How Much Did Nvidia and the Big Winners Fall Along the Way?
The decade’s 20 best large US stocks, and how far each one fell before, during and after its run: in 2018, 2020, 2022 and 2025. Plus the evidence on whether investors actually hold through falls like these.
Key takeaways
- Nvidia fell 66.4% from peak to trough during the decade to Sep 30, 2026, lost 50.3% in calendar 2022 alone and fell 56% in the late-2018 selloff, yet still returned about 13,575% over the 10 years.
- Among the 20 biggest 10-year winners, the median worst fall was 58.8%, and 16 of the 20 fell by half or more; the S&P 500 (SPY) fell 34.1% at worst.
- Nine of the 20 had a calendar year of −40% or worse. Super Micro Computer fell 84.8% at its worst; Tesla 73.6%.
- Investors often fail to hold through volatility: Morningstar estimated fund investors earned 1.2 percentage points a year less than their funds over the 10 years to 2024, mostly from badly timed buying and selling.
- How much did Nvidia fall before it became a huge winner?
- How far did the decade’s 20 biggest winners fall?
- What would a $50,000 Nvidia position have felt like in 2022?
- Would most investors have held the decade’s best stocks through their crashes?
- How do investors hold a winner through a big drawdown?
- Method and caveats
How much did Nvidia fall before it became a huge winner?
A lot, more than once. In the 10 years to Sep 30, 2026, Nvidia’s worst peak-to-trough fall was 66.4%. It dropped 56% in the late-2018 trade-war selloff and 62% in the 2022 inflation bear market, and it lost 50.3% in calendar 2022. The year after, 2023, it gained 239.2%. Before the decade began, it had fallen 85% in the 2007–09 financial crisis.
Over the full decade Nvidia still returned about 13,575% with dividends reinvested. The gain and the falls came together. Anyone who owned it for the whole run had to watch two-thirds of the value disappear at least once, with no way of knowing the recovery would come.
How far did the decade’s 20 biggest winners fall?
Nearly all of them fell by half or more at some point. The table shows each stock’s 10-year return, its worst fall in the decade, its fall in each of four market selloffs and its worst calendar year.
| Stock | 10-year total return | Worst fall in the decade | 2018 selloff | 2020 COVID crash | 2022 bear market | 2025 tariff selloff | Worst calendar year |
|---|---|---|---|---|---|---|---|
| NVIDIA (NVDA) | +13,575% | −66.4% | −56% | −38% | −62% | −33% | −50.3% (2022) |
| Advanced Micro Devices (AMD) | +8,753% | −65.5% | −49% | −34% | −62% | −32% | −55.0% (2022) |
| Micron Technology (MU) | +6,042% | −57.8% | −37% | −43% | −50% | −38% | −45.9% (2022) |
| Comfort Systems USA (FIX) | +5,891% | −50.2% | −30% | −39% | −24% | −25% | No down year |
| Dell Technologies (DELL) | +4,309% | −60.0% | −29% | −46% | −44% | −41% | −26.6% (2022) |
| Lam Research (LRCX) | +3,834% | −56.8% | −23% | −44% | −56% | −34% | −40.7% (2022) |
| Arista Networks (ANET) | +3,727% | −52.2% | −31% | −31% | −37% | −38% | −15.6% (2022) |
| Natera (NTRA) | +3,620% | −77.7% | −51% | −57% | −69% | −23% | −57.0% (2022) |
| FTAI Aviation (FTAI) | +2,607% | −73.5% | −31% | −74% | −42% | −39% | −25.9% (2022) |
