Sector Rotation

Key takeaways · 6

1. 01Sector rotation is the shift of capital between cyclical and defensive sectors as the business cycle and risk appetite turn. The trader-useful version follows observed relative-strength leadership; the calendar-based version has weak academic support after costs.
2. 02Roughly half of a stock's move comes from its industry group. Swing traders fish in leading sectors and skip laggard groups. Minervini, Weinstein, and Qullamägi all enforce this rule, just under different names.
3. 03There are 11 GICS sectors with a SPDR ETF for each (XLK, XLC, XLY, XLI, XLF, XLB, XLE, XLRE, XLP, XLV, XLU). Energy is the classification oddball: MSCI labels it defensive, retail traders treat it cyclical, and in practice it follows crude oil, not the business cycle.
4. 04Defensive leadership without an obvious news catalyst is an early regime warning, often visible one to three months before broad index weakness shows up.
5. 05Mega-cap concentration distorts sector ETFs. XLK is roughly 39% NVDA, AAPL, and MSFT combined. "XLK leadership" can hide a sector where the rest of the names are dead.
6. 06GICS reshuffled twice in living memory. Communication Services was created in September 2018; Visa and Mastercard moved out of Tech into Financials in March 2023. Pre-reshuffle backtests on legacy sector definitions are not directly comparable to today.

The Global Industry Classification Standard (GICS) splits the US stock market into 11 sectors. Every major US sector ETF you will trade (the State Street SPDRs) is built off this list. The tickers are worth memorizing because traders use them as shorthand: XLK is Technology, XLC is Communication Services, XLY is Consumer Discretionary, XLI is Industrials, XLF is Financials, XLB is Materials, XLE is Energy, XLRE is Real Estate, XLP is Consumer Staples, XLV is Health Care, XLU is Utilities.

The cyclical group leads when growth expectations rise and risk appetite is healthy. Technology (XLK) is the biggest sector by market cap and the textbook risk-on bellwether: software, semis, hardware. Communication Services (XLC) is mostly Meta and Alphabet, which makes it trade like growth-tech rather than the legacy telecom block it inherited the name from. Consumer Discretionary (XLY) is autos, retail, restaurants, and travel, the wallet-share gauge. Industrials (XLI) tracks capital expenditure and global PMI. Financials (XLF) lives off the yield curve and credit spreads.

The defensive group holds up better when the cycle rolls over. Consumer Staples (XLP) has inelastic demand for food, drinks, and household goods. Health Care (XLV) is mostly cycle-independent. People get sick on the same schedule regardless of whether the Fed is hiking. Utilities (XLU) is a bond proxy with regulated dividends; it leads when yields fall. Real Estate (XLRE) sits awkwardly between the two (cyclical on yields, defensive on rents) but is classified defensive in MSCI's formal scheme.

Then there is the middle. Materials (XLB) follows inflation and the global commodity cycle, which sometimes lines up with the equity cycle and sometimes does not. Energy (XLE) is the textbook oddball: MSCI classifies it defensive because it has low correlation with broad-market growth, but in practice it follows crude oil. When oil rallies, XLE rallies. When oil collapses, so does XLE, regardless of what the rest of the cyclical block is doing. Treat Energy as a separate animal that occasionally pretends to be a cyclical and occasionally pretends to be defensive depending on the supply situation.

The eleven sectors, the SPDRs that track them, and the three rough personality types. Energy gets its own callout because the textbook treats it defensive and the trading desk treats it cyclical, and they are both partly right.

Sam Stovall, then chief investment strategist at S&P Capital IQ, codified the modern sector-rotation framework in 1995-1996. He divided the business cycle into five phases (early recovery, mid-cycle expansion, late-cycle peak, early contraction, late contraction) and mapped which sectors historically lead each. The framework rests on the empirical regularity that equities lead the real economy by roughly five months on the way up and seven months on the way down. So if you can identify the cycle phase, you can theoretically front-run the rotation that the rest of the market is about to do.

In Stovall's scheme, early recovery favors Consumer Discretionary and Technology as rate cuts and pent-up demand re-emerge. Mid-cycle expansion favors Industrials and Materials as capital expenditure broadens. Late-cycle peak favors Energy and Materials as the inflation print catches up. Early contraction favors Staples and Health Care. Late contraction, with the Fed easing again, favors Utilities and Financials. The diagram below summarises the canonical wheel.

As vocabulary, the framework is useful. When somebody on a finance podcast says "we're late cycle," what they almost always mean is "Energy is leading and Industrials are starting to crack." They are pattern-matching against Stovall whether they know it or not. The framework gives you a shared language for talking about regime.

