Skip to main content

The Cyclical Peak Trap

Cyclical industries—steel, auto manufacturing, construction equipment, semiconductors, oil and gas—move in multi-year waves. During boom phases, earnings expand to multiples of their trough levels. Investors, seduced by attractive valuations at peak earnings, buy cyclical stocks believing they're getting bargains. Then the cycle turns. Earnings collapse. Valuations compress. Stocks fall 50-70%. The mistake was buying at peak earnings with valuation that assumed peak earnings would persist.

Quick definition: The cyclical peak trap is the mistake of valuing a cyclical stock based on its peak-cycle earnings as if that level will sustain, then watching earnings collapse as the cycle turns, destroying shareholder value despite the initial valuation appearing reasonable.

Key Takeaways​

  1. Cyclical stocks appear cheapest (lowest P/E ratios) at peak earnings; this is precisely when they are most expensive on a normalized earnings basis.
  2. Normalized earnings power—adjusted for the full business cycle—is what determines long-term value, not earnings at a particular point in the cycle.
  3. The business cycle phase is paramount in cyclical valuation; buying into the end of an expansion is far more dangerous than buying in the trough.
  4. Mean reversion in earnings is the cyclical investor's worst enemy: what appears to be a permanent 20% return on equity may compress to 8% when conditions normalize.
  5. Industry capacity and supply additions create turning points that are difficult to time; new production capacity entering the market often marks the start of the downturn.
  6. Leverage amplifies cyclical damage; a cyclical company with high debt and peak earnings faces devastation when earnings normalize or compress.

The Earnings Peak as a Valuation Trap​

The cyclical peak trap has a elegant but devastating logic:

Cyclical companies' earnings move in lockstep with their industry. In a steel company's best years—when global demand is booming, capacity is fully utilized, prices are at multi-year highs, and margins are at their peak—earnings might be $10 per share. An investor looking at a $10 P/E ratio on $10 earnings thinks they're buying a cheap stock.

But this is circular reasoning. The stock appears cheap because it's valued on peak earnings. When the cycle turns—demand softens, new capacity enters the market, prices compress, margins normalize—earnings might fall to $5 or $3 per share. The P/E ratio, which looked attractive at 10x peak earnings, expands to 20x normalized earnings. The stock that appeared cheap is now expensive, and it falls accordingly.

The trap is that peak-cycle earnings feel sustainable. When an investor buys a steel company in the 9th year of an expansion with record margins and fully booked capacity, it's psychologically difficult to imagine that earnings will fall 50%. The business looks great. The earnings are real. The problem is that cyclical earnings are never sustainable at their peak.

Distinguishing Cyclical from Structural​

A prerequisite for avoiding the cyclical peak trap is distinguishing genuinely cyclical businesses from those that just happened to have lumpy earnings.

True cyclical industries show several characteristics:

  • Commodity-like output: Differentiation is limited; customers choose based on price and availability.
  • Supply constraints that create cycles: When capacity is constrained, prices and margins spike. New supply eventually floods the market, compressing margins.
  • Highly correlated earnings: All competitors in the industry experience boom and bust together; individual company performance diverges less than in secular growth or decline industries.
  • Historical earnings volatility: The company has experienced multiple full cycles of boom and bust, with earnings swinging 200-300% from trough to peak.

Industries that are textbook cyclical: steel, oil and gas, coal, metals and mining, automotive, construction equipment, semiconductors (though less so in recent decades), chemicals, and shipping. Real estate development is cyclical. Commercial banking can be cyclical (earnings boom during expansions, compress during recessions).

Industries that can be mistaken for cyclical but are actually secular: software (lumpy earnings due to large deals but overall structural growth), biotech (lumpy earnings due to regulatory decisions but secular growth in addressable market), and luxury goods (earnings correlate with consumer confidence but overall growth trend is up).

The mistake is treating a secular grower with lumpy short-term earnings as cyclical. If you buy a software company at peak quarterly sales and assume earnings will stay there, that's not a cyclical trap; that's just a growth extrapolation mistake. But if you buy a steel company at peak industry margins assuming margins will stay at 25%, that's a cyclical peak trap.

Normalized Earnings and Mid-Cycle Valuation​

The correct approach to valuing cyclical businesses is using normalized earnings—an estimate of what the company would earn in a "normal" or mid-cycle period, accounting for the full range of the business cycle.

If a steel company earns:

  • Trough: $2 per share (recession, depressed prices)
  • Mid-cycle: $5 per share (normal demand, normalized margins)
  • Peak: $10 per share (boom, maximum margins)

The normalized earnings for valuation purposes should be $5, not $10. This company at a $50 stock price is trading at 10x normalized earnings, which might be reasonable. But if you valued it on peak earnings of $10, you'd think a $50 stock was trading at only 5x earnings, seemingly cheap—when in reality it's twice as expensive as it appears.