| Broadcom (AVGO) | +2,542% | −48.8% | −15% | −47% | −36% | −36% | −13.3% (2022) |
| Tesla (TSLA) | +2,509% | −73.6% | −22% | −61% | −48% | −38% | −65.0% (2022) |
| Fortinet (FTNT) | +2,319% | −38.3% | −29% | −36% | −31% | −26% | −32.0% (2022) |
| Quanta Services (PWR) | +2,240% | −45.6% | −22% | −41% | −22% | −19% | −22.9% (2018) |
| Lumentum (LITE) | +2,225% | −66.9% | −41% | −29% | −36% | −36% | −50.7% (2022) |
| Marvell Technology (MRVL) | +2,032% | −62.1% | −26% | −32% | −58% | −55% | −57.5% (2022) |
| Teradyne (TER) | +1,855% | −59.7% | −25% | −35% | −57% | −43% | −46.4% (2022) |
| Applied Materials (AMAT) | +1,786% | −55.4% | −27% | −44% | −54% | −28% | −37.5% (2022) |
| Monolithic Power Systems (MPWR) | +1,708% | −51.9% | −23% | −28% | −40% | −34% | −27.8% (2022) |
| Super Micro Computer (SMCI) | +1,655% | −84.8% | −46% | −43% | −31% | −51% | −34.0% (2018) |
| Eli Lilly and Company (LLY) | +1,584% | −34.8% | −10% | −17% | −14% | −22% | No down year |
| S&P 500 (SPY) | +310.7% | −34.1% | −20% | −34% | −25% | −19% | −18.2% (2022) |
- NVIDIA (NVDA)
- 10-year total return
- +13,575%
- Worst fall in the decade
- −66.4%
- 2018 selloff
- −56%
- 2020 COVID crash
- −38%
- 2022 bear market
- −62%
- 2025 tariff selloff
- −33%
- Worst calendar year
- −50.3% (2022)
- Advanced Micro Devices (AMD)
- 10-year total return
- +8,753%
- Worst fall in the decade
- −65.5%
- 2018 selloff
- −49%
- 2020 COVID crash
- −34%
- 2022 bear market
- −62%
- 2025 tariff selloff
- −32%
- Worst calendar year
- −55.0% (2022)
- Micron Technology (MU)
- 10-year total return
- +6,042%
- Worst fall in the decade
- −57.8%
- 2018 selloff
- −37%
- 2020 COVID crash
- −43%
- 2022 bear market
- −50%
- 2025 tariff selloff
- −38%
- Worst calendar year
- −45.9% (2022)
- Comfort Systems USA (FIX)
- 10-year total return
- +5,891%
- Worst fall in the decade
- −50.2%
- 2018 selloff
- −30%
- 2020 COVID crash
- −39%
- 2022 bear market
- −24%
- 2025 tariff selloff
- −25%
- Worst calendar year
- No down year
- Dell Technologies (DELL)
- 10-year total return
- +4,309%
- Worst fall in the decade
- −60.0%
- 2018 selloff
- −29%
- 2020 COVID crash
- −46%
- 2022 bear market
- −44%
- 2025 tariff selloff
- −41%
- Worst calendar year
- −26.6% (2022)
- Lam Research (LRCX)
- 10-year total return
- +3,834%
- Worst fall in the decade
- −56.8%
- 2018 selloff
- −23%
- 2020 COVID crash
- −44%
- 2022 bear market
- −56%
- 2025 tariff selloff
- −34%
- Worst calendar year
- −40.7% (2022)
- Arista Networks (ANET)
- 10-year total return
- +3,727%
- Worst fall in the decade
- −52.2%
- 2018 selloff
- −31%
- 2020 COVID crash
- −31%
- 2022 bear market
- −37%
- 2025 tariff selloff
- −38%
- Worst calendar year
- −15.6% (2022)
- Natera (NTRA)
- 10-year total return
- +3,620%
- Worst fall in the decade
- −77.7%
- 2018 selloff
- −51%
- 2020 COVID crash
- −57%
- 2022 bear market
- −69%
- 2025 tariff selloff
- −23%
- Worst calendar year
- −57.0% (2022)
- FTAI Aviation (FTAI)
- 10-year total return
- +2,607%
- Worst fall in the decade
- −73.5%
- 2018 selloff
- −31%
- 2020 COVID crash
- −74%
- 2022 bear market
- −42%
- 2025 tariff selloff
- −39%
- Worst calendar year
- −25.9% (2022)
- Broadcom (AVGO)
- 10-year total return
