Sell in May 2026 — The 1.7% Statistical Truth, the Three Sectors Where the Effect Is Real, and the Defined-Risk Put-Spread Collar That Keeps You Long Through Summer

Sell in May is real but small. The full long-run dataset, the three sectors that carry the entire seasonal underperformance, the defensive trio that beats the index in summer, and a near-zero-cost SPY put-spread collar that keeps you 100% long through July.

sell in may

stock market seasonality

sector rotation

options strategies

SPY put spread collar

XLP

XLU

XLV

XLY

XLI

S&P 500

2026 market outlook

defensive sectors

calendar effects

Hollis Vance

April 28, 2026

10 min

Sell in May 2026 — The 1.7% Statistical Truth, the Three Sectors Where the Effect Is Real, and the Defined-Risk Put-Spread Collar That Keeps You Long Through Summer

It is the most-quoted six words in all of finance, and they show up on every business-news ribbon the last week of April: Sell in May and go away. The Saint Leger's Day version of the adage dates back to 18th-century London, when traders shut their books for the racing season. The modern version is the one most CNBC viewers absorb by osmosis: stocks underperform from May through October, so move to cash, take the summer off, and come back in November.

It is also one of the most over-applied calendar effects in markets. The seasonal gap is real, but it is much smaller than the headline implies, much more concentrated in a few sectors than the index-level statistic suggests, and — for the second year running — probably the wrong trade for the macro regime we are actually in. Paul Ciana, the chief market technician at Bank of America Securities, just published this week that his read of the six-, three-, and one-month tape says the play is to buy in May and sell in July or August. CNBC's preview of the May 1 trading week is built almost entirely around the question of whether the Iran resolution and the Atlanta Fed's freshly-cut Q1 GDP nowcast (1.2%, down from above 3%) flip the seasonal trade upside down.

This post does three things. It puts the real statistical record next to the headline so you can see what you are actually trading. It breaks the index down into the sectors where the calendar effect has historically been concentrated (and the ones where it is statistical noise). And it lays out three ways to express a defensive May–October stance — including a defined-risk SPX put-spread collar that lets a $50K options account stay 100% invested through the summer with the worst-case outcome capped at a known dollar figure.

The Statistical Truth — What "Sell in May" Actually Means in Numbers

The most-cited dataset is the S&P 500 six-month split since 1950. Across roughly 75 years of data, here is what the calendar actually says:

November through April: average price return of roughly +7.0%, positive in 70% of years.
May through October: average price return of roughly +1.7% to +2.5% depending on the exact window and dividend treatment, positive in 65% of years.
The gap: about 5 percentage points of expected six-month return — meaningful, but not the wholesale exit the slogan implies.

The "go to cash" version of the trade — sell on May 1, hold T-bills, buy back on November 1 every year — has been studied to death. From 1950 through 2024, staying invested across all 12 months returned roughly 8.0% CAGR, while the strict "Sell in May" rule returned closer to 6.9% CAGR. That gap, compounded over 75 years, is the difference between a $1 starting stake growing to roughly $343 versus roughly $156. The strategy underperforms buy-and-hold by half over the long run, even though the seasonal weakness it tries to exploit is statistically real.

Why? Two reasons that matter for how you actually trade this:

The May–October period is positive on average. It is just less positive than November–April. Selling positive expected return to sit in cash is a high bar.
The seasonal gap has compressed dramatically since 2009. Across the post-GFC era, May–October has averaged closer to +4%, well above the 75-year long-run mean. The era of zero-rate liquidity and quarterly buyback flow flattened the calendar effect at the index level.

That second chart bar — May through October being positive in 65–76% of years — is the single most important number in this whole conversation. You are exiting positive expected return. That is not a free trade. To make the exit pay, you either need to (a) believe a specific macro regime is going to push the May–October average decisively negative, or (b) reroute the seasonal-relative-weakness signal into a factor trade or a defined-risk options structure that doesn't require getting flat.

