Phase Space AI

Trade Construction

RB Global [RBA]

RB Global [RBA] — Trade Construction & Risk

The memo issues NO position verdict. Criteria are scored; the book decides.

Vehicle: EQUITY

Default. No LEAP proposed: measured option IV was not retrieved, so the required IV-minus-realised spread is unquantified. Trailing 252-day realised vol 27.7% — the lowest of the three names. Recording this as an explicit non-computation rather than asserting a reason the document cannot support.

The construction question does not really arise

The implied-path test fails by −14.0 to −15.0pp against organic growth and −8.0 to −12.0pp against total revenue including bolt-on M&A. It fails at the screen's own 24.4x exit multiple. It fails across a 9pp band of FCF margin. There is no exit multiple below ~32x — above the 90th percentile of its own six-year history — at which it passes.

The 12-month target is −6.0%. So the near-term and long-horizon instruments agree, which is not always the case and is worth stating when it is.

If the book wants the exposure anyway

The honest version of a long RBA case is a CC&T cycle trade, not a compounder. CC&T lots sold fell 13% in FY2025 and turned +12% in Q1'26. If that is the trough, the higher-take-rate half of the business recovers volume from a depressed base, and RBA's operating leverage on a fixed yard network is real. That case has a specific entry condition and a specific invalidation, both below. It is not the case the screen ranked, which was a 38.3% grower.
Spot $110.76
12-month target (3-yr median EV/Sales 4.52x on NTM revenue) $104, −6.0%
Downside (named: a national salvage carrier re-tenders) $88, −20.5%
Cycle-trade upside (CC&T lots back to FY2023 volume, 18.3% margin held, 3-yr p75 EV/Sales 4.87x) ~$126, +13.8%
Reward : risk on the cycle case 0.67 : 1

Invalidation — three falsifiable tests

  1. Service revenue growth below 4% on any quarterly print. It is the only revenue line unaffected by the principal/agent gross-up, and it is the growth rate the name actually has.
  2. Service take rate below 22.0%. Q1'26 was 22.42%, down 151bp year over year as the late-2024 buyer-fee change annualised out. Below 22.0% the FY2025 take-rate gain has fully reversed.
  3. Inventory sales revenue above 30% of total revenue (Q1'26: 27.3%). Past that the reported revenue line is majority principal gross-up and the consolidated EBIT margin is being diluted faster than the terminal bridge assumes.

Sizing constraint

size_bucket: large, 27.7% realised vol. Evidence grade C — the lowest of the three names, driven by the undisclosed customer concentration, the unquantified organic/inorganic split, and a usable history truncated to three years by the IAA break. Per valuation.md rule 5, uncertainty reduces position size and never the operating assumption.

A specific instruction for the book: if RBA is ranked against other names on valuation_margin_pp, it must be ranked on −14.0pp, not the screen's +23.3pp. Those two numbers differ by 37 percentage points and would place the name at opposite ends of any ranked table.