Skip to main content

When the Rule of 72 Stops Being Accurate

The Rule of 72 is a powerful approximation—so powerful that finance textbooks rarely mention its limitations. Yet like all heuristics, it degrades under stress. When interest rates plummet below 1%, spike above 15%, or when your portfolio contains structured products with exotic compounding, the Rule of 72 begins to lie. This article maps the boundaries of accuracy, explains why those boundaries exist, and teaches you how to recognize when you need exact calculations instead.

Quick definition​

The Rule of 72 accuracy threshold is the range where 72 / r produces doubling time within roughly 10% of the actual result. This range spans approximately 2% to 10% for practical finance; beyond it, error accumulates rapidly. Below 1% or above 20%, the rule should be discarded in favor of exact formulas or logarithmic calculators.

Key takeaways​

  • Rule of 72 degrades below 1% and above 15%, with error compounding as rates diverge from the 5–8% sweet spot.
  • Low rates are dangerous: A 0.5% savings account rate suggests 144 years to double (Rule of 72) when the actual time is ~139 years—a 3.6% error that matters for retirement planning.
  • High rates are equally problematic: A 30% growth rate suggests 2.4 years (Rule of 72) versus the actual 2.64 years (9.9% error)—significant when evaluating startup valuations or emerging-market opportunities.
  • Negative rates and deflation break the rule entirely. When r is negative, the rule's division produces misleading negative doubling times or suggests money will never double.
  • The error source is mathematical, not mysterious: The Rule of 72 relies on approximations (Taylor expansions of logarithms) that lose validity far from r ≈ 0.
  • Exact formulas exist and are simple: t = ln(2) / ln(1 + r) requires only a calculator, not complex derivations.

The Mathematical Roots of Rule Failure​

To understand where Rule of 72 breaks down, revisit its derivation. The exact doubling-time formula is:

t = ln(2) / ln(1 + r)

The Rule of 72 assumes two simplifications:

  1. First approximation: For small r, ln(1 + r) ≈ r. This is a first-order Taylor expansion around r = 0. It works well when r is small (close to zero).

  2. Second approximation: ln(2) ≈ 0.693, so 0.693 / r ≈ 0.72 / r. This rounds the numerator for divisibility.

Both approximations hold near the origin (small r) but diverge as r grows. The further r moves from zero, the worse the approximation becomes.

Let's verify this mathematically. The Taylor series for ln(1 + r) is:

ln(1 + r) = r - r²/2 + r³/3 - r⁴/4 + ...

For small r, the first term (r) dominates. But as r grows, the remaining terms become significant:

  • At r = 0.02 (2%), the error term r²/2 ≈ 0.0002, barely visible.
  • At r = 0.10 (10%), the error term r²/2 ≈ 0.005, now noticeable.
  • At r = 0.20 (20%), the error term r²/2 ≈ 0.02, substantial.
  • At r = 0.50 (50%), the error term r²/2 ≈ 0.125, dominating.

This explains why Rule of 72 stays accurate (within 5–10% error) in the 2–10% band but fails catastrophically at 0.5% or 50%. The approximation error isn't random—it's systematic and mathematically predictable.

The Low-Rate Problem: Below 1%​

Savers know this pain. When your savings account earns 0.5%, the Rule of 72 suggests:

t = 72 / 0.5 = 144 years

The actual time (using the exact formula):

t = ln(2) / ln(1.005) = 0.6931 / 0.004988 ≈ 138.98 years

The Rule of 72 overshoots by 5 years—a 3.6% error. This seems small in absolute terms, but in practice, it's the difference between doubting whether your money will ever compound meaningfully and having a concrete (if depressing) forecast. Over a 50-year planning horizon, this error matters.

The fundamental issue: At very low rates, the Taylor approximation ln(1 + r) ≈ r breaks down because omitted higher-order terms become proportionally larger. When r = 0.005, the term r²/2 = 0.0000125 is 1.25% of the primary term r = 0.005. It's small in absolute value but visible in the quotient.

