Why P/E lies on capital-intensive cyclicals — running the EV/EBITDA comp Quorum's Council uses

P/E is a capital-structure-loaded ratio. If you compare a debt-heavy industrial like Carrier (CARR) to a debt-light peer, the difference in P/E is half "earnings yield" and half "leverage premium" — and you can't tell which without normalizing.

Why P/E breaks on capital-intensive comps

Two firms with identical operating economics, different capital structures:

• Firm A: $100 EBIT, $0 interest, 25% tax → $75 net income. 0 debt. EV = $1,500. P/E at fair value ≈ 20×.
• Firm B: $100 EBIT, $30 interest, 25% tax → $52.5 net income. $500 debt at 6%. EV = $1,500. P/E at the same EV looks like 19× — but only after you back-solve the equity value, which now reflects leverage.

Same EV. Same operating performance. Different P/E. P/E doesn't separate operating from financial — EV/EBITDA does.

The EV/EBITDA build, lease-adjusted

For a capital-intensive industrial post-ASC 842, operating leases sit on the balance sheet as right-of-use assets + lease liabilities. The textbook EV/EBITDA calc that ignores leases understates leverage. The right build:

• Enterprise Value = Market Cap + Total Debt + Operating Lease Liabilities + Pension Underfunding − Cash & Equivalents − Cross-holdings at fair value.
• Adjusted EBITDA = Reported EBITDA + Stock-Based Comp (if you treat SBC as cash, which the Council does for serial dilutors) − Recurring Restructuring (because if it recurs every year it's not "one-time"). Optionally back out lease-interest if you're adjusting EV for leases (don't double-count).

Where the Council's seats actually look

Two seats, two different lenses on the same comp table:

• The moat seat reads margin durability. For CARR: the Viessmann acquisition + AI-data-center cooling exposure could re-rate gross margin from ~27% toward 30%+, which on $25B of revenue is a $750M EBITDA delta — that's where the bull case lives. Named falsifier: "if Viessmann segment gross margin stays below 26% for two consecutive quarters, the moat thesis fails."
• The capital-allocation seat reads reinvestment quality. CARR's flagged concern: ROIC below WACC on the marginal capital deployed. Named falsifier: "if trailing 4Q ROIC clears WACC by >200 bps, the cap-allocation flag drops." Until then, the seat downgrades the bull case even when EV/EBITDA screens cheap.

The verdict shape that survives

Council ran CARR with a 4-1 mode: an Upside Signal with a load-bearing dissent on cap-allocation. The bell-curve shifts visibly left from the unanimous case; the chip stays positive but the consensus width narrows. That's exactly the signal-shape we want — the dissent isn't laundered away, it's surfaced under the chip with the named falsifier.

What this costs to run

Comp screening on EV/EBITDA runs on Quorum Lite — free, 20/day. The comp table, the lease-adjusted multiple, the side-by-side — all rendered on the free tier. Quorum Pro ($39/mo, $29/mo annual) adds the cloud-syndicate council and a 15-day shared-verdict cache so a recent complete run for one user can serve other users until it needs refresh. Quorum Pro+ ($99/mo, $79/mo annual) adds a frontier-model briefing capped at 10/day for the contested calls. For a comp screen, Pro is enough. For the contested CARR call where the cap-allocation dissent is the swing factor — that's a Pro+ ticket.

See the CARR Council verdict →   Tier comparison →