compoundsimNasdaq and SOX Crashes vs the S&P 500: Falls and Recoveries
How far the S&P 500, the Nasdaq 100 and the SOX semiconductor index fell in the same crashes, how long each took to recover, and the SOX's current fall.
Updated
How far the S&P 500, the Nasdaq 100 and the SOX semiconductor index fell in the same crashes, how long each took to recover, and the SOX's current fall.
Updated
In the 2008 crisis and in 2022, the SOX semiconductor index (as tracked by the SOXX ETF) fell further than both the S&P 500 and the Nasdaq 100: its drawdown reached -60% and -41%. In 2024-25 the SOX alone fell hard. The Nasdaq 100 fell further than the S&P 500 in some crashes and less in others. The index that fell further usually took longer to get back to its high, too: the Nasdaq 100 needed more than 14 years after the dot-com bust. 2022 went the other way, with the SOX falling furthest and recovering first.
The figures use the month-end values of an ETF that tracks each index: SPY for the S&P 500, QQQ for the Nasdaq 100 and SOXX for the SOX. The index SOXX tracks has changed over time; it followed the PHLX Semiconductor Sector Index (the SOX) only from October 2010 to June 2021. This article calls SOXX's values the SOX, as the index data page does (the last section has the details). Dividends are reinvested and everything is in US dollars. The fall is the drawdown: how far below the high the index stood at its lowest. The months count until it regained that high.
The high is the one set just before each crash; an older, higher high is not used. The Nasdaq 100 had not regained its 2000 high when the 2008 crisis began, so measured from its all-time high, the 2008 fall would be lost inside the dot-com one. The SOX's 2007 high, too, was below its level in the data's first month (July 2001).
| Crash | S&P 500 | Nasdaq 100 | SOX |
|---|---|---|---|
| Dot-com bust (from 2000) | -45% (75 months) | -81% (175 months) | no data |
| 2008 crisis (from 2007) | -51% (53 months) | -50% (38 months) | -60% (70 months) |
| Covid crash (2020) | -19% (7 months) | -13% (3 months) | -18% (5 months) |
| 2022 rate rises | -24% (24 months) | -33% (24 months) | -41% (19 months) |
| 2024-25 | -8% (5 months) | -10% (5 months) | -25% (14 months) |
The months in parentheses run from the high before the crash until the index was back at it. SOXX data starts in July 2001, so its 2000 high is missing and its dot-com fall cannot be measured from the high. In 2024-25 the SOX started falling first, from a high in mid-2024; the S&P 500 and the Nasdaq 100 fell in 2025.
Months falling and months recovering, by crash (dividends reinvested, US dollars)
High to troughTrough back to the high
Of the three crashes with data for all three indices (2008, Covid and 2022), the SOX fell furthest in 2008 and 2022. In the 2008 crisis the S&P 500 and the Nasdaq 100 each lost about half, while the SOX's drawdown reached -60%. In 2022 the drawdowns were -24% for the S&P 500 and -41% for the SOX.
The 2024-25 highs came at different times, so it is not quite the same crash, but the SOX started falling first, in mid-2024, and its drawdown reached -25%. The S&P 500's and the Nasdaq 100's in 2025 were -8% and -10%.
The Covid crash is the exception: at month ends the S&P 500 fell furthest and the Nasdaq 100 least. The order is not the same in every crash.
The SOX tends to fall further, which fits how much more it swings in ordinary times. The model behind compoundsim's simulators splits markets into a calm regime and a wild regime (stretches of large swings) and estimates each asset's moves in each from its full history. In the wild regime, annualized volatility (the size of the swings) is 21% for the S&P 500, 37% for the Nasdaq 100 and 52% for the SOX. Even in the calm regime the SOX swings 29%, more than the S&P 500 does in the wild one.
In the wild regime, the correlation of the Nasdaq 100's and the SOX's monthly returns is 0.95: they tend to move in the same direction. Partly because chip makers are a large part of the Nasdaq 100, holding both does little to soften a crash.
In three of the four crashes whose highs line up, the index that fell further also took longer to get back to its high:
An index that swings a lot can also climb a lot on the way back, as the SOX did in 2022. In most crashes, though, the deeper fall took longer to make up. When a crash starts, there is no telling which kind it will be.
The SOX has fallen from a high it set in 2026 and, as of the latest month in this article's data, has been below it for 4 months. In the latest month its drawdown was -26% (-26% at its deepest so far).
Nobody knows how deep this fall will go or when the high will be regained. The recovery times in the table are a record of what happened, not a guide to this one. The SOX index data page shows its drawdown chart month by month.
The figures so far are for buying once at the high and holding. Someone putting the same sum in every month from the high kept buying at lower prices, and those purchases grew in the recovery, so the plan got back to what had gone in sooner (US dollars):
| Crash and index | Deepest fall below principal | Back at principal after | For comparison: lump sum back at the high |
|---|---|---|---|
| Dot-com bust, Nasdaq 100 | -51% | 112 months | 175 months |
| 2008 crisis, SOX | -44% | 36 months | 70 months |
| 2022, Nasdaq 100 | -16% | 15 months | 24 months |
| 2022, SOX | -23% | 13 months | 19 months |
The months back at principal count from the high; a plan that touched its principal and then fell under again counts as back only from the month after it was last under. In 2022, a plan started five years before the high never fell below its principal in any of the three indices: the gains built up before absorbed the fall. Being back at principal is not the same as being back at the largest balance before the crash, though. How long the stock market takes to recover from a crash looks at monthly plans in crashes in more detail.
All of these figures are in US dollars. For an investor counting in another currency, the exchange rate's moves come on top. In 2022 the yen weakened, so in yen, for example, the SOX's drawdown was -26% against -41% in dollars. In the 2008 crisis the yen strengthened, and the SOX's drawdown in yen was -69% against -60% in dollars. On the SOX index data page you can switch the currency and see the drawdowns in your own.
Past crashes do not tell how deep or long the next one will be. To see the range of falls the future might hold, open the Monte Carlo simulator set to a lump sum in the SOX and look at the max drawdown across 1,000 paths. With monthly contributions the drawdown comes out shallower than the index's own, and in another currency it includes the exchange rate's moves. The model's wild regime lasts 3.3 months on average, so slides that run for years, like the dot-com bust, tend to come up less often than in history. How Monte Carlo simulation works for investing explains the simulation, and whether diversification works in a crash looks at falls in mixes of stocks and bonds.