The Three Sectors Where the Calendar Effect Is Actually Tradeable

Decomposing the index changes the picture. The May–October weakness in the S&P 500 is not evenly distributed across sectors — it is concentrated in the cyclical and industrial parts of the index. The defensive parts of the index actually trade better on a relative basis during the summer, in part because rotation flow chases them when investors get cautious about cyclical earnings.

| Sector | May–Oct Avg Return | Win Rate | vs SPX Spread (bps) | 2026 Specific Setup | |---|---|---|---|---| | Consumer Staples (XLP) | +3.4% | 74% | +170 bps | Sticky margins as gas rolls back below $4 | | Health Care (XLV) | +3.1% | 71% | +140 bps | Election overhang lifts post-MSC | | Utilities (XLU) | +2.9% | 68% | +120 bps | Rate-cut sensitivity into a dovish Fed | | Real Estate (XLRE) | +2.7% | 64% | +100 bps | REITs lead when 10y holds <4.5% | | Information Tech (XLK) | +2.4% | 65% | +70 bps | Mag 7 Q1 prints divide the cohort | | Communication Services (XLC) | +2.2% | 63% | +50 bps | Streaming/ads earnings catalysts mid-summer | | Financials (XLF) | +1.6% | 61% | -10 bps | Rate-cut tailwind offset by NIM compression | | Energy (XLE) | +1.0% | 58% | -70 bps | Iran ceasefire knocked WTI ~$25 lower | | Materials (XLB) | +0.4% | 55% | -130 bps | Industrial demand softens through summer | | Industrials (XLI) | +0.3% | 54% | -140 bps | The classic "go-away" cohort | | Consumer Discretionary (XLY) | -0.6% | 49% | -230 bps | Largest negative seasonal skew of any sector |

<figcaption>Sector-level "Sell in May" decomposition. May–October average price returns are based on long-run S&P 500 sector data; win rate is the percentage of years the sector posted a positive May–October return; the spread column is the average outperformance vs the S&P 500 over the same window. Defensive sectors (top of the table) actually outperform in the summer; cyclicals and discretionary (bottom) carry the bulk of the seasonal underperformance. Numbers are illustrative composites — verify against your own seasonal tool before sizing.</figcaption> </figure>

Three things to internalize from the table:

Defensive sectors actively outperform. Consumer staples, health care, and utilities have collectively averaged roughly +3% in May–October with a 70%+ win rate, beating the index by 100–170 bps. If you "sell in May" by rotating to the defensive trio rather than to cash, the historical record looks dramatically better than either holding the index or going to T-bills.
The negative skew is concentrated in three sectors. Consumer discretionary, industrials, and materials carry essentially the entire summer underperformance of the index. Discretionary is the only sector with a negative average May–October return and a sub-50% win rate.
Energy is a wildcard. Historically modestly negative in May–October on demand-cycle dynamics. In 2026 the picture is upside down — WTI dropped from a peak near $115 to roughly $88 on the Iran ceasefire, and a stable-to-rolling-over crude tape is itself input-cost relief for the rest of the index.

This is the level at which "Sell in May" stops being a slogan and starts being a portfolio decision. The right question is not should I sell? It is which side of my book is bearing the seasonal weakness? And the historical answer is: the cyclical side, almost exclusively.

Why 2026 Specifically Is the Wrong Year for the Naive Trade

Three macro inputs are conspiring to make the May–October 2026 setup look meaningfully different from the long-run average — and each one rotates the seasonal P&L distribution in the same direction.

1. The Iran ceasefire and oil reset. WTI dropped from a recent peak near $115 to roughly $88 after the April 7 ceasefire framework. Energy is one of the historically negative-skew sectors in May–October, but more importantly, falling crude is a direct margin tailwind for industrials, transports, and discretionary — exactly the cohorts that normally bear the seasonal underperformance. The macro is partially offsetting the calendar effect at the cyclical level.

2. The Fed's dovish pause is intact. The federal funds rate sits at 3.50%–3.75% after the late-2025 cuts, with futures priced flat through Q3 2026 and the next move skewed lower. Two of the four defensive seasonal-outperformers — utilities and real estate — have outsized rate-cut sensitivity. Two of the cyclical seasonal-underperformers — industrials and discretionary — also benefit from lower funding costs. The rate setup leans less negative than the average year.

