How long does it take the S&P 500 to recover after a major crash?

Recovery timelines in this dataset have varied considerably. Some large declines were followed by strong returns within one to two years. Others, particularly those in the 1930s, involved extended periods of volatility before consistent recovery. The chart above shows the full year-by-year pattern for each decade following a major decline, rather than a single recovery date.

What was the worst S&P 500 crash in history?

Using a peak-to-trough methodology on monthly total return data from 1926 to 2025, the largest nominal decline in this dataset was the Great Depression episode beginning in September 1929, which fell roughly 83% from peak to trough. Among episodes since 1940, the October 2007 Subprime Crisis peak produced the deepest nominal drawdown at approximately 49%, followed by the August 2000 dot-com peak at roughly 42%. The August 1987 episode (Black Monday) produced a peak-to-trough decline of approximately 26%, severe in magnitude but largely invisible in calendar-year data because the market had risen significantly earlier that year. Source: Robert J. Shiller, shillerdata.com.

Is it better to invest after a market crash?

Historical data shows that large declines have generally been followed by periods with above-average returns over the subsequent decade. However, the specific timing and magnitude of recovery have varied, and past patterns do not indicate future results. For investors drawing income from a portfolio, the sequence of returns in the years immediately following a decline matters as much as the eventual recovery.

Do all major crashes eventually recover?

In the historical U.S. data from 1926 to 2025, every major decline has been followed by eventual recovery and new highs. However, this dataset reflects U.S. large-cap equity specifically, over a period when the U.S. economy experienced substantial long-run growth. Other markets and time periods have not always followed the same pattern. Past recovery does not guarantee future recovery.

S&P 500 Crashes and the 10-Year Returns That Followed

Summary

The chart shows each major S&P 500 decline episode (filterable by threshold) identified by its pre-crash peak, and the 10 individual calendar-year returns that followed.
In nominal terms, recovery decades have generally produced positive arithmetic averages, but switching to real returns reveals episodes where inflation eroded those gains below zero.
Many recovery periods contained their own negative years, sometimes severe ones.
For accumulating investors and withdrawing investors, the same recovery pattern can have very different practical outcomes.

Reading this data

The chart identifies major decline episodes using a peak-to-trough methodology: it finds each point where the S&P 500 cumulative total return index fell from an all-time high by at least the selected threshold (adjustable from 10% to 35%), then marks the pre-crash peak as the episode start. For each episode, it shows:

The total peak-to-trough drawdown (the red bar extending left from zero).
The arithmetic average annual return for the 10 calendar years following the peak year.
The compound annual growth rate (CAGR) for the same 10-year period.
A sparkline row showing each of the 10 individual annual returns, colored teal for positive and red for negative, on a consistent scale of -60% to +60%.

Episodes peaking after 2015 are excluded because the full 10-year forward window extends beyond the dataset. Episodes are always identified using nominal cumulative returns. Only the drawdown bar and sparkline values change when switching between nominal and real modes.

The arithmetic average and the compound growth rate for the same recovery period are often meaningfully different. A decade with a large early loss followed by strong subsequent gains may show a healthy arithmetic average but a more modest compound growth rate, because the early loss reduced the compounding base. The sparklines make this dynamic visible at a glance.

What followed each major decline

The chart above shows the data. A few observations that tend to hold across the dataset:

In nominal terms, the arithmetic average return for the 10 years following a major decline has been positive across the cases shown. Switching to real returns reveals that some episodes produced negative average real returns for the following decade, particularly those followed by high-inflation periods.
Some recovery decades began with a strong rebound in the first or second year after the decline. Others took longer to establish a positive trend.
The sparklines illustrate that few recovery decades were uniformly positive. Most contained at least one additional negative year within the recovery window, sometimes several.

The chart includes a toggle between nominal and inflation-adjusted (real) returns. The nominal view shows total returns as reported. The real view adjusts each year's return using the CPI series from Robert Shiller's dataset, showing what purchasing power actually experienced. In high-inflation periods such as the 1970s, real recovery profiles differ substantially from nominal ones. Switching to the real view makes this visible across all episodes shown.

