Core answer: Growth rate = (new − old) ÷ old × 100%. CAGR (compound annual growth rate) = (end ÷ start)^(1/years) − 1: revenue growing ¥1M → ¥2M in 5 years has CAGR = 2^(1/5) − 1 = 14.87%/year — NOT (100% ÷ 5 = 20%). Average growth and CAGR diverge wildly in volatile series; CAGR is the only honest multi-year rate. Double-check with the Rule of 72: 72 ÷ 14.87 ≈ 4.8 years to double. ✓
The three rates people confuse
| Metric | Formula | Use | Trap |
|---|---|---|---|
| YoY growth | (this − last) ÷ last | single-period | base effects |
| Average growth | mean of YoY rates | rough feel | ignores compounding |
| CAGR | (end/start)^(1/n) − 1 | multi-year truth | hides volatility |
Why average ≠ CAGR. +100% then −50%: average +25%, actual 0% (¥100 → ¥200 → ¥100). CAGR handles it correctly: (100/100)^(1/2) − 1 = 0%.
Worked examples
Example 1 — Investment check. Fund: ¥100k → ¥161k in 5 years. CAGR = (1.61)^(0.2) − 1 = 10%/yr. The brochure said "61% in 5 years" — true and useless; 10%/yr is comparable against alternatives.
Example 2 — Business planning. Revenue ¥5M, target ¥20M in 4 years: needed CAGR = 4^(1/4) − 1 = 41.4%/yr. Sanity-check: can the market, sales team, and delivery each sustain ~41%? If not, extend the timeline or shrink the target — before promising it to investors.
Example 3 — Negative and recovery. Portfolio ¥1M → ¥600k in 2 years: CAGR = (0.6)^(0.5) − 1 = −22.5%/yr. Recovery to ¥1M from here needs +66.7% total, i.e. CAGR +29.1% over the next 2 years — asymmetric math after losses.
Example 4 — Smooth-looking lie. A company grows +80%, +10%, −30%, +5% (¥1M → ¥1.17M over 4 years): CAGR = 4.0%/yr — respectable-looking, but the path included a near-death year. CAGR says nothing about risk; pair it with max drawdown.
Common mistakes and myths
- Dividing total growth by years — 100% over 5 years is 14.9%/yr, not 20%/yr; compounding is not linear.
- CAGR over cherry-picked dates — start at a crash bottom, end at a bubble top: any number you want. Disclose your endpoints.
- Using CAGR for lumpy cash flows — money in and out mid-period needs XIRR, not CAGR.
- Comparing CAGRs across different risk — 15% from an index fund and 15% from a leveraged scheme are different animals; risk-adjust before ranking.
- Forgetting inflation — 8% nominal CAGR with 3% inflation = ~4.9% real; long-term plans belong in real terms.