It is much less useful as a trading system. The most rigorous recent test (Molchanov and Stangl, "The myth of business cycle sector rotation," International Journal of Finance and Economics, 2024) finds that a strategy of rotating between sectors based on NBER-dated cycle phases earns roughly 0.11% per month in raw alpha. After transaction costs and turnover the gross edge disappears. Earlier work (Conover, Jensen, Johnson, Mercer 2008) found a slightly larger effect when monetary policy was layered in as a counterweight, but even that variant does not survive a 2010-2024 out-of-sample test cleanly. The plain-vanilla calendar-based rotation does not pay.

By the time the cycle phase is identifiable, the rotation has mostly already happened. Equities lead the real economy by months, which means the sector rotation is priced in by the time the macro release confirms the phase. The trader who waits for confirmation is buying after the move. The trader who guesses ahead has no real edge over the published consensus.

Use the cycle as vocabulary. Trade observed leadership.

Stovall's five phases with their textbook leaders. Useful as shared vocabulary for talking about regime. Useless as a trading system once transaction costs come in.

GICS is not static. The committee that maintains the standard (S&P and MSCI jointly) reshuffles the categories whenever the underlying economy drifts far enough that the old labels no longer track. There have been two material reshuffles in living memory and you have to know about both, because backtests that span them will silently be comparing apples to oranges.

September 21, 2018: the Communication Services creation. The old "Telecommunication Services" sector (three or four wireline phone companies) was renamed Communication Services and absorbed several of the largest stocks in Tech and Discretionary. Alphabet (then Google), Meta (then Facebook), Netflix, and the Disney-style media conglomerates all moved over. Tech lost roughly 20% of its weight overnight. Discretionary lost Netflix and a few others. The newly created XLC is mostly Meta and Alphabet, which is why XLC and XLK still move so similarly today even though they are technically different sectors.

March 17, 2023: the payments rebalance. Visa and Mastercard moved from Information Technology into Financials, fitting them next to the banks and insurers they actually do business with. Target, Dollar Tree, and Dollar General moved from Consumer Discretionary into Consumer Staples on the basis that discount retail behaves more defensively than full-line department stores. The reshuffle was small in dollar terms but matters for any sector RS analysis that crosses the date.

For a swing trader the practical implication is short. If you are running a sector momentum strategy and your backtest includes 2017 and 2025, your "Tech sector" is genuinely two different baskets across that window. Either rebuild the basket on the post-2018 definition before testing, or restrict the test to a homogeneous sub-period. Otherwise you are studying a Frankenstein.

The cleanest sector rotation in recent memory played out across calendar 2022. The S&P 500 finished the year down 19.4%, the Nasdaq Composite down 33%, and the Magnificent Seven down roughly 41%. XLE returned +57.6%. The gap between Energy and Tech was almost 90 percentage points in twelve months. If you held one and not the other, you had a year that did not look anything like the one your neighbor had.

The setup. Inflation had been running hot since mid-2021. The Fed had been signalling a hike-fast pivot since early November of 2021 ("transitory" was retired November 30). The S&P 500 hit its all-time high at 4,796.56 on January 3, 2022, and rolled over within days. Tech traded like the long-duration asset it had been for the prior decade. Every basis point of yield revision compressed multiples. The Nasdaq fell 11% in the first three weeks of January alone.

The catalyst. February 24, 2022: Russia invaded Ukraine. Brent crude was around $97 the morning of the invasion; it would peak above $130 by March 8. XLE broke out of a multi-year base on the news and ran almost without interruption through June. By June 8 the ETF was up 62% year-to-date. A single-sector six-month return exceeding the typical full-year return of the S&P 500 in any direction.

The second leg. After XLE topped in June and chopped sideways, the rest of 2022 was about Tech finding a bottom. The S&P 500 made its closing low at 3,577.03 on October 12, the day a hot CPI print briefly knocked the index lower before it reversed sharply intraday. Tech bottomed there alongside it. Energy, having already had its move, drifted but held most of its gains through year-end.

The lesson is not "buy energy." Energy was the obvious trade only in retrospect. The lesson is that the rotation was visible in the relative-strength lines from January, when XLE's ratio against SPY broke clearly above its 200-day moving average and never looked back. Anyone who watched the RS lines instead of the headline indexes had the entire year mapped out by February. The rotation did not require a forecast. It required attention.