Data: Robert J. Shiller, shillerdata.com. Total return series including dividends reinvested; real returns CPI-adjusted using Shiller's CPI series. Past performance does not indicate future results.

Patterns in the recovery data

The recovery data does not follow a single template. Several patterns are worth noting:

Rebound asymmetry: After severe declines, some of the largest single-year gains in the dataset have appeared in the years that followed. This reflects a mathematical property of percentage returns: gains off a depressed base can be large in percentage terms even when the absolute dollar recovery is modest. A 50% rebound after a 33% loss, for example, restores the original level, but the 50% gain is a larger percentage number than the 33% loss.

Internal volatility in recovery decades: The 1930s and 1940s provide a clear illustration. The decade following the worst years of the Great Depression included both some of the largest single-year gains and additional significant declines. A simple summary of the decade as "positive on average" obscures the experience of investors who lived through those years in sequence.

More recent patterns: Three episodes from the past four decades illustrate how different post-crash decades can look. The August 1987 peak (Black Monday) was followed by one of the stronger 10-year periods in the dataset, coinciding with the 1990s technology expansion. The August 2000 dot-com peak was followed by a more challenging decade: the 10 subsequent years (2001-2010) included both the continued dot-com unwinding and the 2008 financial crisis, compressing the arithmetic average and producing a below-average compound growth rate. The October 2007 episode, by contrast, was followed by a decade of above-average returns through the 2010s, despite beginning with the largest nominal drawdown of any post-1940 episode in this dataset at roughly 49%.

What recoveries mean for accumulation vs. withdrawal

The same historical recovery pattern has different implications depending on whether an investor is accumulating or drawing down.

Accumulating investors who continue making contributions during a decline and recovery period purchase shares at lower prices during the downturn. If the subsequent recovery follows the historical pattern, the lower-cost shares purchased during the decline can contribute meaningfully to long-run outcomes. This is the mechanism behind the conventional guidance to "stay invested through downturns" during the accumulation phase.

Withdrawing investors face a different dynamic. An investor drawing a fixed or flexible amount from their portfolio during the year of a major decline, and in the years immediately following, is selling assets at depressed or recovering prices. This reduces the number of shares available to participate in the subsequent recovery. If the recovery arrives but the portfolio has already been materially reduced by early withdrawals at low prices, the investor captures less of it.

This is the core of sequence-of-returns risk: the order in which returns arrive matters, not just the average. A decade with a poor early sequence and a strong late sequence can produce a very different outcome for a withdrawing investor than the same returns in the opposite order. The sparklines in the chart make the sequence visible, which the arithmetic average and compound growth rate summaries do not.

Coordinating withdrawal amounts, sources, and timing to manage this risk is one of the central functions of retirement income planning. Holding fixed-income assets or other lower-volatility reserves that can fund withdrawals during equity downturns is one approach; Roth conversion strategies and tax-efficient sequencing are others. The details depend heavily on the investor's specific portfolio, tax situation, and income needs.

The limits of this data

A few important constraints apply to any conclusions drawn from this dataset:

U.S. large-cap focus: The S&P 500 reflects large-cap U.S. equities specifically. Historical data from other developed markets (including Japan, which experienced a multi-decade period of stagnant equity prices following its 1989 peak) does not always show the same recovery patterns. This dataset does not represent global equity markets or fixed income.

Survivorship and hindsight: The U.S. equity market's recovery record over the past century reflects a period of substantial economic and institutional development. Future conditions may differ. Constructing a retirement plan around the assumption that past recovery patterns will repeat, without accounting for the possibility that they might not, introduces a meaningful source of planning risk.

Individual timing varies: The chart shows patterns across the full dataset. Individual investors whose circumstances require them to reduce equity exposure at a specific point in a downturn may experience outcomes that differ substantially from the dataset averages.