Visit Website →
← Back to blog

Analysts: Pair Cycle Adjusted Comps With Disclosure Scores

September 28, 2026
Analysts: Pair Cycle Adjusted Comps With Disclosure Scores

Cycle-adjusted peer comparisons replace point-in-time earnings with normalized, normalized denominators so that peer multiples better reflect sustainable earnings power rather than where a company sits in its cycle. The technique exists because trailing and forward multiples on cyclical businesses invert at exactly the moments investors most need clarity. The first action for any analyst applying this framework is procedural: compute a normalized or mid-cycle denominator for the subject company, then apply the identical normalization method to every peer in the comparison set, without exception.

TL;DR:

  • Cycle-adjusted multiples better reflect sustainable earnings by using normalized denominators over full cycle windows, avoiding distortions caused by cyclical peaks and troughs.
  • Normalization methods include historical averaging, margin regression, and capacity or utilization adjustments, which account for structural change and asset intensity.
  • Proper peer selection requires matching business models, end-market exposure, growth profiles, and leverage to prevent biased comparisons, with consistent normalization applied across all peers.
  • Presenting multiple valuation metrics side-by-side, such as trailing, mid-cycle, and forward multiples, enhances transparency and helps distinguish genuine value from cyclical timing effects.
  • Disclosure quality impacts the reliability of normalized multiples, necessitating scrutiny of adjusted earnings figures and weighting peers with strong, transparent reporting.
Lacunaindex
Test Valuation Against Actual Delivery
Lacuna Index evaluates public company claims against delivery using forensic reports, sector benchmarks, and execution scores.
Explore Lacuna Index

Table of Contents

Why point-in-time multiples mislead in cyclical industries

A cyclical company trading at its earnings peak often looks cheap on a trailing P/E, because the denominator, current-year earnings, is unusually high relative to what the business can sustain across a full cycle. The same company trading near a trough looks expensive on the identical metric, because earnings have collapsed while the share price has only partially adjusted. An analyst comparing two such companies at different points in their respective cycles, using only trailing multiples, will draw a conclusion that has nothing to do with relative value and everything to do with timing.

Peak and trough earnings distort valuation multiples

This is the core argument behind normalized earnings, as laid out in Damodaran's discussion of normalized earnings: for cyclical and commodity companies, standard point-in-time multiples are distorted by temporary earnings volatility, and averaging across a full economic cycle, commonly a window of five to ten years, produces a denominator that better represents what the business can actually deliver over time.

Three principles follow from this logic:

  • Normalized earnings smooth out the effect of commodity price swings, capacity utilization changes, and demand shocks that have no bearing on long-run unit economics.
  • A mid-cycle multiple is comparable across companies regardless of where each currently sits in its own cycle, provided the normalization method is applied consistently.
  • Moving up the income statement, from net income toward EBITDA or revenue, further reduces the volatility introduced by leverage, depreciation policy, and tax positioning. This point is echoed in Damodaran's chapter on valuing cyclical and commodity companies.

Practical normalization methods analysts can apply

Three techniques dominate practitioner workflows, and the choice between them depends on how much secular change has occurred inside the observation window.

  1. Historical averaging. Take a defensible window spanning a full peak-to-trough cycle and average earnings, margins, or EBITDA across it. Using a window that fully covers a complete cycle is important to avoid partial-cycle distortions.
  2. Margin regression. When a sector has undergone structural change (a shift in cost structure, a permanent change in competitive intensity, or a step change in input prices), regress margins against a cycle indicator such as capacity utilization or a commodity price index, then read off the margin implied by mid-cycle conditions. This method is more defensible than simple averaging when the business has genuinely changed shape rather than merely moved along a repeating cycle.
  3. Capacity or utilization adjustment. For asset-heavy cyclicals, industry utilization rates or operating-rate data can be used to scale reported margins up or down to what they would be at a normalized utilization level, rather than relying purely on historical financials.

Handling inflation and one-off items matters as much as choosing the window. Nominal historical earnings should be deflated to a common price base before averaging, and any year containing a clearly identifiable anomaly, a strike, a natural disaster, an asset impairment unrelated to the cycle, should be flagged and either excluded or footnoted rather than silently blended into the average.

Pro Tip: Record the normalization method and details of excluded periods to ensure transparency and reproducibility, in a footnote attached to the comp table itself, not in a separate memo that readers will never see.