- +2,542%
- Worst fall in the decade
- −48.8%
- 2018 selloff
- −15%
- 2020 COVID crash
- −47%
- 2022 bear market
- −36%
- 2025 tariff selloff
- −36%
- Worst calendar year
- −13.3% (2022)
- Tesla (TSLA)
- 10-year total return
- +2,509%
- Worst fall in the decade
- −73.6%
- 2018 selloff
- −22%
- 2020 COVID crash
- −61%
- 2022 bear market
- −48%
- 2025 tariff selloff
- −38%
- Worst calendar year
- −65.0% (2022)
- Fortinet (FTNT)
- 10-year total return
- +2,319%
- Worst fall in the decade
- −38.3%
- 2018 selloff
- −29%
- 2020 COVID crash
- −36%
- 2022 bear market
- −31%
- 2025 tariff selloff
- −26%
- Worst calendar year
- −32.0% (2022)
- Quanta Services (PWR)
- 10-year total return
- +2,240%
- Worst fall in the decade
- −45.6%
- 2018 selloff
- −22%
- 2020 COVID crash
- −41%
- 2022 bear market
- −22%
- 2025 tariff selloff
- −19%
- Worst calendar year
- −22.9% (2018)
- Lumentum (LITE)
- 10-year total return
- +2,225%
- Worst fall in the decade
- −66.9%
- 2018 selloff
- −41%
- 2020 COVID crash
- −29%
- 2022 bear market
- −36%
- 2025 tariff selloff
- −36%
- Worst calendar year
- −50.7% (2022)
- Marvell Technology (MRVL)
- 10-year total return
- +2,032%
- Worst fall in the decade
- −62.1%
- 2018 selloff
- −26%
- 2020 COVID crash
- −32%
- 2022 bear market
- −58%
- 2025 tariff selloff
- −55%
- Worst calendar year
- −57.5% (2022)
- Teradyne (TER)
- 10-year total return
- +1,855%
- Worst fall in the decade
- −59.7%
- 2018 selloff
- −25%
- 2020 COVID crash
- −35%
- 2022 bear market
- −57%
- 2025 tariff selloff
- −43%
- Worst calendar year
- −46.4% (2022)
- Applied Materials (AMAT)
- 10-year total return
- +1,786%
- Worst fall in the decade
- −55.4%
- 2018 selloff
- −27%
- 2020 COVID crash
- −44%
- 2022 bear market
- −54%
- 2025 tariff selloff
- −28%
- Worst calendar year
- −37.5% (2022)
- Monolithic Power Systems (MPWR)
- 10-year total return
- +1,708%
- Worst fall in the decade
- −51.9%
- 2018 selloff
- −23%
- 2020 COVID crash
- −28%
- 2022 bear market
- −40%
- 2025 tariff selloff
- −34%
- Worst calendar year
- −27.8% (2022)
- Super Micro Computer (SMCI)
- 10-year total return
- +1,655%
- Worst fall in the decade
- −84.8%
- 2018 selloff
- −46%
- 2020 COVID crash
- −43%
- 2022 bear market
- −31%
- 2025 tariff selloff
- −51%
- Worst calendar year
- −34.0% (2018)
- Eli Lilly and Company (LLY)
- 10-year total return
- +1,584%
- Worst fall in the decade
- −34.8%
- 2018 selloff
- −10%
- 2020 COVID crash
- −17%
- 2022 bear market
- −14%
- 2025 tariff selloff
- −22%
- Worst calendar year
- No down year
- S&P 500 (SPY)
- 10-year total return
- +310.7%
- Worst fall in the decade
- −34.1%
- 2018 selloff
- −20%
- 2020 COVID crash
- −34%
- 2022 bear market
- −25%
- 2025 tariff selloff
- −19%
- Worst calendar year
- −18.2% (2022)
The 2022 bear market hit the chip stocks hardest: Nvidia, AMD, Lam Research, Marvell, Teradyne and Applied Materials all fell more than 50% within that selloff. Eli Lilly is the outlier, a top-20 winner that never had a down calendar year in the period and whose worst fall was 34.8%. Each selloff is covered in detail in the 2022 bear market and the 2018 trade-war selloff.
What would a $50,000 Nvidia position have felt like in 2022?
