Approximating Exponential Growth
The Rule of 70 and its relatives turn 'how long to double at rate r' into a one-line mental calculation, using the small-rate approximation that underlies compound growth.
Prerequisites: Estimating Logs and Powers of Two
An investment grows 7% a year. Roughly how many years to double? You could set up and solve with logs, but there's a shortcut that skips the calculator entirely and gets within a fraction of a year: the Rule of 70.
Where the rule comes from
Growth at rate (as a decimal) for years multiplies your value by . For small , — a standard small-number approximation. Doubling means , i.e. . Substituting the approximation:
In plain English: the number of periods to double is roughly a fixed constant (about 0.693, or 69.3%) divided by the growth rate — and using 70 instead of 69.3 as the constant makes the arithmetic even easier while staying accurate for typical interview rates (roughly 1% to 20%).
Worked example 1: doubling at 7%
Check against the exact answer: , close to 2 — the Rule of 70 slightly underestimates here but lands within a few percent, well inside what's needed for a mental estimate. The approximation gets slightly worse as grows larger, since the underlying step is only exact in the limit as .
Worked example 2: doubling at 2% vs 20%
At 2%: years. At 20%: years. Notice the relationship is inversely proportional, not linear — going from 2% to 20% (a 10x increase in rate) cuts the doubling time by 10x too (35 down to 3.5), a direct consequence of the form. This inverse relationship is worth internalizing on its own: doubling the growth rate roughly halves the time to double, and it's exactly this pattern that makes compounding at even modestly higher rates dramatically more powerful over decades.
What this means in practice
The Rule of 70 (and its cousins — Rule of 72 for rates that divide nicely by whole numbers, like 6% or 8%) is a workhorse for any quick compounding estimate: how long until a fund doubles, how fast a liability grows, or roughly how many compounding periods a strategy needs to reach a target. It's built on the same small-rate log approximation that underlies continuous compounding more generally, so understanding why it works (not just the number 70) transfers directly to other exponential-growth mental-math questions.
Doubling time at rate is approximately years, derived from for small and . The relationship is inverse: doubling the rate roughly halves the doubling time.
Use 72 instead of 70 when divides evenly into it (6, 8, 9, 12) — the extra divisibility makes the arithmetic even cleaner with negligible loss of accuracy for typical rates.
Practice in interviews
Further reading
- Common quant interview prep guides (mental math drills)