Valuing Cyclical Companies

A miner trading at 4x earnings appears absurdly cheap. A steel company at 3x earnings screams opportunity. But cyclical businesses are traps for the unwary: they look cheapest at peak cycles and most expensive at troughs. Buffett famously said, "Investors should be wary of businesses that split their attention between multiple goals and whose earnings change with economic cycles." Valuing a cyclical business requires understanding where you are in the cycle, estimating normalized earnings, and asking whether the cycle will turn in your favor.

Quick definition: Cyclical companies have earnings that fluctuate dramatically with economic conditions or commodity prices. Normalized cyclical earnings are the average of earnings across a full economic or commodity cycle, used to determine intrinsic value independent of current cycle position.

Key Takeaways

• Peak cycle earnings look cheap but are unsustainable; trough earnings are expensive but temporary
• Normalized earnings for cyclical businesses are best estimated by averaging a full cycle
• Cycle timing is nearly impossible to predict; value investors rely on mean reversion
• Strong balance sheets matter more for cyclical companies because downturns are severe
• Capital discipline matters: cyclical companies that reinvest at peaks destroy value
• True opportunities in cyclicals come when the cycle turns and few see it

The Cycle Problem: Why Peak Earnings Are Deceptive

A steel company at peak cycle earns $100M; trades at $800M market cap = 8x earnings (cheap!). But at trough, it earns $20M; market cap of $200M = 10x earnings (expensive!).

This counterintuitive outcome occurs because earnings fall faster than market caps during downturns. Investors gradually recognize that peak earnings are unsustainable and mark down valuations in advance of actual trough earnings.

The Peak-to-Trough Sequence:

1. Boom: Earnings surge to $100M; stock rises to $800M (8x, cheap on peak earnings)
2. Recognition: Investors realize peak is unsustainable; stock falls to $600M (6x, looks cheap)
3. Early Trough: Earnings fall to $40M; stock at $400M (10x current earnings, but…)
4. Trough: Earnings bottom at $20M; stock stabilizes at $200M (10x trough earnings)
5. Recovery: Earnings rise to $60M; stock at $450M (7.5x, finally a value opportunity?)

The valuation multiple is highest at the trough, when earnings are lowest. A value investor who buys at "8x earnings" at the peak has paid the most money for the least sustainable earnings.

How to Normalize Cyclical Earnings

Method 1: Historical Cycle Averaging

Identify a full cycle (peak to trough to recovery to peak) and average.

Example: Oil & Gas Services Company

Year	Earnings	Phase
2006	$500M	Boom
2007	$600M	Peak
2008	$200M	Crash
2009	$100M	Trough
2010	$250M	Recovery
2011	$400M	Boom
2012	$450M	Peak
2013	$150M	Crash
2014	$80M	Trough
2015	$200M	Recovery

Average: (~$500 + $600 + $200 + $100 + $250 + $400 + $450 + $150 + $80 + $200) / 10 = $293M

This normalized earnings figure is the starting point for valuation, not peak ($600M) or trough ($80M).

Method 2: Structural Earnings Power

Estimate the earnings a company can sustain indefinitely, independent of cycle position.

A mining company with high fixed costs and commodity-price-dependent revenues might have:

• Production capacity: 1M units annually
• Commodity price at normalized levels: $50 per unit
• Variable cost: $30 per unit
• Fixed cost: $5M annually

Normalized earnings = (1M × $50) - (1M × $30) - $5M = $15M

This structural approach works when you can decompose the business into units produced × price per unit minus costs. For commodity businesses, this is more reliable than averaging historical earnings if the cycle is changing (e.g., new production capacity, new competing sources).

Method 3: Peer Comparison at Normalized Cycles

Look at how peers' earnings have behaved and estimate where you are in the cycle.

If three shipbuilding companies all report 50% lower earnings year-over-year, and all have $1B+ order backlogs, you're likely at a trough. Average their peak-year earnings as a proxy for normalized earnings.

