The S&P 500 closed at 6,910 on Friday 20 November 2026, according to Yahoo Finance daily closes, 11.6% below its high of 7,819 on 6 October. Since 17 November the index has traded below its 200-day moving average, the average of the last 200 daily closes, which now stands at 7,349. Apple is the exception. The stock closed at $342.56, only 3.8% below its record of 13 November and 13.0% above its own 200-day moving average.
The usual focus in such phases is whether the index can reclaim its 200-day moving average. Alongside it, a second signal has appeared at every low without exception: in every correction of 10% or more since 2015, Apple broke below its 200-day moving average before the index bottomed. This time the heaviest stock in the index is holding up. As long as that remains the case, the signal that accompanied all eight lows of the past eleven years is missing.
What follows is derived from 15 completed corrections since 1998 and from the index weights published by State Street for 19 November. It is not a forecast. The analysis tests the thesis in two strengths, looks for today's set-up in earlier episodes, breaks the decline down by the seven largest technology stocks and derives the index levels that an Apple correction along historical lines would imply.
Thesis: In 8 of 8 corrections since 2015, Apple closed below its 200-day moving average in the 20 sessions before the index low. Between 1998 and 2014 that held in only 2 of 7 cases. Apple's own low came no later than ten sessions after the index low in 6 of the 8 cases since 2015.
Set-up: When the S&P 500 was at least 10% below its high while Apple still traded above its 200-day moving average, the index went on to fall further in 8 of 9 earlier episodes, by a median 9.9%. Since 2015 the set-up has occurred twice, in 2020 and 2022, with further declines of 24.9% and 16.9%.
Magic 7: Apple, Microsoft, NVIDIA, Amazon, Alphabet, Meta and Tesla have lost 18.8% on a cap-weighted basis since the high and account for 56% of the index decline. Apple is the only one of the seven that rose. When the other six were weak and Apple strong, Apple followed them down within 120 sessions in 7 of 8 phases since 2013.
Projection: If Apple catches up to its typical correction low since 2015, around $262, the base case for the S&P 500 is 6,060 to 6,360. If instead the rest of the Magic 7 recover while Apple holds, the index would rise to 7,250 to 7,440, and the thesis would be refuted.
Apple and the S&P 500 have been moving apart since October
According to Yahoo Finance price data, Apple has risen 2,361-fold since January 1998 and the S&P 500 just over sevenfold, so the two are hard to compare in scale. On a log scale, where equal distances mean equal percentage moves, both charts plot the daily close and the 200-day moving average. Dotted verticals mark the lows of the index's 15 completed corrections since 1998, drawn darker from 2015 and lighter before, and show where Apple stood at each.
The view since 2024 shows how the two have diverged in recent weeks. Since the high of 6 October, the S&P 500 has fallen first through its 50-day moving average at 7,625 and then through its 200-day line. Over the same period Apple set a new record and still trades just above its 50-day moving average of $341.81. It would have to fall more than 11% to reach its 200-day line at $303.24. That gap separates today from the pattern of every index low since 2015.
To compare the two over almost 30 years, both prices are expressed as their distance from their own 200-day moving average. Apple stands at +13.0% today, the S&P 500 at −6.0%. At each of the eight index lows since 2015, Apple's line sat below zero on this measure, while the index itself was marginally above it once, in February 2018. The test that follows uses this measure, because it captures price level and trend in one number.
Since 2015, Apple was below its 200-day line at every S&P 500 low
A correction here is any phase in which the S&P 500 traded at least 10% below its high of the preceding 252 sessions. It ends when the index comes back to within 3% of that rolling high. Measuring against the rolling rather than the all-time high also captures the corrections of 2010 and 2011, which fell within the long recovery after 2009 and would vanish from a count against the all-time high. Both are counterexamples to the thesis and are therefore included.
The thesis can be tested in two strengths. The weak version asks whether Apple had visibly corrected before the index low, measured as a close below its 200-day moving average on at least one of the 20 sessions before the low. It held in 8 of 8 cases since 2015, in only 2 of 7 between 1998 and 2014, and in 10 of 15 overall. The strict version also requires Apple's own low to come no later than ten sessions after the index low. It held in 6 of 8 cases since 2015. In 2016 and 2022, Apple bottomed 65 and 61 sessions later respectively, and the index low held anyway.