3. Growth is slowing. The Atlanta Fed's GDPNow tracker for Q1 2026 has been cut to 1.2%, down from estimates above 3% earlier in the quarter. Slower growth without recession is the textbook regime for defensive outperformance. Staples, health care, and utilities historically gain relative strength when nominal GDP cools but stays positive. That maps cleanly onto the strongest seasonal cohort.

Stack the three together. The macro is leaning for the defensive trio, neutralizing the energy drag, and partly offsetting the cyclical drag. None of those inputs were present in an average post-1950 year. The textbook seasonal trade is being run against an atypical macro tape. That is the foundation of Paul Ciana's BofA call to buy in May and exit in July or August, and it is why CNBC is leading the May 1 preview with the "may not work this year" framing.

Three Ways to Express the Trade

Just like with any factor or macro thesis, the trade has a clean spectrum from full-discretion to retail-defined-risk. Pick the one that matches your account size, options access, and tolerance for whip risk through the summer.

Version 1 — The Pod-Trader Sector Pair

Long an equal-weight basket of XLP, XLV, XLU, XLRE. Short an equal-dollar basket of XLY, XLI, XLB. Run market-neutral on dollar exposure and beta-adjust if your platform requires it. Holding period: May 1 entry, August 8 exit (the BofA "buy May, sell July/August" overlay is roughly the historical optimum for the defensive-outperforms-cyclical leg). Expected gross spread historically averages 2.5%–3.5% over a three-month window with ~70% directional hit rate. Carry the book through earnings on each leg and reduce gross when sector implied vol pops above its 12-month median.

Version 2 — The Long-Only Investor Tilt

For a cash-equity portfolio, the right interpretation of the data is not to go to cash. It is to underweight the three structurally weak summer sectors (XLY, XLI, XLB) by 200–300 bps each versus benchmark, overweight the three structurally strong defensive sectors (XLP, XLV, XLU) by the same amount, and let the rest of the book run unchanged. Rebalance back to neutral on October 31. The expected lift is in the 70–120 bps range over the six-month window, with markedly lower drawdown profile than benchmark. Tax-aware investors should run the rebalance through new contributions where possible to avoid realizing gains on the trim.

Version 3 — The Defined-Risk Put-Spread Collar (Stay 100% Long)

This is the version retail traders should focus on, because it preserves the full positive expected return of staying invested while capping drawdown to a known dollar figure. The structure on SPY (ETF tracking S&P 500), assumed spot near $580 entering May:

Long leg: Buy 1x SPY July 30 -2.5%-OTM put (strike ~$565) to hedge a moderate drawdown. Cost roughly $5.40 debit per contract.
Short put leg: Sell 1x SPY July 30 -7%-OTM put (strike ~$540) to finance the long put. Credit roughly $2.80.
Short call leg: Sell 1x SPY July 30 +4%-OTM call (strike ~$603) to fully finance the spread. Credit roughly $2.65.

Net debit: approximately -$0.05 to +$0.30 per share depending on skew on the day of entry. Protected zone: any drawdown between roughly -2.5% and -7% on SPY. Capped upside: ~4% in the underlying through July expiry. This is the classic "summer collar" — you give up the right tail of a screaming summer rally in exchange for capping the left tail at -7%, for free.

Worked example for a $50,000 account running 60% long SPY (~52 shares):

Position size: 52 shares of SPY ≈ $30,160 notional.
Collar: 1x put-spread (long 565, short 540 puts) + 1x short 603 call, July 30 expiry.
Net cost: roughly $0–$30 net per spread.
Worst-case loss on the hedged equity leg through July expiry: ~$2.5% + ~$25 of additional spread loss below SPY 540 = roughly -$780 dollars on $30,160 notional. Compare to the unhedged worst case of an SPY -10% drawdown = -$3,016.

