A snowy mountain summit in winter evening light

52-Week Highs in December: What the S&P 500 Did Next

Data through Decembers 1929-2025; forward returns through October 2026. Past market behavior doesn't guarantee future results.

When the S&P 500 sets a 52-week closing high in December, what follows has looked a lot like what follows any December. From 1929 to 2025, 36 of 97 Decembers had at least one 52-week closing high. The next calendar year averaged 8.59% after those Decembers (median 11.39%, up 71% of the time, n = 35), against 8.00% for all Decembers. The one horizon where they stood out was the next three months.

One thing to be clear about up front: this page uses the index’s own 52-week high, not a count of how many stocks made new highs. More on why below. Every number was calculated from daily S&P 500 closes (price index, no dividends). Related studies are on the Market Stats hub.

What the original post measured, and what this one measures

The December 2011 post on this URL looked at market breadth: days when more than 4.5% of NYSE-listed issues closed at a fresh 52-week high for four sessions in a row, the first time that happened in a given December, going back to 1970. It then tracked the S&P 500 over the next two sessions, to year-end, and to the 12th session of January.

A daily history of NYSE new-high counts back to 1970 isn’t available to us as a free, downloadable series, so that exact study can’t be rebuilt here. The closest honest version with public data is the S&P 500’s own high: a daily close above every close in the previous 365 calendar days. That’s a different signal. It says the index as a whole hit a one-year high, not that many individual stocks did, and the two can diverge.

The index series starts on December 30, 1927, so the first December with a full year of history behind it is 1929.

After a December 52-week high: 1, 3, 6 and 12 months

Returns run from December’s last close to the last close of January (1 month), March (3), June (6) and the next December (12, which is simply the next calendar year). December 2025’s 12-month figure needs December 2026, so it isn’t counted yet.

Group Horizon from Dec 31 n Average Median % positive
December with a 52-week high 1 mo 36 1.16% 1.42% 67%
3 mo 36 2.29% 2.24% 69%
6 mo 36 4.33% 5.43% 72%
12 mo 35 8.59% 11.39% 71%
…with 3+ new-high closes in a row 1 mo 22 0.44% 0.57% 59%
3 mo 22 0.88% 1.65% 64%
6 mo 22 1.95% 4.21% 64%
12 mo 22 6.53% 8.74% 68%
December without one 1 mo 61 1.20% 1.81% 61%
3 mo 61 0.90% 0.77% 52%
6 mo 61 3.42% 3.28% 61%
12 mo 61 7.66% 11.78% 66%
All Decembers 1 mo 97 1.18% 1.55% 63%
3 mo 97 1.42% 1.67% 59%
6 mo 97 3.76% 5.01% 65%
12 mo 96 8.00% 11.59% 68%

The three-month row is the only one with a clear gap: 2.29% and 69% up after a December high, against 0.90% and 52% after Decembers without one. At 12 months the two groups are within a point of each other. The “3+ in a row” group, the closest thing here to the old post’s four-day streak, actually did worse than the plain new-high group at every horizon. For scale, every month-end from 1929 to 2025 (n = 1,162 at 12 months) was followed by a 7.95% average gain over the next year.

The result flips depending on the era

Splitting the sample at 1970, the year the original study started, shows how fragile the pattern is.

Decembers n 3-mo average 3-mo % positive n (12 mo) 12-mo average 12-mo median 12-mo % positive
1929-1969, with a high 10 3.28% 80% 10 10.57% 14.66% 70%
1929-1969, without 31 -0.60% 45% 31 4.41% 2.62% 55%
1970-2025, with a high 26 1.91% 65% 25 7.79% 8.99% 72%
1970-2025, without 30 2.46% 60% 30 11.01% 15.20% 77%

Before 1970, a December high came before better results at both horizons. Since 1970, Decembers without a new high were followed by the stronger next year. Ten pre-1970 cases is a very small group, so the early edge rests on a handful of years. New highs have been common lately: 26 of the 56 Decembers since 1970 had one, against 10 of 41 before.

The short windows the old post looked at

Measured from the close of the first December 52-week high each year, the S&P 500’s next session was a coin flip, while the stretch into mid-January was up more often than usual. The base row starts from every December session in the same years.

