Seasonal Trading Patterns: Fact or Fiction?

By Marcel Hambálek · Senior Trader, For Traders

Seasonal trading patterns are real but uneven: a 217-year study across 68 markets by Erasmus University Rotterdam found persistent calendar effects, yet roughly half the famous retail patterns have decayed since 2000. Gold's January strength and the turn-of-the-month effect still hold; the January Effect in small caps is largely dead.

Key takeaways

• Seasonality is a recurring calendar-linked tendency in returns, volume or volatility — a probability tilt, never a signal on its own.
• The turn-of-the-month effect, gold's January and August legs and the natural gas heating-season cycle still show up in 2026 data; the January small-cap effect and a naive 'sell in May' exit largely do not.
• Futures seasonality is the least-crowded corner, but contract rolls and contango can invent or invert an apparent seasonal edge on continuous charts.
• A 12-year backtest is 12 data points — demand 25+ observations, a Sharpe above roughly 0.6 and a max drawdown you could actually survive under a daily loss limit.
• Holiday windows cut liquidity hard: US volume can fall toward 70% of normal by Thanksgiving Wednesday, widening spreads and worsening fills.
• Trade seasonal windows with technical confirmation, 1.5× ATR stops and fractional sizing so one failed pattern can't breach a max drawdown.

Watch: related video

Are Seasonal Trading Patterns Real? The Short Verdict

Yes, seasonal trading patterns are real — but only some of them, and never as a guarantee. Seasonality is a repeating tendency for a market to behave a certain way during a specific calendar window, driven by structural flows rather than chance. That's the whole definition. No mysticism, no lunar cycles — just recurring supply, demand, and capital-flow events that happen to line up with the calendar.

What seasonality actually means

A seasonal trader isn't betting on a date. They're betting on the recurring event behind the date — a harvest, a fiscal year-end, a heating season. The seasonality definition that matters for your P&L is narrow: a statistically observable tendency, not a rule. Gold tends to firm into January on Asian jewelry demand and portfolio rebalancing; that's a tendency built from decades of repeated flow, not a superstition.

Why some patterns persist and others die

Researchers at Erasmus University Rotterdam ran the toughest test anyone's applied to this idea — 68 markets, 217 years of data — and still found calendar anomalies that survived out-of-sample. That's the strongest evidence we have that seasonal trading isn't purely a data-mined artefact dressed up in a spreadsheet.

But persistence isn't universal, and the line is instructive:

• Structural causes survive. Heating demand, fiscal year-end flows, harvest supply cycles — these are physical or institutional facts. They don't arbitrage away because a hedge fund read about them.
• Behavioural anomalies decay. The classic January Effect in small caps — driven by tax-loss selling and January reinvestment — has been fading for two decades. Once enough capital front-runs the crowd, the edge gets traded out of existence.

The difference between an edge and a coincidence

A real seasonal edge shifts probabilities by a few percentage points — it might tilt a 50/50 coin toss to 54/46. That's meaningful over hundreds of trades and worthless on any single one. If someone's selling you a seasonal pattern as a guaranteed direction — "gold always rallies in January" — they're not selling seasonality, they're selling certainty that doesn't exist in any market, calendar-based or not.

The honest way to use seasonal trading patterns is as one input among several: a nudge on bias, a reason to size slightly differently, never a standalone entry signal. Treat it the way you'd treat any other statistical edge — real, measurable, and still capable of losing on any given month.

The 2026 Verdict Table: Grading Every Famous Seasonal Pattern

Here's the direct answer: of the eight most-cited seasonal trading patterns, three still hold up as fact, three are fading with shrinking or unstable edges, and two have decayed into fiction that keeps getting repeated on trading forums without evidence behind it. Below is the grading breakdown, with sample size and historical hit rate for each.

