Chapter 7: Growth, value, quality, and other labels
For the thirteen years ending in 2021, the winning label in US stocks was "growth." In the academic sorts that Ken French publishes, cheap stocks beat expensive ones in only four of those thirteen calendar years, and trailed by about 4 percentage points a year on average. Then came 2022. The Russell 1000 Growth index lost 29.14% that calendar year while the Russell 1000 Value index lost 7.54%, per FTSE Russell: a gap of nearly 22 points, in value's favor, arriving immediately after growth's greatest decade. The decade's winning label became the year's losing one in twelve months.
This chapter exists so that whipsaw never happens to you on purpose. Fund names, screeners, and pundits throw style words around as if they described a company's character. They describe a math sort. Once you can see the sort underneath each label, you can use the labels for what they are good at, which is understanding a stock, and stop using them for what they are terrible at, which is telling you where to move your money next.
Every label is a sorting rule
A growth stock is not a company certified to grow, and a value stock is not a certified bargain. Both labels come off an assembly line. An index provider takes a list of companies, computes a ratio or two for each, and draws lines through the ranking. FTSE Russell, which runs the indexes in the opening paragraph, sorts the 1,000 largest US companies using price-to-book ratios and forecast growth: companies with higher ratios and faster forecasts land in the growth index, companies with lower ratios and slower forecasts land in the value index, and many companies land partly in both. A blend label simply means the sort came out near the middle.
Chapter 5 already taught you to read the underlying numbers, so translate the labels back into that language. A growth label means the market is paying a high multiple, which means big promises are already in the price; at 60 times earnings, the growth is the entry fee, not the bonus. A value label means a low multiple, which Chapter 5 showed is the market's forecast of decline, a forecast that is sometimes wrong and sometimes right. Hugo's department-store chain at 8 times earnings wore the value label all the way down.
Notice what this implies: a company can change labels without changing anything about itself. Cut a high-flyer's price in half and the sort may quietly reclassify it from growth to value, even though the same people are selling the same product from the same buildings. The label tracks the ratio, and the ratio has a price in its numerator.
The style box, and where Cascade Coffee lives
In 1992 Morningstar organized the two most common sorts into a nine-square grid called the style box, and it remains the standard map. The columns run value, blend, growth, using sorts like the ones above. The rows run by size, measured by market capitalization, the share price times the share count from Chapter 1. The bands FINRA publishes give the common convention: large caps sit above $10 billion, mid caps between $2 billion and $10 billion, and small caps between $250 million and $2 billion.
Place Cascade Coffee from Chapter 4 on the map, with every step shown. Cascade has 120 million shares outstanding, and at $48 a share its market cap is 120 million times $48, which is $5.76 billion. That lands between $2 billion and $10 billion, so Cascade sits in the mid-cap row. For the column, its P/E of 20 is neither the deep discount of Steady Mills at 8 nor the hot promise of Lumen Labs at 60, and Chapter 5 called 20 roughly an ordinary price for an ordinary good business. The sort comes out near the middle, so Cascade takes the blend column. One coffee company, one cell: mid-cap blend, dead center.
The map is useful, and it has the same limit as every label here. Ten years ago a smaller, faster-growing Cascade might have sat in small-cap growth. A rough stretch that knocked the price to $30 could slide it toward mid-cap value. Throughout, it roasts the same coffee.
One company, several labels at once
The style box is only one labeling system, and a stock wears tags from all of them at once. A sector label says what business a company is in: the GICS system, run by MSCI and S&P, divides every public company into 11 sectors, from energy to utilities. A cyclical label means profits swing with the economy, the way they do for airlines and homebuilders; a defensive label means people keep paying for the product in a recession, the way they do for electricity and toothpaste. A quality label, in MSCI's widely used definition, requires high return on equity, stable year-over-year earnings growth, and low financial leverage. A momentum label means the price rose recently. An income label means the dividend yield runs above average, with all of Chapter 6's warnings attached.
Elaine's inherited utility shows why the stacking matters. On a fund screen it earns two comforting labels, income and defensive, and both are accurate. The third label never appears on any screen: 800 shares of a single company is a concentrated position, whatever the company sells. Labels describe the stock. None of them describes her risk, which lives in the size of the position, and that distinction is Chapter 13's whole subject. Mara ran the same exercise as homework, sorting her 14 tickers into the nine boxes, and found ten of them crowded into two growth cells. The variety she felt while buying them was never on the map.
What a century of scorekeeping shows
The oldest style question is whether cheap stocks beat expensive ones, and the cleanest scoreboard is the one Eugene Fama and Ken French built by sorting US stocks on price-to-book and measuring the gap, published free in the Ken French data library. From 1927 through 2025, that value-minus-growth gap, the value premium, averaged about 4 percentage points a year. A century of data, a positive average, and a real finding: that is the honest case for value.
