not an automatic fail — some excellent industrial suppliers live with it — but it changes the risk you are underwriting, and it belongs in the shortlist notes.

Free cash flow conversion, after stock-based compensation. The ratio of free cash flow to net income is one of the most underused screens in retail investing, and one of the most revealing. A company that consistently converts a high proportion of its reported earnings into cash is, broadly, telling the truth about its profitability. A company where net income has been growing while free cash flow has been flat or declining is worth hard questions: aggressive revenue recognition, working-capital games, or capital expenditure the income statement is not fully capturing.

A conversion ratio above roughly 70 to 80 per cent is a reasonable starting gate. It is not magic. It will not catch a one-off working-capital release that flatters a single year, capitalised costs that should have been expensed, or — the gap that matters most in US growth names — stock-based compensation that never touches free cash flow as many data vendors define it. Add two companion checks. Stock-based compensation as a percentage of revenue, and the trend in diluted share count. A business can convert earnings to cash and still hand a rising share of that engine to employees rather than to owners. For US technology and growth names, a rising share count is part of the quality test, not a footnote.

Balance sheet resilience. Net debt relative to EBITDA, and interest coverage, separate the compounders that can survive a bad cycle from the ones that need the good cycle to keep going. This is not about avoiding debt. Plenty of excellent businesses use leverage sensibly. It is about screening out balance sheets where a downturn or a rate shock could threaten the business, not just the share price. A starting gate of net debt below about 2.5 times EBITDA, with interest cover comfortably above 4 times, is conservative enough for a first pass. Banks, insurers, and property trusts do not fit this gate. Screen them with a different balance-sheet test, or leave them out of this universe.

Part Two: The Metrics That Mislead