Determining Cycle Position

Ask: Where am I in the cycle?

Indicators:

• Capacity utilization: If all mills are operating at 80%+ capacity, near peak
• Industry pricing power: If companies are cutting prices to fill capacity, likely in trough/early recovery
• Management commentary: "Utilization improving," "Demand stabilizing," "We're optimistic about 2025" → recovery
• Order books: Growing backlogs suggest cycle strength; shrinking backlogs suggest weakness
• Capital spending: Heavy capex spending → companies expect sustained demand (late cycle warning)
• Peer stock prices: If entire sector is at multi-year lows, you're likely near trough

Valuation Framework for Cyclicals

Step 1: Estimate normalized earnings across the full cycle (method above).

Step 2: Apply an appropriate multiple.

Cyclical businesses deserve lower multiples than non-cyclicals because earnings are uncertain:

• Non-cyclical stable company: 15–20x normalized earnings
• Cyclical company: 8–12x normalized earnings (discount for uncertainty)

Step 3: Identify cycle position and margin of safety.

If normalized earnings are $100M and you use 10x multiple = $1,000M fair value:

• At trough ($20M earnings, -80% from normalized): Stock at $200M is CHEAP
• At peak ($100M earnings, 0% from normalized): Stock at $800M is FAIR
• 20% above peak ($120M earnings, +20% from normalized): Stock at $1,200M is EXPENSIVE

Step 4: Only buy if the cycle has likely turned or will turn soon.

Buying a cyclical at trough with improving indicators (backlogs, utilization) is a value play. Buying at peak because it "looks cheap" on peak earnings is a trap.

The Capital Discipline Problem

The Curse of Peak Cycle Capital Spending

At peak cycle earnings, companies feel rich and invest heavily—exactly when they should be conserving. A steel company earning $100M at peak capacity might spend $50M on new mills, betting on sustained demand. The cycle turns; demand collapses; new mills operate at 30% utilization.

Value investors in cyclicals look for management that:

• Cuts capex during booms
• Returns capital to shareholders when earnings peak
• Preserves balance sheet strength for downturns
• Invests heavily during downturns (when iron is cheap and distressed assets are available)

This discipline is rare. Most cyclical managers reinvest at peaks and cut at troughs—destroying value.

Real-World Examples

Steel (2010–2015)

Iron ore prices collapsed from $150/ton to $40/ton. Steel companies' earnings fell 80%. A company earning $500M in 2010 earned $100M in 2015.

By reported 2015 earnings, the company looked expensive at any price. But by normalized earnings (averaging 2010–2015 cycle), the company was undervalued. After 2016 recovery, earnings rebounded to $400M. Investors who bought in 2015 on normalized earnings made 4–5x returns.

Banking (2008)

Banks at peak 2007 earned $50B collectively; traded at $3T market cap (60x earnings!). In 2009 at trough, banks earned $5B (losses); traded at $800B (but on trough earnings, "expensive").

Normalized earnings (averaging 2005–2009 cycle) were ~$20B. Fair value was ~$1.2T. The $800B valuation in 2009 was deeply undervalued. Those who invested in 2009–2010 tripled their money by 2012.

Airlines (2019–2022)

Pre-COVID normalized earnings: ~$15B across US carriers. COVID collapse: earnings turned negative. Stock prices fell 70%.

Recovery in 2021–2022 brought earnings back toward $10B (slightly below normalized, due to increased fuel costs). But pilots bargained for 40% wage increases, cutting future earnings. Normalized earnings fell to ~$8B. Airlines trading at 12x 2022 earnings looked cheap, but on normalized earnings, they were fair at best.

When NOT to Buy Cyclicals

Avoid cyclicals when:

1. The cycle has structural headwinds. Airlines have secular margin pressure from labor, fuel, and competition. Their cycle hasn't changed fundamentally; they're just less profitable at each peak. Avoid unless buying at genuine distress with a catalyst.