The honest rest of the story is how the premium arrives, which is in long, lurching eras. From 2000 through 2006, value won all seven calendar years, by roughly 15 points a year on average, while the survivors of the dot-com bust learned what 60-times-earnings prices had assumed. From 2009 through 2021, the pendulum swung back: value trailed by about 4 points a year on average, and 2020 was value's worst single year in the entire series back to 1927, roughly 47 points behind. Then 2022 flipped again, with value ahead by about 26 points in the same sorts, the year the Russell numbers in the opening paragraph were printed.
Should you move your money into whichever style is winning?
No, and the reason is that switching styles is market timing wearing a costume. Chapter 3 showed that prices fold public information in fast; by the time a style sits on top of the leaderboard, the repricing that put it there has already happened, and you would be buying the band in Figure 7.2 after it was painted. The eras only look obvious in hindsight. Value spent thirteen years losing before its 2022 win, and growth delivered its worst year in decades immediately after its best stretch, so a switcher has to be right twice, once on the way in and once on the way out, which is the same double bet Chapter 3 attached to any trade. The professionals who try full time do not clear the bar: in the SPIVA scorecard covered in Chapter 8, growth funds and value funds both sat inside the 22 US fund categories in which, over the 15 years ending 2024, not one category had a majority of active managers beat their own benchmark. A total-market index fund already owns all nine boxes at market weight, every sector and every style, which means it held the 2022 winners before anyone knew 2022 would pick them. Owning everything, always, is the position that never needs the pendulum forecast.
Where labels cost real money
Beyond rotation, the labels have four cheaper failure modes worth naming. Chasing last decade's box is the big one, covered above. Reading "growth" as "good company" is the second: the label reports a high price, and Chapter 5 showed that at high multiples the growth is required just to break even. The third is the value trap, Hugo's tuition receipt from Chapter 5, where the low multiple was a correct forecast of decline rather than a discount. The fourth is treating income labels as safety, when Chapter 6 showed a 9% yield is usually a collapsed price forecasting a dividend cut.
There is also a label that arrives with an invoice, so the seller's-pitch test applies.
Funds marketed as "smart beta" or factor funds package a sorting rule, value, quality, momentum, into a product that usually charges several times what a plain index fund costs. Take a factor fund charging 0.40% a year against a broad total-market fund charging 0.04%: on a $100,000 balance that is $400 a year against $40, every year, compounding against you, for a sort whose recipe sits free in the Ken French data library. Some factor funds are honest products at fair prices. The test is the same as everywhere else in this guide: find the fee before you admire the label, and ask what the extra charge buys that the sort itself does not explain.
Treat style labels as lenses, never as buy lists. A total-market fund already owns all nine boxes, so any tilt toward one of them is a deliberate bet that the box will beat the rest, and it belongs inside the same explore sleeve and sizing caps Chapter 13 puts on every conviction.
In December 2021 Dev read a long thread arguing that value investing was dead, that rates and software had retired the old sorts for good. The author was smart, the charts went up and to the right, and Dev caught himself drafting the sell orders that would tilt his $140,000 three-fund portfolio hard into a growth index fund for the new year. Then he did what he does with stocks and read the underlying numbers instead of the story. Every chart in the thread stopped at the present; every era in the data eventually ended. He kept the boring core. Over calendar 2022, $100,000 in the Russell 1000 Growth index became about $70,860, while the same money in the Value index became about $92,460. Dev still likes growth companies, and he owns his three favorites inside his explore sleeve, sized so that being wrong about them, or about the whole label, changes nothing he cares about.
Key takeaways
- Style labels are sorting rules on ratios, not judgments of quality. Index providers sort on measures like price-to-book and forecast growth, and a company can switch labels just because its price moved.
- The style box has nine cells: value, blend, and growth columns against large, mid, and small rows. Cascade Coffee, with 120M shares at $48, carries a $5.76B market cap and a P/E of 20, landing it in the center cell, mid-cap blend.
- One stock wears many labels at once: one of 11 GICS sectors, cyclical or defensive, quality, momentum, income. No label measures the risk of position size, as Elaine's triple-labeled utility shows.
- The value premium averaged about 4 points a year from 1927 through 2025 in the Ken French data, but it arrives in eras: value led 2000–2006 by about 15 points a year, growth led 2009–2021 by about 4, and in 2022 Russell 1000 Growth lost 29.14% while Value lost 7.54%.
- Style rotation is market timing in a costume, and SPIVA's 15-year scorecard shows the growth and value fund categories sitting inside the 0-of-22 result from Chapter 8.
- A total-market fund owns every box already. A tilt is a conviction bet, priced like one (watch the fees) and sized like one, inside Chapter 13's caps.
Sources: Ken French data library · FTSE Russell (LSEG) · Morningstar · MSCI · Investor.gov: Market capitalization · FINRA: For investors · SPIVA U.S. scorecard