Your portfolio gained 20% one year and lost 20% the next. The arithmetic average return is 0%. But your balance is not back to where it started. It is 4% lower.
Start with 100,000. Gain 20%: 120,000. Lose 20%: 96,000.
That missing 4,000 did not go to fees or taxes. It disappeared into a mathematical phenomenon called variance drain, also known as volatility drag. And it compounds relentlessly over a multi-decade retirement.
How Variance Drain Works
The core mechanic is simple: percentage losses are more damaging than percentage gains of the same size.
A 50% loss requires a 100% gain to recover. A 30% loss requires a 43% gain. A 10% loss requires an 11% gain. The relationship is always asymmetric, and the asymmetry grows with the size of the move.
This means that a portfolio with volatile returns will always compound at a lower rate than its arithmetic average suggests. The gap between the arithmetic mean and the actual compound growth rate (the geometric mean) is approximately:
Variance drain is about half the variance (or equivalently, half the square of standard deviation)
For U.S. equities with a standard deviation of roughly 16%, the drain is about 0.16 squared / 2 = 1.28% per year. That means if the arithmetic average return is 10%, the geometric (compound) return is closer to 8.7%.
Over 30 years, compounding at 10% turns 1,000,000 into 17,449,000. Compounding at 8.7% turns it into 12,138,000. The difference of over 5,000,000 is entirely due to volatility eating into compound growth.
Why It Matters More in Retirement
During accumulation, variance drain reduces your ending balance but you have decades to compensate with additional savings. During retirement, you are withdrawing from a portfolio that is simultaneously being eroded by volatility drag. The two forces compound each other.
Consider a retiree with 1,000,000, withdrawing 40,000 per year (inflation-adjusted), with an arithmetic average return of 8% and a standard deviation of 16%.
Without variance drain (compounding at the arithmetic mean of 8%), the portfolio grows in most years even after withdrawals. With variance drain (actual compound rate closer to 6.7%), the portfolio treads water at best. Add a bad sequence of returns in the first few years, and the combination of withdrawals, sequence risk, and variance drain can be fatal.
This is why portfolios with the same average return but different volatilities produce very different retirement outcomes. A 60/40 portfolio with an 8% arithmetic mean and 10% standard deviation has a variance drain of about 0.5%. An all-equity portfolio with the same 8% arithmetic mean but 16% standard deviation has a drain of 1.28%. The balanced portfolio compounds faster despite having the same average return.
The Hidden Case for Diversification
This is one of the strongest arguments for diversification and it is rarely discussed.
Most people think of diversification as "reducing risk" in the sense of smaller drawdowns and smoother returns. That is true, but it undersells the benefit. Diversification also increases compound growth by reducing variance drain.
If you combine two assets with similar returns but imperfect correlation, the portfolio's volatility is lower than the weighted average of the individual volatilities. Lower volatility means less variance drain. Less drain means higher geometric returns. You end up with more money, not just less risk.
This is why a diversified portfolio often compounds at a higher rate than its "best" component over long periods, even though that component has a higher arithmetic average return. The volatility penalty of the concentrated position eats into its advantage.
Variance Drain Under Fat Tails
The standard formula (drain is about half the variance) assumes a normal distribution. Under fat-tailed distributions, the effective variance is higher because extreme returns contribute disproportionately.
A Student's t-distribution with the same scale parameter as a normal distribution has higher actual variance (the tails add to it). This means the true variance drain under fat tails is worse than the simple formula suggests.
How much worse depends on the degree of tail thickness (the degrees of freedom parameter). With 5 degrees of freedom (moderate fat tails), the effective variance drain can be 30-50% higher than the Gaussian estimate. With thicker tails, the discrepancy grows.
This is yet another way that assuming normal returns leads to overoptimistic projections. You underestimate tail risk and underestimate variance drain simultaneously.
What This Means for Your Retirement Plan
Do not use arithmetic returns in simple calculators. If you plug 10% into a compound growth calculator, you will overestimate your ending balance. Use the geometric return instead (roughly: arithmetic return minus half the variance). Or better yet, use a Monte Carlo retirement calculator that handles the math automatically.
Volatility is not just about comfort. Many investors think of volatility as a psychological issue: can you stomach the drawdowns? But it is also a mathematical issue. Higher volatility means lower compound growth, period. A portfolio that "returns the same on average" but with higher volatility will leave you with less money.
The case for a balanced portfolio is stronger than it looks. When you compare an 80/20 portfolio to a 60/40 portfolio, the arithmetic return difference might be 1%. But after accounting for variance drain, the geometric return difference might be only 0.3-0.5%. The balanced portfolio gives up less compound growth than the arithmetic means suggest, while substantially reducing drawdown risk and sequence-of-returns risk.
Rebalancing helps. Regular rebalancing keeps the portfolio at its target allocation, which tends to sell high and buy low at the margin. This reduces the effective volatility of the portfolio over time, which reduces variance drain. The "rebalancing bonus" is partly a variance drain reduction effect.
Variance drain is not dramatic. It does not show up as a single bad year or a market crash headline. It is a slow, steady erosion that takes 1-2% per year off your compound growth rate, invisible in any single period but devastating over decades. The difference between a retirement plan that works and one that falls short can be exactly this invisible force.