XLE and XLK across 2022, both rebased to 100 at the start. The two lines diverge from January 3 (XLK's all-time high) and never reconnect. Energy ends the year up 57.6%; Tech ends it down 28.4%.

Some rotations take a year. Others take a single session. Monday, November 9, 2020 was the second kind.

The setup. Tech had led off the COVID lows in March 2020 and continued to dominate through the summer. The Magnificent Seven names (the original FAANG plus Tesla and Nvidia) compounded at a pace that pulled the S&P 500 to fresh all-time highs by August. XLK peaked at $125.94 on September 2, then traded sideways for two months while the rest of the market caught up. XLE, by contrast, was still grinding through the worst year of its modern history. Crude had collapsed in March, briefly going negative on the WTI futures roll in April, and energy stocks had not participated in the rebound. By the end of October XLE was down roughly 50% on the year while XLK was up around 28%. The gap was generational.

The catalyst. Pre-market on November 9, Pfizer and BioNTech announced their COVID-19 vaccine had achieved 90% efficacy in late-stage trials. The implications were enormous and immediate: a path back to normal economic activity, a re-acceleration of energy demand, a return to spending on services that had been frozen for eight months. Within minutes of the open, the trade flipped. XLE closed up 14.3% on the day. XLK closed down 0.7%. The single-session relative move was almost 15 percentage points.

What followed. The rotation persisted. From November 9 through the end of March 2021, XLE returned roughly 48% from its pre-Pfizer level while XLK returned roughly 10%. The Pfizer Day was the visible inflection, but the regime change had been priced into the bond market for weeks before. Long-duration Treasuries had been selling off since early August, which is the macro equivalent of a defensive bid going away. A trader watching the 10-year yield could have anticipated the rotation. A trader watching only the headline index saw it as a one-day news shock.

Some leadership changes drift in over months. Some flip in a single session. Pfizer Day was the cleanest single-session flip in living memory. The lesson: when the cross-asset signals (long bonds selling off, value catching a bid against growth on the relative line) start to line up, the news catalyst is just the trigger. The setup had been forming for months.

September 2020 through March 2021, the Pfizer-Day rotation. The vertical separation on November 9 is the single largest one-session leadership flip on this chart. The trend persisted for the next four months.

Two charts cannot cover the patterns the framework describes. Four more episodes worth knowing, compressed.

2000-2003 dot-com bust. The Nasdaq fell 78% from March 2000 to October 2002. Defensive sectors held up: Consumer Staples returned roughly 11% annualized over the same window, Utilities held positive, Health Care lost less than the index. Energy actually rallied on a crude move from $20 to over $35. This is the textbook example of defensive leadership during a tech-led drawdown.

2003-2007 financials and materials lead. The Fed cut from 6.5% to 1% by mid-2003. Financials, Real Estate, and Materials/Energy led a five-year run while Tech lagged badly off its bubble peak. The commodity supercycle plus the housing-and-credit boom did the heavy lifting until they cracked together in 2007-2008.

2008 GFC defensives only. During the Global Financial Crisis drawdown, Staples, Health Care, and Utilities held up best in relative terms. XLF lost roughly 55%, XLE roughly 35%, while XLP fell only about 15%. Classic recession-quadrant behavior. Worth flagging as a counter-example: 2008 is a teaching chart for how rotation reduces losses, not for how rotation creates returns. Both lines went down. The question was which went down less.

2009-2021 secular tech run. Post-GFC zero-interest-rate policy plus the FAANG (later Magnificent Seven) earnings juggernaut produced 12-plus years of growth and tech leadership. Brief value flares (mid-2013, late 2016) appeared and faded within months. The era ended with the November 2021 turn that became the 2022 rotation in the previous section.

2023-2024 narrow then broad. The seven mega-caps drove roughly 62% of the S&P 500's 26.3% return in 2023. By mid-2024 their share dropped toward 40% as Financials, Industrials, and Health Care joined the rally. The textbook breadth thrust setup that often precedes a durable bull market continuation. Equal-weight S&P (RSP) ended 2023 at +11.7% versus cap-weighted at +26.3%, a 14-point gap that was one of the widest in 30 years. By late 2024 RSP had closed about half of that gap, which was itself the rotation.

When you say "Tech is leading," you are usually saying something more specific than you mean. The cap-weighted SPDR sector ETFs are dominated by their largest names to a degree that has gotten worse, not better, over the last decade.

XLK is the worst case. Nvidia is roughly 15.4% of the ETF as of recent rebalancing, Apple is about 13.5%, Microsoft is about 9.8%. Three names, 38.7% of the basket. The other 65 holdings split the remaining 61% pro rata. When you "buy XLK," you are buying a 38.7% bet on three megacaps and a 61% bet on a 65-stock basket whose weights run from roughly 2.7% (Broadcom) down to under 0.4% at the long tail.

