The formula, and why the exponent is 1 over n
The ratio end ÷ begin is your total growth over the whole period. Taking the n-th root (that's the 1/n power) spreads it evenly across the n years, and subtracting 1 turns the multiplier into a percentage. Run it backwards to project forward: future value = present × (1 + CAGR)n.
CAGR smooths what an average can't
A simple average treats each year as independent, so it ignores compounding and can flatter a bumpy record. CAGR answers the honest question: what steady rate would have produced this exact result?
A year of +50% then a year of −30% averages to +10%, but $100 becomes $150 then $105 — a real 5% loss. CAGR reports the loss; the average hides it.
What counts as a good CAGR
| Where | Typical CAGR |
|---|---|
| Savings account | 3–5% |
| Residential real estate | 5–8% |
| S&P 500 (30-year) | 10–11% |
| Growth stocks | 15%+ |
| Startup revenue | 25–100%+ |
CAGR only sees the first and last points, so it says nothing about the ride in between. Two investments can share a 10% CAGR while one climbs steadily and the other lurches. Below three years the figure is too noisy to trust.
Common questions
How do I calculate CAGR?
Divide the ending value by the beginning value, raise that to the power of 1 divided by the number of years, then subtract 1. Growing $10,000 to $15,000 over 5 years gives (15000 / 10000) to the power 0.2, minus 1, which is 8.45%.
How is CAGR different from an average return?
A simple average adds the yearly percentages and divides. CAGR compounds them, so it reflects what you actually kept. Returns of +50% then -30% average to +10%, but the real compounded result is a 5% loss over the two years.
Can CAGR be negative?
Yes. If the ending value is below the beginning value the rate is negative. A fall from $100,000 to $80,000 over 5 years is a CAGR of about -4.4%.
Is a 10% CAGR good?
For a stock portfolio it roughly matches the long-run S&P 500 and is strong. For a startup it may be weak, since early companies often target 25% or more. Always compare against the right benchmark.
