What Is Seasonality? Why Seasonal Patterns Work (Examples)

Seasonality is the tendency of a stock, ETF or commodity to move in a similar way at the same time each year. A retailer may tend to rise before the holidays, or a chipmaker around its earnings months. This article explains why these seasonal patterns happen, using real examples from 15 years of adjusted prices (2011 to 2025), and how often they actually keep working. For the fuller introduction, see our stock seasonality guide.

Seasonal patterns, seasonal trends and seasonal charts

• Seasonal pattern: a specific date range with a repeating record. For example, XRT, the retail ETF, rose from November 20 to 30 in 14 of the last 15 years.
• Seasonal trend: the broader direction over part of the year. For example, many stocks have tended to be strong from October into the new year.
• Seasonal chart: a picture of a stock’s average year, which shows those trends at a glance.
• Win rate: how often a pattern repeated. A window that rose in 14 of 15 years has a 93% win rate.

How to read a seasonal chart

This seasonal chart shows XRT’s average year. Each year starts at 0% on January 1, and the years are averaged day by day. Where the line climbs, retail stocks have tended to gain. Where it falls, they have tended to lose.

A seasonal chart is a starting point, because an average can hide a lot. One huge year can create a bump that most years never had. So always check the win rate and every individual year too. You can drag across any stock’s seasonal chart to test a date range in the Chart Explorer.

Example: AAPL’s seasonal pattern

Apple shows how this works in practice. Here is AAPL’s seasonal chart over the same 15 years:

The seasonal trend: July is Apple’s best month, averaging +7.14% and rising in 14 of 15 years, while September is its weakest, averaging −1.75% and rising in only 5 of 15. The seasonal pattern: the June 20 to July 20 window rose in all 15 years, and again in 2026 at +9.59%. Its median is +4.39%, well below its +7.09% average, so a few big years lift the average, including a best of +22.70%.

It also shows the limits of easy explanations. Apple usually launches new iPhones in September, yet that’s when the stock has been weakest, and the widely expected launch may already be priced in. See the full record on the AAPL seasonality page or in our AAPL seasonality article.

Seasonal patterns in other popular stocks

Here is one of the strongest seasonal patterns for each of these well-known stocks, with what happened in 2026, after the 15-year record (JPM’s window ended in January 2026):

Look at the last column. These windows rose in nearly every one of the previous 15 years, yet only 3 of the 8 rose in 2026. One year isn’t much evidence, but it’s a useful reminder that a long winning record can’t promise the next result. Each stock’s full history is on its seasonality page: AMZN, JPM, GOOGL, MSFT, NVDA, TSLA, WMT.

Why seasonal patterns happen: five real examples

These explanations are widely cited and make sense, but historical data can only show that a pattern existed, not prove why. Treat each reason as plausible, not proven.

1. Consumer spending. Retail’s best stretch lines up with the holiday shopping season. XRT’s best month is November, averaging +5.46% and rising in 11 of 15 years, and its 10-day window around Thanksgiving and Black Friday rose in 14 of 15. Home improvement has its own calendar: Home Depot rose from April 15 to 25 in 13 of 15 years, during the spring season.

2. Booking and travel seasons. Cruise lines take many of their bookings early in the year, in what the industry calls “wave season”. Royal Caribbean rose from December 19 to January 18 in 14 of 15 years, averaging +6.24%, although its worst year fell 13.53%.

3. Earnings cycles. Companies report on a roughly fixed calendar, so earnings reactions can repeat at the same time each year. Deere usually reports its full-year results in November, and its October 26 to November 25 window rose in 13 of 15 years, with a worst year of just −1.75%. Nvidia’s most reliable months are also its usual earnings months.

4. Commodity demand. Gold has been strongest in January, averaging +2.99%, a period often linked to buying around the Lunar New Year. It has been weak in September, rising in only 4 of 15 years.

5. Fund flows. Money moves on a schedule: new-quarter allocations, year-end rebalancing and tax-related selling. That’s one common explanation for early-summer strength, when SPY rose from June 21 to July 21 in all 15 years. Seasonal trends like this often run across the whole market, as sector seasonality shows.

When the obvious reason fails

Sometimes the seasonal trend is the opposite of what you’d expect, because markets look ahead. Winter is peak demand for natural gas, yet UNG, a natural gas fund, fell from November 26 to December 16 in 13 of 15 years, averaging −11.28%. Everyone knows winter is coming, so it may already be in the price. UNG also has a built-in drag from the cost of rolling futures, so it has fallen in most months. Likewise, Amazon’s December has been its weakest month, despite the holidays.

Does seasonality actually work? The honest answer

Partly. Here is the evidence on both sides.

Win rates can flatter a pattern. Stocks rise most of the time. SPY rose in 69% of all 30-day calendar windows over these 15 years. At that rate, a window would rise in 12 or more of 15 years about 1 time in 4 by luck alone. So “up 12 of 15” isn’t special for SPY, and a win rate only means something compared with the normal rate. A perfect record is rarer: nine SPY windows rose in all 15 years, against about one expected by chance. But they overlap heavily, all starting between June 13 and June 21, so they’re closer to one finding than nine.

Our own test found a real but modest edge. In a walk-forward study of 9,292 windows across 65 large-cap stocks from 2014 to 2025, seasonal patterns went in their historical direction about 70% of the time the next time around, against about 55% for random calendar windows. That’s useful, but it also means about 3 in 10 didn’t repeat.

The flashiest patterns repeat least. In that study, the calmest patterns repeated about 14 percentage points more often than the most extreme. In a separate test of patterns with a perfect win rate and an average above 4% over a six-year look-back, only about 57% repeated. A spectacular record is often partly luck, so it’s a reason for caution, not confidence.

Even strong patterns break or fade. QQQ’s June 22 to July 22 window rose in all 15 years, then fell 4.4% in 2026. The classic January effect has faded: small caps beat the S&P 500 in only 6 of the last 15 Januaries. And the Santa Claus rally fell in each of the last four seasons.

How to use seasonal patterns sensibly

• Compare with the normal rate. An 80% win rate means less for a stock that rises in most windows anyway.
• Check the size of the losing years. A window can win often and still lose heavily when it fails.
• Compare the average with the median. A big gap usually means a few outsized years are doing the work, as with Nvidia’s January.
• Prefer patterns with a reason. A plausible driver, such as earnings timing, is more believable than a date range with no explanation.
• Be wary of perfection. The most impressive records have tended to repeat less often.
• Use it as context. Seasonality can inform timing, but it’s one input alongside your own research and risk management.

Past seasonal performance does not guarantee future results. Win rates and average returns are historical base rates, not predictions. This is for informational purposes only and is not financial advice.