Why this matters: During deflationary periods (like parts of the 2008 recession) or in countries with near-zero rates (Japan, 2010–2019), Rule of 72 produces misleading guidance. A Japanese investor with money in a 0.1% account might conclude, "This will never double in my lifetime" (720 years by Rule of 72), when the actual answer is "Yes, in roughly 693 years"—still depressing, but not quite apocalyptic.

For precise planning at low rates, use the exact formula. Or better: recognize that below 1%, you're not actually planning around "doubling time"—you're acknowledging that purchasing power is eroding faster than money is compounding. The rule's failure is less important than the economic reality it reveals.

The High-Rate Problem: Above 15%​

Startup valuations, emerging-market growth, and leveraged positions all operate above 15%. Here, Rule of 72 fails in the opposite direction—it becomes overly optimistic.

At 25% annual growth (not uncommon for successful tech startups):

Rule of 72: t = 72 / 25 = 2.88 years Actual: t = ln(2) / ln(1.25) = 0.6931 / 0.2231 ≈ 3.11 years

The Rule of 72 understates doubling time by 0.23 years (about 2.8 months). For a 7-year VC fund evaluating if an investment will 2x before exit, this error compounds. If the fund assumes Rule of 72 for all 25% positions, it systematically overestimates exit multiples.

At 40% growth (venture capital's aspirational target):

Rule of 72: t = 72 / 40 = 1.8 years Actual: t = ln(2) / ln(1.40) = 0.6931 / 0.3365 ≈ 2.06 years

The error is now 0.26 years (about 3 months), or 14% of the actual time. A VC expecting a 2x in 1.8 years will be disappointed.

Why this happens: At high rates, the terms ln(1 + r) deviates significantly from r itself. The series ln(1 + r) = r - r²/2 + ... shows that the negative term r²/2 becomes substantial. For r = 0.40, r²/2 = 0.08, meaning ln(1.4) ≈ 0.40 - 0.08 + ... ≈ 0.32 (actual: 0.336). The rule's denominator should be 0.336, not 0.40 (its underlying assumption), making the quotient smaller and doubling time longer.

The Boundary Map: Error by Rate​

Here's a quantitative map of where Rule of 72 breaks down:

RateActual (yrs)Rule 72 (yrs)Error (yrs)% Error
0.5%138.98144.00+5.02+3.6%
1%69.6672.00+2.34+3.4%
2%35.0036.00+1.00+2.9%
3%23.4524.00+0.55+2.3%
5%14.2114.40+0.19+1.3%
7%10.2410.29+0.05+0.5%
8%9.019.00–0.01–0.1%
10%7.277.20–0.07–1.0%
15%4.964.80–0.16–3.2%
20%3.803.60–0.20–5.3%
25%3.112.88–0.23–7.4%
30%2.642.40–0.24–9.1%
40%2.061.80–0.26–12.6%
50%1.711.44–0.27–15.8%

The table reveals a clear pattern:

  • Below 8%: Rule of 72 overshoots (positive error). The lower the rate, the larger the error percentage.
  • At 8%: Rule of 72 is nearly perfect (it was designed around this benchmark).
  • Above 8%: Rule of 72 undershoots (negative error). The higher the rate, the larger the error percentage.
  • Tolerable error zone: 2–15% rates keep error below 5%, acceptable for most financial decisions.
  • High-error zone: Below 1% or above 20% produces errors exceeding 5%, requiring exact calculations.

Negative Growth: When the Rule Collapses​

What happens when you apply Rule of 72 to negative growth—deflation, currency devaluation, or a shrinking company?

