All articles

Inside the Ranking System: Six Questions Every Stock Has to Answer

A detailed walk through the multi-factor ranking behind The Small Edge: quality, value, growth, low volatility, sentiment and momentum, with every weight explained.

November 16, 2026·9 min read

A rules-based strategy is only as good as its rules. This article opens up the ranking system behind The Small Edge screener and backtest: what it measures, how much each part counts, and why it is built the way it is.

One score, six questions

Every stock in the universe gets a single score from 0 to 100. A score of 95 means the stock ranks better than 95% of its peers. The score is the weighted sum of six factor families, and each family answers one simple question:

Factor familyWeightThe question it answers
Quality30%Is this a good business?
Value25%Is it cheap?
Growth15%Is it growing, and faster than before?
Low volatility10%Is it calm?
Sentiment10%Are expectations improving?
Momentum10%Is the market starting to notice?

Quality and value together carry 55% of the weight. That is deliberate. In small caps, the biggest risk is not missing a rocket but owning a broken business that only looks cheap. The other four families help separate cheap stocks that are about to recover from cheap stocks that deserve to be cheap.

Two design principles run through the whole system:

  • Many factors, each with a small weight. No single ratio can push a stock to the top. A one-off accounting quirk in one number gets outvoted by the rest.
  • Short and long horizons side by side. Most families look at the latest quarter or year and at a five-year history. A business that has been good for five years and is still good today scores higher than one with a single strong quarter.

Quality (30%): Is this a good business?

Quality is the largest block. It has four equal parts:

  • Margins (25%): operating margin over the last twelve months and as a five-year average, plus how far the gross margin sits above the industry average.
  • Asset turnover (25%): how much revenue the company generates per dollar of assets, today and on a five-year average.
  • Return on capital (25%): return on investment and return on equity, again for the last twelve months and the five-year average. The long-term figures get the larger share (60% vs. 40%), because consistency matters more than one good year.
  • Finances (25%): interest coverage (can operating profit comfortably pay the interest?) for the last year and five years, and debt relative to total capital.

Most quality factors are ranked within the company's industry, not across the whole market. A software company will always have higher margins than a grocery chain; what matters is whether this grocery chain is better than other grocery chains.

Value (25%): Is it cheap?

Value looks at price from four angles, so the system does not depend on a single accounting line:

  • Earnings (30%): the P/E ratio relative to the industry, the forward earnings yield based on analyst estimates, and EBITDA relative to enterprise value.
  • Sales (35%): price-to-sales, EV-to-sales, forward price-to-sales and gross profit relative to enterprise value. Sales are harder to manipulate than earnings and also work for companies that are temporarily unprofitable.
  • Free cash flow (25%): free cash flow relative to the share price and unlevered free cash flow relative to enterprise value. Cash is the hardest number to fake.
  • Assets (10%): price-to-book and price-to-tangible-book. A small weight, because book value says less about modern businesses than it used to.

Enterprise value (market cap plus debt minus cash) shows up repeatedly. It stops heavily indebted companies from looking cheap just because their equity is small.

Growth (15%): Is it growing, and faster than before?

Growth is measured on three lines of the income statement, each with the same weight: earnings per share, sales and operating income. For each one, the system looks at two things:

  • Growth itself (50%): growth versus the same quarter last year, over the last twelve months, and over five years.
  • Acceleration (50%): is recent growth faster than the twelve-month rate, and is the twelve-month rate faster than the five-year rate?

The acceleration part is what makes this block interesting. A company growing at 10% after years at 3% is often more attractive than one that has grown at 15% for a decade and is slowing down. Markets tend to react late to that kind of change.

Low volatility (10%): Is it calm?

The low-volatility anomaly is one of the most robust findings in finance: calmer stocks have historically delivered similar or better returns than wild ones, with much less pain along the way. Two equal parts:

  • Beta (50%): sensitivity to the overall market over one, three and five years. Lower is better.
  • Price volatility (50%): the swing of the share price over one, three and five years. Lower is better.

In small caps this block also works as a filter against lottery tickets: stocks that move 10% a day on news flow rarely belong in a rules-based portfolio.

Sentiment (10%): Are expectations improving?

Sentiment captures what analysts and other investors are doing right now:

  • Estimate revisions (45%): how much the earnings estimates for the current year and current quarter changed over the last four weeks, plus how much analysts disagree with each other (less disagreement is better).
  • Earnings surprises (25%): did the company beat expectations in the last two quarters? The most recent quarter counts more.
  • Recommendations (15%): are analyst recommendations improving compared with four weeks ago, and how positive are they overall?
  • Short interest (15%): a low share of stock sold short is better. Heavy short interest often means informed investors see a problem.

Many micro caps have little or no analyst coverage. For them, the other families carry the score, which is exactly how it should be.

Momentum (10%): Is the market starting to notice?

Momentum looks at the direction of the share price:

  • Price change (35%): performance over roughly six to nine months, measured over four overlapping windows so the score does not hinge on one start date.
  • Industry momentum (35%): how the company's industry has performed over six and twelve months. Stocks in strong industries tend to keep doing well.
  • Up/down ratio (15%): the ratio of up days to down days over one, three and six months. Steady gains score better than one big jump.
  • Quarterly returns (15%): returns in each of the last four quarters, with the most recent quarter weighted highest. This rewards consistent strength over a single spike.

At 10%, momentum is a confirming signal, not a driver. The system wants good, cheap businesses that the market is starting to recognise, not stocks that have already run.

From score to portfolio

The rules for turning scores into a portfolio are short:

  1. Universe: primary listings in the US, Canada and Europe.
  2. Buy rules: market cap between $10 million and $2 billion, and an average of more than $30,000 traded per day over the last 20 days.
  3. Selection: the ~20 best-ranked stocks that pass the buy rules, equally weighted.
  4. Selling: the ranks are checked once a week. A stock is held as long as its rank is 95 or higher. When it drops below 95, it is sold and the money goes into the best-ranked stock not yet held. Then the cycle starts again.

Holding until the rank drops below 95, instead of selling as soon as a stock leaves the top 20, keeps turnover and trading costs down. A stock ranked 97 is still a very good stock.

The backtest from January 2009 applies variable slippage between 0.08% and 4.00% per trade, depending on each stock's liquidity. All results, the yearly returns and the caveats are on the backtesting page.

Why this mix?

None of the factors here are secret. Value, quality, momentum and low volatility have been studied for decades. The edge does not come from a hidden formula. It comes from three things:

  • Combining factors that disagree. Value and momentum often point in opposite directions. Quality and growth do not always overlap. A stock that scores well on all of them at once is rare, and those are the stocks the system is looking for.
  • Applying it where coverage is thin. In small and micro caps, the same factors work better than in large caps, because fewer professionals are looking.
  • Following the rules when it is uncomfortable. The ranking does not know about headlines. That is the whole point.

This article is for informational purposes only and is not investment advice. Backtested results are hypothetical and do not guarantee future returns.

Small caps, picked by rules, not by stories.

Get access