Why We Score Stocks Within Their Sector

#The Problem: Sector Bias in Global Screening

Imagine ranking all 3,000 stocks on a single, global scale using our six factors. What happens?

Value Factor Bias

Banks and insurance companies trade at low price-to-book ratios by nature — their assets are mostly financial instruments carried near fair value. Technology companies trade at high price-to-book ratios because their most valuable assets (software, IP, talent) don't appear on the balance sheet.

A global value screen would systematically rank: - Banks: Score 80+ on value (they're always "cheap") - Tech companies: Score 20 or below on value (they're always "expensive")

This doesn't mean banks are better investments. It means the value factor has a structural sector bias.

Volatility Factor Bias

Utilities and Consumer Staples are inherently less volatile than Biotech and Semiconductor stocks. A global low-volatility screen would fill your portfolio with utility companies — not because they have superior risk-adjusted returns within their sector, but because the sector itself is low-volatility.

The Result of Global Screening

A globally screened factor portfolio would be permanently overweight in Financials, Energy, and Utilities — and permanently underweight in Technology and Healthcare. You'd miss the best-performing sectors of the last two decades.

#The Solution: Sector-Relative Scoring

Instead of comparing Apple to JPMorgan, we compare: - Apple to other Technology stocks - JPMorgan to other Financial stocks - UnitedHealth to other Healthcare stocks

How It Works

For each of the 11 GICS sectors, we independently:

1. 1Calculate raw factor values for every stock in that sector
2. 2Z-score within the sector — the mean and standard deviation are calculated using only stocks in that sector
3. 3Winsorize at ±3 standard deviations within the sector
4. 4Combine into a composite score using the same fixed weights
5. 5Convert to a percentile (0-100) within the sector

This means a stock's score tells you how it ranks relative to its sector peers, not relative to the entire universe.

What the Scores Mean

A 75 in Technology means "this tech stock is better than 75% of other tech stocks." A 75 in Healthcare means "this healthcare stock is better than 75% of other healthcare stocks." The scores are directly comparable across sectors because they both represent the same relative position.

#Cross-Sector Comparison: Apples to Apples

This is the most common question we receive: "Why is this bank rated higher than this tech stock?"

The answer is that they're not being compared to each other. They're each being compared to their own sector.

Example

In the global screen, JPMorgan appears best simply because Financials score well on value and low volatility by default. In our sector-relative screen, Apple ranks highest because it's truly exceptional within its sector — dominant profitability, strong momentum, reasonable valuation for a tech company.

Our scoring answers the right question: not "Is this stock cheap in absolute terms?" but "Is this the best stock in its sector?"

#The GICS Classification System

We use the Global Industry Classification Standard (GICS), developed jointly by S&P and MSCI. It's the most widely used sector taxonomy in professional finance.

The 11 GICS Sectors

Why GICS and Not Something Else?

GICS strikes the right balance: granular enough to group genuine peers, broad enough to have statistically meaningful sample sizes within each sector (minimum ~50 stocks per sector in our universe).

#Edge Cases and Nuances

Conglomerates

Companies like Berkshire Hathaway or General Electric span multiple sectors. GICS assigns them to a primary sector based on their largest revenue source. This is imperfect but consistent — and consistency matters more than perfection in a systematic model.

Sector Reclassifications

GICS is updated annually. When a company moves sectors (as Facebook did when "Communication Services" was created in 2018), its scores are recalculated relative to the new peer group. This can cause temporary score changes that don't reflect any change in the company's fundamentals.

Small Sectors

Sectors with fewer stocks (Materials, Real Estate) have less statistical power in their z-score calculations. We monitor for this but have found the sample sizes adequate (50+ stocks per sector in our universe).

#Why This Matters for Your Portfolio

Sector-relative scoring has practical implications for portfolio construction:

1. 1Every sector contributes Strong Buy stocks. You won't be forced into a few sectors.
2. 2Sector allocation is your choice. Our scores don't push you toward or away from any sector — that decision is yours.
3. 3Cross-sector comparisons are valid. A 5-star Technology stock and a 5-star Healthcare stock are both top-quartile within their peer group.
4. 4No hidden sector bets. A globally screened portfolio makes implicit sector bets. Ours doesn't.

#The Bottom Line

Sector-relative scoring is one of the most important design decisions in our methodology. It ensures that our rankings reflect genuine stock quality — not sector membership. When you see a stock rated Strong Buy, you know it's truly exceptional among its peers, regardless of whether it's a bank, a tech company, or a utility.

Explore sector-relative rankings →

#Further Reading

• How Composite Scores Work →
• How to Build a Portfolio Using Our Rankings →
• Why We Use Exactly Six Factors →
• Our Complete Methodology →