XLC and XLY are not far behind. Communication Services is about 45% Meta and Alphabet combined. Discretionary is about 40% Amazon and Tesla. The 2018 GICS reshuffle accidentally created a problem: by carving the most concentrated mega-caps into their own sectors, the SPDRs ended up with two-stock or three-stock proxies in everything but name.

For a swing trader the implication is direct. "Sector leadership" inside XLK can mean three megacaps lifting the ETF while the rest of Tech is flat or down. This is what happened for most of the first half of 2024: NVDA was up 150% year-to-date, the rest of XLK was up roughly 7%, and the headline ETF showed about 18%. Anyone reading "XLK leadership" as "Tech breadth is healthy" was getting the wrong signal.

The fix is a cross-check. Equal-weight ETFs solve the concentration problem by capping every name at roughly 1/N of the basket. RSPT (Invesco S&P 500 Equal Weight Technology) holds the same Tech names as XLK but sizes each to about 1.5% to 2.2% of the basket. When XLK is rising sharply but RSPT is flat, the leadership is concentrated. When both are rising together, the leadership is broad. The same pairing exists for every sector (RSPC for Communications, RSPD for Discretionary, and so on). The divergence between cap-weight and equal-weight is one of the cleanest breadth signals available without subscribing to anything.

XLK versus RSPT. Same Tech sector, two very different bets. The right-hand column is what "Tech sector exposure" looked like before cap-weighting concentrated everything into three names.

Four practical tools for watching rotation as it happens. None require a paid subscription.

Sector RS lines. Each sector ETF plotted as a ratio against SPY. A rising line means the sector is outperforming; a falling line means it is lagging. StockCharts and TradingView both show this directly with "$XLK:$SPY" or "XLK/SPY" syntax. The simplest read is which lines are rising, which are falling, and which just turned. Stan Weinstein's Mansfield Relative Strength is a refinement that normalizes the RS line against its own moving average so a strong but trendless ratio reads neutral; for most retail traders the raw line is fine.

Pairwise ratio charts. XLY divided by XLP, Consumer Discretionary over Consumer Staples, is the canonical risk-on / risk-off gauge. A rising ratio means traders are paying up for cyclicals and risk appetite is healthy. A falling ratio means capital is rotating defensive. XLK over XLE captures growth versus commodity. XLF over XLU captures cyclical versus bond-proxy. Watching three ratio charts gives you a fuller cross-section than any single one.

IBD Industry Group rankings. Investor's Business Daily ranks 197 industry groups by 6-month price performance and publishes the table daily. The top 20 is the working hunting ground for CANSLIM-style traders. The math behind why this matters is O'Neil's claim that roughly half of a stock's move comes from its industry group. Buy the leading stock in a leading group. Ignore breakouts in laggard groups regardless of how clean the chart looks.

Defensive bid as an early warning. XLU, XLP, and XLV outperforming SPY without an obvious news catalyst is a regime warning that often shows up one to three months before the broad index actually rolls over. The early-2018 defensive bid foreshadowed the Q4 correction. The early-2022 defensive bid foreshadowed the Tech-led drawdown that accelerated through the year. Pay attention when staples and utilities lead a market that is making new highs.

Three of the most-studied retail trading methodologies all use sector rotation. They just call it different things.

Mark Minervini's SEPA. The Specific Entry Point Analysis framework explicitly demands that the stock's industry group be in a Stage 2 advance before any long entry is considered. Minervini quantifies this through his Trend Template: industry group RS rating of 70 or higher, group price above its rising 50-day moving average. A clean breakout in a stock whose group fails the template is rejected before chart structure is even examined.

Stan Weinstein's stages. The original Weinstein rule, from "Secrets for Profiting in Bull and Bear Markets" (1988): never buy a Stage 2 stock if its sector is in Stage 4 (decline). The sector context overrides the individual chart. A stock breaking out from a perfect Stage 1 base inside a Stage 4 sector is simply a setup with bad base rates; the win rate is too low to bother. Weinstein's sector-first rule is the one that survived the longest into the modern era.

Kristjan Kullamägi (Qullamaggie). The Estonian trader who compounded a small account into nine figures is publicly explicit that he treats the leading sector as a force multiplier. He hunts breakouts and episodic pivots inside the strongest groups and ignores everything else. His filter is brutal: sector RS rank in the top three, stock RS rank above 90, no exceptions. Three different systems, same rule: the sector tailwind matters.