At –2% annual decline (like slowly shrinking purchasing power):

Rule of 72: t = 72 / –2 = –36 years

What does a negative time mean? The rule breaks down conceptually. The exact formula still works:

t = ln(2) / ln(1 – 0.02) = ln(2) / ln(0.98) = 0.6931 / –0.0202 ≈ –34.3 years

The negative sign indicates that instead of doubling, the money halves. But the phrasing becomes confusing. You can't say money "doubles in negative time." Instead, rephrase as: "Money halves in 34.3 years at a 2% annual loss."

Real example: During the 2008 financial crisis, some homeowners saw property values decline 5% annually. Rule of 72 would give –72/5 = –14.4 years, meaning a $500,000 home would become $250,000 in value in 14.4 years. The exact formula confirms: t = ln(2) / ln(0.95) ≈ 14.2 years. Here, both rule and exact formula agree because the rate is close to zero in mathematical terms (5% is relatively small), so the Taylor approximation still holds.

Deflationary scenarios: If inflation becomes deflation at –3% per year, savers are happy (their money buys more tomorrow), but Rule of 72 again struggles:

Rule of 72: 72 / –3 = –24 years (nonsensical phrasing) Exact: t = ln(0.5) / ln(0.97) ≈ 22.8 years until savings have lost half their nominal value (though purchasing power has increased)

The rule's failure here is largely semantic, but it highlights that Rule of 72 is optimized for positive growth scenarios. Once you cross into decline, the approximation becomes confusing and should be abandoned in favor of exact formulas.

Time Horizon Effects: Short Intervals​

A subtler failure mode appears when analyzing short time periods. Suppose a crypto asset rises 200% in a month (an annualized 1,260%, not uncommon in 2017–2021). Rule of 72 suggests:

t = 72 / 1260% = 0.057 years ≈ 3 weeks

Actual: t = ln(2) / ln(3.00) = 0.6931 / 1.0986 ≈ 0.631 years ≈ 7.6 months

At extreme rates, Rule of 72 is off by a factor of more than 10. This isn't surprising—the rule was designed for annual compounding over years, not intramonth or intraday moves.

Why this matters: Traders and analysts sometimes casually apply Rule of 72 to ultra-high-frequency scenarios (hourly or daily returns) and get wild misguidance. A stock rallying 5% per day would suggest (72 / 5% = 1,440 days ≈ 3.9 years) by Rule of 72, but the exact formula gives t = ln(2) / ln(1.05) ≈ 14.2 days. The rule is 97 times too conservative because it was never meant for intraday volatility.

The deeper issue: Rule of 72 assumes steady-state compounding, where the same rate repeats every period. Real markets violate this constantly. A stock that gains 5% one day and loses 4% the next isn't compounding at a steady 2.5% daily rate. This is why applying Rule of 72 to individual stock movements or short-term trading is fundamentally misguided—not because the approximation is weak, but because the underlying assumption (constant rate) is violated.

Discrete vs. Continuous Mismatch​

Rule of 72 implicitly assumes discrete annual compounding:

FV = PV × (1 + r)^t

But many financial products compound more frequently (daily, continuous) or less frequently (quarterly, monthly). When you misapply Rule of 72 to a product with different compounding, error emerges.

A bond paying 6% compounded semiannually (twice per year):

FV = PV × (1 + 0.06/2)^(2t) = PV × (1.03)^(2t)

To find doubling time:

2 = (1.03)^(2t) ln(2) = 2t × ln(1.03) t = ln(2) / (2 × ln(1.03)) ≈ 0.6931 / (2 × 0.02956) ≈ 11.72 years

Rule of 72 (assuming annual compounding) gives: 72 / 6 = 12 years, an error of about 2.4%.

A bond paying 6% compounded continuously:

2 = e^(0.06t) t = ln(2) / 0.06 ≈ 11.55 years

Rule of 72 overshoots again (12 years vs. 11.55 actual). The approximation error compounds when compounding frequency doesn't match the rule's implicit assumptions.

