Cyclical Stocks: Industrials and the 3.5-Year Business Cycle (3M, Parker-Hannifin, Emerson, Eaton vs XLI and SPY)

3M, Parker-Hannifin, Emerson and Eaton share the market's 3.5-year business-cycle rhythm. We compare their cycles and drawdowns with the industrial sector fund XLI and with SPY.

Industrial stocks are the textbook cyclical stocks: when the economy grows, companies order machinery, components and freight, and when it slows those orders dry up first. But are their price swings actually cyclical in the sense of a repeating rhythm? We measured the multi-year cycles of 25 long-history industrial stocks and compared them with the Industrial Select Sector SPDR fund (XLI) and with SPY. Four of them share the same beat as the broad market.

  • 3M, Parker-Hannifin, Emerson Electric and Eaton all show a measurable 3.4–3.5-year cycle, the same length as the Dow’s and close to the 40-month Kitchin business cycle.
  • The sector fund XLI shows the same 3.4-year rhythm, but less reliably (about a 12% chance of luck) because its members peak at slightly different times.
  • XLI fell more than SPY in 7 of its 8 major drawdowns since 1999, including −58% against −52% in 2007–09 and −28% against −18% in early 2020.
  • Parker-Hannifin amplifies the sector: it fell more than both XLI and SPY in every one of its recent drawdowns, then rebounded far more (+637% from the March 2020 low).
  • Not all industrials follow the business cycle. Defence stocks such as Lockheed Martin barely move with the market, and railroads, Caterpillar and Deere show no reliable period.

Why industrial stocks follow the business cycle

Industrial companies sell capital goods, components and services whose demand depends on other companies’ spending plans. When order books fill, manufacturers add inventory and capacity; when demand cools, customers run down inventory and orders fall faster than end demand. That inventory swing is exactly the mechanism economists use to explain the short Kitchin cycle of around 40 months, and it is why investors watch manufacturing surveys (PMIs) and industrial production to judge where the business cycle stands. In sector rotation frameworks, industrials are expected to lead in early recoveries and lag late in the cycle.

The 3.5-year rhythm in four industrial stocks

The chart shows how far each stock’s monthly close sat above or below its own long-term trend. Peaks and troughs line up with each other and with the grey S&P 500 sell-offs: 1987, 1990, 1998, 2000–02, 2008–09, 2011, 2018, 2020 and 2022. Use the legend to switch symbols on and off; XLI and SPY are included for comparison.

The table compares the strongest cycle of each stock with the sector fund and the market. “Moves with XLI” and “Moves with SPY” are correlations of the swings around trend (1 = identical, 0 = unrelated).

Industrial stocks: multi-year cycle statistics, with the Industrial Select Sector SPDR (XLI) and SPY. "Chance it's luck" = share of random-walk simulations producing a cycle at least as strong; correlations compare the swings around each symbol's long-term trend.
SymbolNameComplete yearsStrongest cycleChance it's luckCycle ratingSwing around trendTypical peak spacingDeclines during S&P sell-offsMoves with XLIMoves with SPY
MMM3M Company533.5 yrs<1%Moderate±9%2.3 yrs10 of 160.570.40
PHParker-Hannifin Corporation533.4 yrs2%Moderate±10%2.4 yrs12 of 130.820.62
EMREmerson Electric Co533.4 yrs2%Moderate±7%2.1 yrs15 of 170.850.74
ETNEaton Corporation plc533.4 yrs4%Moderate±11%2.8 yrs12 of 170.760.59
BAThe Boeing Company535.6 yrs2%Moderate±19%4.0 yrs10 of 110.690.43
UPSUnited Parcel Service Inc268.0 yrs3%Moderate±11%1.8 yrs6 of 80.480.46
HONHoneywell International Inc416.8 yrs9%Weak±11%2.5 yrs10 of 110.760.82
GDGeneral Dynamics Corporation534.3 yrs10%Weak±13%2.3 yrs10 of 160.630.46
PCARPACCAR Inc535.2 yrs12%Weak±11%2.7 yrs12 of 180.590.46
CATCaterpillar Inc536.0 yrs27%None±12%2.9 yrs11 of 150.650.49
DEDeere & Company533.4 yrs18%None±11%2.8 yrs10 of 170.640.42
ITWIllinois Tool Works Inc523.4 yrs36%None±6%1.9 yrs14 of 170.660.65
GEGE Aerospace568.0 yrs34%None±12%3.8 yrs9 of 100.610.71
UNPUnion Pacific Corporation535.4 yrs26%None±8%2.8 yrs11 of 160.540.35
LMTLockheed Martin Corporation305.3 yrs18%None±13%2.6 yrs8 of 90.240.02

