RB Global [RBA]
Spot $110.76 · 187.5m diluted shares · market cap $20,635m · net debt $4,024m (lease-INCLUSIVE, preferred-INCLUSIVE) · EV $24,659m · TTM revenue $4,736m · EV/Sales 5.21x · EV/EBIT 28.4x
EV composition, stated explicitly because the screen's was wrong by 10.6%:
Market cap (187.5m x $110.76) 20,634.6
+ Short-term debt 282.0
+ Current portion of long-term debt 51.2
+ Long-term debt 2,271.0
+ Operating lease liabilities (132.8 + 1,472.3) 1,605.1 <- EV is LEASE-INCLUSIVE
+ Series A Senior Preferred (temporary equity) 482.0 <- omitted by the screen entirely
- Cash and cash equivalents (667.2) <- EXCLUDES $192.2m restricted cash
= Enterprise value 24,658.7
Test applied. State A requires no structural regime change; RBA has had two — the IAA acquisition (20 March 2023, which doubled revenue in two quarters) and the SYNETIQ deconsolidation (21 June 2025) — plus two bolt-ons inside the last twelve months. State A is unavailable.
State B requires revenue or margin driven by an exogenous cycle. Evidence, all from the company: - Insurance-salvage volume depends on catastrophic weather. FY2025 automotive GTV growth was reduced by "the non-recurrence of significant catastrophic events in the prior year", and Q1'26 automotive was reduced by "the non-repeat of prior year catastrophic events in the United States." - Used-equipment supply is in a downcycle. CC&T lots sold −13% in FY2025, Other −8%. - The 10-K's own forward-looking-statements list names "the supply trend of equipment and vehicles in the market and the anticipated price environment, as well as the resulting effect on ... GTV" and "the behavior of commercial assets and vehicle pricing" as the primary uncertainties.
State B is declared, and its consequence is binding: the terminal margin must be MID-CYCLE and normalised on BOTH revenue and margin, and the cycle position must be DATED.
Cycle position, dated: RBA is at a CC&T trough and an automotive plateau. CC&T lots have fallen 13% and Other 8% from FY2024; automotive lots grew 7% but average price per lot fell "primarily due to a shift in customer mix, with a greater proportion of remarketed vehicles relative to those from insurance providers." Q1'26 shows the first CC&T recovery (+12% lots) but on a base that includes two acquisitions. The forecast therefore begins at or just past a CC&T trough, which is the flattering starting point for a five-year growth extrapolation and must be said out loud.
Evidence grade: C. Three things hold it here: customer concentration is undisclosed while the MD&A attributes three separate line movements to unnamed "significant customer contracts" (§4c of the research); the organic/inorganic split of Q1'26's +13% GTV is not quantified by the company; and the IAA regime break truncates the usable history to three years.
The screen carried 15.5% as "own trailing operating margin". The trailing margin is 18.3%, so the stored value sat 2.8pp below the company's own actual — putting RBA into the population of 21 of 84 names the terminal-margin audit found doing exactly this.
| period | operating margin |
|---|---|
| FY2023 (part-year IAA) | ~11.6% |
| FY2024 | 17.9% |
| FY2025 | ~18.2% (filed basis) |
| TTM to 2026-03-31 | 18.3% |
| Q1'26 | 17.6% (filed: $217.5m on $1,234.6m) |
Bridge and the two offsetting forces:
| direction | size | |
|---|---|---|
| CC&T volume recovery from trough toward mid-cycle | + | CC&T is the higher-take-rate sector; a return to FY2023 lot volumes adds roughly +0.8pp |
| Inventory (principal) mix dilution | − | Inventory revenue is growing at 32% against service at 5%. Principal revenue books gross GTV with a 5.3% inventory rate, so every point of mix shift into it dilutes the consolidated EBIT margin mechanically. Roughly −1.1pp at the current rate of shift |
| Terminal | 18.0% |
Hard constraint m_EBIT,T ≤ m_gross,T: 18.0% ≤ 45.6%. SATISFIED.
