Skip to content

How Much Gain You Need to Recover a Loss

A 20% loss needs a 25% gain; a 50% loss needs 100%. The recovery formula, the years it implies at 4/6/8% growth, and why the average return on your statement is not the return you got.

By the editorial team at The Long View How we check our guides 9 min read
A desk at evening with two monitors showing blurred line charts and a notebook with handwritten figures

A 20 percent loss needs a 25 percent gain to get back to even. That single sentence contains the most under-appreciated piece of arithmetic in investing, and it is the reason two portfolios with identical average returns can end up decades apart in value. The asymmetry is not a trick of presentation; it falls directly out of how percentages work against a shrinking base. Once you internalise it, several pieces of standard advice - don't sell in a panic, rebalance on schedule, hold enough in bonds to sleep - stop sounding like platitudes and start sounding like the only sensible response to the math.

What follows is the calculation, the time it implies, and the one input most people forget: contributions. Nothing here recommends a specific fund or allocation. It is arithmetic you can reproduce on a calculator.

The recovery formula

If a portfolio falls by a fraction d, the gain needed to return to the starting value is:

recovery gain = 1 / (1 - d) - 1

For a 20 percent loss, that is 1/0.8 - 1 = 0.25, or 25 percent. It is not 20 percent, and it is not 22 percent. The reason is that the 20 percent was calculated on the larger, original base, while the recovery is calculated on the smaller, remaining base. Losing 20 percent of $100 leaves $80, and 20 percent of $80 is only $16 - so a symmetrical percentage leaves you at $96.

Drawdown Gain needed to recover Remaining value of $100
-10%+11.1%$90.00
-20%+25.0%$80.00
-30%+42.9%$70.00
-40%+66.7%$60.00
-50%+100.0%$50.00
-60%+150.0%$40.00
-70%+233.3%$30.00

The curve steepens as the loss deepens, which is the whole point. The first 10 percent of loss costs 11 percent of recovery; the loss from 60 to 70 percent costs an additional 83 points of required gain. This is why risk management is cheapest early and most expensive late, and why "I'll sell if it drops another 10 percent" is a much worse plan than it sounds.

The same percentage, different years

Percentages tell you nothing about time. To convert the recovery gain into years, divide the log of the required multiple by the log of your growth rate:

years = ln(1 / (1 - d)) / ln(1 + g)

Drawdown at 4% growth at 6% growth at 8% growth
-10%2.7 yr1.8 yr1.4 yr
-20%5.7 yr3.8 yr2.9 yr
-30%9.1 yr6.1 yr4.6 yr
-50%17.7 yr11.9 yr9.0 yr

Two useful readings. First, a 30 percent drawdown in a portfolio compounding at 6 percent takes six years to undo - not six months. If you are 55 and planning to retire at 60, that is your entire remaining window. Second, the growth rate you use should be the real one if you care about purchasing power: a 6 percent nominal return with 3 percent inflation recovers in nominal terms on schedule but leaves you 17 percent behind in real terms after six years.

A worked example with round numbers

Start with $100,000. A bad year takes 18 percent, leaving $82,000. To get back to $100,000 you need 100/82 - 1 = 21.95 percent, not 18. At a 6 percent compound return that takes 3.4 years; at 8 percent, 2.6 years.

Now the part that most illustrations skip. During those 3.4 years, most people are still contributing.

Annual contribution Months to get back to $100,000 Years
$0403.3
$5,000211.8
$10,000151.2
$20,00090.8

Assumptions: contributions made monthly, 6 percent annual return compounded monthly, no taxes or fees. The magnitude of the effect is the story. At $10,000 a year, the portfolio is back above its old high in fifteen months instead of forty - the drawdown is largely refilled by new money buying at lower prices, which is precisely what "buying the dip" means in practice when you are in accumulation.

The same math is brutal in reverse for withdrawals. A retiree taking money out during a 20 percent decline sells shares at depressed prices and locks in the loss. That is why sequence-of-returns risk dominates retirement outcomes and why the recovery table above is a different problem for a 35-year-old than for a 68-year-old drawing 4 percent a year. The arithmetic is identical; the ability to wait it out is not.

When losses stack

Two consecutive 20 percent losses are not a 40 percent loss. They are a 36 percent loss (0.8 × 0.8 = 0.64), which requires a 56.25 percent gain to recover. Three consecutive 20 percent losses leave 51.2 percent of the original value and require a 95.3 percent gain - essentially a doubling - just to break even.