The challenge is estimating normalized earnings, which requires:

  1. Understanding the industry cycle and where it stands
  2. Knowing the company's cost structure at various utilization levels
  3. Estimating medium-term pricing, capacity, and demand

This requires industry expertise. Generic screens that identify cheap stocks by low P/E ratios fail catastrophically at cyclical companies because they catch companies at peak earnings.

The Capacity Turning Point​

The most dangerous moment in a cyclical industry is typically 3-6 months before a major downturn begins. At this point, the cycle has peaked, all capacity is spoken for, margins are at their highs, and momentum appears strongest. Investors are most confident. Then capacity additions come online or demand softens, and the turn happens suddenly.

For example, in semiconductors, the cycle often peaks when all leading-edge capacity is fully booked and new plants are being announced. But as announcements become capacity additions, oversupply follows within months. By the time the oversupply is visible in industry data, the stock has already fallen 30%.

In oil and gas, peaks often coincide with surging upstream capital spending and announcements of new production. These new projects come online 2-3 years later, right as demand has normalized. By then, the industry is in downturn, and companies that seemed wonderfully profitable are generating losses.

Understanding industry capacity dynamics—how much new supply is planned to enter the market and when—is essential. An industry that looks fully booked but has 30% additional capacity coming online in 18 months is not a good buy at peak margins.

Leverage: The Cyclical Killer​

Cyclical companies with high debt face devastation when cycles turn. This is where the trap becomes truly dangerous.

A steel company with $10 peak earnings and $60 billion in debt might feel manageable when earnings are strong. The debt-to-EBITDA ratio might be 3-4x, "reasonable" for the industry. But when earnings normalize to $5 billion (a 50% decline), the same debt becomes 12x EBITDA. When earnings fall to $2 billion during recessions, it becomes 30x EBITDA—often above covenant levels.

Many highly leveraged cyclical companies are forced to:

  • Cut dividends, destroying capital for income investors
  • Sell assets at distressed prices to repair balance sheets
  • Dilute shareholders through equity raises to reduce debt
  • Violate covenants and face restructuring

An investor who bought a seemingly cheap, highly leveraged cyclical company at peak earnings discovers that the debt risk was masked by peak profitability. The leverage was always there; peak earnings just hid it.

Before buying any cyclical stock, examine debt-to-normalized-EBITDA and interest coverage at various cycle points. If debt looks manageable only at peak earnings, the company is too leveraged for a cyclical business.

Real-World Examples​

Energy Sector (2010–2016). Oil majors traded at 8–10x P/E ratios in 2013-2014, with oil above $100/barrel. Investors saw "cheap" oil companies and bought. But this was peak-cycle earnings. Oil supply was expanding, and demand growth was slowing. By 2016, oil had crashed to $35/barrel. Companies that reported $5 earnings per share at $100 oil dropped to $1 earnings at $35 oil. Stocks fell 50-70%. The "cheap" 10x multiple suddenly looked like 50x. Companies with debt saw major covenant concerns. The cycle peak trap was deadly.

Steel Industry (2007–2008). Steel stocks traded at 5-6x earnings in early 2008, seemingly cheap. But this was peak iron ore prices, peak global demand, and maximum margin expansion. When financial crisis hit, demand collapsed 30-40% within months. Earnings fell 80-90%. Stocks that seemed cheap at 6x earnings fell 70% as earnings crashed. Companies had borrowed heavily in 2005-2007 at peak profitability to expand, then faced debt troubles when earnings compressed.

Commercial Real Estate (2019–2023). Commercial real estate companies traded at attractive valuations in 2019-2021 with strong NOI growth. But this was at the peak of office occupancy and the height of commercial real estate expansion. As remote work accelerated, occupancy rates fell, and new supply came online, NOI growth reversed to decline. Companies that seemed fairly valued on peak NOI dropped 40-60% as the cycle turned. Leverage in the sector amplified losses.

Semiconductors (2021–2022). Semiconductor stocks roared higher in 2020-2021 on peak demand for chips, leading to 5x P/E ratios despite historically elevated earnings. But this was peak-cycle capacity utilization. Massive new fabrication plants were under construction globally. By 2023, oversupply had arrived and margins were compressing. Stocks that seemed cheap at peak margins faced significant headwinds as normalized earnings fell.