The corollary: if you are taking a setup inside a laggard sector because the chart looks particularly good, you are betting that the sector will catch up rather than that the stock will continue to outperform. Both bets can win. The second one has a 50-year body of evidence behind it.

The Leadership subscore on the dashboard rolls five rotation signals into a single 0-to-100 score that feeds the headline Stance number. The anchor at 40% weight is sector RS, which is the share of US sectors beating SPY over a rolling window. 30% beating maps to a narrow regime read; 70% beating maps to a broad-participation read. Small versus large (IWM minus SPY 60-day spread, 20% weight) and growth versus broad (QQQ minus SPY, 15% weight) refine the read across cap and style. Cyclicals minus defensives (15%) and equal-weight versus cap-weight (10%) round it out.

The dashboard does not predict which sector must lead next. It shows whether leadership is broadening, narrowing, or shifting toward defensive areas, so the trader checking before market open knows whether to press, pause, or step back. The sector heatmap surfaces the rotation in real time, with every sector's RS line color-coded for trend direction. The Leadership subscore feeds the headline Stance, so a defensive rotation drags the regime read down rather than getting averaged away by an otherwise healthy Trend score.

A trader who has read this far does not need TickerStance to read sector rotation. Watching XLY:XLP and the IBD industry group leaderboard for fifteen minutes a day will give you the same signal. The dashboard consolidates that signal into a single reproducible number, weighted against three other regime lenses, with the ingredients exposed so you can disagree with the weighting. The methodology page documents every signal and weight, so the same inputs always produce the same Stance.

GICS — the Global Industry Classification Standard, jointly maintained by S&P and MSCI. Defines the 11 sector categories and the underlying industry sub-groups. Reshuffled in September 2018 (Communication Services creation) and March 2023 (Visa/Mastercard moved to Financials, discount retail moved to Staples).

Cyclical sector — a sector whose earnings rise and fall with the business cycle: Technology, Communication Services, Discretionary, Industrials, Financials, Materials, plus Real Estate on yields. Outperforms in expansions and underperforms in contractions.

Defensive sector — a sector whose earnings are relatively cycle-independent: Consumer Staples, Health Care, Utilities. Holds up better than the index in downturns and lags in strong expansions.

SPDR sector ETFs — the State Street family of single-sector ETFs that traders use as sector proxies: XLK, XLC, XLY, XLI, XLF, XLB, XLE, XLRE, XLP, XLV, XLU. One per GICS sector.

Equal-weight sector ETF — a sector ETF that caps every constituent at roughly 1/N of the basket regardless of market cap. RSPT (Tech), RSPC (Communications), RSPD (Discretionary), and so on. Used as a cross-check against cap-weight ETF concentration.

Stovall cycle — Sam Stovall's five-phase framework (early recovery, mid-expansion, late peak, early contraction, late contraction) mapping cycle phase to historical sector leaders. Useful as vocabulary, weak as a trading system after costs.

Sector RS line — the ratio of a sector ETF to a broad-market benchmark (typically SPY), plotted as a line. Rising = sector outperforming; falling = sector lagging. The simplest leadership read available.

Mansfield Relative Strength — Stan Weinstein's normalized RS line, adjusted against its own moving average so a flat-but-strong ratio reads neutral instead of bullish. Reduces noise in trendless RS lines.

Ratio chart — a chart of one ticker divided by another, used to read rotation between two assets directly. XLY:XLP is the canonical risk-on / risk-off ratio. XLK:XLE is the growth-versus-commodity ratio.

IBD Industry Group rank — Investor's Business Daily's daily ranking of 197 industry groups by 6-month price performance. Top 20 is the standard working hunting ground for CANSLIM traders.

Defensive bid — quiet outperformance by Utilities, Staples, and Health Care without an obvious news catalyst. An early-warning signal of regime stress, often visible one to three months before broad-index weakness.

Mega-cap concentration — the share of a sector ETF concentrated in its largest holdings. XLK is roughly 39% NVDA + AAPL + MSFT combined. Concentration distorts the ETF's price away from the average sector stock.

Frequently asked questions

Keep reading

• Market RegimeWhy sector leadership is one of the four lenses for reading market regime.
• Relative StrengthThe same RS math powers sector leadership and stock leaderboards.
• Stock LeadersOnce you know the leading sectors, the leaderboards surface the strongest names inside them.
• MethodologyEvery weight, normalization rule, and signal that powers the Leadership subscore.