Note what the screen's gross margin would have done. The screen recorded a 16.7% gross margin against a 15.5% operating margin — 1.2pp of headroom for every operating expense the company has. Had a terminal margin of 18.0% been set against that, the model would have reported an arithmetic violation on a business that has no such problem. The real gross margin is 45.6%. A 28.9pp data error nearly manufactured a hard-constraint failure.
steady_state_check.py — what it flaggedpython3 steady_state_check.py --ticker RBA --terminal-margin 0.187 --exit-multiple 24.4
| margin ramp 2020→2025 | 19.1% → 17.7%, −1.4pp; latest is window peak: False |
| terminal assumed | 18.7% (+1.0pp vs latest, −7.5pp vs peak) |
| ROIC measured | 6.2% |
| warranted multiple at a 6.2% steady-state ROIC | 5.8x |
| terminal reinvestment rate implied by g/ROIC | 48.4% |
| exit multiple used vs warranted | 4.19x |
| FINDING | EXIT_MULTIPLE_UNWARRANTED_AT_STEADY_STATE — 24.4x is 4.2x the 5.8x warranted at a 6% steady-state ROIC. |
This is the largest warranted-multiple gap of the three names by a wide margin, and the ROIC is the reason. RBA earns 6.2% on invested capital because IAA left $4,662.3m of goodwill and $2,388.0m of intangibles on a $12,397.2m balance sheet — 57% of assets are acquisition accounting. A shareholder paid for that capital, so it is invested capital, and the identity is right to charge for it. At a 6.2% ROIC and 9% WACC, RBA destroys value on incremental invested capital, which is why the identity produces 5.8x.
The tool did not flag MARGIN_STILL_RAMPING — correct, the margin has fallen 1.4pp over the window
— and did not flag ROIC_IS_SCALING_PHASE, also correct: 6.2% is not a scaling-phase return, it is a
low one.
I act on the flag by setting the base exit multiple at 15.0x, below RBA's own three-year p25 of 24.53x and above the identity's 5.8x, and by reporting the whole surface.
Fixed: terminal margin 18.0%, WACC 9.0%, 5 years, EBIT basis, net debt $4,024m, 187.5m shares. Solved for: revenue CAGR. FCF margin: 16.0% base — see the tool-defect note below.
| exit EV/EBIT → | 5.8x (warranted) | 11x | 15x | 18x | 20x | 24.4x (screen's) | 27x |
|---|---|---|---|---|---|---|---|
| FCF margin 11.0% | ~41.3 | 27.48 | 20.63 | 16.69 | 14.46 | 10.32 | 8.25 |
| FCF margin 16.0% (BASE) | ~38.0 | 25.41 | 18.99 | 15.27 | 13.15 | 9.19 | 7.21 |
| FCF margin 19.9% | ~35.7 | 23.86 | 17.75 | 14.19 | 12.14 | 8.33 | 6.41 |
| Required CAGR (base: 15.0x, 18.0%, 16.0% FCF) | 18.99% |
| Demonstrated CAGR — organic service revenue | 4–5% |
| Demonstrated CAGR — total revenue incl. bolt-on M&A | 7–11% |
| MARGIN vs organic | −14.0 to −15.0pp — FAIL |
| MARGIN vs total revenue incl. M&A | −8.0 to −12.0pp — FAIL |
| Screen's margin | +23.3pp |
| Swing | −31 to −38pp |
At the screen's own exit multiple of 24.4x, required CAGR is still 9.19% against 4–5% organic — a −4.2 to −5.2pp FAIL. The verdict does not depend on my having lowered the multiple. It flips on the demonstrated CAGR alone.
For RBA to PASS at its 4–5% organic rate, the exit multiple must exceed roughly 32x — above the 90th percentile of its own six-year EV/EBIT history and 5.5x the identity-warranted 5.8x. That is the single-sentence version of the case.
Sensitivity is over the exit multiple, never over scenario probabilities. The name fails at every multiple below ~32x and at every FCF margin in a 9pp band.
| current | 6-year (n=1,506 days) | 3-year window | |
|---|---|---|---|
| EV/Sales | 4.88x | 38th percentile (min 3.46, p25 4.52, med 6.25, max 11.17) | 76th percentile (p25 4.22, med 4.52, p75 4.87) |
| EV/EBIT | 26.09x | 24th percentile (min 21.05, p25 26.24, med 31.18, max 68.09) | 40th percentile (p25 24.53, med 27.12, p75 31.72) |
(These percentiles are computed on the screen's EV; on the corrected 10.6%-higher EV the current EV/Sales is 5.21x and EV/EBIT 28.4x, which moves RBA further up its own range in both windows.)