Documentation discipline is not a formality. A normalized multiple that cannot be reproduced by another analyst from the same public filings is not a defensible input to a valuation call, and the guide to comparable company analysis for cyclical businesses treats this reproducibility as a precondition for a valid comp set, not an optional refinement.

Building a peer set that survives cycle scrutiny

Peer selection for cycle-adjusted comparisons is a filtering exercise before it is a calculation exercise. A comp set assembled purely on sector classification, with no further screening, will mix companies that have nothing in common except a shared industry code, and that mismatch will corrupt the normalized multiples no matter how carefully each individual company is normalized.

A working peer screen typically covers:

  • Business model similarity: distribution versus manufacturing, contract versus spot pricing, and integrated versus pure-play exposure change what "normal" earnings look like.
  • End-market exposure: two companies in the same industry code can face entirely different demand drivers if one sells to residential construction and the other to industrial capital projects.
  • Growth profile: a peer with structurally higher growth is not comparable at the same multiple even after cycle normalization, since growth deserves its own premium.
  • Capital intensity and leverage: differences in asset intensity and balance sheet structure change free cash flow conversion and should be flagged, segmented, or adjusted for rather than ignored.
  • Secular versus cyclical mix: a company mid-transition away from a declining product line will show earnings patterns that look cyclical but are actually secular, and that mismatch will bias a mid-cycle estimate. This screening logic follows the peer-selection framework described in guidance on choosing comparable companies for valuation, which recommends assigning peers to core, secondary, or context-only roles rather than treating the set as homogeneous.

The single most consequential error in this stage is asymmetric normalization: adjusting the subject company's earnings to a mid-cycle basis while leaving peer earnings on a trailing or as-reported basis. This produces an implied multiple gap that has nothing to do with relative valuation and everything to do with which company got the adjustment. As the analysis of valuing cyclical and commodity companies notes, a mathematically valid cycle-adjusted comparison requires identical normalization logic applied to every company in the peer set, without exception, or the resulting multiples carry a built-in bias.

Calculating and presenting cycle-adjusted multiples

Once normalized denominators exist for the subject company and every peer, the calculation proceeds in a fixed sequence.

  • Compute the normalized denominator for each company (mid-cycle EBITDA, normalized net income, or a normalized revenue base, depending on the chosen multiple).
  • Divide each peer's enterprise value, or equity value for earnings-based multiples, by its normalized denominator to produce the cycle-adjusted multiple.
  • Aggregate the peer multiples into percentile bands (25th, median, 75th) rather than a single average, since a point estimate implies a precision the peer set does not actually support, a practice recommended in the comparable company analysis guide.
  • Apply the resulting multiple to the subject company's own normalized denominator to derive an implied enterprise value, then bridge to equity value by subtracting net debt and dividing by fully diluted shares outstanding to reach an implied per-share figure.

Multiple choice should follow the business, not habit. EV/EBITDA is generally preferred for cyclicals with variable leverage and depreciation policy, since it sits above those distortions on the income statement. EV/Revenue serves as a floor check for companies with currently negative or depressed normalized earnings, where an earnings multiple cannot be computed at all. Normalized P/E remains usable but carries the caveat that it embeds capital structure and tax effects that EV-based multiples avoid, so it works best as a cross-check rather than the primary metric. Presenting trailing, next-twelve-months, and mid-cycle multiples side by side, with each column clearly labeled, lets a reader see immediately whether a peer looks cheap because of genuine value or because it happens to sit near a cycle trough, a presentation discipline emphasized in the analysis of comps for cyclical stocks.

Footnotes should disclose the normalization window used, the data vintage for each peer's financials, and the sensitivity range produced when the window or growth assumption is varied, so that a reader can judge how much of the implied value depends on modeling choices rather than observed fundamentals.

Validation checks and the pitfalls that undermine the analysis

A cycle-adjusted comp set can look rigorous and still be wrong in ways that only surface under a checklist review.

  1. Data vintage consistency. Every peer's financials should be pulled as of a comparable reporting date, and the LTM versus NTM policy should be applied uniformly. Mixing a peer's trailing twelve months with another peer's forward estimate produces a spread that reflects reporting timing, not fundamentals, a pitfall specifically flagged in the Pomegra comps analysis.
  2. One-time item treatment. Restructuring charges, impairments, and litigation settlements should be stripped using the same criteria across every peer, and the specific adjustments made for each company should be documented rather than absorbed silently into an adjusted EBITDA figure supplied by management.
  3. Leverage sensitivity. Since EV-based multiples partially neutralize leverage differences but do not eliminate them entirely, segmenting peers into leverage bands, or explicitly testing how the implied multiple shifts under alternative capital structures, catches distortions that a single blended peer average would hide.
  4. Scenario testing. Rerunning the normalization with a shorter window, a longer window, and an adjusted secular growth assumption shows how sensitive the implied valuation is to modeling choices, and that sensitivity range belongs in the final output, not just in a working file.