| S&P high | S&P low | S&P from high | S&P vs 200d | Apple from high | Apple vs 200d | Apple below 200d | Apple low | Lag (sessions) |
|---|---|---|---|---|---|---|---|---|
| 17 Jul 1998 | 31 Aug 1998 | −19.3% | −9.6% | −27.5% | +21.1% | no | – | – |
| 16 Jul 1999 | 15 Oct 1999 | −12.1% | −4.8% | −5.7% | +54.0% | no | – | – |
| 24 Mar 2000 | 14 Apr 2000 | −11.2% | −1.9% | −22.4% | +24.4% | no | – | – |
| 24 Mar 2000 | 9 Oct 2002 | −49.1% | −24.3% | −81.1% | −33.1% | yes | 9 Oct 2002 | 0 |
| 9 Oct 2007 | 9 Mar 2009 | −56.8% | −36.1% | −58.4% | −34.2% | yes | 20 Jan 2009 | −34 |
| 23 Apr 2010 | 2 Jul 2010 | −16.0% | −8.0% | −9.9% | +12.6% | no | – | – |
| 29 Apr 2011 | 3 Oct 2011 | −19.4% | −14.1% | −9.4% | +6.0% | no | – | – |
| 21 May 2015 | 25 Aug 2015 | −12.4% | −10.1% | −21.8% | −14.6% | yes | 24 Aug 2015 | −1 |
| 21 May 2015 | 11 Feb 2016 | −14.2% | −10.1% | −29.4% | −19.2% | yes | 12 May 2016 | 65 |
| 26 Jan 2018 | 8 Feb 2018 | −10.2% | +1.7% | −13.4% | −2.9% | yes | 8 Feb 2018 | 0 |
| 20 Sep 2018 | 24 Dec 2018 | −19.8% | −14.6% | −36.7% | −24.2% | yes | 3 Jan 2019 | 8 |
| 19 Feb 2020 | 23 Mar 2020 | −33.9% | −26.3% | −31.4% | −9.8% | yes | 23 Mar 2020 | 0 |
| 3 Jan 2022 | 12 Oct 2022 | −25.4% | −14.3% | −24.0% | −12.8% | yes | 5 Jan 2023 | 61 |
| 31 Jul 2023 | 27 Oct 2023 | −10.3% | −2.9% | −14.4% | −1.4% | yes | 26 Oct 2023 | −1 |
| 19 Feb 2025 | 8 Apr 2025 | −18.9% | −13.4% | −33.4% | −24.6% | yes | 8 Apr 2025 | 0 |
| 6 Oct 2026 | 20 Nov 2026 (open) | −11.6% | −6.0% | −3.8% | +13.0% | no | – | – |
There is a plausible, untested explanation for the break around 2015: the weight of the large technology stocks in the index and the share of passively managed money have both grown markedly since then. According to State Street, Apple now makes up 8.24% of the S&P 500. In a broad correction, index funds, futures hedgers and risk models sell the heavyweights pro rata along with everything else. Institutional investors typically cut losing positions first and sell holdings still in profit last. If the heaviest constituent has not yet been sold, that last phase may still be to come.
Today's set-up has marked the low only once before
A second test starts not from the known low but from the situation in place today, because the low is only ever known in hindsight. It looks for every correction in which the S&P 500 was at least 10% below its high while Apple still traded above its 200-day moving average, and measures from the first day of that set-up to the later index low. This applied to 9 of the 15 episodes, and in 8 of them the index went on to fall further, by a median 9.9%. Applied to today's level, that would mean around 6,225.
The dispersion is wide, however, and argues for caution with any single number. The range runs from zero, in spring 2000, when the first day of the set-up was also the low, to −51.9% in the financial crisis. Since 2015 the set-up has occurred only twice: in 2020 the index fell a further 24.9%, in 2022 a further 16.9%. Two cases do not make a law. In both, most of the decline still lay ahead after the first day of the set-up.
| S&P high | First day of set-up | Apple vs 200d | S&P low | Further decline |
|---|---|---|---|---|
| 17 Jul 1998 | 14 Aug 1998 | +64.9% | 31 Aug 1998 | −9.9% |
| 16 Jul 1999 | 29 Sep 1999 | +26.7% | 15 Oct 1999 | −1.7% |
| 24 Mar 2000 | 14 Apr 2000 | +24.4% | 14 Apr 2000 | 0.0% |
| 24 Mar 2000 | 14 Apr 2000 | +24.4% | 9 Oct 2002 | −42.7% |
| 9 Oct 2007 | 26 Nov 2007 | +36.4% | 9 Mar 2009 | −51.9% |
| 23 Apr 2010 | 20 May 2010 | +15.4% | 2 Jul 2010 | −4.6% |
| 29 Apr 2011 | 4 Aug 2011 | +11.4% | 3 Oct 2011 | −8.4% |
| 19 Feb 2020 | 27 Feb 2020 | +13.3% | 23 Mar 2020 | −24.9% |
| 3 Jan 2022 | 22 Feb 2022 | +8.6% | 12 Oct 2022 | −16.9% |
The Magic 7 account for 56% of the decline, and Apple is the only one that rose
Looking at Apple alone is not enough, because the market depends on a small group of stocks. Apple, Microsoft, NVIDIA, Amazon, Alphabet, Meta and Tesla, the Magic 7, made up 34.8% of the index at the high of 6 October and 31.9% today. Since that high they have lost 18.8% on a cap-weighted basis, taking 6.5% off the S&P 500's 11.6% decline, or 56% of it. The weights at the high are derived from State Street's figures for 19 November and the subsequent price moves, an approximation rather than a figure from the index provider.
