QuantitudeScreeners and back-tests
Updated 17 Oct 26
Your screener gave you a hundred stocks. Now what?
From a long list to a tested shortlist
Plenty of screeners show every stock that meets your criteria. Some return 50, 75, 100 or more. That is a list, not a portfolio. ScreenLab is built for the next step.
How ScreenLab works, in four steps
- Bring. Write your own screening rules. ScreenLab keeps them exactly as you wrote them. A criterion it cannot apply is shown as "Not applied" rather than silently dropped.
- Rank. Every name that passes is ranked by the Quantitude composite, a blend of factors such as price strength, earnings acceleration, capital efficiency and margin inflection.
- Choose. You set the dials: how many names to hold (N, from 5 to 25; a smaller N is more concentrated, a larger N is more spread out), the market-cap band, how often to rebalance (monthly, quarterly, half-yearly or yearly), and whether stops are on.
- Test. ScreenLab back-tests your full list against the ranked top-N shortlist over the same rules and the same dates, next to a benchmark.
What the three lines mean
- Your screen (raw): every name that passes your rules.
- Quantitude edge: only the top N of those names by the composite score.
- Benchmark: an index for context.
- Ranking lift is the edge line's compound annual growth rate (CAGR) minus the raw line's. It answers one question: in that test, did ranking your own rules by the Quantitude score change the result?
Why the factor weights matter: the dots on the line
The composite does not use one fixed formula. Its factor weights are re-fitted on past data every quarter (a walk-forward approach), and each rebalance uses only the weights that were known on that date, never a later set.
In the chart, each dot on the Quantitude line marks a re-tune. Hovering a dot shows the mix in force: which factors the score leaned on, and by how many percentage points each changed. In the sample video, three hover views each highlight the biggest shift.
This is the design idea behind the score: weights that adjust to conditions, instead of staying fixed. Whether it helps depends on the rules you bring and the period tested, and a back-test cannot show that it will help next time.
Reading the sample result honestly
In the invented sample, the edge line grows faster (14.4% CAGR against 12.0% for the raw list and 10.7% for the benchmark), so the sample ranking lift is +2.4 points. But the edge line's worst fall is deeper too (-19.1%, against -12.1% and -8.2%).
Both halves matter. A back-test reports what happened in a particular history; it cannot say what comes next.
Glossary
- CAGR: compound annual growth rate, the steady yearly rate that would turn the start value into the end value.
- Worst fall (maximum drawdown): the largest peak-to-trough decline in the period.
- Rebalance: the dates on which the holdings are refreshed against the latest ranking.
- Walk-forward: fitting on past data and applying the result only to later dates.
Bring your screen
ScreenLab is a back-testing tool on Quantitude. Signing in with Google is required to use it.
Open ScreenLabFull transcript
Narration, word for word
Your screener gave you a hundred stocks. Now what?
A screen can hand you fifty, seventy-five, even a hundred names. That's a list, not a portfolio.
So bring your own screening logic. ScreenLab keeps your rules exactly as you wrote them.
Then it ranks every name that passes using the Quantitude composite, and keeps just the top N.
You choose N, plus caps, stops and rebalancing. Bolder or steadier, your call.
It back-tests your full list against the ranked shortlist. Same rules, same dates.
See the dots on the Quantitude line? Each one is a re-tune: the factor weights changed.
Point at one, and you see the mix in force: which factors the score leaned on, and by how much.
The idea: weights re-fit on past data every quarter, not one fixed formula. Each rebalance only uses the weights known on that date.
Whether that rewards your rules shows up in the gap between the lines: the ranking lift, in that test.
In this made-up sample, the Quantitude line grows faster. But its worst fall is deeper, too.
A back-test is history. It can't tell you what comes next.
Bring your screen to Quantitude.
Sources
- How ScreenLab behaves is described from the tool itself and its on-screen methodology notes. All figures shown are invented sample data.