Start Window n Average Median % positive
First December 52-week high Next session 36 0.02% -0.03% 50%
To year-end 35 1.01% 0.83% 71%
To 12th January session 36 2.00% 1.87% 81%
Any December session Next session 2046 0.06% 0.05% 53%
To year-end 1949 0.91% 0.79% 70%
To 12th January session 2046 1.61% 2.01% 67%

The year-end window has one fewer case because one first high (1944) fell on the last session of December. The 81% figure for the run into mid-January is the strongest number on this page, but it is 29 of 36 cases, overlapping heavily with December’s normal seasonal strength (see December seasonality).

Every December 52-week high since 2000

December of New-high closes First one 1 mo 3 mo 6 mo 12 mo
2003 11 Dec 01 1.73% 1.29% 2.60% 8.99%
2004 8 Dec 01 -2.53% -2.59% -1.70% 3.00%
2005 1 Dec 14 2.55% 3.73% 1.76% 13.62%
2006 4 Dec 04 1.41% 0.18% 6.00% 3.53%
2009 5 Dec 14 -3.70% 4.87% -7.57% 12.78%
2010 11 Dec 08 2.26% 5.42% 5.01% -0.00%
2013 7 Dec 09 -3.56% 1.30% 6.05% 11.39%
2014 6 Dec 03 -3.10% 0.44% 0.20% -0.73%
2016 4 Dec 07 1.79% 5.53% 8.24% 19.42%
2017 5 Dec 08 5.62% -1.22% 1.67% -6.24%
2019 9 Dec 12 -0.16% -20.00% -4.04% 16.26%
2020 7 Dec 01 -1.11% 5.77% 14.41% 26.89%
2021 4 Dec 10 -5.26% -4.95% -20.58% -19.44%
2023 11 Dec 01 1.59% 10.16% 14.48% 23.31%
2024 4 Dec 02 2.70% -4.59% 5.50% 16.39%
2025 3 Dec 11 1.37% -4.63% 9.55% n/a

December 2021 is the reminder that a new high says nothing about what’s coming: it was followed by a 19.44% drop in 2022, the worst next year after any December high in the full sample. December 2019’s high came three months before the 2020 crash took the index down 20% from its year-end level by the end of March.

What this does and doesn’t show

A December 52-week high tells you the index was in an uptrend at year-end. Six months later it was higher 72% of the time, close to the 65% for all Decembers. It has not been a reliable edge over a typical December or a typical month. Over 1970-2025, the span the original study started from, the 12-month advantage reversed. If you study it alongside other calendar patterns, such as the presidential cycle’s third year, keep the sample sizes in view: 36 cases spread over nearly a century is not much to go on.

Frequently asked questions

What happens to the S&P 500 after a 52-week high in December?

Historically, about what happens after any December. From 1929 to 2025, the 36 Decembers with at least one S&P 500 52-week closing high were followed by an average 8.59% gain the next year (71% up, n = 35), against 8.00% for all Decembers (n = 96). The clearest difference was over the next three months.

Are new highs a bearish sign?

Not in this data. Years whose December set a 52-week high did not do worse afterward on average than other years over 1 to 12 months. From 1970 to 2025, though, Decembers without a new high were followed by a stronger next year (11.01% vs 7.79%), so a new high has not been a reliable bullish sign either.

Why not use the number of NYSE stocks making new highs?

The original 2011 version of this study used the share of NYSE issues at new 52-week highs. That daily breadth count is not available to us as a free, downloadable series back to 1970, so this page uses the S&P 500's own 52-week closing high, which can be computed from daily index closes.

How is a 52-week high defined here?

A daily close above every S&P 500 close in the previous 365 calendar days. It uses closing prices, not intraday highs, and the price index without dividends.

Sources

  1. S&P 500 index daily closes (^GSPC), Yahoo Finance (checked November 8, 2026)
  2. Trading the Odds, original post of December 28, 2011 (Wayback Machine snapshot, January 5, 2012) (checked November 8, 2026)

Jordan Whitfield

Editor, Trading the Odds

Jordan Whitfield edits Trading the Odds, which covers stock market statistics (seasonality, volatility and how markets behave around recurring events) and how prediction markets work and where they are legal. Market statistics on the site are calculated from public index data and say when the data runs through; every legal claim carries the date it was checked and a link to its source. How we check facts.