Illustrative, using Nvidia’s calendar-year total returns. A $50,000 position at the start of 2022 was worth about $24,850 at the end of that year. The same $50,000 in the S&P 500 (SPY) was worth about $40,900. Holding on through 2023 and 2024 turned the Nvidia position into about $228,700. Selling at the end of 2022 locked in a $25,150 loss.
| Illustrative $50,000 at the start of 2022 | End of 2022 | End of 2023 | End of 2024 |
|---|---|---|---|
| Nvidia (NVDA) | $24,850 | $84,291 | $228,682 |
| S&P 500 (SPY) | $40,900 | — | — |
The math of a fall is unforgiving. A 66.4% decline needs a gain of about 198% just to get back to even; see loss recovery math. On a $1,000,000 portfolio with 15% in one such stock, the same fall would remove about $100,000, which is where many investors sell.
Would most investors have held the decade’s best stocks through their crashes?
The evidence says many would not. Morningstar’s Mind the Gap 2025 study estimated that the average dollar invested in US mutual funds and ETFs earned 7.0% a year over the 10 years to Dec 31, 2024, against 8.2% a year for the funds themselves. The 1.2-point gap, about 15% of the return, comes from investors buying and selling at the wrong times. Sector equity funds had the widest gap of the fund groups: investors earned 7.0% a year against the funds’ 8.5%.
Those are funds, which are far less volatile than a single stock that falls 60%. With an individual winner the pressure to sell is greater, and the reasons to sell sound better at the time: a cycle peaking, a competitor, a guidance cut. Most of the decade’s winners also spent whole years behind the index; the median top-20 winner trailed the S&P 500 in 3 of the 9 calendar years from 2017 to 2025.
How do investors hold a winner through a big drawdown?
Mostly by deciding before the fall. A position small enough to survive a 60% drop is one you can keep; a position that would wreck the portfolio is one you will sell. Investors who held through these crashes often had three things in common: a sized position, a written reason for owning it, and a rule for what would change their mind.
- Size each stock so that its worst historical fall would cost an amount you can accept.
- Write down why you own it and which facts, not price moves, would make you sell.
- Check how the whole portfolio would have done in 2018, 2020 and 2022, not just the stock.
- Decide in advance whether you will trim a winner that grows past a set weight.
More in drawdown tolerance, single-stock concentration limits and how to find the next Nvidia.
Method and caveats
Data: the 20 largest 10-year total returns among the largest US companies listed on the NYSE or Nasdaq (common shares, roughly $17 billion and up at the October 2026 export), from Exent’s market data, 10 years to Sep 30, 2026. “Worst fall” is the largest peak-to-trough decline in the period. Selloff columns are each stock’s peak-to-trough fall within the selloff window ending Dec 24, 2018, Mar 23, 2020, Oct 12, 2022 and Apr 8, 2025. Calendar years are 2017 to 2026 year to date.
- Survivorship bias: these are the stocks that recovered. Many stocks that fell as far never came back and are not in the table.
- Past drawdowns do not set a limit on future ones.
- Illustrative dollar figures use calendar-year returns and ignore taxes and costs.
How to see your own crash exposure with Exent
- Upload a screenshot of your holdings (or a CSV, Excel, PDF or Google Sheet); no brokerage login.
- Exent replays your portfolio through 2008, 2018, 2020, 2022 and 2025 and shows how far it would have fallen in each.
- Use the what‑if simulator to change a winner’s weight and see how much smaller the next crash would have been before you trade.
Frequently asked questions
Nvidia lost 50.3% in calendar 2022 and fell about 62% from peak to trough in the 2022 bear market. In 2023 it gained 239.2%.
Its worst peak-to-trough fall in the 10 years to Sep 30, 2026 was 66.4%. Earlier, in the 2007–09 financial crisis, it fell about 85%.
Yes. Among the 20 biggest 10-year winners among large US stocks, the median worst fall was 58.8% and 16 of the 20 fell by half or more, against a worst fall of 34.1% for the S&P 500.
Evidence suggests many do not. Morningstar estimated that fund investors earned 1.2 percentage points a year less than their funds over the 10 years to 2024 because of badly timed buying and selling, and the gap was widest in sector funds.
See what your portfolio is really made of.
Upload a screenshot of your holdings and get concentration, overlap, crash exposure and a risk-matched benchmark in about a minute.
Educational content, not investment advice. Examples are illustrative; past performance does not guarantee future results.