financial-metrics
CAGR is the constant annual growth rate that takes a starting value to an ending value over n years. It is a geometric average — it ignores the path between the two points.
Boards and teasers need a single growth statistic. The arithmetic average of yearly growth rates overstates a volatile path. CAGR is the geometric rate that compounds smoothly from start to finish:
CAGR = (Ending value / Beginning value) ^ (1 / n) − 1
where n is the number of periods. For three years of data you usually have two growth intervals if you go year 0 → year 2; count the intervals, not the number of annual reports on the table.
Illustration only. Hypothetical subscription ARR.
| Year-end | ARR |
|---|---|
| 2022 | €4.00m |
| 2023 | €6.20m |
| 2024 | €6.50m |
| 2025 | €8.10m |
2022 → 2025 is three years:
CAGR = (8.10 / 4.00) ^ (1/3) − 1 = 26.5%
Yearly rates were +55%, +5%, +25%. The arithmetic mean is ~28% — close here, but in a down-year path the gap widens. A buyer who underwrites 26% forever has not read 2024.
CAGR is a summary of two points. It cannot see a 2024 acquisition, a 2023 price spike, or a 2025 contract pull-forward. Always pair it with a yearly bridge (organic vs inorganic, price vs volume vs FX).
For negative starts or sign changes (loss to profit), CAGR is the wrong tool; use absolute bridges.
Revenue and EBITDA in teasers; public-comp growth screens; IRR is not CAGR — IRR is money-weighted with interim cash flows.
Endpoint bias: choose a trough start and a peak end and you manufacture a growth company. Different fiscal year-ends and 52/53-week retailers distort n. CAGR on adjusted EBITDA can compound the add-backs.
Dazur shows CAGR next to the yearly organic bridge. Investors matching on “20% CAGR” without that bridge are matching on a slogan.