2. Balance sheet can't survive the trough. High leverage means bankruptcy risk in downturns. A mining company at trough earning $50M but carrying $500M debt might not survive the next downturn. Avoid unless leverage is being paid down.

3. Peak cycle is arriving, not leaving. If backlogs are surging and capacity is fully utilized, the cycle is peaking. This is when cyclicals "look cheap." Avoid.

4. Management has a history of poor capital allocation. If they've overinvested at peaks and cut at troughs repeatedly, they'll do it again. Avoid unless there's new management.

5. You have no idea where you are in the cycle. Better to skip than to guess wrong. Uncertainty about cycle position is a legitimate reason to avoid.

The Catalyst for Cyclicals

Cyclicals need a catalyst: A turning point where the cycle shifts, few investors see it, and your contrarian view is about to be validated.

• Steel: New mine comes online, crushing prices to $40/ton; capacity utilization falls below 60%
• Banks: Interest rates rise sharply, blowing out credit spreads; equity prices fall 40%
• Oil: OPEC cuts production; supply tightens; prices rebound from $40 to $70 per barrel
• Airlines: Fuel prices crash; labor costs stabilize; margins expand

Without visibility to a turning cycle, cyclical investing is speculation, not value investing.

Common Mistakes

Using peak earnings for valuation. A company is never "cheap" on peak cycle earnings. Cheap comes at trough.

Ignoring the cycle entirely. Assuming a cyclical company's earnings will grow steadily destroys capital. Always model cycle reversion.

Buying on balance sheet strength alone. A miner with $500M cash and $200M debt might seem safe, but if earnings are negative for three years, cash depletes fast.

Underestimating cycle severity. Your normalized earnings estimate might be too high. Industries are more cyclical than you think.

Perfect timing. You can't call the exact bottom. Buy with patience and scale in over quarters, not all-in at the first sign of recovery.

FAQ

Q: How long is a typical cycle? A: 3–5 years for most industries. Mining and oil can have 7–10 year cycles. Financial cycles: 4–7 years. There's no fixed rule.

Q: Should I avoid cyclicals entirely? A: Not if you can identify the cycle phase and have a margin of safety. Cyclicals at troughs with strong balance sheets and improving indicators are among the best value opportunities.

Q: How do I know if an industry is becoming permanently impaired? A: Watch the peak cycle earnings. If peaks are declining year-over-year (peak 2010 $500M, peak 2015 $400M, peak 2020 $300M), the industry is in structural decline. Avoid.

Q: What leverage level is safe for cyclicals? A: Debt/EBITDA should be below 2x at normalized earnings levels. At trough, it might reach 4–5x; if it exceeds that, bankruptcy risk rises sharply.

Q: Is normalized earnings the same as free cash flow? A: No. Normalized earnings exclude taxes and capex needs. Normalized free cash flow is more reliable: normalized earnings - taxes - maintenance capex.

Related Concepts

• Economic Cycle: The macroeconomic phases that drive cyclical earnings
• Commodity Prices: The driver of earnings for mining, oil, shipping, agriculture
• Capacity Utilization: An indicator of where you are in an industrial cycle
• Balance Sheet Strength: Critical during downturns to avoid bankruptcy
• Mean Reversion: The tendency of cycles to return to historical averages
• Capital Discipline: Management's ability to restrain spending at peak and invest at trough

Summary

Cyclical companies offer tremendous opportunities at troughs if you can identify where you are in the cycle and normalize earnings properly. The trap is buying "cheap" at peaks, when earnings are at their highest. Value investors in cyclicals succeed by averaging earnings across the full cycle, identifying when the cycle has likely turned or is about to turn, and buying only when the balance sheet is strong enough to survive the next downturn. The most important skill is resisting the siren song of "cheap" peak-cycle multiples and waiting for the trough where earnings are lowest, sentiment is terrible, and the margin of safety is genuine.

Next

Proceed to the next article: Adjusting for Stock Options and Dilution.