What to do: If the compounding frequency is stated, use the exact formula. If you must use a rule, apply:

  • Rule of 69 for continuous compounding.
  • Rule of 70 for quarterly or semiannual compounding.
  • Rule of 72 only when annual compounding is confirmed or when quick mental math is more valuable than precision.

Structural Products and Exotic Terms​

Modern finance has created investment structures that violate Rule of 72's assumptions entirely. Consider:

  • Inverse ETFs (go down when the market goes up): Applying Rule of 72 to an inverse product gives negative "doubling" times, which is nonsensical.
  • Leveraged ETFs (amplify daily moves by 2x, 3x, etc.): These don't compound linearly. A 2x leveraged ETF in a trending market can double faster than Rule of 72 suggests because of path effects; in a choppy market, it can underperform due to volatility drag.
  • Convertible bonds (fixed income with embedded equity option): The "interest rate" isn't a fixed r—it's dynamic based on the underlying stock. Rule of 72 assumes a stable, predictable rate.
  • Callable bonds (issuer can redeem early): The doubling time is capped; if rates fall and the bond is called, you never reach 2x. Rule of 72 ignores this embedded optionality.

For these products, Rule of 72 is not merely inaccurate—it's misleading because the underlying assumptions don't hold at all. You need scenario analysis, simulation, or exact present-value calculations.

When to Abandon the Rule Entirely​

Stop using Rule of 72 (or any approximation rule) and reach for exact calculations in these scenarios:

  1. Rates below 1% or above 20%: Error exceeds 5%. Use t = ln(2) / ln(1 + r).
  2. Negative or zero rates: The rule produces nonsensical results. Use exact formulas.
  3. High-stakes decisions: A retirement plan, long-term funding strategy, or major capital allocation deserves precision, not approximation.
  4. Compounding frequencies other than annual: Verify the stated frequency and use the exact formula tailored to it.
  5. Volatile or path-dependent returns: Crypto daily moves, leveraged ETFs, or derivatives where the "constant rate" assumption breaks.
  6. Very short or very long time horizons: The rule is optimized for 10–30 year forecasts; beyond that, small rate changes matter enormously.
  7. Regulatory or audit contexts: Regulators and auditors expect precision. Approximations are for planning, not reporting.

Rule of 72 Accuracy Decision Tree​

Real-World Examples of Rule Failure​

Example 1: Japanese Savers, 1995–2015

Japan's interest rates hovered near 0.5% for decades. A saver with ¥10 million in a 0.5% account would think (using Rule of 72): "I'll never see this double—144 years!" The exact answer (139 years) is barely better, but the psychological difference matters. Better insight: "Your money will double once in your retirement planning horizon, just slowly." Rule of 72 magnified despair.

Example 2: Emerging Market Boom

India's real GDP growth averaged 6.5% from 2003–2008. Rule of 72 suggested doubling in 72/6.5 = 11.1 years. The exact formula: t = ln(2) / ln(1.065) ≈ 11.0 years. Here Rule of 72 was spot-on, validating its use in normal economic growth forecasts.

Example 3: Crypto Bubble

Bitcoin rallied from $1 to $19,000 in seven years. That's roughly 3,000x growth. Rule of 72 is useless here (it applies to doubling times, not 3000x). Even knowing the annualized return (~212%), Rule of 72 would suggest 72/212 = 0.34 years ≈ 4.1 months to double, when the actual average (including 2018 crash) was more volatile. The rule's failure reveals a deeper lesson: in speculative markets, "doubling time" is the wrong metric; volatility and sequence-of-returns matter more.

Example 4: Deflation in the 2008 Crisis

Property values in Las Vegas fell 5% annually from 2006–2012. Rule of 72 suggested home values would halve in 72/5 = 14.4 years. The exact formula: t = ln(2) / ln(0.95) ≈ 14.2 years. Both methods agreed because 5% is moderate-sized relative to the r ≈ 0 Taylor approximation. Rule of 72 held up reasonably here.

