Why do some sectors swing with the economy and others barely notice?
The economy breathes — expanding, peaking, slowing, recovering. Sectors don't all breathe with it equally. Some rise and crash with the economic tide; others plod along almost regardless. Sorting sectors into cyclical and defensive is the first tool of cycle analysis.
Two temperaments
Cyclical sectors boom in good times and slump in bad ones, because their demand is deferrable. People buy new cars, take holidays, and build factories when they feel confident — and postpone all of it when they don't. The cyclical crowd: Consumer Discretionary, Industrials, Materials, Energy, and much of Financials and Technology.
Defensive (non-cyclical) sectors hold up in downturns because their demand is non-negotiable. Recession or boom, people still buy food, take medicine, and keep the lights on. The defensive crowd: Consumer Staples, Health Care, Utilities.
The dividing question is simple: can the customer postpone the purchase? If yes, cyclical. If no, defensive.
Why this matters for reading a business
The same company quality means different things depending on temperament. A cyclical company posting record profits at the top of a boom may be at its most dangerous moment, not its safest — those earnings could be a cyclical peak about to roll over. A defensive company's steadier earnings deserve a different mental model entirely.
This also reframes the word "risk." A defensive utility rarely doubles, but rarely halves either. A cyclical carmaker can do both. Neither is "better" — they're different shapes of ride, and knowing the shape is the point.
A worked example
Run two companies through a recession in your head. FeastFoods (Consumer Staples) sells everyday groceries — in a downturn, sales dip only slightly; people trade down but still eat. LuxAuto (Consumer Discretionary) sells premium cars — in the same downturn, buyers simply wait, and sales can fall 30-40%. Same economy, opposite experience. If you compared their latest earnings without noticing their temperaments, you'd misread both: LuxAuto's crash is normal for its type, not a company failure.
In the data
The revenue history that separates the two temperaments is one revenue figure per fiscal year, and its depth is the catch. The listing date on a company profile is only a hint at it:
Apple listed in December 1980, yet its annual history in this data begins at fiscal 1985; Intel listed in 1971 and also begins in 1985. At the young end, Rivian listed on 10 November 2021 and its history begins at fiscal 2019, with zero revenue in both 2019 and 2020, so the one downturn inside the window meets a company with nothing yet to sell (all measured 28 September 2026). Testing a company's cyclicality against a downturn only works when the downturn falls inside the years the data actually spans.
Try it now
- List the 11 GICS sectors and tag each cyclical or defensive using the "can the customer postpone it?" test. Do this from memory first.
- Then test two of your tags against filings: a carmaker, which sells something customers can postpone, and a household-products maker, which does not. Four fiscal years each, pinned to the 2008-09 recession:
Compute each company's revenue change from its highest year to its lowest. One should visibly buckle across the downturn and the other should barely notice. Then read the net income rows: which company's profit was the more predictable of the two? 3. Before you trust a comparison like that, check that the downturn is inside the window. Both tables above reach back to 2007 because both companies have filed for decades. The IPO date in the profile under "In the data" reads 1980, yet that company's history begins at fiscal 1985, and the same section gives a company whose history begins two years before its listing. Testing cyclicality against a recession only works when the recession falls inside the years the data actually spans, with revenue in them. Take the section's young company and say what its first two fiscal years rule out. 4. A defensive business, read on published figures, for the shape you are looking for:
Thin margins, enormous revenue, and demand nobody postpones. Note which company's earnings you would trust to look similar three years from now. That difference in predictability is the whole cyclical/defensive distinction in one observation.