NVIDIA alone cost the index 4.1%, more than the other five weak names combined. According to Yahoo Finance closing prices, the stock has fallen 47.5% since the index high, 14% on 20 November alone. Four of the six names next to Apple now trade below their 200-day moving averages, while Microsoft and Amazon remain above theirs. At +2.7%, Apple is the only one of the seven to have risen since the high, which makes it the only one of the seven to have supported the index in this correction.
| Stock | Weight 20 Nov | Since 6 Oct 2026 | vs 200d | Contribution |
|---|---|---|---|---|
| NVIDIA | 5.1% | −47.5% | −38.3% | −4.08% |
| Alphabet | 4.7% | −23.7% | −21.3% | −1.30% |
| Meta | 2.1% | −24.1% | −10.8% | −0.58% |
| Microsoft | 6.0% | −8.5% | +9.2% | −0.49% |
| Amazon | 4.1% | −3.7% | +0.7% | −0.14% |
| Tesla | 1.5% | −9.2% | −9.4% | −0.14% |
| Apple | 8.4% | +2.7% | +13.0% | +0.19% |
The rest of the market also fell mainly in its larger names. The other 493 stocks lost 7.8% on a cap-weighted basis and added 5.1% to the decline, while the equal-weight S&P 500 lost only 2.3% over the same period, according to prices for the Invesco S&P 500 Equal Weight ETF (RSP). The weakness is concentrated in the larger stocks, including those outside the Magic 7. This is consistent with the thesis: the heavyweights are driving the correction, and Apple is the only member of the Magic 7 still holding.
How the "rest of the Magic 7 weak, Apple strong" set-up resolved in the past can be measured from 2013. It cannot go back further, because Meta has only been listed since 2012 and a 200-day moving average needs a run-up. The test looks for every phase in which at least four of the other six traded below their 200-day moving averages while Apple stayed above. In 7 of 8 completed phases, Apple followed and broke its own line within 120 sessions, after a median of eleven sessions and no more than 34. Only in the phase starting on 27 January 2015 did Apple stay above it. The phase that began on 22 June 2026 is still open and so far runs against the pattern: Apple has closed above its 200-day moving average on all 108 sessions since. In the 120 sessions that followed, the S&P 500 traded at its low a median 11.6% below where it started.
| First day | Below 200d | S&P from high | S&P +60 sessions | Lowest S&P in 120 | Apple below 200d from |
|---|---|---|---|---|---|
| 27 Jan 2015 | 4 of 6 | −2.9% | +4.1% | −1.7% | no |
| 10 Oct 2018 | 4 of 6 | −4.9% | −7.6% | −15.6% | 13 Nov 2018 (24 sessions) |
| 12 Mar 2020 | 4 of 6 | −26.7% | +30.3% | −9.8% | 16 Mar 2020 (2 sessions) |
| 25 Jan 2022 | 4 of 6 | −9.2% | +0.9% | −15.8% | 14 Mar 2022 (34 sessions) |
| 29 Jul 2022 | 6 of 6 | −13.9% | −8.1% | −13.4% | 30 Aug 2022 (22 sessions) |
| 21 Feb 2023 | 4 of 6 | −13.7% | +4.0% | −3.5% | 24 Feb 2023 (3 sessions) |
| 4 Mar 2025 | 4 of 6 | −6.0% | +2.3% | −13.8% | 11 Mar 2025 (5 sessions) |
| 13 Mar 2026 | 4 of 6 | −5.0% | +11.4% | −4.3% | 30 Mar 2026 (11 sessions) |
| 22 Jun 2026 | 4 of 6 | −1.8% | +1.1% | – | open |
| 5 Nov 2026 | 4 of 6 | −5.9% | – | – | open |
If Apple catches up, the base case is 6,060 to 6,360
Three methods produce an index level. Route A applies the position of the S&P 500 at its lows since 2015 to today, without using Apple: at those lows the index stood a median 16.5% below its high and 11.8% below its 200-day moving average, which would correspond to 6,526 and 6,483 today. Route B starts from Apple, infers the index level and is the base case. Route C is the set-up from the previous section, at around 6,225. Routes A and B draw on the same eight episodes, so they are not independent of each other.