Common Mistakes​

Mistake 1: Using Rule of 72 for sub-1% interest. A savings account at 0.4% will double in ~173 years (exact), not 180 (Rule of 72). Worse, inexperienced savers often conclude "my money never grows"—mathematically wrong and psychologically destructive.

Mistake 2: Applying Rule of 72 across different compounding frequencies. A CD compounded daily, a Treasury bond compounded semiannually, and a stock ETF's reinvested dividends (approximately continuous) are not comparable using Rule of 72 alone. The rule assumes annual compounding, and violations produce systematic bias.

Mistake 3: Treating Rule of 72 as precise in high-rate scenarios. A startup growing 40% annually is told it will 2x in 1.8 years (Rule of 72). Actual: 2.06 years. The 3-month error might seem small, but it compounds across multiple doubling cycles and affects exit timing in VC fund modeling.

Mistake 4: Using Rule of 72 for volatile, short-term moves. A day trader seeing 10% daily gains might compute "Rule of 72 suggests 7.2 days to double." This assumes the 10% rate persists (unlikely in real trading) and ignores volatility drag. The rule is designed for years, not days.

Mistake 5: Forgetting that the rule is context-dependent. Students often memorize Rule of 72 as a timeless truth, forgetting it's an approximation born from Taylor series and divisibility constraints. It's not physics—it's convenience.

FAQ​

Q: Why is 8% special for Rule of 72?

A: Because 72 / 8 = 9 years, and the exact formula gives t = ln(2) / ln(1.08) ≈ 9.01 years. The rule was reverse-engineered (or discovered) to match this common historical stock return, so it's optimized for that point. Rates above or below 8% accumulate error in opposite directions.

Q: Should I memorize Rule of 70 instead of 72?

A: If you're studying macroeconomics or institutional finance, Rule of 70 is more universally useful. If you're doing back-of-envelope financial planning, Rule of 72 is sufficient. In modern practice with calculators ubiquitous, neither is essential—knowing they exist and when to use them matters more than memorization.

Q: Can I fix Rule of 72 errors by using a different numerator?

A: Yes. Rule of 69 targets continuous compounding. Rule of 70 balances multiple compounding regimes. But these are different approximations for different scenarios, not "fixes" to 72. Each works best in its designed range.

Q: What's the exact formula I should use if I don't trust the rules?

A: t = ln(2) / ln(1 + r), where r is the decimal interest rate (so 5% becomes 0.05). Plug into any calculator or spreadsheet. This is exact for annual discrete compounding. For other frequencies, adjust the formula according to the compounding method.

Q: Does Rule of 72 work for inflation?

A: Yes, if you use the inflation rate as r. At 3% inflation, purchasing power halves (rather than your money doubles) in 72/3 = 24 years. Exact: t = ln(2) / ln(1.03) ≈ 23.45 years. Rule of 72 is acceptable but Rule of 70 is better (23.3 years).

Q: Why do some sources cite different "magic numbers" like 69.3 or 70?

A: Different numerators reflect different assumptions (continuous vs. discrete compounding, different optimization targets). They're not errors—they're variations on a theme, each best for specific scenarios.

Summary​

The Rule of 72 breaks down when interest rates leave the 2–10% sweet spot, when compounding frequencies diverge from the annual assumption, or when negative growth enters the picture. The mathematical root cause is the Taylor approximation on which it's built—valid only for small values of r. Understanding where and why the rule fails transforms it from a magic number into a useful tool with clear boundary conditions. For retirement planning, mortgage analysis, or general financial intuition, Rule of 72 remains excellent. For precision, emerging markets, deflationary scenarios, or high-frequency trading, the exact formula is worth the extra keystroke. The choice between approximation and precision should be deliberate, not unconscious.

Next​

Using the Rule of 72 for Inflation — Apply the rule to purchasing power erosion and see how inflation transforms personal savings over time.