Three groups stand out. The machinery and electrical-equipment makers (Parker-Hannifin, Emerson, Eaton, and 3M until its recent troubles) have the cleanest 3.4–3.5-year cycles and track XLI closely (correlations of 0.76–0.85 for Emerson, Parker-Hannifin and Eaton). Boeing and UPS have longer 5.6- and 8-year waves tied to their own aircraft and parcel cycles. Defence contractors are the exception: Lockheed Martin’s swings are essentially unrelated to SPY’s (correlation 0.02) because government budgets, not the business cycle, drive their orders.

Spotlight: Parker-Hannifin, a leveraged play on the industrial cycle

Parker-Hannifin makes motion-control and filtration components used across manufacturing and aerospace. Its fitted cycle lasts about 3.4 years (around a 2% chance of luck), and 12 of its 13 major drawdowns happened during broad S&P 500 sell-offs, the highest share in the group.

Lining up each Parker-Hannifin drawdown with XLI and SPY over the same months shows how much the stock amplifies the cycle. In early 2020 it fell 37% while XLI fell 28% and SPY 20%; in 2008–09 it lost 61% against 56% and 47%. The rebounds were just as outsized: +184% after 2009, +219% after 2011 and +637% (so far) after March 2020.

PH: major declines (monthly closes, reversals of 25% or more), latest first, with other symbols over the same months. "–" = no data for that period.
Peak – troughMonthsPHXLI same monthsSPY same monthsPH recovery to next peakDuring S&P sell-off
Dec 2019 – Mar 20203−37%−28%−20%+637% (so far)Yes
Jan 2018 – Dec 201811−26%−19%−11%+38%Yes
Mar 2011 – Sep 20116−33%−22%−15%+219%Yes
May 2008 – Feb 20099−61%−56%−47%+184%Yes
May 2001 – Sep 20014−29%−23%−17%+270%Yes
Apr 2000 – Sep 20005−27%+3%−1%+43%Yes
Dec 1999 – Feb 20002−29%−13%−6%+28%Yes
Mar 1998 – Aug 19985−43%–−13%+77%Yes

3M: when a cyclical stock breaks from the cycle

3M has one of the most statistically reliable cycles on the site (under a 1% chance of luck over 53 years), yet its most recent drawdowns show the limits of any cycle model. From May 2021 to May 2023 the stock fell 54% while XLI slipped only 8% and SPY was flat. That decline was driven by company-specific problems, large product-liability settlements and a restructuring, not by the industrial cycle. The lesson: a stock’s historical rhythm tells you how it usually behaves around the market cycle, not what happens when its own fundamentals change.

MMM: major declines (monthly closes, reversals of 17% or more), latest first, with other symbols over the same months. "–" = no data for that period.
Peak – troughMonthsMMMXLI same monthsSPY same monthsMMM recovery to next peakDuring S&P sell-off
Jul 2023 – Oct 20233−18%−11%−9%+142% (so far)No
May 2021 – May 202324−54%−8%−1%+19%Yes
Jan 2018 – Mar 202026−46%−26%−9%+49%Yes
Apr 2011 – Sep 20115−26%−24%−17%+249%Yes
Sep 2007 – Feb 200917−51%−58%−52%+114%Yes
Apr 2006 – Jul 20063−18%−8%−3%+33%No

XLI against SPY: how the market behaves around the industrial cycle

The sector fund itself is the best guide to the group. XLI’s swings around trend move closely with SPY’s (correlation 0.80), and every major XLI drawdown except a short one in 2010 happened during a broad market sell-off. Industrials fell harder than the market in 7 of 8 cases; the exception was 2022, when growth and technology stocks led the decline and XLI fell 22% against SPY’s 25%.