Automotive (1995–2008). Auto stocks traded cheaply (8-10x earnings) in 2005-2007, but this was peak auto production and peak leverage in the industry. Companies had borrowed heavily, thinking the good times would continue. When financial crisis hit and auto sales collapsed, leverage nearly destroyed the entire industry. GM and Chrysler required government bailouts. Investors who bought on "cheap" valuations at peak earnings faced total wipeouts.

Common Mistakes​

  1. Using peak-cycle earnings for valuation without adjusting for the cycle. Always normalize earnings. If a company has never earned more than $6 per share in any recent full cycle, don't value it on $10 earnings just because that's the current number.

  2. Ignoring industry capacity announcements and planned supply additions. If the industry is adding 20% capacity in the next 18 months, margins will compress. Don't ignore forward capacity data just because current capacity is tight.

  3. Assuming leverage is manageable at peak earnings. Calculate debt ratios at normalized and trough earnings, not peak earnings. If debt looks dangerous at normalized levels, the company is too leveraged.

  4. Buying cyclical stocks on momentum at the end of expansions. The best time to buy cyclicals is early in the recovery, when despair is high and earnings are rebounding. The worst time is late in the expansion when earnings are peaking and momentum is strongest.

  5. Treating all earnings volatility as cyclical. Some earnings volatility is one-time, some is structural, some is cyclical. Distinguish before buying. A company with lumpy one-time gains in a growth industry isn't cyclical.

  6. Neglecting mean reversion in returns on equity. When a cyclical company posts 25% ROE at peak cycle, assume it will normalize to 10-12% in a normal cycle. Don't extrapolate peak returns into perpetuity.

FAQ​

Q: How do I estimate normalized earnings for a cyclical company? A: Look at the company's historical earnings through multiple cycles (ideally 2-3 full cycles). Identify peak, trough, and mid-point earnings. The mid-point is a reasonable normalized estimate. Alternatively, use analyst estimates for "normalized" earnings, though these are often optimistic. Compare to industry margins at normalized capacity utilization.

Q: What's the best time to buy a cyclical stock? A: Early in the recovery phase when earnings are rising but multiples are still depressed because the cycle is still recent in investors' memories. Buy when pessimism is high. Avoid buying at the end of expansions when momentum is strong but cycle maturity is evident.

Q: How much should I discount a stock based on cycle position? A: There's no formula, but a stock that appears 8x normalized earnings but is at peak cycle earnings might be worth only 80-90% as much as it appears. Conservative investors apply a 20-30% discount to peak-cycle valuations relative to normalized valuations.

Q: Should I ever buy a cyclical stock at peak earnings? A: Only if: (1) you have high conviction the cycle will extend longer than market expects, backed by specific evidence (new demand drivers, constrained supply beyond the plan); (2) the company has minimal debt; or (3) you're buying for a multi-year hold expecting multiple cycles. Otherwise, wait for weakness.

Q: How do I know when a cycle is peaking? A: Watch for: (1) full capacity utilization across the industry; (2) announcements of major new capacity additions; (3) peak margins in company and peer results; (4) maximum leverage in company balance sheets; (5) extreme optimism in analyst estimates. When you see all five, the peak is usually 3-12 months away.

Q: Can I use ratios like EV/EBITDA to avoid the cyclical trap? A: EV/EBITDA helps less than you'd think because EBITDA is also at peak levels. Instead, use EV/(normalized or mid-cycle EBITDA). Better: use EV/trough-EBITDA as a worst-case reality check. If the company doesn't survive a trough scenario, it's too risky.

  • Chapter 9: Discounted Cash Flow Analysis — DCF for cyclicals requires normalized earnings as a starting point, not peak earnings.
  • Chapter 10: Earnings Quality and Cash Flow — Understanding the durability of earnings is essential in cyclical contexts.
  • Chapter 12: Relative Valuation Metrics — Multiple compression and expansion in cyclicals amplify valuation mistakes.
  • Chapter 14: Debt and Financial Leverage — Leverage in cyclical industries is particularly dangerous.

Summary​

Cyclical stocks are cheapest by P/E ratio at peak earnings, which is precisely when they are most expensive on a normalized basis. The correct approach is estimating normalized earnings across the full business cycle, not valuing on current peak earnings. Capacity additions coming online often mark cycle peaks, making late-expansion purchases especially dangerous. Leverage amplifies cyclical damage; high debt combined with peak earnings creates the potential for covenant violations and shareholder dilution when the cycle turns. The best cyclical investments are made early in recovery, not late in expansion. Understanding where an industry stands in its cycle is as important as understanding the company itself when investing in cyclical businesses.

Next​

Over-optimistic DCF Assumptions