Fact: patterns that still show up in the data

The turn of the month effect is the tendency for equity indices to post outsized returns in the last one to two trading sessions of a calendar month and the first three of the next — a window that's produced a disproportionate share of total monthly returns for decades. It survives because it's structural, not psychological: pension contributions, 401(k) auto-investments, and payroll-linked buying all cluster into this window, and that plumbing hasn't changed. Gold's January and August strength holds for similar structural reasons — Indian wedding-season demand and Western institutional rebalancing both land in those months. Natural gas winter seasonality persists because heating demand is a physical, non-negotiable driver that no amount of arbitrage capital can erase.

Fading: patterns with shrinking edge or unstable samples

The Santa Claus Rally is only seven trading sessions — too small a sample to lean on with conviction, and recent years have delivered flat or negative readings that dent the historical hit rate. Sell in May and go away, the Halloween indicator's flip side, still shows a return gap between the two six-month halves, but the gap has compressed sharply since 2000 as algorithmic capital arbitrages the seasonality away. Crude's driving-season strength has weakened too, diluted by US shale flexibility that smooths out summer demand spikes that used to move price on a schedule.

Fiction: patterns you should stop trading

The January Effect — small caps outperforming large caps in early January on tax-loss-selling rebound — is the clearest fiction on this list. It was real when retail access to small caps was thin and tax-loss selling created genuine December liquidity gaps. Post-2000, ETF access flattened that friction, and tax-loss automation front-runs the rebound before retail traders can act on it. If you're still building seasonal trading strategies around this one, the data says stop.

Gold Seasonality: Why January and August Matter for XAUUSD

Gold seasonality is one of the few calendar patterns that survives contact with modern markets: XAUUSD has historically posted its strongest average returns in January and again from mid-August into September, driven by physical demand flows that don't care what a chart pattern says. Unlike the January Effect in small caps, gold January strength is anchored in real buying, not a tax-quirk that algorithms have since arbitraged away.

The demand calendar behind gold's seasonal legs

Two physical demand cycles put a genuine bid under gold at predictable points in the year. Indian wedding season and the run-up to Diwali (typically October-November) drive jewellery fabrication demand that builds through late summer — Indian and Chinese households historically account for a large share of global physical gold offtake. Ahead of Chinese New Year, jewellers and central bank-adjacent institutions restock inventory, and that restocking usually front-runs the holiday by six to eight weeks, landing squarely in January. Layer in Q4 fabrication demand for Western jewellery retail (holiday gifting), and you get two distinct windows — January and mid-August through September — where physical buyers are net accumulators rather than net sellers. This is the gold January strength and gold August rally traders reference, and it's the physical gold demand cycle underneath both legs.

Summer softness and the June-July trough

The flip side is the June-July trough — the gold summer doldrums. Indian wedding buying has typically wound down, Chinese New Year restocking is long spent, and Western investment desks are running reduced summer books with less appetite for fresh macro positioning. Physical demand thins at exactly the point where speculative flow would need to pick up the slack, and it usually doesn't. This is the seasonal low point for XAUUSD in a typical year — not a crash, just a genuine air pocket in demand.

How reliable is the gold seasonal window in practice

Be honest about the distribution here: "average January is positive" hides a wide spread of outcomes. Some Januaries rally hard, others chop sideways or dip before recovering, and a hawkish Fed repricing or a sudden dollar spike (check the Federal Reserve's rate decisions and the DXY) can override the calendar with zero warning. The seasonal edge is a tilt, not a signal you trade in isolation.

Use this as a bias, not a trigger: in the seasonally strong windows, favor long setups on pullbacks and give breakout longs a bit more benefit of the doubt on your XAUUSD seasonal read; in the June-July trough, tighten your conviction on longs and let short-side or range setups carry more weight. The calendar tells you which side of the book to lean toward — your entry still has to earn its own risk-reward.

Index Seasonality: US100, S&P 500 and the Six-Month Effect

The Halloween effect is real in the data but weaker than the folklore: November-April has historically beaten May-October on the S&P 500 by a few percentage points a year on average, yet the gap has compressed hard since 2010 and simply staying invested has often outperformed the "sell in May and go away" rotation once you account for missed rallies during the supposedly weak half.