XLI: major declines (monthly closes, reversals of 15% or more), latest first, with other symbols over the same months. "–" = no data for that period.
Peak – troughMonthsXLISPY same monthsXLI recovery to next peakDuring S&P sell-off
Dec 2021 – Sep 20229−22%−25%+104% (so far)Yes
Nov 2019 – Mar 20204−28%−18%+79%Yes
Jan 2018 – Dec 201811−19%−11%+27%Yes
Apr 2011 – Sep 20115−24%−17%+173%Yes
Apr 2010 – Jun 20102−16%−13%+41%No
Sep 2007 – Feb 200917−58%−52%+90%Yes
Dec 2001 – Feb 200314−31%−26%+114%Yes
May 2001 – Sep 20014−23%−17%+17%Yes

In other words, the industrial cycle and the market cycle are largely the same cycle, with industrials as the higher-beta version. Their troughs have coincided with market bottoms (March 2009, September 2011, December 2018, March 2020, September 2022), and their recoveries have been stronger: XLI gained 173% from the 2011 low and has more than doubled since September 2022.

What this means for investors

  • Expect deeper drawdowns than the index. Over the last 25 years a typical XLI decline was a few percentage points deeper than SPY’s over the same months, and individual machinery stocks fell further still.
  • Industrial troughs have been buying opportunities when they coincided with market-wide sell-offs, as in 2009, 2011, 2018, 2020 and 2022.
  • Watch the business cycle, not the calendar. The 3.5-year rhythm is an average; actual gaps between peaks ranged from under two years to over eight.
  • Separate cyclical from company-specific risk, as 3M showed in 2021–23.

How we measure a cycle: each symbol’s monthly closes are compared with their own long-term trend (a 10-year local regression of the log price). We then test every cycle length from 2 to 8 years and ask how often random price wandering with the same persistence produces a cycle at least as strong (120 simulations per symbol; this is the “chance it’s luck” column). Turning points are peaks and troughs separated by a reversal scaled to each symbol’s volatility. Prices exclude dividends. Full method in our guide to stock market cycles.

Frequently asked questions

Are industrial stocks cyclical?

Yes. Industrial companies sell equipment and services whose demand rises and falls with business spending, so their earnings and share prices follow the business cycle. In our data, 3M, Parker-Hannifin, Emerson and Eaton show a repeating 3.4–3.5-year price cycle, the same length as the Dow’s.

How long is the industrial stock cycle?

About 3.4–3.5 years from peak to peak for the clearest examples, close to the 40-month Kitchin inventory cycle. The sector fund XLI shows the same 3.4-year rhythm, but individual gaps between peaks have ranged from under two years to more than eight.

Do industrial stocks fall more than the S&P 500?

Usually. XLI fell more than SPY in 7 of its 8 major drawdowns since 1999, for example −58% against −52% in 2007–09 and −28% against −18% in early 2020. The exception was 2022, when technology stocks led the decline.

Which industrial stocks are not cyclical?

Defence contractors such as Lockheed Martin move largely independently of the market because their orders depend on government budgets. Railroads, Caterpillar and Deere are economically cyclical but show no reliable repeating period in their prices.

What is the best ETF for industrial stocks?

The Industrial Select Sector SPDR Fund (XLI), which holds the industrial companies in the S&P 500, is the most widely used benchmark and the one we compare against here. This is not a recommendation; compare costs and holdings before investing.

More market-cycle research

Figures are historical, based on monthly closing prices to mid-November 2026, and are updated as new data arrives. Past cycles do not predict future returns; this is research, not investment advice.