Regime-overlap check. The regime break is unambiguous and dated: IAA closed 20 March 2023. The six-year window opens 2020-07-27, so only 56% of it is post-IAA — the earlier 44% describes a company less than half the size with a different sector mix and no automotive-salvage business at all. The three-year window is 100% post-IAA. Status: IDENTIFIED on the 3-year window; the 6-year figures are disclosed but not used.
And the two windows disagree in the direction that matters. On the contaminated six-year window RBA looks cheap (38th / 24th percentile). On the regime-consistent three-year window it is 76th percentile on EV/Sales — i.e. expensive within its own post-IAA history. A memo that anchored on the longer series would have reached the opposite conclusion.
Build. NTM revenue = 5/12 × $4,896m + 7/12 × $5,144m = $5,041m (consensus, time-weighted). Applied at the 3-year median EV/Sales of 4.52x (50th percentile of the regime-consistent window): EV $22,785m − $4,024m net debt = $18,761m equity → $100.06. At the corrected EV basis, holding the 3-year median: $104.08, −6.0% to spot.
Target is BELOW spot, and the reason is stated rather than defaulted to. Item B16 records that 16 of 16 house targets sat below spot, which is a house view rather than valuation. This one is below spot because the name trades at the 76th percentile of its own regime-consistent multiple history while growing service revenue at 5% — a specific, falsifiable claim about RBA, not a market view.
Revision breadth is mildly positive (1 up / 0 down on both fiscal years over 30 days) — the only one of the three names with positive breadth, and it argues against the target. Stated here.
Named events inside 12 months: Q2'26 results (early May pattern suggests early August 2026 — INFERRED, see calendar), which will show whether the CC&T lot recovery of Q1'26 persists once the J.M. Wood and Smith Broughton contributions annualise into the comparative from Q3'26 and Q4'26.
Not shared with MELI or BKNG. RBA is a $4.7bn asset-disposition marketplace growing service revenue at 5% with a 6.2% ROIC; MELI is a $31.8bn LatAm marketplace-plus-lender growing 39%. One multiple across the two would be the defect valuation.md documents.
| Comparator | Why | Growth bracket at exit |
|---|---|---|
| Copart (CPRT) | the direct automotive-salvage auction competitor; the closest comparable in existence | 8–12% |
| OPENLANE (KAR) | wholesale vehicle marketplace; brackets the digital-auction take rate | 3–7% |
| ACV Auctions (ACVA) | digital wholesale vehicle auctions; brackets the higher-growth end | 15–20% |
| IAA (pre-acquisition trading history) | RBA's own acquired segment as a standalone | 5–10% |
| Ashtead / United Rentals | industrial-equipment cycle read-across for the CC&T half | 3–8% |
Bracketing test: RBA's exit-year growth under any credible path (4–8%) sits inside the KAR / IAA /
Ashtead bracket and below CPRT and ACVA. The set brackets the subject at exit. ADMISSIBLE. Note
the contrast with the screen's exit_multiple_peer_n: 106 on SIC 7389 — 106 "peers" sharing a
services SIC code with an auction house is not a growth-matched set, it is a sector average.
Named cause: a large insurance carrier re-tenders its salvage contract. RBA discloses no customer concentration (§4c) yet explains three separate line movements by reference to unnamed "significant customer contracts", and CC&T GTV fell in FY2025 partly on "the non-recurrence of certain significant customer contracts." Automotive is 53% of GTV and insurance-carrier salvage is a small-N market.
The observable loss of one national carrier would remove low-to-mid single-digit percent of GTV directly, and more through the operating leverage of a fixed yard-and-tow network. At a 16.0% operating margin instead of 18.3%, on flat GTV, applied at the 3-year p25 EV/Sales of 4.22x: roughly $88, −20.5% to spot.
Type: MEASURED. Logged and scored; it does not reject the name — the name is already rejected by the implied-path test on its own arithmetic.