The November-April vs May-October split

"Sell in May and go away" is one of the oldest seasonal stock market patterns on record, and the underlying logic — thinner summer liquidity, institutional desks running lighter books, less fresh capital deployment until autumn — still holds some water. But the historical S&P 500 seasonality edge from this split has narrowed from a wide gap in the pre-2000 sample to something much closer to a coin flip over the last decade. A trader who went to cash every May and re-entered every November would have sat out several of the strongest summer rallies of the 2020s, including sharp AI-driven and rate-pivot-driven legs higher in the "weak" half. The pattern is a tilt, not a rule you trade mechanically.

Nasdaq's own seasonal profile and earnings clustering

US100 seasonality runs on a different clock than the broad S&P 500 because the index is concentrated in mega-cap tech, and tech has its own earnings rhythm. The Nasdaq seasonal profile tends to show strength around January guidance season as companies reset full-year forecasts, and two distinct volatility clusters around April and October earnings — when a handful of megacaps move the whole index on a single print. Late summer, particularly August into early September, has historically carried a volatility uptick as thinner holiday liquidity meets the first pre-earnings positioning of the quarter. If you trade US100 setups, the earnings calendar matters as much as the seasonal calendar — often more.

September: the one month with a genuinely negative skew

September effect isn't a myth traders tell each other around the campfire — it's the most consistently negative single month across a long sample of S&P 500 and broader developed-market data, spanning decades and multiple market regimes. Nobody has a fully clean explanation: quarter-end rebalancing, post-summer institutional repositioning, and the historical clustering of macro shocks (Fed meetings, fiscal-year-end selling) around this window all get blamed. It's the one calendar effect in this section that shows up with enough consistency to actually adjust size around, rather than just note and move on.

For a prop trader, the practical translation is narrow but useful: index seasonality should nudge your position sizing and setup selection — smaller size and tighter risk management heading into September, more patience for pullback longs across November-April — but it never tells you to sit out. The calendar adjusts how much risk you put behind an idea, not whether you show up to trade it.

Trading Seasonalities in the Futures Markets

Futures seasonality is a different animal from index seasonality, and it's more reliable in principle because it's anchored to something physical: a tank of natural gas, a barrel of crude, a bushel of corn in the ground. You're not trading a calendar quirk in fund flows — you're trading storage cycles, refinery schedules, and pollination weather. That structural anchor is exactly why futures seasonality rewards traders who study the underlying commodity, and punishes anyone backtesting a continuous chart without understanding how that chart was built.

Natural gas and the heating-season demand cycle

Natural gas winter seasonality is the cleanest physical story in commodities. From April through October, utilities inject gas into underground storage ahead of winter demand — prices typically soften or chop as inventory builds. From November through March, storage draws down to heat homes and run power plants, and any cold snap against a tight storage print can spike the front month hard. The pattern isn't superstition; it's the withdrawal season showing up in EIA storage reports every Thursday. The catch: a mild winter or a shale supply surge can override the seasonal tendency completely, so this is a bias to lean on, not a signal to trade blind.

WTI crude: driving season and refinery maintenance

Crude oil's summer driving season — roughly Memorial Day through Labor Day in the US — lifts gasoline demand and historically firms crack spreads and WTI itself. But the setup and unwind around that window matter more than the season itself: refiners run maintenance ("turnaround") in spring before driving season ramps up, and again in autumn before winter-grade fuel production, temporarily cutting crude demand and gasoline supply in ways that swing the WTI-RBOB relationship independent of the outright price trend.

Grains: planting, WASDE reports and harvest lows

Corn and soybean seasonality tracks the crop calendar, not the trading calendar. Prices price in weather risk from planting through pollination in July — the "weather market" window when a dry forecast in Iowa moves more than any technical level. Once the crop is made, cash grain floods the market at harvest and prices frequently bleed into harvest lows through September-October. Layered on top: the monthly WASDE report (World Agricultural Supply and Demand Estimates) resets the supply/demand picture and routinely produces the sharpest single-day moves of the month, seasonal bias or not.

