Top-Down vs Bottom-Up

In short

These are two directions of reasoning, not two philosophies.

Canonical data. Figures tie to Wexford Instruments (USD millions).

Top-down starts with the economy, narrows to industries likely to do well in those conditions, and then selects within them. Bottom-up starts with individual businesses, judges them on their own merits, and treats the economy as context rather than as the starting question. Neither is better, and the interesting thing about them is that they fail in opposite ways.

How top-down fails

It depends on macroeconomic forecasting, which has a poor record. Growth, inflation, and rate paths are forecast badly and consistently, by well-resourced institutions, and the errors are largest exactly when they matter most — at turning points.

And the second failure is subtler and worse: even a correct macro call does not reliably produce the expected market outcome. The chain from a macroeconomic variable to a company's cash flows to its share price is long, and each link is uncertain — because prices already embed a consensus view of the macro outlook, so what moves them is the difference between what happens and what was expected. Being right about the economy and wrong about the market is an ordinary outcome rather than an ironic one.

How bottom-up fails

It is blind to common factors. A collection of individually excellent businesses can share a single exposure — to interest rates, to one currency, to one customer industry, to a regulatory regime — and nothing in the individual analysis reveals it, because the exposure only becomes visible when the holdings are viewed together.

This maps precisely onto a distinction Pillar 22 established. Diversifiable risk is company-specific; systematic risk is shared and cannot be diversified away. Bottom-up analysis is excellent at the first and structurally blind to the second — and since correlations between holdings rise in stressed conditions, that blindness is worst precisely when it costs most.

What practitioners actually do

Almost everyone blends the two, and the labels describe emphasis rather than exclusivity. A bottom-up analyst who ignores what interest rates do to a leveraged company is not being purist; they are being incomplete. A top-down investor who never examines a business is expressing a view about a sector average that no individual holding may resemble.

The useful question is not which direction to reason in but which risks each direction leaves you unable to see — and then deliberately checking those.

Worked example

Worked example: both directions on the same company (canonical figures, USD millions). Bottom-up. Wexford is a capital-intensive manufacturer with revenue growth of 13.6%, an improving operating margin of 10.0%, capital expenditure at 1.86 times depreciation, free cash flow of 20.3, and net debt of 273.7. That description is complete on its own terms and mentions the economy nowhere. Top-down. What happens to industrial demand, to input costs, and to interest rates? And here is what only the combination reveals. Wexford is exposed to rates twice over. Directly, through borrowing: a one-point rise in its cost of debt cuts net income by 4.2%. And indirectly, through valuation: a one-point rise in the discount rate moves a valuation by more than 12%. The bottom-up analysis produced the first exposure and would never have framed the second as a risk; the top-down view would identify rates as a driver without knowing that this particular balance sheet doubles the effect. The lesson. Neither direction was wrong. Each was incomplete in a way the other happens to cover — which is the practical argument for using both rather than choosing. (Canonical figures; the 4.2% is one point on gross debt of 340.0, tax-effected, against net income of 62.3; the 12.5% is the Pillar 22 growing-perpetuity sensitivity.)

Frequently asked

6 questions

What's the difference between top-down and bottom-up?

Direction of reasoning. Top-down starts with the economy and narrows to sectors and then securities; bottom-up starts with individual businesses and treats the economy as context. They're not competing philosophies.

Why does top-down analysis fail?

Two ways. Macro forecasting has a poor record, with the largest errors at turning points. And even a correct call doesn't reliably produce the expected market outcome, because prices already embed a consensus view — what moves them is the difference between what happens and what was expected.

Why does bottom-up analysis fail?

It's blind to common factors. Individually excellent businesses can share one exposure — rates, a currency, a customer industry — that's invisible in single-company analysis and only appears when holdings are viewed together.

How does that relate to systematic risk?

Directly. Bottom-up analysis is excellent at company-specific risk and structurally blind to shared risk — and since correlations rise in stressed conditions, that blindness is worst when it costs most.

Which should I use?

Most practitioners blend them, and the labels describe emphasis rather than exclusivity. The useful question isn't which direction to reason in, but which risks each direction leaves you unable to see.

What did the worked example show?

That the illustrative company is exposed to interest rates twice — directly through borrowing costs and indirectly through valuation. Bottom-up analysis surfaces the first and wouldn't frame the second as a risk; top-down identifies rates as a driver without knowing this balance sheet doubles the effect.

References

Investor.gov (SEC) — How to Read Financial Statements —
Investor.gov (SEC) — Diversification (the shared-exposure point) —