Pro Tip: Treat any comp set where trailing and mid-cycle multiples move in opposite directions for more than half the peers as a signal to re-examine the cycle-position assumptions before trusting the output.

Reading disclosure quality into the peer comparison

Numeric normalization corrects for cycle timing, but it cannot correct for a peer's reported adjustments being unreliable in the first place. Non-GAAP add-backs, recurring items labeled as one-time, and selective segment carve-outs can distort the same "adjusted EBITDA" figure that a normalization model treats as raw input, which means disclosure quality has to be assessed before the numbers are trusted, not after.

A cycle-adjusted multiple built on an unreliable adjusted-earnings figure inherits that unreliability no matter how carefully the cycle math is done.

This is the layer that Lacuna Index's disclosure-quality benchmarking approach is built to address, by scoring how a company's disclosed adjustments compare against its underlying public filings. Folding a disclosure-quality score into peer weighting, down-weighting or excluding companies whose adjusted figures show a wide aspiration-to-execution gap, can keep a cycle-adjusted comp set from mistaking engineered numbers for normalized ones.

When cycle-adjusted comps actually change the call

When cycle-adjusted comps actually change the call — overview diagram

The method matters most at inflection points. In commodity cycles, a trailing multiple that looks expensive near a trough is often the signal that a stock is undervalued on a mid-cycle basis, while the same trailing multiple near a peak looks deceptively cheap. In capital goods, order backlogs create a lag between demand turns and reported earnings, so a normalized multiple built on backlog-adjusted revenue often flips a "hold" into a "buy" well before trailing figures catch up.

Two rules hold up across both cases: always show trailing and mid-cycle multiples side by side rather than picking one, and document every normalization choice as if another analyst will need to reproduce it without asking a question.

*— Glen

How Lacuna Index supports this workflow

Cycle-adjusted peer work depends on reliable inputs as much as correct math, and that is where Lacuna Index's sector benchmarks fit into the process: they provide public, disclosure-scored baselines for sectors including financials, giving analysts a starting reference before they build a normalized peer set of their own. For a subject company or a specific peer whose adjusted figures deserve closer scrutiny, forensic analysis can quantify the gap between what management claims and what the filings actually support. Analysts running this work at scale can review Lacuna's Cohort Pulse and Sector Sweep plans for ongoing cohort-level access, or start with a single per-company report on the name currently under review.

Sources

FAQ

What makes cycle-adjusted peer comparisons different from standard comps?

Standard comps use trailing or forward multiples as reported, which can be distorted by where each company sits in its earnings cycle. Cycle-adjusted comps replace those raw denominators with normalized, mid-cycle figures, typically averaged over a five-to-ten-year window, so the comparison reflects sustainable earnings power rather than timing.

How long should the normalization window be?

Practitioners commonly use a normalization period that spans a full peak-to-trough cycle appropriate to the industry's rhythm for the sector in question. The Damodaran framework for cyclical and commodity companies treats this window length as a starting point, to be adjusted for the specific industry's cycle length.

Which multiple works best for cyclical companies?

EV/EBITDA is generally preferred because it sits above leverage and depreciation distortions, while EV/Revenue serves as a backstop when normalized earnings are depressed or negative. Analysts commonly present both alongside a normalized P/E as a cross-check, following the approach outlined for comparable company analysis in cyclical industrials.

What is the biggest mistake analysts make when normalizing peer comps?

The most common error is normalizing the subject company's earnings while leaving peer earnings on a trailing or as-reported basis. This asymmetry introduces a mathematical bias into the implied multiple that has nothing to do with relative valuation, a risk specifically addressed in the analysis of cyclical and commodity company valuation.

How does disclosure quality affect cycle-adjusted comps?

Even a correctly normalized multiple inherits any unreliability in the underlying adjusted-earnings figures a company reports, since non-GAAP add-backs and selective carve-outs can distort the raw inputs before normalization even begins. Weighting or excluding peers based on disclosure-quality scoring, such as the approach described in Lacuna Index's disclosure-quality benchmarking, helps keep the comp set from treating engineered figures as normalized ones.