This is not a theoretical curiosity. It is why the 2000-2002 and 2007-2009 bear markets were so damaging to portfolios that were 100 percent equities: the cumulative drawdown in each was roughly 45 to 50 percent at the index level, requiring a doubling to recover, and the recovery took years rather than months. A portfolio holding 30 to 40 percent in bonds through those periods lost considerably less and needed far less gain to get back, which is the entire mechanical argument for diversification - not that bonds return more, but that they shrink the number you have to climb back from.

The drag hiding inside your expected return

The recovery math also appears inside long-run returns, in the difference between the arithmetic average return and the compound return you actually experience. For a return series with arithmetic mean a and standard deviation s, the compound return is approximately:

geometric ≈ a - s² / 2

At an 8 percent arithmetic mean with 18 percent volatility - typical of an all-equity portfolio - the compound return is about 6.4 percent. The missing 1.6 percentage points per year is volatility drag, and over 30 years that gap is enormous: $1 growing at 8 percent becomes $10.06, while at 6.4 percent it becomes $6.44.

A 60/40 portfolio with a 10 percent arithmetic mean and 12 percent volatility compounds at about 9.3 percent, losing only 0.7 points to drag. At 7 percent mean and 5 percent volatility - a bond-heavy mix - drag is just 0.12 points. Lower volatility does not just feel better; it keeps more of the return you were promised, which compounds into the recovery arithmetic from the other direction.

Why the average return lies

Here is the cleanest demonstration of the same arithmetic. A portfolio gains 25 percent one year and loses 20 percent the next, and repeats that pattern for ten years. Multiply it out: 1.25 × 0.80 = 1.00 exactly. After a decade the portfolio is worth precisely what it started with, and the compound annual return is 0.0 percent.

Now compute the arithmetic average of those ten annual returns: five years at +25 and five at -20 gives +2.5 percent a year. A brochure quoting the average would show a positive number, and the statement would show no growth at all. The gap is not a scam; it is the recovery formula applied repeatedly. The 20 percent loss is measured on a bigger base each time it happens than the 25 percent gain was measured on, so the pair cancels rather than compounds upward.

Two practical consequences follow. When you compare two funds or two strategies, compare compound returns (CAGR or the annualised figure on the factsheet) rather than average annual returns, because only the compound figure describes what your money did. And when you evaluate your own portfolio, ignore the yearly percentages and look at the balance at five-year intervals - the arithmetic mean of your annual returns has almost no relationship to the outcome you actually experienced.

What the math actually implies for decisions

Position size is a recovery decision. Choosing your stock allocation is choosing the drawdown you are willing to climb back from. If a 40 percent decline would take seven years to undo at your expected return, and you cannot tolerate seven years, the allocation is wrong - regardless of what the long-run average says.

Selling converts a temporary number into a permanent one. An unrealised 30 percent drawdown needs a 42.9 percent recovery. A realised one needs the same 42.9 percent, but you now have to be right twice: right to sell, and right to buy back lower. The 2020 recovery took about five months from the trough; investors who sold in March and waited for clarity missed most of it.

Rebalancing is the mechanical version of buying low. Selling the asset that fell to buy the one that held up restores your target risk and, in the process, positions the portfolio to recover from a smaller hole. It is the one action that improves the arithmetic without requiring a forecast.

Cash flow changes everything - in both directions. Contributions shrink recovery time dramatically, as the table above shows. Withdrawals lengthen it just as dramatically. Anyone within five years of drawing on the portfolio should be sizing the recovery math with withdrawals included, not contributions.

For the retirement-specific version of that last point, sequence-of-returns risk runs the same drawdown through a withdrawal schedule. If you want to see why investors consistently earn less than their funds, the behaviour gap quantifies the cost of selling at the wrong moment. And for how allocation itself is chosen rather than assumed, the stock-bond mix covers the trade-off in the same worked-example form.

This article is for general educational purposes only and is not investment, tax, or financial advice. It does not recommend any security, fund, or allocation. Expected returns and volatility figures are illustrative inputs for arithmetic, not forecasts; actual results will differ. Consider consulting a fee-only fiduciary adviser before making decisions based on your own circumstances.