The Analyst’s Course · Quant SQL
Screener Traps — How Every Quantitative Screen Lies
Survivorship bias
Your live screener contains today's constituents — the winners. Screen today's S&P 500 for "cheap quality" and you are reading the roster after the losing teams were cut. The honest version needs yesterday's universe including the bankruptcies and delistings that no longer appear. This is why historical backtests without point-in-time data overstate returns by whole percentage points annually — the dead are excluded from the average precisely because they died. The ClearView screener is explicitly a today-tool; the moment you mentally treat its output as a historical backtest, the bias is in your process.
Lookahead bias
A subtle cousin: using data that was not actually knowable at the decision time. Q4 earnings are filed in January–February; a "screen run in November" that uses Q4 numbers could not have existed. The SEC-filing timestamps in this platform's pipeline (the accounting track) exist partly to make this check possible. The habit: for every column in your query, ask "when did this number actually become public?" Financial data is born stale — the filing date, not the fiscal period, is what a real process could have acted on.
The value trap
A cheap stock is cheap for a reason, and the screen cannot see reasons. Sears Holdings screened as a deep-value superstar for a decade — low P/E (when it had earnings), P/B far below 1, generous dividend — while the accounting track's instruments told the other story: negative retained earnings, Z-Score in distress, deteriorating fundamentals. The screen said cheap; the Z-Score said does-this-survive. The full process (Track A → this track) exists precisely because the two layers answer different questions and the second vetoes the first.
Sears: the decade-long lesson in two layers
2010–2018, Sears screened beautifully on static value metrics — while the Altman Z sat in distress territory for most of it, receivables and payables stretched, and the store base monetized to fund the hole. Value screens kept flagging it; the survival screen kept rejecting it. The stock lost >95% into the 2018 bankruptcy. The two-layer lesson is not that value investing failed — it is that valuation without solvency is an answer to a question nobody should be asking first.
What you'll practise
Screening today's index members for historical "cheapness" suffers from…
3 graded checkpoints · certification exam at the end of the track
Sources
Fama & French (1992, 2015); Daniel & Moskowitz (2016); screener traps per standard practice
Learn content is for education only — not individualized financial advice, a recommendation, or a solicitation to buy or sell any security. Options involve substantial risk. Examples are simplified and historical patterns never guarantee future results.