Why contract rolls and contango break naive seasonal backtests

Here's the part almost nobody in retail content covers properly. Most seasonal backtests run on continuous back-adjusted charts — a single price series stitched together across CME futures roll dates as one contract expires and the next becomes the front month. That stitching is fine for spotting a chart pattern, but it quietly launders the roll yield into your "seasonal edge."

If a market sits in contango — deferred contracts priced above the front month, common in natural gas and crude during oversupply — a continuous chart rolled forward will bleed value every roll, manufacturing a fake downtrend that no trader ever actually captured, because you'd have been paying that spread on every roll, not collecting it. In backwardation — deferred months priced below front month, common in tight physical markets — the reverse happens: the continuous chart shows a seasonal uptrend inflated by roll yield, not by the calendar effect you think you're trading.

The fix is straightforward: test seasonality on individual contract months (the actual March natural gas contract, year after year) or on roll-adjusted spreads rather than a naive continuous series. Platforms like Seasonax and SpreadCharts are built specifically to handle CME futures roll dates and contango/backwardation correctly, which is why serious futures traders use them instead of eyeballing a generic continuous chart. If you're trading futures seasonality on a For Traders challenge, that distinction isn't academic — it's the difference between a backtested edge and a roll-yield mirage.

Holiday and Summer Liquidity: The Pattern That Never Died

This is the seasonal effect you don't need a 200-year dataset to trust: trading volume mechanically collapses around specific calendar dates, every single year, because the humans and algos that provide liquidity are literally not at their desks. Unlike statistical curiosities that decay once enough traders crowd them, holiday and summer liquidity gaps are structural — they show up whether or not anyone is trying to trade them.

Thanksgiving

US equity and equity-index volume can fall toward 70% of a normal session by the Wednesday before Thanksgiving, and Friday's half-day session often trades on a skeleton book. If you hold US100 or S&P-linked positions into that window, you're trading in a market where the same order size moves price further than it would on a regular Tuesday. Thanksgiving trading volume thinning isn't a rumor — it's visible on any volume histogram you pull up, and it bleeds into futures and FX pairs correlated with US risk sentiment too.

Christmas and the volume cliff

The stretch from December 24 through January 2 runs on skeleton desks across every major financial center. Christmas liquidity doesn't taper gently — it drops off a cliff, then snaps back just as abruptly once desks reopen in the new year. Gold, indices, and majors all see this, but it's most dangerous in instruments that already have wider natural spreads, because holiday spread widening compounds on top of an already-thin book.

August doldrums and European school holidays

Roughly the first three weeks of August, a huge chunk of European institutional desks are on holiday, and US desks quiet down too. This is the classic August summer doldrums window: early London-hours price action in August behaves nothing like early London hours in October. The same 8:00 London open that reliably produces a clean directional leg in autumn can chop sideways on fumes in August, then suddenly gap on a headline because there's no depth to absorb it.

Trading thin markets: spreads, slippage and fake breakouts

Three practical rules keep thin-liquidity periods from eating your account:

• Widen expected slippage, don't ignore it. A fill that costs you 2 pips on a normal Tuesday can cost you 6-8 pips in the Christmas week — build that into your R:R before you enter, not after you get filled.
• Cut position size rather than stop distance. Tightening your stop to compensate for a wider spread just gets you stopped out by noise. Reduce lot size instead and keep your stop logic intact.
• Treat thin-volume breakouts as suspect. A level that breaks cleanly on August 5th with a quarter of normal volume behind it deserves confirmation from the next full session before you trust it. Fading illiquid moves is equally risky — with no depth on the book, price can run far past where the order flow "should" stop it.

The Month-by-Month Seasonal Calendar for 2026

A seasonal trading calendar is only useful if you can see it collide with the macro calendar — because a "seasonal" entry stacked on an NFP CPI FOMC print isn't seasonality, it's an event trade wearing a costume. Below is the 2026 map across XAUUSD, US100 (Nasdaq) and S&P 500, plus the CME futures contracts that share the same institutional flows.