Apple's typical correction low since 2015 sat a median 13.7% below its 200-day moving average, which today means roughly $262, or 23.6% below Friday's close. On days when Apple fell, the S&P 500 moved on average 0.33 times (last two years) to 0.52 times (since 2015) as much, measured as the down-beta, the slope of a regression over those days only. Applied to Apple's typical low, that gives 6,060 to 6,360. If Apple only falls back to its 200-day moving average at $303, the same route ends at 6,500 to 6,650.
The row for Apple's index weight shows what the projection does not rest on. If Apple alone fell to its typical low and every other stock stayed unchanged, the index would lose only around 2% in arithmetic terms. The projection rests on co-movement, because Apple sell-offs have so far coincided with broad sell-offs. The hardest case since 2015, the slide to 23 March 2020 with a drop of 33.9% from the high, would correspond to 5,166 today. It lies far outside the base case but belongs to the range of the data.
The index only rises if the rest of the Magic 7 recover while Apple holds
Three continuations follow from the set-up. In the first, Apple corrects below its 200-day moving average, catches up to its typical low and drags the index with it. That is the base case of 6,060 to 6,360. Apple broke below its line in 7 of 8 comparable phases since 2013. In the second, the rest of the Magic 7 recover while Apple holds steady. If the six regain their 200-day moving averages, the index gains 4.8% to 7,250 in arithmetic terms, and if they return to their levels of 6 October, it gains 7.6% to 7,440. The low would then already be in, as in 1998, 1999, 2010 and 2011, and the thesis would be refuted.
The third continuation is a handover. The rest of the Magic 7 recover to their 200-day moving averages while Apple gives back 11.5% to 23.6%. The two moves largely offset each other, and the index lands at 7,110 to 7,180. Apple would then formally satisfy the thesis without the index having to fall any further. Continuations two and three work only through index weights and hold all other stocks unchanged, so they understate the co-movement that past corrections have shown.
Correlation is at its weakest since 2014
The strongest argument against a mechanical read-across comes from correlation. Over 250 sessions, the correlation of daily returns between Apple and the S&P 500 currently stands at 0.35, against a median of 0.56 since 1998, and a reading this low was last seen in 2014. Apple is following its own drivers more than usual. The down-betas in Route B, by contrast, come from periods of closer co-movement. If investors are using Apple as a place to hide while NVIDIA sells off, the stock can hold up for months, as it did in 2015, when it stayed above its line while the rest of the Magic 7 were weak.
The argument carries weight, yet it undermines the analysis less than it first appears. The phase in which four of the other six Magic 7 stocks traded below their 200-day moving averages began on 5 November, and in seven of eight completed earlier cases Apple's special role lasted no more than 34 sessions. Trading practice suggests that low correlation in a correction tends to be temporary and often ends in a shared sell-off. This analysis does not measure that directly, and the open phase since June shows that Apple can hold up for longer. Confidence in the thesis therefore stays at medium.
Three markers will show which continuation unfolds
The analysis points to three markers of which continuation is unfolding. They are observation points for portfolio and risk managers, not a trading recommendation, and they do not replace a firm's own review of its positioning, hedging and time horizon.
The level currently sits at around $303 and edges higher with every session. A close below it would be the signal that has preceded every index low since 2015 and would make the first continuation more likely. For risk managers with a high index allocation, the level is an additional observation point alongside the index's own 200-day moving average, which has already been breached.
If NVIDIA, Alphabet, Meta and Tesla regain their 200-day moving averages while Apple holds, that would be the second continuation and a contradiction of the thesis. An index with 31.9% of its weight in seven stocks reacts more strongly to their recovery than to a move in market breadth. The number of the six trading below their line, four today, is therefore the simplest daily gauge of whether the low is already in.
If the S&P 500 reaches this range while Apple closes below its 200-day moving average, the set-up would match the pattern of the index lows since 2015. If it gets there while Apple stays above its line, the analysis loses its basis. The zone is derived from eight episodes rather than forecast, and it shifts with the 200-day moving average if the correction drags on.
Keep reading – every 14 days in your inbox.
Capital-markets insights, regulatory updates and AI trends. Concise, well-founded, free.
GDPR-compliant. Unsubscribe at any time.