The whole point: you are not selling in May. You are staying long, paying near-zero for protection over the historically weak six-week window, and giving up a strip of upside that the data says is the least likely part of the summer-return distribution. If the BofA "buy May, sell July/August" call is right and the rally extends, you cap upside but participate to +4%. If the seasonal fade is real this year, the put-spread covers two-thirds of a moderate drawdown.

Watch This Before You Position

A short, focused walkthrough of the seasonal setup and the technical posture going into May. The video below frames the slogan against current price action and the key levels you should have on a chart this week.

The takeaway you should screen for as you watch: any May–October trade has to survive a credit-spread shock and a Fed-week surprise. If the macro stays where it is — Fed paused, oil rolling lower, growth cooling but positive — the defensive overweight is the trade. If credit spreads widen abruptly or the Fed signals a hawkish pause, that is when the seasonal weakness historically does show up at the index level, not just in the cyclical cohort.

The Decision Tree — Run This Before You Adjust the Book

A May–October repositioning has to pass each row of the matrix below in order. A "no" anywhere in the chain kills the trade or downsizes it sharply. The retail edge here is not faster execution — it is not entering the seasonal trade in the wrong macro regime.

| Step | Question | Action If Yes | Action If No | |---|---|---|---| | 1. Trend | Is SPX above its 50-day moving average? | Proceed | Tilt smaller — momentum already weak; calendar effect overstates the move | | 2. Credit | Is HY-IG OAS spread under 350 bps? | Proceed | Lean more defensive — calendar effect amplifies under credit stress | | 3. Rates | Is the 10y yield contained between 3.75% and 4.75%? | Proceed | Adjust XLU/XLRE allocation — both are rate-sensitive at the extremes | | 4. Oil regime | Is WTI between $75 and $100? | Proceed | Re-evaluate energy underweight — extreme moves break the seasonal pattern | | 5. Hedging cost | Can you buy a SPY put-spread collar for ≤ $0.50 net debit? | Layer the collar on top | Skip the collar, run the sector tilt only | | 6. Exit catalyst | Do you have a defined exit at the July FOMC and a kill switch on credit blowout? | Hold the book | Define both before adding risk |

<figcaption>The "Sell in May" repositioning decision tree. The trade is not a binary cash/equity flip. It is a sector tilt and an optional defined-risk collar, gated on credit, rates, and oil regime. Answer each row in order; one "no" stops the sequence or modifies the size.</figcaption> </figure>

The Bottom Line

The "Sell in May" calendar effect is real, but it is a 5-percentage-point relative effect at the index level, not a license to go to cash. The strict cash version of the rule has underperformed buy-and-hold by ~110 bps annually over 75 years, which compounds into more than half the wealth lost over a long career. The honest version of the data says the summer underperformance lives almost entirely in consumer discretionary, industrials, and materials, while staples, health care, and utilities have outperformed the index in the same window 70%+ of the time.

For 2026 specifically, three macro inputs are leaning against the textbook seasonal trade: the Iran ceasefire's oil reset, the Fed's dovish pause, and a slowing-but-positive GDP backdrop that historically powers defensive outperformance rather than cyclical breakdown. That is why the BofA technical desk is calling for buy May, sell July/August, and it is why every serious long-only strategist this week is recommending tilt-and-hedge rather than wholesale exit.

If you are going to act on the slogan, do it the way the data actually supports it. Underweight the three structurally weak summer sectors, overweight the defensive trio, and — for the retail accounts that have options access — wrap the equity book in a near-zero-cost SPY put-spread collar through July expiry. You stay long, you stay invested, and you cap the only part of the distribution the calendar effect has historically touched at the index level. That is the version of "Sell in May" that survives a 30-year career; the cash version is the version that wrecks one.

Educational content; not investment advice. Options trading involves substantial risk of loss and is not appropriate for all investors. Past performance is not indicative of future results. Always verify metrics, chain pricing, and macro data on your own broker's platform immediately before any trade, and consult a licensed financial professional. Read the OCC's Characteristics and Risks of Standardized Options before acting on any of the ideas discussed here.

Ready to Automate Your Trading?

Put these strategies into action with our AI-powered automation platform.

Start Trading