Q1: January Gold, Japanese Fiscal Year End, March Quad Witching

January has historically been gold's strongest single month — new-year allocation flows into safe-haven assets, reinforced by Lunar New Year physical demand out of Asia. Layer that against the first NFP Friday of the year and you get exaggerated moves that fade fast once the print is digested.

Late March brings the Japanese fiscal year end on 31 March. Japanese institutions and exporters repatriate offshore holdings back into yen ahead of book close, a flow that has historically pressured USD/JPY lower into month-end regardless of the interest-rate backdrop. It's one of the few seasonal patterns with a clean structural cause — not superstition, an actual balance-sheet deadline.

The third Friday of March is also quad witching — stock index futures, index options, stock options and single-stock futures all expire simultaneously. Volume and volatility spike into the close as funds roll and rebalance; US100 and S&P 500 often see erratic intraday swings that have nothing to do with fundamentals and everything to do with expiring contracts.

Q2 and Q3: Driving Season, Summer Thinning, September Risk

April through August is "driving season" for crude oil (CME's WTI contract) — US gasoline demand rises seasonally, and refiners bid up crack spreads ahead of the summer travel peak. June through August also brings the well-documented summer thinning across FX and equities: lower volume, wider spreads, more false breakouts. This is where the reduced-size, wait-for-confirmation approach from thin-liquidity sessions matters most.

September has the worst average seasonal return for the S&P 500 going back decades — and it's also loaded with FOMC, quarter-end rebalancing, and the return of full institutional volume after summer desks empty out. Don't assume September weakness is "just seasonality" — check whether the move you're seeing is riding a September FOMC decision instead.

Q4: Harvest Lows, Heating Season, Tax-Loss Selling and Window Dressing

Agricultural futures (corn, soybeans, wheat) often print harvest lows in September-October as physical supply floods the market post-harvest, then seasonally recover into year-end. Natural gas flips from injection to withdrawal season in November as US storage draws down for winter heating demand — this switch has historically been one of the more reliable calendar effects in the CME NYMEX complex.

December is dominated by two institutional flows working in opposite directions: tax-loss harvesting, where investors sell losing positions before year-end to realize losses against gains, pressures already-weak names lower into mid-December; then window dressing — fund managers buying winners and dumping losers before year-end statements go out — tends to lift the same broad indices into the final week. This is the mechanical backbone behind the "Santa Claus rally" narrative.

How to Backtest a Seasonal Pattern Without Fooling Yourself

A seasonal pattern only earns a spot in your playbook if it survives a hostile backtest — most don't. Run the numbers properly and you'll find the majority of "reliable" seasonal setups posted in trading forums fall apart the moment you apply real statistical scrutiny. Here's how to tell the survivors from the noise.

Sample size: why 12 years is 12 data points

If you backtest "buy Gold on October 15, sell November 15" over 12 years, you don't have thousands of data points — you have 12. One seasonal window per year is one independent observation, no matter how many candles sit inside it. A 60% hit rate across 12 years means 7 wins and 5 losses. Flip a coin 12 times and you'll land on 7 heads a meaningful chunk of the time. That's not an edge, that's variance dressed up as a strategy. Sample size seasonality is the first filter, and it kills more "proven" patterns than any other single factor.

Multiple testing and the pattern you found by looking

Scan 12 calendar months across 40 markets and you've run 480 separate tests. At a standard 95% confidence threshold, roughly 24 of those will look statistically significant purely by chance — before you've found a single real edge. This is multiple testing bias, and it's the quiet reason so many seasonal patterns look incredible in a backtest and evaporate in live trading. If you scanned for it, assume it's noise until proven otherwise.

What a credible Sharpe and max drawdown look like

Real seasonal edges don't look like miracle equity curves. Quantpedia's published seasonality strategy research — a widely cited academic-style database — shows metrics like a 7.62% CAGR, 9.43% volatility, a 0.81 Sharpe ratio, and a -11.98% max drawdown. That's the honest shape of a genuine, published, non-hyped seasonal edge: modest, survivable, unglamorous. If a pattern someone's selling you claims a 3+ Sharpe with single-digit drawdown over a seasonal window, be skeptical — that's a red flag, not an edge.

For the research itself, tools like Seasonax (built by Dimitri Speck) and SpreadCharts (developed by Pavel Hála) let you visualize decades of seasonal data and test window stability directly — both are built specifically for this kind of seasonal backtest work, and both make it easy to see whether your edge survives a shifted window before you risk simulated capital on it.

Trading a Seasonal Window Inside Prop Firm Rules

A seasonal bias tells you which direction to lean, not when to pull the trigger — you still need price structure to confirm the trade, and you still need to size it so one bad thesis doesn't wreck your evaluation. Prop firm risk rules don't care how many decades of data back your pattern. A daily loss limit breach on day three ends the account regardless of whether gold's January strength shows up on schedule three weeks later.

Confirmation first: never trade the calendar blindly

The window opening is not your entry signal. If the seasonal edge points long, you wait for the market to show you a pullback into a level, a break-and-retest of structure, or some other technical trigger before committing simulated capital. Buying the moment the calendar flips just because "this is historically the strong week" is how you end up entering on the worst possible tick, right before a shakeout. Treat the seasonal window as a filter that narrows your watchlist and sets your directional bias — the actual entry still comes from the chart, the same way it would on any other setup you'd take under your prop firm risk rules.

1.5× ATR stops and fractional risk sizing

Once structure confirms, place your stop roughly 1.5× ATR beyond the swing, not on the round number just below it. Round numbers get swept first — everyone's stop is sitting there, and market makers know it. ATR stop placement forces you to size the position around actual volatility instead of an arbitrary price level, which matters more in seasonal setups because the whole idea is holding through some chop while the pattern plays out.

Widen that ATR buffer during holiday sessions — the week around Christmas, early January, and other thin-liquidity stretches where spreads inflate and a normal-sized stop gets tagged on noise alone. Your position sizing should shrink as your stop distance grows, so risk per trade stays constant regardless of how wide the ATR buffer needs to be that week.

Protecting the daily loss limit and max drawdown

Size every seasonal trade so a full thesis failure costs a small slice of your daily loss limit — not most of it. Seasonal trades often need weeks to resolve, not hours, and that time horizon is exactly what makes them dangerous under evaluation rules built around daily and max drawdown limits. You can be completely right about the pattern and still get disqualified if one oversized entry blows through your daily cap before the thesis has time to work.

Two habits protect you here. First, never stack a seasonal entry directly on FOMC, NFP, or CPI release dates — let the volatility spike pass, then re-enter on confirmation, because a seasonal edge measured in weeks doesn't need you exposed to a five-minute data-driven whipsaw. Second, stagger entries across the window instead of committing full size on day one. Scaling in as confirmation builds keeps your average risk lower and means a false start doesn't consume the daily loss limit you need for the rest of the window to play out.

Seasonal Trading: Where It Helps and Where It Hurts

Pros

• Gives you a directional tilt and a reason to stand aside in historically hostile windows
• Futures seasonality is anchored in physical supply and demand, so it decays more slowly than behavioural equity anomalies
• Holiday and summer liquidity effects are mechanical and observable in real time, not just in backtests
• Pairs cleanly with technical confirmation as a filter rather than competing with it
• Encourages planning trades weeks ahead, which cuts impulsive screen-time decisions

Cons / risks

• Small samples — a 20-year window is 20 observations, easily mistaken for significance
• Continuous futures charts and contango can manufacture seasonal returns that were never tradeable
• Crowding has visibly decayed popular patterns such as the January Effect since 2000
• One macro shock, policy shift or geopolitical event overrides any calendar bias instantly
• Seasonal trades often need weeks to work, which sits awkwardly with daily loss limits and evaluation time pressure

Frequently Asked Questions